{
  "best_metric": 2.4781594276428223,
  "best_model_checkpoint": "outputs/checkpoint-3125",
  "epoch": 0.025,
  "eval_steps": 3125,
  "global_step": 3125,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 8e-06,
      "grad_norm": 1.671875,
      "learning_rate": 4e-05,
      "loss": 2.2353,
      "step": 1
    },
    {
      "epoch": 1.6e-05,
      "grad_norm": 1.6171875,
      "learning_rate": 8e-05,
      "loss": 2.2466,
      "step": 2
    },
    {
      "epoch": 2.4e-05,
      "grad_norm": 1.8125,
      "learning_rate": 0.00012,
      "loss": 2.0724,
      "step": 3
    },
    {
      "epoch": 3.2e-05,
      "grad_norm": 1.40625,
      "learning_rate": 0.00016,
      "loss": 2.273,
      "step": 4
    },
    {
      "epoch": 4e-05,
      "grad_norm": 1.65625,
      "learning_rate": 0.0002,
      "loss": 2.1555,
      "step": 5
    },
    {
      "epoch": 4.8e-05,
      "grad_norm": 1.484375,
      "learning_rate": 0.00024,
      "loss": 2.0545,
      "step": 6
    },
    {
      "epoch": 5.6e-05,
      "grad_norm": 1.34375,
      "learning_rate": 0.00028000000000000003,
      "loss": 2.2388,
      "step": 7
    },
    {
      "epoch": 6.4e-05,
      "grad_norm": 1.578125,
      "learning_rate": 0.00032,
      "loss": 2.1948,
      "step": 8
    },
    {
      "epoch": 7.2e-05,
      "grad_norm": 2.25,
      "learning_rate": 0.00035999999999999997,
      "loss": 2.1208,
      "step": 9
    },
    {
      "epoch": 8e-05,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004,
      "loss": 2.1646,
      "step": 10
    },
    {
      "epoch": 8.8e-05,
      "grad_norm": 1.53125,
      "learning_rate": 0.00044,
      "loss": 2.3298,
      "step": 11
    },
    {
      "epoch": 9.6e-05,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048,
      "loss": 2.2796,
      "step": 12
    },
    {
      "epoch": 0.000104,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0005200000000000001,
      "loss": 2.1952,
      "step": 13
    },
    {
      "epoch": 0.000112,
      "grad_norm": 1.765625,
      "learning_rate": 0.0005600000000000001,
      "loss": 2.2962,
      "step": 14
    },
    {
      "epoch": 0.00012,
      "grad_norm": 1.484375,
      "learning_rate": 0.0006,
      "loss": 2.2255,
      "step": 15
    },
    {
      "epoch": 0.000128,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00064,
      "loss": 2.2761,
      "step": 16
    },
    {
      "epoch": 0.000136,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00068,
      "loss": 2.0533,
      "step": 17
    },
    {
      "epoch": 0.000144,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0007199999999999999,
      "loss": 1.9831,
      "step": 18
    },
    {
      "epoch": 0.000152,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00076,
      "loss": 2.03,
      "step": 19
    },
    {
      "epoch": 0.00016,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0008,
      "loss": 2.2509,
      "step": 20
    },
    {
      "epoch": 0.000168,
      "grad_norm": 1.75,
      "learning_rate": 0.00084,
      "loss": 2.1374,
      "step": 21
    },
    {
      "epoch": 0.000176,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00088,
      "loss": 2.1474,
      "step": 22
    },
    {
      "epoch": 0.000184,
      "grad_norm": 1.5625,
      "learning_rate": 0.00092,
      "loss": 2.317,
      "step": 23
    },
    {
      "epoch": 0.000192,
      "grad_norm": 1.65625,
      "learning_rate": 0.00096,
      "loss": 2.2492,
      "step": 24
    },
    {
      "epoch": 0.0002,
      "grad_norm": 1.546875,
      "learning_rate": 0.001,
      "loss": 1.9702,
      "step": 25
    },
    {
      "epoch": 0.000208,
      "grad_norm": 1.640625,
      "learning_rate": 0.0009999999841452775,
      "loss": 2.177,
      "step": 26
    },
    {
      "epoch": 0.000216,
      "grad_norm": 1.484375,
      "learning_rate": 0.0009999999365811114,
      "loss": 2.1614,
      "step": 27
    },
    {
      "epoch": 0.000224,
      "grad_norm": 1.4375,
      "learning_rate": 0.0009999998573075042,
      "loss": 2.3617,
      "step": 28
    },
    {
      "epoch": 0.000232,
      "grad_norm": 1.75,
      "learning_rate": 0.0009999997463244613,
      "loss": 1.9932,
      "step": 29
    },
    {
      "epoch": 0.00024,
      "grad_norm": 1.796875,
      "learning_rate": 0.0009999996036319896,
      "loss": 2.1507,
      "step": 30
    },
    {
      "epoch": 0.000248,
      "grad_norm": 1.546875,
      "learning_rate": 0.0009999994292300982,
      "loss": 2.057,
      "step": 31
    },
    {
      "epoch": 0.000256,
      "grad_norm": 1.8046875,
      "learning_rate": 0.000999999223118798,
      "loss": 2.1284,
      "step": 32
    },
    {
      "epoch": 0.000264,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009999989852981024,
      "loss": 2.1191,
      "step": 33
    },
    {
      "epoch": 0.000272,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0009999987157680262,
      "loss": 2.0103,
      "step": 34
    },
    {
      "epoch": 0.00028,
      "grad_norm": 1.640625,
      "learning_rate": 0.0009999984145285866,
      "loss": 1.7998,
      "step": 35
    },
    {
      "epoch": 0.000288,
      "grad_norm": 1.8125,
      "learning_rate": 0.0009999980815798028,
      "loss": 1.9909,
      "step": 36
    },
    {
      "epoch": 0.000296,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009999977169216957,
      "loss": 2.24,
      "step": 37
    },
    {
      "epoch": 0.000304,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0009999973205542885,
      "loss": 1.9391,
      "step": 38
    },
    {
      "epoch": 0.000312,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0009999968924776063,
      "loss": 1.8929,
      "step": 39
    },
    {
      "epoch": 0.00032,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0009999964326916765,
      "loss": 2.0016,
      "step": 40
    },
    {
      "epoch": 0.000328,
      "grad_norm": 1.921875,
      "learning_rate": 0.000999995941196528,
      "loss": 2.1935,
      "step": 41
    },
    {
      "epoch": 0.000336,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009999954179921921,
      "loss": 2.1559,
      "step": 42
    },
    {
      "epoch": 0.000344,
      "grad_norm": 3.015625,
      "learning_rate": 0.000999994863078702,
      "loss": 1.9318,
      "step": 43
    },
    {
      "epoch": 0.000352,
      "grad_norm": 1.828125,
      "learning_rate": 0.0009999942764560925,
      "loss": 1.9298,
      "step": 44
    },
    {
      "epoch": 0.00036,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0009999936581244013,
      "loss": 1.9234,
      "step": 45
    },
    {
      "epoch": 0.000368,
      "grad_norm": 28.5,
      "learning_rate": 0.0009999930080836674,
      "loss": 1.9601,
      "step": 46
    },
    {
      "epoch": 0.000376,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009999923263339322,
      "loss": 2.0047,
      "step": 47
    },
    {
      "epoch": 0.000384,
      "grad_norm": 2.375,
      "learning_rate": 0.0009999916128752388,
      "loss": 2.0174,
      "step": 48
    },
    {
      "epoch": 0.000392,
      "grad_norm": 2.0,
      "learning_rate": 0.0009999908677076324,
      "loss": 1.7112,
      "step": 49
    },
    {
      "epoch": 0.0004,
      "grad_norm": 2.625,
      "learning_rate": 0.0009999900908311602,
      "loss": 2.126,
      "step": 50
    },
    {
      "epoch": 0.000408,
      "grad_norm": 2.125,
      "learning_rate": 0.0009999892822458716,
      "loss": 1.8007,
      "step": 51
    },
    {
      "epoch": 0.000416,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009999884419518179,
      "loss": 1.7955,
      "step": 52
    },
    {
      "epoch": 0.000424,
      "grad_norm": 3.109375,
      "learning_rate": 0.0009999875699490523,
      "loss": 2.2763,
      "step": 53
    },
    {
      "epoch": 0.000432,
      "grad_norm": 1.90625,
      "learning_rate": 0.0009999866662376302,
      "loss": 1.7904,
      "step": 54
    },
    {
      "epoch": 0.00044,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009999857308176088,
      "loss": 1.9445,
      "step": 55
    },
    {
      "epoch": 0.000448,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009999847636890475,
      "loss": 1.5594,
      "step": 56
    },
    {
      "epoch": 0.000456,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009999837648520075,
      "loss": 1.6301,
      "step": 57
    },
    {
      "epoch": 0.000464,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009999827343065524,
      "loss": 2.0072,
      "step": 58
    },
    {
      "epoch": 0.000472,
      "grad_norm": 2.375,
      "learning_rate": 0.0009999816720527475,
      "loss": 1.8644,
      "step": 59
    },
    {
      "epoch": 0.00048,
      "grad_norm": 2.171875,
      "learning_rate": 0.00099998057809066,
      "loss": 1.7375,
      "step": 60
    },
    {
      "epoch": 0.000488,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009999794524203592,
      "loss": 2.0267,
      "step": 61
    },
    {
      "epoch": 0.000496,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009999782950419168,
      "loss": 2.1871,
      "step": 62
    },
    {
      "epoch": 0.000504,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0009999771059554059,
      "loss": 1.6625,
      "step": 63
    },
    {
      "epoch": 0.000512,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009999758851609022,
      "loss": 1.6938,
      "step": 64
    },
    {
      "epoch": 0.00052,
      "grad_norm": 1.984375,
      "learning_rate": 0.000999974632658483,
      "loss": 1.6608,
      "step": 65
    },
    {
      "epoch": 0.000528,
      "grad_norm": 1.65625,
      "learning_rate": 0.0009999733484482277,
      "loss": 1.5917,
      "step": 66
    },
    {
      "epoch": 0.000536,
      "grad_norm": 6.375,
      "learning_rate": 0.0009999720325302177,
      "loss": 3.0894,
      "step": 67
    },
    {
      "epoch": 0.000544,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009999706849045365,
      "loss": 1.7119,
      "step": 68
    },
    {
      "epoch": 0.000552,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009999693055712695,
      "loss": 2.0265,
      "step": 69
    },
    {
      "epoch": 0.00056,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009999678945305044,
      "loss": 1.8266,
      "step": 70
    },
    {
      "epoch": 0.000568,
      "grad_norm": 2.25,
      "learning_rate": 0.0009999664517823304,
      "loss": 1.6983,
      "step": 71
    },
    {
      "epoch": 0.000576,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009999649773268391,
      "loss": 2.1469,
      "step": 72
    },
    {
      "epoch": 0.000584,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009999634711641242,
      "loss": 1.6582,
      "step": 73
    },
    {
      "epoch": 0.000592,
      "grad_norm": 2.4375,
      "learning_rate": 0.000999961933294281,
      "loss": 2.2505,
      "step": 74
    },
    {
      "epoch": 0.0006,
      "grad_norm": 2.375,
      "learning_rate": 0.0009999603637174071,
      "loss": 1.6958,
      "step": 75
    },
    {
      "epoch": 0.000608,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0009999587624336022,
      "loss": 1.4917,
      "step": 76
    },
    {
      "epoch": 0.000616,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009999571294429673,
      "loss": 2.3751,
      "step": 77
    },
    {
      "epoch": 0.000624,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009999554647456067,
      "loss": 1.9417,
      "step": 78
    },
    {
      "epoch": 0.000632,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009999537683416254,
      "loss": 1.9714,
      "step": 79
    },
    {
      "epoch": 0.00064,
      "grad_norm": 1.921875,
      "learning_rate": 0.0009999520402311313,
      "loss": 2.0728,
      "step": 80
    },
    {
      "epoch": 0.000648,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009999502804142338,
      "loss": 2.105,
      "step": 81
    },
    {
      "epoch": 0.000656,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009999484888910447,
      "loss": 1.8865,
      "step": 82
    },
    {
      "epoch": 0.000664,
      "grad_norm": 1.90625,
      "learning_rate": 0.0009999466656616775,
      "loss": 1.7919,
      "step": 83
    },
    {
      "epoch": 0.000672,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009999448107262479,
      "loss": 1.9277,
      "step": 84
    },
    {
      "epoch": 0.00068,
      "grad_norm": 1.921875,
      "learning_rate": 0.0009999429240848733,
      "loss": 1.7681,
      "step": 85
    },
    {
      "epoch": 0.000688,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009999410057376736,
      "loss": 1.9214,
      "step": 86
    },
    {
      "epoch": 0.000696,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009999390556847704,
      "loss": 2.3393,
      "step": 87
    },
    {
      "epoch": 0.000704,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009999370739262876,
      "loss": 2.0304,
      "step": 88
    },
    {
      "epoch": 0.000712,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009999350604623504,
      "loss": 2.2334,
      "step": 89
    },
    {
      "epoch": 0.00072,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009999330152930868,
      "loss": 1.7739,
      "step": 90
    },
    {
      "epoch": 0.000728,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009999309384186264,
      "loss": 2.692,
      "step": 91
    },
    {
      "epoch": 0.000736,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009999288298391008,
      "loss": 2.4189,
      "step": 92
    },
    {
      "epoch": 0.000744,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009999266895546442,
      "loss": 2.9039,
      "step": 93
    },
    {
      "epoch": 0.000752,
      "grad_norm": 2.734375,
      "learning_rate": 0.000999924517565392,
      "loss": 2.078,
      "step": 94
    },
    {
      "epoch": 0.00076,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009999223138714816,
      "loss": 2.3612,
      "step": 95
    },
    {
      "epoch": 0.000768,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009999200784730534,
      "loss": 2.2161,
      "step": 96
    },
    {
      "epoch": 0.000776,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009999178113702487,
      "loss": 1.9707,
      "step": 97
    },
    {
      "epoch": 0.000784,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0009999155125632116,
      "loss": 2.1359,
      "step": 98
    },
    {
      "epoch": 0.000792,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009999131820520877,
      "loss": 2.4184,
      "step": 99
    },
    {
      "epoch": 0.0008,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009999108198370248,
      "loss": 2.6635,
      "step": 100
    },
    {
      "epoch": 0.000808,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009999084259181728,
      "loss": 2.153,
      "step": 101
    },
    {
      "epoch": 0.000816,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009999060002956835,
      "loss": 2.164,
      "step": 102
    },
    {
      "epoch": 0.000824,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009999035429697105,
      "loss": 2.0433,
      "step": 103
    },
    {
      "epoch": 0.000832,
      "grad_norm": 1.90625,
      "learning_rate": 0.0009999010539404102,
      "loss": 2.3213,
      "step": 104
    },
    {
      "epoch": 0.00084,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009998985332079398,
      "loss": 2.3252,
      "step": 105
    },
    {
      "epoch": 0.000848,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009998959807724596,
      "loss": 2.6247,
      "step": 106
    },
    {
      "epoch": 0.000856,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009998933966341313,
      "loss": 2.4031,
      "step": 107
    },
    {
      "epoch": 0.000864,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009998907807931186,
      "loss": 2.4303,
      "step": 108
    },
    {
      "epoch": 0.000872,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009998881332495878,
      "loss": 2.3953,
      "step": 109
    },
    {
      "epoch": 0.00088,
      "grad_norm": 1.828125,
      "learning_rate": 0.0009998854540037066,
      "loss": 2.0928,
      "step": 110
    },
    {
      "epoch": 0.000888,
      "grad_norm": 2.015625,
      "learning_rate": 0.000999882743055645,
      "loss": 2.4892,
      "step": 111
    },
    {
      "epoch": 0.000896,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0009998800004055745,
      "loss": 2.3083,
      "step": 112
    },
    {
      "epoch": 0.000904,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009998772260536696,
      "loss": 2.4337,
      "step": 113
    },
    {
      "epoch": 0.000912,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009998744200001063,
      "loss": 2.0558,
      "step": 114
    },
    {
      "epoch": 0.00092,
      "grad_norm": 2.03125,
      "learning_rate": 0.000999871582245062,
      "loss": 2.0343,
      "step": 115
    },
    {
      "epoch": 0.000928,
      "grad_norm": 2.25,
      "learning_rate": 0.0009998687127887168,
      "loss": 2.2093,
      "step": 116
    },
    {
      "epoch": 0.000936,
      "grad_norm": 1.84375,
      "learning_rate": 0.000999865811631253,
      "loss": 2.3889,
      "step": 117
    },
    {
      "epoch": 0.000944,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009998628787728546,
      "loss": 2.6206,
      "step": 118
    },
    {
      "epoch": 0.000952,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009998599142137072,
      "loss": 2.3169,
      "step": 119
    },
    {
      "epoch": 0.00096,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009998569179539992,
      "loss": 2.3796,
      "step": 120
    },
    {
      "epoch": 0.000968,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009998538899939203,
      "loss": 2.3898,
      "step": 121
    },
    {
      "epoch": 0.000976,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009998508303336625,
      "loss": 1.8769,
      "step": 122
    },
    {
      "epoch": 0.000984,
      "grad_norm": 2.125,
      "learning_rate": 0.0009998477389734203,
      "loss": 2.646,
      "step": 123
    },
    {
      "epoch": 0.000992,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009998446159133897,
      "loss": 2.5595,
      "step": 124
    },
    {
      "epoch": 0.001,
      "grad_norm": 2.375,
      "learning_rate": 0.000999841461153768,
      "loss": 2.4175,
      "step": 125
    },
    {
      "epoch": 0.001008,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009998382746947562,
      "loss": 2.0392,
      "step": 126
    },
    {
      "epoch": 0.001016,
      "grad_norm": 2.125,
      "learning_rate": 0.0009998350565365557,
      "loss": 2.2891,
      "step": 127
    },
    {
      "epoch": 0.001024,
      "grad_norm": 2.234375,
      "learning_rate": 0.000999831806679371,
      "loss": 2.5708,
      "step": 128
    },
    {
      "epoch": 0.001032,
      "grad_norm": 2.078125,
      "learning_rate": 0.000999828525123408,
      "loss": 2.0987,
      "step": 129
    },
    {
      "epoch": 0.00104,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009998252118688749,
      "loss": 2.3873,
      "step": 130
    },
    {
      "epoch": 0.001048,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009998218669159818,
      "loss": 2.6298,
      "step": 131
    },
    {
      "epoch": 0.001056,
      "grad_norm": 2.171875,
      "learning_rate": 0.000999818490264941,
      "loss": 2.2247,
      "step": 132
    },
    {
      "epoch": 0.001064,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009998150819159664,
      "loss": 2.1807,
      "step": 133
    },
    {
      "epoch": 0.001072,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0009998116418692743,
      "loss": 2.1879,
      "step": 134
    },
    {
      "epoch": 0.00108,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009998081701250827,
      "loss": 2.2223,
      "step": 135
    },
    {
      "epoch": 0.001088,
      "grad_norm": 1.9375,
      "learning_rate": 0.0009998046666836117,
      "loss": 2.8348,
      "step": 136
    },
    {
      "epoch": 0.001096,
      "grad_norm": 2.015625,
      "learning_rate": 0.000999801131545084,
      "loss": 2.3764,
      "step": 137
    },
    {
      "epoch": 0.001104,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009997975647097232,
      "loss": 2.6021,
      "step": 138
    },
    {
      "epoch": 0.001112,
      "grad_norm": 2.484375,
      "learning_rate": 0.000999793966177756,
      "loss": 2.5467,
      "step": 139
    },
    {
      "epoch": 0.00112,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009997903359494102,
      "loss": 2.397,
      "step": 140
    },
    {
      "epoch": 0.001128,
      "grad_norm": 3.390625,
      "learning_rate": 0.0009997866740249165,
      "loss": 1.8168,
      "step": 141
    },
    {
      "epoch": 0.001136,
      "grad_norm": 2.0,
      "learning_rate": 0.0009997829804045063,
      "loss": 2.2066,
      "step": 142
    },
    {
      "epoch": 0.001144,
      "grad_norm": 2.03125,
      "learning_rate": 0.000999779255088415,
      "loss": 2.0214,
      "step": 143
    },
    {
      "epoch": 0.001152,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009997754980768778,
      "loss": 1.8775,
      "step": 144
    },
    {
      "epoch": 0.00116,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009997717093701336,
      "loss": 2.1563,
      "step": 145
    },
    {
      "epoch": 0.001168,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009997678889684222,
      "loss": 2.2833,
      "step": 146
    },
    {
      "epoch": 0.001176,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009997640368719866,
      "loss": 2.5408,
      "step": 147
    },
    {
      "epoch": 0.001184,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009997601530810705,
      "loss": 2.1147,
      "step": 148
    },
    {
      "epoch": 0.001192,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009997562375959202,
      "loss": 2.1716,
      "step": 149
    },
    {
      "epoch": 0.0012,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009997522904167844,
      "loss": 2.5352,
      "step": 150
    },
    {
      "epoch": 0.001208,
      "grad_norm": 2.46875,
      "learning_rate": 0.000999748311543913,
      "loss": 2.3243,
      "step": 151
    },
    {
      "epoch": 0.001216,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009997443009775587,
      "loss": 2.997,
      "step": 152
    },
    {
      "epoch": 0.001224,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009997402587179755,
      "loss": 2.0119,
      "step": 153
    },
    {
      "epoch": 0.001232,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009997361847654202,
      "loss": 2.3479,
      "step": 154
    },
    {
      "epoch": 0.00124,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009997320791201507,
      "loss": 1.9742,
      "step": 155
    },
    {
      "epoch": 0.001248,
      "grad_norm": 2.0625,
      "learning_rate": 0.0009997279417824278,
      "loss": 2.156,
      "step": 156
    },
    {
      "epoch": 0.001256,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0009997237727525136,
      "loss": 2.1854,
      "step": 157
    },
    {
      "epoch": 0.001264,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009997195720306724,
      "loss": 1.9683,
      "step": 158
    },
    {
      "epoch": 0.001272,
      "grad_norm": 2.40625,
      "learning_rate": 0.000999715339617171,
      "loss": 2.6756,
      "step": 159
    },
    {
      "epoch": 0.00128,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009997110755122773,
      "loss": 1.9821,
      "step": 160
    },
    {
      "epoch": 0.001288,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009997067797162624,
      "loss": 2.911,
      "step": 161
    },
    {
      "epoch": 0.001296,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009997024522293981,
      "loss": 2.2004,
      "step": 162
    },
    {
      "epoch": 0.001304,
      "grad_norm": 2.3125,
      "learning_rate": 0.000999698093051959,
      "loss": 2.353,
      "step": 163
    },
    {
      "epoch": 0.001312,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009996937021842219,
      "loss": 1.9116,
      "step": 164
    },
    {
      "epoch": 0.00132,
      "grad_norm": 3.375,
      "learning_rate": 0.0009996892796264648,
      "loss": 2.4662,
      "step": 165
    },
    {
      "epoch": 0.001328,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009996848253789685,
      "loss": 2.2417,
      "step": 166
    },
    {
      "epoch": 0.001336,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009996803394420153,
      "loss": 2.1415,
      "step": 167
    },
    {
      "epoch": 0.001344,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009996758218158899,
      "loss": 2.3052,
      "step": 168
    },
    {
      "epoch": 0.001352,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009996712725008786,
      "loss": 2.1729,
      "step": 169
    },
    {
      "epoch": 0.00136,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009996666914972698,
      "loss": 2.0963,
      "step": 170
    },
    {
      "epoch": 0.001368,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009996620788053543,
      "loss": 2.5435,
      "step": 171
    },
    {
      "epoch": 0.001376,
      "grad_norm": 2.78125,
      "learning_rate": 0.0009996574344254246,
      "loss": 2.5244,
      "step": 172
    },
    {
      "epoch": 0.001384,
      "grad_norm": 2.125,
      "learning_rate": 0.0009996527583577752,
      "loss": 1.8031,
      "step": 173
    },
    {
      "epoch": 0.001392,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009996480506027025,
      "loss": 2.3148,
      "step": 174
    },
    {
      "epoch": 0.0014,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009996433111605053,
      "loss": 2.4068,
      "step": 175
    },
    {
      "epoch": 0.001408,
      "grad_norm": 2.21875,
      "learning_rate": 0.000999638540031484,
      "loss": 2.5376,
      "step": 176
    },
    {
      "epoch": 0.001416,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009996337372159413,
      "loss": 1.9859,
      "step": 177
    },
    {
      "epoch": 0.001424,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009996289027141814,
      "loss": 2.5014,
      "step": 178
    },
    {
      "epoch": 0.001432,
      "grad_norm": 1.75,
      "learning_rate": 0.0009996240365265114,
      "loss": 2.3672,
      "step": 179
    },
    {
      "epoch": 0.00144,
      "grad_norm": 2.890625,
      "learning_rate": 0.00099961913865324,
      "loss": 2.8156,
      "step": 180
    },
    {
      "epoch": 0.001448,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009996142090946771,
      "loss": 2.1967,
      "step": 181
    },
    {
      "epoch": 0.001456,
      "grad_norm": 1.734375,
      "learning_rate": 0.0009996092478511361,
      "loss": 1.7635,
      "step": 182
    },
    {
      "epoch": 0.001464,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009996042549229311,
      "loss": 2.5308,
      "step": 183
    },
    {
      "epoch": 0.001472,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009995992303103792,
      "loss": 2.8292,
      "step": 184
    },
    {
      "epoch": 0.00148,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009995941740137985,
      "loss": 2.4503,
      "step": 185
    },
    {
      "epoch": 0.001488,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009995890860335103,
      "loss": 2.2086,
      "step": 186
    },
    {
      "epoch": 0.001496,
      "grad_norm": 2.625,
      "learning_rate": 0.0009995839663698367,
      "loss": 2.4392,
      "step": 187
    },
    {
      "epoch": 0.001504,
      "grad_norm": 1.796875,
      "learning_rate": 0.000999578815023103,
      "loss": 1.655,
      "step": 188
    },
    {
      "epoch": 0.001512,
      "grad_norm": 2.03125,
      "learning_rate": 0.000999573631993635,
      "loss": 2.6982,
      "step": 189
    },
    {
      "epoch": 0.00152,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0009995684172817623,
      "loss": 1.8588,
      "step": 190
    },
    {
      "epoch": 0.001528,
      "grad_norm": 3.203125,
      "learning_rate": 0.0009995631708878152,
      "loss": 2.2669,
      "step": 191
    },
    {
      "epoch": 0.001536,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009995578928121262,
      "loss": 2.0772,
      "step": 192
    },
    {
      "epoch": 0.001544,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009995525830550306,
      "loss": 2.3521,
      "step": 193
    },
    {
      "epoch": 0.001552,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009995472416168647,
      "loss": 2.3464,
      "step": 194
    },
    {
      "epoch": 0.00156,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009995418684979674,
      "loss": 2.2214,
      "step": 195
    },
    {
      "epoch": 0.001568,
      "grad_norm": 2.828125,
      "learning_rate": 0.0009995364636986794,
      "loss": 2.4397,
      "step": 196
    },
    {
      "epoch": 0.001576,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009995310272193434,
      "loss": 2.3908,
      "step": 197
    },
    {
      "epoch": 0.001584,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009995255590603046,
      "loss": 2.3377,
      "step": 198
    },
    {
      "epoch": 0.001592,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009995200592219091,
      "loss": 2.3603,
      "step": 199
    },
    {
      "epoch": 0.0016,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009995145277045061,
      "loss": 2.6503,
      "step": 200
    },
    {
      "epoch": 0.001608,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009995089645084465,
      "loss": 2.1431,
      "step": 201
    },
    {
      "epoch": 0.001616,
      "grad_norm": 1.9296875,
      "learning_rate": 0.000999503369634083,
      "loss": 1.8555,
      "step": 202
    },
    {
      "epoch": 0.001624,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009994977430817702,
      "loss": 2.3319,
      "step": 203
    },
    {
      "epoch": 0.001632,
      "grad_norm": 2.25,
      "learning_rate": 0.0009994920848518652,
      "loss": 2.4978,
      "step": 204
    },
    {
      "epoch": 0.00164,
      "grad_norm": 2.25,
      "learning_rate": 0.0009994863949447268,
      "loss": 2.4384,
      "step": 205
    },
    {
      "epoch": 0.001648,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009994806733607158,
      "loss": 2.4791,
      "step": 206
    },
    {
      "epoch": 0.001656,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009994749201001952,
      "loss": 2.2863,
      "step": 207
    },
    {
      "epoch": 0.001664,
      "grad_norm": 2.125,
      "learning_rate": 0.0009994691351635295,
      "loss": 2.4172,
      "step": 208
    },
    {
      "epoch": 0.001672,
      "grad_norm": 1.78125,
      "learning_rate": 0.0009994633185510858,
      "loss": 2.5061,
      "step": 209
    },
    {
      "epoch": 0.00168,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009994574702632331,
      "loss": 2.8813,
      "step": 210
    },
    {
      "epoch": 0.001688,
      "grad_norm": 2.28125,
      "learning_rate": 0.000999451590300342,
      "loss": 2.2217,
      "step": 211
    },
    {
      "epoch": 0.001696,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0009994456786627858,
      "loss": 2.3392,
      "step": 212
    },
    {
      "epoch": 0.001704,
      "grad_norm": 2.328125,
      "learning_rate": 0.000999439735350939,
      "loss": 2.237,
      "step": 213
    },
    {
      "epoch": 0.001712,
      "grad_norm": 2.921875,
      "learning_rate": 0.000999433760365179,
      "loss": 2.4355,
      "step": 214
    },
    {
      "epoch": 0.00172,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009994277537058842,
      "loss": 1.9892,
      "step": 215
    },
    {
      "epoch": 0.001728,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009994217153734357,
      "loss": 2.0212,
      "step": 216
    },
    {
      "epoch": 0.001736,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009994156453682166,
      "loss": 2.2348,
      "step": 217
    },
    {
      "epoch": 0.001744,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009994095436906118,
      "loss": 2.3287,
      "step": 218
    },
    {
      "epoch": 0.001752,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009994034103410082,
      "loss": 2.3509,
      "step": 219
    },
    {
      "epoch": 0.00176,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009993972453197948,
      "loss": 2.4201,
      "step": 220
    },
    {
      "epoch": 0.001768,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009993910486273624,
      "loss": 2.3065,
      "step": 221
    },
    {
      "epoch": 0.001776,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009993848202641045,
      "loss": 2.0683,
      "step": 222
    },
    {
      "epoch": 0.001784,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009993785602304154,
      "loss": 2.229,
      "step": 223
    },
    {
      "epoch": 0.001792,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009993722685266927,
      "loss": 2.2344,
      "step": 224
    },
    {
      "epoch": 0.0018,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009993659451533353,
      "loss": 2.3085,
      "step": 225
    },
    {
      "epoch": 0.001808,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009993595901107437,
      "loss": 2.4873,
      "step": 226
    },
    {
      "epoch": 0.001816,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009993532033993216,
      "loss": 2.0693,
      "step": 227
    },
    {
      "epoch": 0.001824,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009993467850194738,
      "loss": 2.2549,
      "step": 228
    },
    {
      "epoch": 0.001832,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009993403349716071,
      "loss": 2.0618,
      "step": 229
    },
    {
      "epoch": 0.00184,
      "grad_norm": 2.21875,
      "learning_rate": 0.000999333853256131,
      "loss": 1.9574,
      "step": 230
    },
    {
      "epoch": 0.001848,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009993273398734561,
      "loss": 2.568,
      "step": 231
    },
    {
      "epoch": 0.001856,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009993207948239958,
      "loss": 2.3608,
      "step": 232
    },
    {
      "epoch": 0.001864,
      "grad_norm": 2.625,
      "learning_rate": 0.000999314218108165,
      "loss": 2.5201,
      "step": 233
    },
    {
      "epoch": 0.001872,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009993076097263812,
      "loss": 2.1417,
      "step": 234
    },
    {
      "epoch": 0.00188,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009993009696790628,
      "loss": 2.3575,
      "step": 235
    },
    {
      "epoch": 0.001888,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009992942979666314,
      "loss": 1.9917,
      "step": 236
    },
    {
      "epoch": 0.001896,
      "grad_norm": 2.265625,
      "learning_rate": 0.00099928759458951,
      "loss": 2.4808,
      "step": 237
    },
    {
      "epoch": 0.001904,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009992808595481236,
      "loss": 2.2889,
      "step": 238
    },
    {
      "epoch": 0.001912,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009992740928428997,
      "loss": 2.1614,
      "step": 239
    },
    {
      "epoch": 0.00192,
      "grad_norm": 11.75,
      "learning_rate": 0.0009992672944742669,
      "loss": 2.3334,
      "step": 240
    },
    {
      "epoch": 0.001928,
      "grad_norm": 3.890625,
      "learning_rate": 0.0009992604644426567,
      "loss": 2.3416,
      "step": 241
    },
    {
      "epoch": 0.001936,
      "grad_norm": 2.75,
      "learning_rate": 0.000999253602748502,
      "loss": 2.6133,
      "step": 242
    },
    {
      "epoch": 0.001944,
      "grad_norm": 2.5,
      "learning_rate": 0.0009992467093922384,
      "loss": 2.4849,
      "step": 243
    },
    {
      "epoch": 0.001952,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009992397843743026,
      "loss": 2.4091,
      "step": 244
    },
    {
      "epoch": 0.00196,
      "grad_norm": 2.390625,
      "learning_rate": 0.000999232827695134,
      "loss": 1.8876,
      "step": 245
    },
    {
      "epoch": 0.001968,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009992258393551738,
      "loss": 2.4843,
      "step": 246
    },
    {
      "epoch": 0.001976,
      "grad_norm": 2.625,
      "learning_rate": 0.0009992188193548653,
      "loss": 2.4377,
      "step": 247
    },
    {
      "epoch": 0.001984,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009992117676946534,
      "loss": 1.8956,
      "step": 248
    },
    {
      "epoch": 0.001992,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009992046843749853,
      "loss": 2.4234,
      "step": 249
    },
    {
      "epoch": 0.002,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009991975693963108,
      "loss": 2.4962,
      "step": 250
    },
    {
      "epoch": 0.002008,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009991904227590802,
      "loss": 2.6841,
      "step": 251
    },
    {
      "epoch": 0.002016,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009991832444637475,
      "loss": 2.4064,
      "step": 252
    },
    {
      "epoch": 0.002024,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009991760345107676,
      "loss": 2.5219,
      "step": 253
    },
    {
      "epoch": 0.002032,
      "grad_norm": 2.421875,
      "learning_rate": 0.000999168792900598,
      "loss": 2.5258,
      "step": 254
    },
    {
      "epoch": 0.00204,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009991615196336975,
      "loss": 2.8346,
      "step": 255
    },
    {
      "epoch": 0.002048,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009991542147105277,
      "loss": 2.685,
      "step": 256
    },
    {
      "epoch": 0.002056,
      "grad_norm": 2.40625,
      "learning_rate": 0.000999146878131552,
      "loss": 2.5313,
      "step": 257
    },
    {
      "epoch": 0.002064,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009991395098972353,
      "loss": 2.9855,
      "step": 258
    },
    {
      "epoch": 0.002072,
      "grad_norm": 2.75,
      "learning_rate": 0.0009991321100080452,
      "loss": 2.438,
      "step": 259
    },
    {
      "epoch": 0.00208,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009991246784644507,
      "loss": 2.2729,
      "step": 260
    },
    {
      "epoch": 0.002088,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009991172152669235,
      "loss": 2.4812,
      "step": 261
    },
    {
      "epoch": 0.002096,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009991097204159365,
      "loss": 2.5951,
      "step": 262
    },
    {
      "epoch": 0.002104,
      "grad_norm": 2.359375,
      "learning_rate": 0.000999102193911965,
      "loss": 2.3268,
      "step": 263
    },
    {
      "epoch": 0.002112,
      "grad_norm": 2.703125,
      "learning_rate": 0.000999094635755487,
      "loss": 2.5413,
      "step": 264
    },
    {
      "epoch": 0.00212,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009990870459469809,
      "loss": 2.094,
      "step": 265
    },
    {
      "epoch": 0.002128,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009990794244869286,
      "loss": 2.3004,
      "step": 266
    },
    {
      "epoch": 0.002136,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009990717713758133,
      "loss": 2.2431,
      "step": 267
    },
    {
      "epoch": 0.002144,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009990640866141204,
      "loss": 2.0509,
      "step": 268
    },
    {
      "epoch": 0.002152,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009990563702023372,
      "loss": 2.2746,
      "step": 269
    },
    {
      "epoch": 0.00216,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009990486221409531,
      "loss": 2.6621,
      "step": 270
    },
    {
      "epoch": 0.002168,
      "grad_norm": 2.5,
      "learning_rate": 0.0009990408424304595,
      "loss": 1.9638,
      "step": 271
    },
    {
      "epoch": 0.002176,
      "grad_norm": 2.203125,
      "learning_rate": 0.00099903303107135,
      "loss": 2.3514,
      "step": 272
    },
    {
      "epoch": 0.002184,
      "grad_norm": 36.25,
      "learning_rate": 0.0009990251880641194,
      "loss": 1.9819,
      "step": 273
    },
    {
      "epoch": 0.002192,
      "grad_norm": 2.921875,
      "learning_rate": 0.0009990173134092656,
      "loss": 2.5405,
      "step": 274
    },
    {
      "epoch": 0.0022,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009990094071072877,
      "loss": 2.0224,
      "step": 275
    },
    {
      "epoch": 0.002208,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009990014691586874,
      "loss": 2.2874,
      "step": 276
    },
    {
      "epoch": 0.002216,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0009989934995639678,
      "loss": 1.9016,
      "step": 277
    },
    {
      "epoch": 0.002224,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009989854983236345,
      "loss": 2.2984,
      "step": 278
    },
    {
      "epoch": 0.002232,
      "grad_norm": 2.5625,
      "learning_rate": 0.000998977465438195,
      "loss": 2.505,
      "step": 279
    },
    {
      "epoch": 0.00224,
      "grad_norm": 3.25,
      "learning_rate": 0.0009989694009081584,
      "loss": 2.3128,
      "step": 280
    },
    {
      "epoch": 0.002248,
      "grad_norm": 2.921875,
      "learning_rate": 0.0009989613047340366,
      "loss": 2.26,
      "step": 281
    },
    {
      "epoch": 0.002256,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0009989531769163428,
      "loss": 2.1726,
      "step": 282
    },
    {
      "epoch": 0.002264,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009989450174555927,
      "loss": 2.462,
      "step": 283
    },
    {
      "epoch": 0.002272,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009989368263523034,
      "loss": 1.7459,
      "step": 284
    },
    {
      "epoch": 0.00228,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009989286036069945,
      "loss": 2.8299,
      "step": 285
    },
    {
      "epoch": 0.002288,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009989203492201876,
      "loss": 2.2145,
      "step": 286
    },
    {
      "epoch": 0.002296,
      "grad_norm": 3.28125,
      "learning_rate": 0.000998912063192406,
      "loss": 2.3917,
      "step": 287
    },
    {
      "epoch": 0.002304,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009989037455241754,
      "loss": 2.0512,
      "step": 288
    },
    {
      "epoch": 0.002312,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009988953962160232,
      "loss": 2.2264,
      "step": 289
    },
    {
      "epoch": 0.00232,
      "grad_norm": 2.875,
      "learning_rate": 0.0009988870152684787,
      "loss": 2.1718,
      "step": 290
    },
    {
      "epoch": 0.002328,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009988786026820737,
      "loss": 2.2672,
      "step": 291
    },
    {
      "epoch": 0.002336,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009988701584573417,
      "loss": 2.0606,
      "step": 292
    },
    {
      "epoch": 0.002344,
      "grad_norm": 4.96875,
      "learning_rate": 0.0009988616825948179,
      "loss": 2.4,
      "step": 293
    },
    {
      "epoch": 0.002352,
      "grad_norm": 2.625,
      "learning_rate": 0.0009988531750950403,
      "loss": 2.4624,
      "step": 294
    },
    {
      "epoch": 0.00236,
      "grad_norm": 1.90625,
      "learning_rate": 0.0009988446359585482,
      "loss": 2.0189,
      "step": 295
    },
    {
      "epoch": 0.002368,
      "grad_norm": 2.046875,
      "learning_rate": 0.000998836065185883,
      "loss": 2.3329,
      "step": 296
    },
    {
      "epoch": 0.002376,
      "grad_norm": 3.0,
      "learning_rate": 0.0009988274627775885,
      "loss": 2.4574,
      "step": 297
    },
    {
      "epoch": 0.002384,
      "grad_norm": 2.15625,
      "learning_rate": 0.00099881882873421,
      "loss": 2.3093,
      "step": 298
    },
    {
      "epoch": 0.002392,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009988101630562954,
      "loss": 2.2495,
      "step": 299
    },
    {
      "epoch": 0.0024,
      "grad_norm": 2.34375,
      "learning_rate": 0.000998801465744394,
      "loss": 2.0472,
      "step": 300
    },
    {
      "epoch": 0.002408,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009987927367990574,
      "loss": 2.8981,
      "step": 301
    },
    {
      "epoch": 0.002416,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009987839762208392,
      "loss": 2.6346,
      "step": 302
    },
    {
      "epoch": 0.002424,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009987751840102952,
      "loss": 2.5962,
      "step": 303
    },
    {
      "epoch": 0.002432,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009987663601679828,
      "loss": 2.0378,
      "step": 304
    },
    {
      "epoch": 0.00244,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009987575046944614,
      "loss": 2.4527,
      "step": 305
    },
    {
      "epoch": 0.002448,
      "grad_norm": 3.375,
      "learning_rate": 0.000998748617590293,
      "loss": 2.3003,
      "step": 306
    },
    {
      "epoch": 0.002456,
      "grad_norm": 2.234375,
      "learning_rate": 0.000998739698856041,
      "loss": 2.4695,
      "step": 307
    },
    {
      "epoch": 0.002464,
      "grad_norm": 2.859375,
      "learning_rate": 0.000998730748492271,
      "loss": 2.6476,
      "step": 308
    },
    {
      "epoch": 0.002472,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009987217664995506,
      "loss": 2.0852,
      "step": 309
    },
    {
      "epoch": 0.00248,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009987127528784496,
      "loss": 2.4339,
      "step": 310
    },
    {
      "epoch": 0.002488,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009987037076295395,
      "loss": 2.1669,
      "step": 311
    },
    {
      "epoch": 0.002496,
      "grad_norm": 1.828125,
      "learning_rate": 0.000998694630753394,
      "loss": 2.2129,
      "step": 312
    },
    {
      "epoch": 0.002504,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009986855222505887,
      "loss": 2.6125,
      "step": 313
    },
    {
      "epoch": 0.002512,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009986763821217012,
      "loss": 2.0916,
      "step": 314
    },
    {
      "epoch": 0.00252,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0009986672103673112,
      "loss": 1.9549,
      "step": 315
    },
    {
      "epoch": 0.002528,
      "grad_norm": 3.234375,
      "learning_rate": 0.0009986580069880004,
      "loss": 2.5368,
      "step": 316
    },
    {
      "epoch": 0.002536,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009986487719843525,
      "loss": 2.4054,
      "step": 317
    },
    {
      "epoch": 0.002544,
      "grad_norm": 2.453125,
      "learning_rate": 0.000998639505356953,
      "loss": 2.1829,
      "step": 318
    },
    {
      "epoch": 0.002552,
      "grad_norm": 2.578125,
      "learning_rate": 0.00099863020710639,
      "loss": 2.448,
      "step": 319
    },
    {
      "epoch": 0.00256,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009986208772332526,
      "loss": 2.6432,
      "step": 320
    },
    {
      "epoch": 0.002568,
      "grad_norm": 2.5,
      "learning_rate": 0.000998611515738133,
      "loss": 2.2065,
      "step": 321
    },
    {
      "epoch": 0.002576,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009986021226216244,
      "loss": 2.2516,
      "step": 322
    },
    {
      "epoch": 0.002584,
      "grad_norm": 2.3125,
      "learning_rate": 0.000998592697884323,
      "loss": 2.4531,
      "step": 323
    },
    {
      "epoch": 0.002592,
      "grad_norm": 2.421875,
      "learning_rate": 0.000998583241526826,
      "loss": 2.4651,
      "step": 324
    },
    {
      "epoch": 0.0026,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009985737535497337,
      "loss": 1.9364,
      "step": 325
    },
    {
      "epoch": 0.002608,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009985642339536474,
      "loss": 2.1175,
      "step": 326
    },
    {
      "epoch": 0.002616,
      "grad_norm": 2.78125,
      "learning_rate": 0.000998554682739171,
      "loss": 3.052,
      "step": 327
    },
    {
      "epoch": 0.002624,
      "grad_norm": 2.640625,
      "learning_rate": 0.00099854509990691,
      "loss": 2.5822,
      "step": 328
    },
    {
      "epoch": 0.002632,
      "grad_norm": 2.125,
      "learning_rate": 0.0009985354854574724,
      "loss": 2.2052,
      "step": 329
    },
    {
      "epoch": 0.00264,
      "grad_norm": 2.296875,
      "learning_rate": 0.000998525839391468,
      "loss": 2.0189,
      "step": 330
    },
    {
      "epoch": 0.002648,
      "grad_norm": 2.265625,
      "learning_rate": 0.000998516161709508,
      "loss": 2.2192,
      "step": 331
    },
    {
      "epoch": 0.002656,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009985064524122066,
      "loss": 2.536,
      "step": 332
    },
    {
      "epoch": 0.002664,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009984967115001795,
      "loss": 2.2632,
      "step": 333
    },
    {
      "epoch": 0.002672,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009984869389740446,
      "loss": 2.5079,
      "step": 334
    },
    {
      "epoch": 0.00268,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009984771348344212,
      "loss": 1.8112,
      "step": 335
    },
    {
      "epoch": 0.002688,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009984672990819316,
      "loss": 2.3602,
      "step": 336
    },
    {
      "epoch": 0.002696,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009984574317171992,
      "loss": 2.281,
      "step": 337
    },
    {
      "epoch": 0.002704,
      "grad_norm": 2.265625,
      "learning_rate": 0.00099844753274085,
      "loss": 2.2476,
      "step": 338
    },
    {
      "epoch": 0.002712,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009984376021535118,
      "loss": 2.4334,
      "step": 339
    },
    {
      "epoch": 0.00272,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009984276399558141,
      "loss": 2.6812,
      "step": 340
    },
    {
      "epoch": 0.002728,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009984176461483888,
      "loss": 2.2421,
      "step": 341
    },
    {
      "epoch": 0.002736,
      "grad_norm": 2.5625,
      "learning_rate": 0.00099840762073187,
      "loss": 2.5509,
      "step": 342
    },
    {
      "epoch": 0.002744,
      "grad_norm": 2.140625,
      "learning_rate": 0.000998397563706893,
      "loss": 2.5669,
      "step": 343
    },
    {
      "epoch": 0.002752,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009983874750740963,
      "loss": 1.9478,
      "step": 344
    },
    {
      "epoch": 0.00276,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009983773548341188,
      "loss": 2.7325,
      "step": 345
    },
    {
      "epoch": 0.002768,
      "grad_norm": 2.390625,
      "learning_rate": 0.000998367202987603,
      "loss": 2.4125,
      "step": 346
    },
    {
      "epoch": 0.002776,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009983570195351925,
      "loss": 2.2774,
      "step": 347
    },
    {
      "epoch": 0.002784,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009983468044775332,
      "loss": 2.6438,
      "step": 348
    },
    {
      "epoch": 0.002792,
      "grad_norm": 2.5625,
      "learning_rate": 0.000998336557815273,
      "loss": 2.801,
      "step": 349
    },
    {
      "epoch": 0.0028,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009983262795490613,
      "loss": 2.4298,
      "step": 350
    },
    {
      "epoch": 0.002808,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009983159696795503,
      "loss": 2.9898,
      "step": 351
    },
    {
      "epoch": 0.002816,
      "grad_norm": 2.328125,
      "learning_rate": 0.000998305628207394,
      "loss": 2.5354,
      "step": 352
    },
    {
      "epoch": 0.002824,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009982952551332478,
      "loss": 2.5802,
      "step": 353
    },
    {
      "epoch": 0.002832,
      "grad_norm": 2.0,
      "learning_rate": 0.0009982848504577696,
      "loss": 1.9932,
      "step": 354
    },
    {
      "epoch": 0.00284,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009982744141816197,
      "loss": 2.4366,
      "step": 355
    },
    {
      "epoch": 0.002848,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009982639463054597,
      "loss": 2.34,
      "step": 356
    },
    {
      "epoch": 0.002856,
      "grad_norm": 3.3125,
      "learning_rate": 0.0009982534468299534,
      "loss": 2.6942,
      "step": 357
    },
    {
      "epoch": 0.002864,
      "grad_norm": 2.875,
      "learning_rate": 0.0009982429157557667,
      "loss": 2.2945,
      "step": 358
    },
    {
      "epoch": 0.002872,
      "grad_norm": 3.390625,
      "learning_rate": 0.0009982323530835673,
      "loss": 2.4701,
      "step": 359
    },
    {
      "epoch": 0.00288,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009982217588140254,
      "loss": 2.1866,
      "step": 360
    },
    {
      "epoch": 0.002888,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009982111329478126,
      "loss": 2.2042,
      "step": 361
    },
    {
      "epoch": 0.002896,
      "grad_norm": 1.96875,
      "learning_rate": 0.000998200475485603,
      "loss": 1.9329,
      "step": 362
    },
    {
      "epoch": 0.002904,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009981897864280726,
      "loss": 1.9892,
      "step": 363
    },
    {
      "epoch": 0.002912,
      "grad_norm": 2.28125,
      "learning_rate": 0.000998179065775899,
      "loss": 2.2071,
      "step": 364
    },
    {
      "epoch": 0.00292,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009981683135297618,
      "loss": 2.4648,
      "step": 365
    },
    {
      "epoch": 0.002928,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009981575296903436,
      "loss": 2.3775,
      "step": 366
    },
    {
      "epoch": 0.002936,
      "grad_norm": 3.546875,
      "learning_rate": 0.0009981467142583277,
      "loss": 2.7459,
      "step": 367
    },
    {
      "epoch": 0.002944,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009981358672344004,
      "loss": 2.6901,
      "step": 368
    },
    {
      "epoch": 0.002952,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0009981249886192498,
      "loss": 2.4016,
      "step": 369
    },
    {
      "epoch": 0.00296,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009981140784135652,
      "loss": 2.4385,
      "step": 370
    },
    {
      "epoch": 0.002968,
      "grad_norm": 2.046875,
      "learning_rate": 0.0009981031366180387,
      "loss": 2.2802,
      "step": 371
    },
    {
      "epoch": 0.002976,
      "grad_norm": 2.75,
      "learning_rate": 0.0009980921632333644,
      "loss": 2.5682,
      "step": 372
    },
    {
      "epoch": 0.002984,
      "grad_norm": 2.265625,
      "learning_rate": 0.000998081158260238,
      "loss": 2.4764,
      "step": 373
    },
    {
      "epoch": 0.002992,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009980701216993578,
      "loss": 2.8371,
      "step": 374
    },
    {
      "epoch": 0.003,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009980590535514234,
      "loss": 2.3063,
      "step": 375
    },
    {
      "epoch": 0.003008,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009980479538171368,
      "loss": 2.2352,
      "step": 376
    },
    {
      "epoch": 0.003016,
      "grad_norm": 2.375,
      "learning_rate": 0.0009980368224972017,
      "loss": 2.6703,
      "step": 377
    },
    {
      "epoch": 0.003024,
      "grad_norm": 2.625,
      "learning_rate": 0.0009980256595923248,
      "loss": 2.422,
      "step": 378
    },
    {
      "epoch": 0.003032,
      "grad_norm": 2.640625,
      "learning_rate": 0.000998014465103213,
      "loss": 2.331,
      "step": 379
    },
    {
      "epoch": 0.00304,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009980032390305771,
      "loss": 2.4668,
      "step": 380
    },
    {
      "epoch": 0.003048,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009979919813751287,
      "loss": 2.5207,
      "step": 381
    },
    {
      "epoch": 0.003056,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009979806921375815,
      "loss": 2.0195,
      "step": 382
    },
    {
      "epoch": 0.003064,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009979693713186516,
      "loss": 2.4653,
      "step": 383
    },
    {
      "epoch": 0.003072,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009979580189190575,
      "loss": 2.4737,
      "step": 384
    },
    {
      "epoch": 0.00308,
      "grad_norm": 2.5,
      "learning_rate": 0.0009979466349395185,
      "loss": 2.107,
      "step": 385
    },
    {
      "epoch": 0.003088,
      "grad_norm": 5.84375,
      "learning_rate": 0.0009979352193807568,
      "loss": 1.9972,
      "step": 386
    },
    {
      "epoch": 0.003096,
      "grad_norm": 2.828125,
      "learning_rate": 0.0009979237722434963,
      "loss": 2.4289,
      "step": 387
    },
    {
      "epoch": 0.003104,
      "grad_norm": 2.96875,
      "learning_rate": 0.000997912293528463,
      "loss": 2.7841,
      "step": 388
    },
    {
      "epoch": 0.003112,
      "grad_norm": 2.421875,
      "learning_rate": 0.000997900783236385,
      "loss": 2.2794,
      "step": 389
    },
    {
      "epoch": 0.00312,
      "grad_norm": 2.296875,
      "learning_rate": 0.000997889241367992,
      "loss": 2.2905,
      "step": 390
    },
    {
      "epoch": 0.003128,
      "grad_norm": 2.40625,
      "learning_rate": 0.000997877667924016,
      "loss": 2.5872,
      "step": 391
    },
    {
      "epoch": 0.003136,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009978660629051913,
      "loss": 2.3872,
      "step": 392
    },
    {
      "epoch": 0.003144,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009978544263122536,
      "loss": 2.66,
      "step": 393
    },
    {
      "epoch": 0.003152,
      "grad_norm": 2.5,
      "learning_rate": 0.0009978427581459409,
      "loss": 2.4671,
      "step": 394
    },
    {
      "epoch": 0.00316,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009978310584069932,
      "loss": 2.3653,
      "step": 395
    },
    {
      "epoch": 0.003168,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009978193270961527,
      "loss": 2.2136,
      "step": 396
    },
    {
      "epoch": 0.003176,
      "grad_norm": 2.09375,
      "learning_rate": 0.000997807564214163,
      "loss": 1.9739,
      "step": 397
    },
    {
      "epoch": 0.003184,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009977957697617703,
      "loss": 2.2714,
      "step": 398
    },
    {
      "epoch": 0.003192,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009977839437397228,
      "loss": 2.3418,
      "step": 399
    },
    {
      "epoch": 0.0032,
      "grad_norm": 2.09375,
      "learning_rate": 0.00099777208614877,
      "loss": 2.4377,
      "step": 400
    },
    {
      "epoch": 0.003208,
      "grad_norm": 1.953125,
      "learning_rate": 0.0009977601969896643,
      "loss": 2.4357,
      "step": 401
    },
    {
      "epoch": 0.003216,
      "grad_norm": 2.875,
      "learning_rate": 0.0009977482762631594,
      "loss": 2.0659,
      "step": 402
    },
    {
      "epoch": 0.003224,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009977363239700118,
      "loss": 2.1965,
      "step": 403
    },
    {
      "epoch": 0.003232,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009977243401109788,
      "loss": 2.1372,
      "step": 404
    },
    {
      "epoch": 0.00324,
      "grad_norm": 2.3125,
      "learning_rate": 0.000997712324686821,
      "loss": 2.1244,
      "step": 405
    },
    {
      "epoch": 0.003248,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009977002776983,
      "loss": 2.7695,
      "step": 406
    },
    {
      "epoch": 0.003256,
      "grad_norm": 2.28125,
      "learning_rate": 0.00099768819914618,
      "loss": 2.1499,
      "step": 407
    },
    {
      "epoch": 0.003264,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009976760890312271,
      "loss": 2.2522,
      "step": 408
    },
    {
      "epoch": 0.003272,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009976639473542092,
      "loss": 1.9964,
      "step": 409
    },
    {
      "epoch": 0.00328,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009976517741158962,
      "loss": 2.4306,
      "step": 410
    },
    {
      "epoch": 0.003288,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009976395693170603,
      "loss": 2.1884,
      "step": 411
    },
    {
      "epoch": 0.003296,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009976273329584753,
      "loss": 1.679,
      "step": 412
    },
    {
      "epoch": 0.003304,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009976150650409174,
      "loss": 1.7586,
      "step": 413
    },
    {
      "epoch": 0.003312,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009976027655651646,
      "loss": 2.3981,
      "step": 414
    },
    {
      "epoch": 0.00332,
      "grad_norm": 2.765625,
      "learning_rate": 0.000997590434531997,
      "loss": 2.499,
      "step": 415
    },
    {
      "epoch": 0.003328,
      "grad_norm": 4.125,
      "learning_rate": 0.0009975780719421964,
      "loss": 2.2521,
      "step": 416
    },
    {
      "epoch": 0.003336,
      "grad_norm": 3.46875,
      "learning_rate": 0.000997565677796547,
      "loss": 2.3066,
      "step": 417
    },
    {
      "epoch": 0.003344,
      "grad_norm": 4.4375,
      "learning_rate": 0.0009975532520958346,
      "loss": 2.208,
      "step": 418
    },
    {
      "epoch": 0.003352,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009975407948408475,
      "loss": 2.041,
      "step": 419
    },
    {
      "epoch": 0.00336,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009975283060323756,
      "loss": 1.9143,
      "step": 420
    },
    {
      "epoch": 0.003368,
      "grad_norm": 2.078125,
      "learning_rate": 0.0009975157856712108,
      "loss": 2.0154,
      "step": 421
    },
    {
      "epoch": 0.003376,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009975032337581474,
      "loss": 2.601,
      "step": 422
    },
    {
      "epoch": 0.003384,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009974906502939814,
      "loss": 2.5768,
      "step": 423
    },
    {
      "epoch": 0.003392,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009974780352795104,
      "loss": 2.4689,
      "step": 424
    },
    {
      "epoch": 0.0034,
      "grad_norm": 2.921875,
      "learning_rate": 0.000997465388715535,
      "loss": 2.2722,
      "step": 425
    },
    {
      "epoch": 0.003408,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009974527106028567,
      "loss": 2.8003,
      "step": 426
    },
    {
      "epoch": 0.003416,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009974400009422802,
      "loss": 2.4932,
      "step": 427
    },
    {
      "epoch": 0.003424,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009974272597346107,
      "loss": 2.493,
      "step": 428
    },
    {
      "epoch": 0.003432,
      "grad_norm": 3.203125,
      "learning_rate": 0.000997414486980657,
      "loss": 2.4744,
      "step": 429
    },
    {
      "epoch": 0.00344,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009974016826812288,
      "loss": 2.3724,
      "step": 430
    },
    {
      "epoch": 0.003448,
      "grad_norm": 2.671875,
      "learning_rate": 0.000997388846837138,
      "loss": 2.9139,
      "step": 431
    },
    {
      "epoch": 0.003456,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009973759794491988,
      "loss": 2.2189,
      "step": 432
    },
    {
      "epoch": 0.003464,
      "grad_norm": 2.5,
      "learning_rate": 0.0009973630805182273,
      "loss": 1.8075,
      "step": 433
    },
    {
      "epoch": 0.003472,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009973501500450414,
      "loss": 2.6154,
      "step": 434
    },
    {
      "epoch": 0.00348,
      "grad_norm": 2.65625,
      "learning_rate": 0.000997337188030461,
      "loss": 2.2914,
      "step": 435
    },
    {
      "epoch": 0.003488,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009973241944753086,
      "loss": 2.3142,
      "step": 436
    },
    {
      "epoch": 0.003496,
      "grad_norm": 3.03125,
      "learning_rate": 0.000997311169380408,
      "loss": 2.4818,
      "step": 437
    },
    {
      "epoch": 0.003504,
      "grad_norm": 2.875,
      "learning_rate": 0.0009972981127465848,
      "loss": 2.1067,
      "step": 438
    },
    {
      "epoch": 0.003512,
      "grad_norm": 2.671875,
      "learning_rate": 0.000997285024574668,
      "loss": 2.1949,
      "step": 439
    },
    {
      "epoch": 0.00352,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009972719048654867,
      "loss": 2.3398,
      "step": 440
    },
    {
      "epoch": 0.003528,
      "grad_norm": 2.265625,
      "learning_rate": 0.000997258753619873,
      "loss": 2.0794,
      "step": 441
    },
    {
      "epoch": 0.003536,
      "grad_norm": 4.84375,
      "learning_rate": 0.0009972455708386618,
      "loss": 2.4958,
      "step": 442
    },
    {
      "epoch": 0.003544,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0009972323565226883,
      "loss": 1.6019,
      "step": 443
    },
    {
      "epoch": 0.003552,
      "grad_norm": 2.921875,
      "learning_rate": 0.000997219110672791,
      "loss": 2.1453,
      "step": 444
    },
    {
      "epoch": 0.00356,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009972058332898095,
      "loss": 2.4853,
      "step": 445
    },
    {
      "epoch": 0.003568,
      "grad_norm": 2.5,
      "learning_rate": 0.0009971925243745863,
      "loss": 2.0103,
      "step": 446
    },
    {
      "epoch": 0.003576,
      "grad_norm": 3.453125,
      "learning_rate": 0.000997179183927965,
      "loss": 2.0358,
      "step": 447
    },
    {
      "epoch": 0.003584,
      "grad_norm": 3.6875,
      "learning_rate": 0.0009971658119507924,
      "loss": 2.811,
      "step": 448
    },
    {
      "epoch": 0.003592,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009971524084439155,
      "loss": 2.0178,
      "step": 449
    },
    {
      "epoch": 0.0036,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009971389734081848,
      "loss": 2.2803,
      "step": 450
    },
    {
      "epoch": 0.003608,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009971255068444527,
      "loss": 1.9865,
      "step": 451
    },
    {
      "epoch": 0.003616,
      "grad_norm": 2.5,
      "learning_rate": 0.0009971120087535726,
      "loss": 2.4114,
      "step": 452
    },
    {
      "epoch": 0.003624,
      "grad_norm": 3.921875,
      "learning_rate": 0.000997098479136401,
      "loss": 2.1527,
      "step": 453
    },
    {
      "epoch": 0.003632,
      "grad_norm": 2.5,
      "learning_rate": 0.0009970849179937956,
      "loss": 1.8789,
      "step": 454
    },
    {
      "epoch": 0.00364,
      "grad_norm": 2.28125,
      "learning_rate": 0.000997071325326617,
      "loss": 2.3195,
      "step": 455
    },
    {
      "epoch": 0.003648,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009970577011357265,
      "loss": 2.0212,
      "step": 456
    },
    {
      "epoch": 0.003656,
      "grad_norm": 3.109375,
      "learning_rate": 0.0009970440454219888,
      "loss": 2.1487,
      "step": 457
    },
    {
      "epoch": 0.003664,
      "grad_norm": 2.625,
      "learning_rate": 0.0009970303581862692,
      "loss": 2.4279,
      "step": 458
    },
    {
      "epoch": 0.003672,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009970166394294364,
      "loss": 1.5254,
      "step": 459
    },
    {
      "epoch": 0.00368,
      "grad_norm": 2.53125,
      "learning_rate": 0.00099700288915236,
      "loss": 2.3411,
      "step": 460
    },
    {
      "epoch": 0.003688,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009969891073559123,
      "loss": 2.4888,
      "step": 461
    },
    {
      "epoch": 0.003696,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009969752940409671,
      "loss": 2.3925,
      "step": 462
    },
    {
      "epoch": 0.003704,
      "grad_norm": 2.75,
      "learning_rate": 0.0009969614492084005,
      "loss": 2.0726,
      "step": 463
    },
    {
      "epoch": 0.003712,
      "grad_norm": 2.96875,
      "learning_rate": 0.0009969475728590907,
      "loss": 3.1414,
      "step": 464
    },
    {
      "epoch": 0.00372,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009969336649939175,
      "loss": 2.2196,
      "step": 465
    },
    {
      "epoch": 0.003728,
      "grad_norm": 2.390625,
      "learning_rate": 0.000996919725613763,
      "loss": 2.3339,
      "step": 466
    },
    {
      "epoch": 0.003736,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009969057547195112,
      "loss": 2.2078,
      "step": 467
    },
    {
      "epoch": 0.003744,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009968917523120483,
      "loss": 2.2207,
      "step": 468
    },
    {
      "epoch": 0.003752,
      "grad_norm": 2.4375,
      "learning_rate": 0.000996877718392262,
      "loss": 2.4416,
      "step": 469
    },
    {
      "epoch": 0.00376,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009968636529610426,
      "loss": 1.9991,
      "step": 470
    },
    {
      "epoch": 0.003768,
      "grad_norm": 1.9453125,
      "learning_rate": 0.000996849556019282,
      "loss": 1.9126,
      "step": 471
    },
    {
      "epoch": 0.003776,
      "grad_norm": 2.578125,
      "learning_rate": 0.000996835427567874,
      "loss": 2.2536,
      "step": 472
    },
    {
      "epoch": 0.003784,
      "grad_norm": 2.578125,
      "learning_rate": 0.000996821267607715,
      "loss": 2.0077,
      "step": 473
    },
    {
      "epoch": 0.003792,
      "grad_norm": 3.125,
      "learning_rate": 0.0009968070761397028,
      "loss": 2.1747,
      "step": 474
    },
    {
      "epoch": 0.0038,
      "grad_norm": 3.25,
      "learning_rate": 0.0009967928531647372,
      "loss": 2.5177,
      "step": 475
    },
    {
      "epoch": 0.003808,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009967785986837207,
      "loss": 2.4961,
      "step": 476
    },
    {
      "epoch": 0.003816,
      "grad_norm": 2.296875,
      "learning_rate": 0.000996764312697557,
      "loss": 1.9746,
      "step": 477
    },
    {
      "epoch": 0.003824,
      "grad_norm": 4.0,
      "learning_rate": 0.000996749995207152,
      "loss": 2.2243,
      "step": 478
    },
    {
      "epoch": 0.003832,
      "grad_norm": 2.1875,
      "learning_rate": 0.000996735646213414,
      "loss": 1.9488,
      "step": 479
    },
    {
      "epoch": 0.00384,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009967212657172528,
      "loss": 2.1615,
      "step": 480
    },
    {
      "epoch": 0.003848,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009967068537195805,
      "loss": 3.115,
      "step": 481
    },
    {
      "epoch": 0.003856,
      "grad_norm": 3.65625,
      "learning_rate": 0.0009966924102213107,
      "loss": 2.2303,
      "step": 482
    },
    {
      "epoch": 0.003864,
      "grad_norm": 2.40625,
      "learning_rate": 0.00099667793522336,
      "loss": 1.8144,
      "step": 483
    },
    {
      "epoch": 0.003872,
      "grad_norm": 2.46875,
      "learning_rate": 0.000996663428726646,
      "loss": 2.1236,
      "step": 484
    },
    {
      "epoch": 0.00388,
      "grad_norm": 3.171875,
      "learning_rate": 0.000996648890732089,
      "loss": 2.6643,
      "step": 485
    },
    {
      "epoch": 0.003888,
      "grad_norm": 2.921875,
      "learning_rate": 0.0009966343212406104,
      "loss": 2.3852,
      "step": 486
    },
    {
      "epoch": 0.003896,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009966197202531347,
      "loss": 2.2391,
      "step": 487
    },
    {
      "epoch": 0.003904,
      "grad_norm": 2.25,
      "learning_rate": 0.000996605087770588,
      "loss": 2.1362,
      "step": 488
    },
    {
      "epoch": 0.003912,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009965904237938977,
      "loss": 2.7153,
      "step": 489
    },
    {
      "epoch": 0.00392,
      "grad_norm": 2.640625,
      "learning_rate": 0.000996575728323994,
      "loss": 1.8512,
      "step": 490
    },
    {
      "epoch": 0.003928,
      "grad_norm": 2.671875,
      "learning_rate": 0.000996561001361809,
      "loss": 2.6052,
      "step": 491
    },
    {
      "epoch": 0.003936,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009965462429082769,
      "loss": 2.6124,
      "step": 492
    },
    {
      "epoch": 0.003944,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009965314529643333,
      "loss": 2.0278,
      "step": 493
    },
    {
      "epoch": 0.003952,
      "grad_norm": 2.828125,
      "learning_rate": 0.000996516631530916,
      "loss": 2.4138,
      "step": 494
    },
    {
      "epoch": 0.00396,
      "grad_norm": 2.5,
      "learning_rate": 0.0009965017786089654,
      "loss": 2.2343,
      "step": 495
    },
    {
      "epoch": 0.003968,
      "grad_norm": 3.921875,
      "learning_rate": 0.0009964868941994233,
      "loss": 2.0331,
      "step": 496
    },
    {
      "epoch": 0.003976,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009964719783032334,
      "loss": 2.5999,
      "step": 497
    },
    {
      "epoch": 0.003984,
      "grad_norm": 2.3125,
      "learning_rate": 0.000996457030921342,
      "loss": 2.6078,
      "step": 498
    },
    {
      "epoch": 0.003992,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009964420520546969,
      "loss": 2.2722,
      "step": 499
    },
    {
      "epoch": 0.004,
      "grad_norm": 2.28125,
      "learning_rate": 0.000996427041704248,
      "loss": 2.6602,
      "step": 500
    },
    {
      "epoch": 0.004008,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009964119998709473,
      "loss": 2.2992,
      "step": 501
    },
    {
      "epoch": 0.004016,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009963969265557487,
      "loss": 2.3706,
      "step": 502
    },
    {
      "epoch": 0.004024,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009963818217596082,
      "loss": 2.6383,
      "step": 503
    },
    {
      "epoch": 0.004032,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009963666854834838,
      "loss": 2.4167,
      "step": 504
    },
    {
      "epoch": 0.00404,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009963515177283352,
      "loss": 2.2844,
      "step": 505
    },
    {
      "epoch": 0.004048,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009963363184951244,
      "loss": 2.0252,
      "step": 506
    },
    {
      "epoch": 0.004056,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009963210877848155,
      "loss": 2.4154,
      "step": 507
    },
    {
      "epoch": 0.004064,
      "grad_norm": 3.0,
      "learning_rate": 0.0009963058255983742,
      "loss": 2.3374,
      "step": 508
    },
    {
      "epoch": 0.004072,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009962905319367682,
      "loss": 2.8243,
      "step": 509
    },
    {
      "epoch": 0.00408,
      "grad_norm": 2.453125,
      "learning_rate": 0.000996275206800968,
      "loss": 2.251,
      "step": 510
    },
    {
      "epoch": 0.004088,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009962598501919453,
      "loss": 2.3201,
      "step": 511
    },
    {
      "epoch": 0.004096,
      "grad_norm": 3.125,
      "learning_rate": 0.0009962444621106736,
      "loss": 2.5702,
      "step": 512
    },
    {
      "epoch": 0.004104,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009962290425581293,
      "loss": 2.6587,
      "step": 513
    },
    {
      "epoch": 0.004112,
      "grad_norm": 2.28125,
      "learning_rate": 0.00099621359153529,
      "loss": 2.0824,
      "step": 514
    },
    {
      "epoch": 0.00412,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009961981090431356,
      "loss": 2.2835,
      "step": 515
    },
    {
      "epoch": 0.004128,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009961825950826483,
      "loss": 2.1895,
      "step": 516
    },
    {
      "epoch": 0.004136,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009961670496548114,
      "loss": 2.4409,
      "step": 517
    },
    {
      "epoch": 0.004144,
      "grad_norm": 2.3125,
      "learning_rate": 0.000996151472760611,
      "loss": 2.9021,
      "step": 518
    },
    {
      "epoch": 0.004152,
      "grad_norm": 4.96875,
      "learning_rate": 0.0009961358644010353,
      "loss": 2.2175,
      "step": 519
    },
    {
      "epoch": 0.00416,
      "grad_norm": 2.53125,
      "learning_rate": 0.000996120224577074,
      "loss": 2.4303,
      "step": 520
    },
    {
      "epoch": 0.004168,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009961045532897188,
      "loss": 2.6017,
      "step": 521
    },
    {
      "epoch": 0.004176,
      "grad_norm": 2.75,
      "learning_rate": 0.0009960888505399638,
      "loss": 2.2876,
      "step": 522
    },
    {
      "epoch": 0.004184,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009960731163288043,
      "loss": 2.3491,
      "step": 523
    },
    {
      "epoch": 0.004192,
      "grad_norm": 3.125,
      "learning_rate": 0.000996057350657239,
      "loss": 2.1683,
      "step": 524
    },
    {
      "epoch": 0.0042,
      "grad_norm": 2.203125,
      "learning_rate": 0.000996041553526267,
      "loss": 2.0598,
      "step": 525
    },
    {
      "epoch": 0.004208,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009960257249368905,
      "loss": 2.0296,
      "step": 526
    },
    {
      "epoch": 0.004216,
      "grad_norm": 3.5,
      "learning_rate": 0.0009960098648901133,
      "loss": 2.6361,
      "step": 527
    },
    {
      "epoch": 0.004224,
      "grad_norm": 3.28125,
      "learning_rate": 0.000995993973386941,
      "loss": 2.6909,
      "step": 528
    },
    {
      "epoch": 0.004232,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009959780504283818,
      "loss": 2.0399,
      "step": 529
    },
    {
      "epoch": 0.00424,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009959620960154454,
      "loss": 2.5895,
      "step": 530
    },
    {
      "epoch": 0.004248,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009959461101491432,
      "loss": 2.373,
      "step": 531
    },
    {
      "epoch": 0.004256,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009959300928304895,
      "loss": 2.3215,
      "step": 532
    },
    {
      "epoch": 0.004264,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009959140440605001,
      "loss": 2.3436,
      "step": 533
    },
    {
      "epoch": 0.004272,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009958979638401925,
      "loss": 2.0572,
      "step": 534
    },
    {
      "epoch": 0.00428,
      "grad_norm": 2.125,
      "learning_rate": 0.0009958818521705866,
      "loss": 2.1372,
      "step": 535
    },
    {
      "epoch": 0.004288,
      "grad_norm": 2.375,
      "learning_rate": 0.0009958657090527042,
      "loss": 2.4991,
      "step": 536
    },
    {
      "epoch": 0.004296,
      "grad_norm": 2.5,
      "learning_rate": 0.000995849534487569,
      "loss": 2.5641,
      "step": 537
    },
    {
      "epoch": 0.004304,
      "grad_norm": 2.34375,
      "learning_rate": 0.000995833328476207,
      "loss": 2.0741,
      "step": 538
    },
    {
      "epoch": 0.004312,
      "grad_norm": 2.25,
      "learning_rate": 0.0009958170910196458,
      "loss": 2.2448,
      "step": 539
    },
    {
      "epoch": 0.00432,
      "grad_norm": 2.4375,
      "learning_rate": 0.000995800822118915,
      "loss": 2.3521,
      "step": 540
    },
    {
      "epoch": 0.004328,
      "grad_norm": 2.625,
      "learning_rate": 0.0009957845217750468,
      "loss": 2.1974,
      "step": 541
    },
    {
      "epoch": 0.004336,
      "grad_norm": 2.875,
      "learning_rate": 0.0009957681899890746,
      "loss": 2.8252,
      "step": 542
    },
    {
      "epoch": 0.004344,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009957518267620343,
      "loss": 2.6459,
      "step": 543
    },
    {
      "epoch": 0.004352,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009957354320949635,
      "loss": 1.9224,
      "step": 544
    },
    {
      "epoch": 0.00436,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0009957190059889018,
      "loss": 1.8312,
      "step": 545
    },
    {
      "epoch": 0.004368,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009957025484448913,
      "loss": 2.3098,
      "step": 546
    },
    {
      "epoch": 0.004376,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009956860594639758,
      "loss": 2.3105,
      "step": 547
    },
    {
      "epoch": 0.004384,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009956695390472006,
      "loss": 2.217,
      "step": 548
    },
    {
      "epoch": 0.004392,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009956529871956133,
      "loss": 2.0734,
      "step": 549
    },
    {
      "epoch": 0.0044,
      "grad_norm": 2.609375,
      "learning_rate": 0.000995636403910264,
      "loss": 2.5188,
      "step": 550
    },
    {
      "epoch": 0.004408,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009956197891922045,
      "loss": 1.8601,
      "step": 551
    },
    {
      "epoch": 0.004416,
      "grad_norm": 2.84375,
      "learning_rate": 0.000995603143042488,
      "loss": 2.7357,
      "step": 552
    },
    {
      "epoch": 0.004424,
      "grad_norm": 2.5,
      "learning_rate": 0.0009955864654621707,
      "loss": 2.0502,
      "step": 553
    },
    {
      "epoch": 0.004432,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009955697564523096,
      "loss": 1.8601,
      "step": 554
    },
    {
      "epoch": 0.00444,
      "grad_norm": 2.71875,
      "learning_rate": 0.000995553016013965,
      "loss": 2.4712,
      "step": 555
    },
    {
      "epoch": 0.004448,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009955362441481984,
      "loss": 2.1726,
      "step": 556
    },
    {
      "epoch": 0.004456,
      "grad_norm": 2.859375,
      "learning_rate": 0.0009955194408560733,
      "loss": 2.7408,
      "step": 557
    },
    {
      "epoch": 0.004464,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009955026061386554,
      "loss": 2.6763,
      "step": 558
    },
    {
      "epoch": 0.004472,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009954857399970124,
      "loss": 2.3973,
      "step": 559
    },
    {
      "epoch": 0.00448,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009954688424322138,
      "loss": 2.1744,
      "step": 560
    },
    {
      "epoch": 0.004488,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009954519134453316,
      "loss": 2.2592,
      "step": 561
    },
    {
      "epoch": 0.004496,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009954349530374388,
      "loss": 2.0077,
      "step": 562
    },
    {
      "epoch": 0.004504,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009954179612096114,
      "loss": 2.263,
      "step": 563
    },
    {
      "epoch": 0.004512,
      "grad_norm": 3.671875,
      "learning_rate": 0.0009954009379629272,
      "loss": 2.0888,
      "step": 564
    },
    {
      "epoch": 0.00452,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009953838832984655,
      "loss": 2.1235,
      "step": 565
    },
    {
      "epoch": 0.004528,
      "grad_norm": 3.6875,
      "learning_rate": 0.0009953667972173077,
      "loss": 2.4951,
      "step": 566
    },
    {
      "epoch": 0.004536,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009953496797205376,
      "loss": 2.1148,
      "step": 567
    },
    {
      "epoch": 0.004544,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009953325308092412,
      "loss": 2.5784,
      "step": 568
    },
    {
      "epoch": 0.004552,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009953153504845054,
      "loss": 2.3503,
      "step": 569
    },
    {
      "epoch": 0.00456,
      "grad_norm": 2.296875,
      "learning_rate": 0.00099529813874742,
      "loss": 2.1618,
      "step": 570
    },
    {
      "epoch": 0.004568,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009952808955990765,
      "loss": 1.9721,
      "step": 571
    },
    {
      "epoch": 0.004576,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009952636210405685,
      "loss": 2.085,
      "step": 572
    },
    {
      "epoch": 0.004584,
      "grad_norm": 3.0,
      "learning_rate": 0.0009952463150729918,
      "loss": 2.3833,
      "step": 573
    },
    {
      "epoch": 0.004592,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009952289776974435,
      "loss": 2.1663,
      "step": 574
    },
    {
      "epoch": 0.0046,
      "grad_norm": 3.0625,
      "learning_rate": 0.0009952116089150232,
      "loss": 2.682,
      "step": 575
    },
    {
      "epoch": 0.004608,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009951942087268325,
      "loss": 2.4682,
      "step": 576
    },
    {
      "epoch": 0.004616,
      "grad_norm": 3.640625,
      "learning_rate": 0.000995176777133975,
      "loss": 3.0183,
      "step": 577
    },
    {
      "epoch": 0.004624,
      "grad_norm": 2.859375,
      "learning_rate": 0.000995159314137556,
      "loss": 2.0004,
      "step": 578
    },
    {
      "epoch": 0.004632,
      "grad_norm": 2.75,
      "learning_rate": 0.000995141819738683,
      "loss": 2.5587,
      "step": 579
    },
    {
      "epoch": 0.00464,
      "grad_norm": 2.5,
      "learning_rate": 0.0009951242939384658,
      "loss": 2.4154,
      "step": 580
    },
    {
      "epoch": 0.004648,
      "grad_norm": 2.625,
      "learning_rate": 0.0009951067367380154,
      "loss": 2.3805,
      "step": 581
    },
    {
      "epoch": 0.004656,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009950891481384455,
      "loss": 2.3163,
      "step": 582
    },
    {
      "epoch": 0.004664,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009950715281408715,
      "loss": 2.3556,
      "step": 583
    },
    {
      "epoch": 0.004672,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009950538767464108,
      "loss": 2.1463,
      "step": 584
    },
    {
      "epoch": 0.00468,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009950361939561829,
      "loss": 2.4619,
      "step": 585
    },
    {
      "epoch": 0.004688,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009950184797713093,
      "loss": 3.4063,
      "step": 586
    },
    {
      "epoch": 0.004696,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009950007341929133,
      "loss": 1.9453,
      "step": 587
    },
    {
      "epoch": 0.004704,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009949829572221203,
      "loss": 2.14,
      "step": 588
    },
    {
      "epoch": 0.004712,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009949651488600577,
      "loss": 2.8374,
      "step": 589
    },
    {
      "epoch": 0.00472,
      "grad_norm": 2.03125,
      "learning_rate": 0.0009949473091078551,
      "loss": 1.8569,
      "step": 590
    },
    {
      "epoch": 0.004728,
      "grad_norm": 2.78125,
      "learning_rate": 0.0009949294379666433,
      "loss": 2.5642,
      "step": 591
    },
    {
      "epoch": 0.004736,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009949115354375565,
      "loss": 2.4275,
      "step": 592
    },
    {
      "epoch": 0.004744,
      "grad_norm": 2.375,
      "learning_rate": 0.0009948936015217293,
      "loss": 1.9503,
      "step": 593
    },
    {
      "epoch": 0.004752,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009948756362202996,
      "loss": 2.5101,
      "step": 594
    },
    {
      "epoch": 0.00476,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009948576395344062,
      "loss": 2.4073,
      "step": 595
    },
    {
      "epoch": 0.004768,
      "grad_norm": 2.84375,
      "learning_rate": 0.000994839611465191,
      "loss": 2.1728,
      "step": 596
    },
    {
      "epoch": 0.004776,
      "grad_norm": 2.8125,
      "learning_rate": 0.000994821552013797,
      "loss": 2.0661,
      "step": 597
    },
    {
      "epoch": 0.004784,
      "grad_norm": 2.25,
      "learning_rate": 0.0009948034611813694,
      "loss": 2.0198,
      "step": 598
    },
    {
      "epoch": 0.004792,
      "grad_norm": 2.078125,
      "learning_rate": 0.000994785338969056,
      "loss": 2.1655,
      "step": 599
    },
    {
      "epoch": 0.0048,
      "grad_norm": 2.140625,
      "learning_rate": 0.0009947671853780054,
      "loss": 2.0912,
      "step": 600
    },
    {
      "epoch": 0.004808,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009947490004093694,
      "loss": 1.86,
      "step": 601
    },
    {
      "epoch": 0.004816,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009947307840643012,
      "loss": 2.2482,
      "step": 602
    },
    {
      "epoch": 0.004824,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009947125363439558,
      "loss": 2.8052,
      "step": 603
    },
    {
      "epoch": 0.004832,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009946942572494905,
      "loss": 2.2962,
      "step": 604
    },
    {
      "epoch": 0.00484,
      "grad_norm": 2.03125,
      "learning_rate": 0.000994675946782065,
      "loss": 2.1352,
      "step": 605
    },
    {
      "epoch": 0.004848,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009946576049428399,
      "loss": 1.8982,
      "step": 606
    },
    {
      "epoch": 0.004856,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009946392317329788,
      "loss": 2.1391,
      "step": 607
    },
    {
      "epoch": 0.004864,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009946208271536469,
      "loss": 2.183,
      "step": 608
    },
    {
      "epoch": 0.004872,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009946023912060112,
      "loss": 2.7096,
      "step": 609
    },
    {
      "epoch": 0.00488,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009945839238912411,
      "loss": 2.4638,
      "step": 610
    },
    {
      "epoch": 0.004888,
      "grad_norm": 2.96875,
      "learning_rate": 0.0009945654252105076,
      "loss": 2.2467,
      "step": 611
    },
    {
      "epoch": 0.004896,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009945468951649838,
      "loss": 2.041,
      "step": 612
    },
    {
      "epoch": 0.004904,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009945283337558452,
      "loss": 2.5474,
      "step": 613
    },
    {
      "epoch": 0.004912,
      "grad_norm": 2.25,
      "learning_rate": 0.0009945097409842687,
      "loss": 2.1981,
      "step": 614
    },
    {
      "epoch": 0.00492,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009944911168514336,
      "loss": 2.069,
      "step": 615
    },
    {
      "epoch": 0.004928,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009944724613585208,
      "loss": 2.4677,
      "step": 616
    },
    {
      "epoch": 0.004936,
      "grad_norm": 2.921875,
      "learning_rate": 0.0009944537745067133,
      "loss": 2.4839,
      "step": 617
    },
    {
      "epoch": 0.004944,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009944350562971966,
      "loss": 2.5064,
      "step": 618
    },
    {
      "epoch": 0.004952,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009944163067311575,
      "loss": 2.257,
      "step": 619
    },
    {
      "epoch": 0.00496,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009943975258097852,
      "loss": 2.1787,
      "step": 620
    },
    {
      "epoch": 0.004968,
      "grad_norm": 3.109375,
      "learning_rate": 0.000994378713534271,
      "loss": 2.9127,
      "step": 621
    },
    {
      "epoch": 0.004976,
      "grad_norm": 2.09375,
      "learning_rate": 0.0009943598699058073,
      "loss": 1.9914,
      "step": 622
    },
    {
      "epoch": 0.004984,
      "grad_norm": 3.546875,
      "learning_rate": 0.0009943409949255896,
      "loss": 2.2802,
      "step": 623
    },
    {
      "epoch": 0.004992,
      "grad_norm": 2.984375,
      "learning_rate": 0.000994322088594815,
      "loss": 2.5863,
      "step": 624
    },
    {
      "epoch": 0.005,
      "grad_norm": 2.0,
      "learning_rate": 0.0009943031509146824,
      "loss": 2.0677,
      "step": 625
    },
    {
      "epoch": 0.005008,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009942841818863927,
      "loss": 2.4079,
      "step": 626
    },
    {
      "epoch": 0.005016,
      "grad_norm": 3.46875,
      "learning_rate": 0.000994265181511149,
      "loss": 2.0485,
      "step": 627
    },
    {
      "epoch": 0.005024,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009942461497901563,
      "loss": 2.8185,
      "step": 628
    },
    {
      "epoch": 0.005032,
      "grad_norm": 2.015625,
      "learning_rate": 0.0009942270867246215,
      "loss": 2.1898,
      "step": 629
    },
    {
      "epoch": 0.00504,
      "grad_norm": 2.125,
      "learning_rate": 0.0009942079923157534,
      "loss": 2.1861,
      "step": 630
    },
    {
      "epoch": 0.005048,
      "grad_norm": 2.5,
      "learning_rate": 0.0009941888665647633,
      "loss": 1.9236,
      "step": 631
    },
    {
      "epoch": 0.005056,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009941697094728642,
      "loss": 2.185,
      "step": 632
    },
    {
      "epoch": 0.005064,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009941505210412705,
      "loss": 2.6772,
      "step": 633
    },
    {
      "epoch": 0.005072,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009941313012711995,
      "loss": 2.1723,
      "step": 634
    },
    {
      "epoch": 0.00508,
      "grad_norm": 3.109375,
      "learning_rate": 0.00099411205016387,
      "loss": 2.828,
      "step": 635
    },
    {
      "epoch": 0.005088,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009940927677205026,
      "loss": 2.5504,
      "step": 636
    },
    {
      "epoch": 0.005096,
      "grad_norm": 2.625,
      "learning_rate": 0.0009940734539423208,
      "loss": 2.3441,
      "step": 637
    },
    {
      "epoch": 0.005104,
      "grad_norm": 3.859375,
      "learning_rate": 0.0009940541088305492,
      "loss": 2.4745,
      "step": 638
    },
    {
      "epoch": 0.005112,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009940347323864143,
      "loss": 2.4442,
      "step": 639
    },
    {
      "epoch": 0.00512,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009940153246111453,
      "loss": 2.7647,
      "step": 640
    },
    {
      "epoch": 0.005128,
      "grad_norm": 2.1875,
      "learning_rate": 0.000993995885505973,
      "loss": 2.275,
      "step": 641
    },
    {
      "epoch": 0.005136,
      "grad_norm": 4.375,
      "learning_rate": 0.00099397641507213,
      "loss": 1.8479,
      "step": 642
    },
    {
      "epoch": 0.005144,
      "grad_norm": 2.546875,
      "learning_rate": 0.000993956913310851,
      "loss": 2.4127,
      "step": 643
    },
    {
      "epoch": 0.005152,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009939373802233733,
      "loss": 2.1718,
      "step": 644
    },
    {
      "epoch": 0.00516,
      "grad_norm": 2.28125,
      "learning_rate": 0.0009939178158109353,
      "loss": 2.2501,
      "step": 645
    },
    {
      "epoch": 0.005168,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009938982200747778,
      "loss": 2.7814,
      "step": 646
    },
    {
      "epoch": 0.005176,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009938785930161437,
      "loss": 2.2868,
      "step": 647
    },
    {
      "epoch": 0.005184,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009938589346362773,
      "loss": 2.4013,
      "step": 648
    },
    {
      "epoch": 0.005192,
      "grad_norm": 2.96875,
      "learning_rate": 0.0009938392449364256,
      "loss": 2.2646,
      "step": 649
    },
    {
      "epoch": 0.0052,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009938195239178375,
      "loss": 2.7358,
      "step": 650
    },
    {
      "epoch": 0.005208,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009937997715817632,
      "loss": 2.1794,
      "step": 651
    },
    {
      "epoch": 0.005216,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009937799879294559,
      "loss": 2.3841,
      "step": 652
    },
    {
      "epoch": 0.005224,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00099376017296217,
      "loss": 1.7848,
      "step": 653
    },
    {
      "epoch": 0.005232,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009937403266811618,
      "loss": 2.3441,
      "step": 654
    },
    {
      "epoch": 0.00524,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009937204490876904,
      "loss": 2.3864,
      "step": 655
    },
    {
      "epoch": 0.005248,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009937005401830165,
      "loss": 1.9804,
      "step": 656
    },
    {
      "epoch": 0.005256,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009936805999684025,
      "loss": 2.5116,
      "step": 657
    },
    {
      "epoch": 0.005264,
      "grad_norm": 3.125,
      "learning_rate": 0.000993660628445113,
      "loss": 2.4067,
      "step": 658
    },
    {
      "epoch": 0.005272,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009936406256144143,
      "loss": 1.773,
      "step": 659
    },
    {
      "epoch": 0.00528,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009936205914775754,
      "loss": 1.8227,
      "step": 660
    },
    {
      "epoch": 0.005288,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009936005260358664,
      "loss": 2.6003,
      "step": 661
    },
    {
      "epoch": 0.005296,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009935804292905603,
      "loss": 1.5483,
      "step": 662
    },
    {
      "epoch": 0.005304,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009935603012429313,
      "loss": 2.2882,
      "step": 663
    },
    {
      "epoch": 0.005312,
      "grad_norm": 3.15625,
      "learning_rate": 0.0009935401418942561,
      "loss": 2.7045,
      "step": 664
    },
    {
      "epoch": 0.00532,
      "grad_norm": 1.9609375,
      "learning_rate": 0.000993519951245813,
      "loss": 1.7143,
      "step": 665
    },
    {
      "epoch": 0.005328,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009934997292988825,
      "loss": 2.2947,
      "step": 666
    },
    {
      "epoch": 0.005336,
      "grad_norm": 2.359375,
      "learning_rate": 0.000993479476054747,
      "loss": 2.1091,
      "step": 667
    },
    {
      "epoch": 0.005344,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009934591915146912,
      "loss": 2.3429,
      "step": 668
    },
    {
      "epoch": 0.005352,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009934388756800013,
      "loss": 1.9753,
      "step": 669
    },
    {
      "epoch": 0.00536,
      "grad_norm": 3.640625,
      "learning_rate": 0.000993418528551966,
      "loss": 2.479,
      "step": 670
    },
    {
      "epoch": 0.005368,
      "grad_norm": 2.46875,
      "learning_rate": 0.000993398150131875,
      "loss": 2.3421,
      "step": 671
    },
    {
      "epoch": 0.005376,
      "grad_norm": 2.75,
      "learning_rate": 0.0009933777404210213,
      "loss": 2.5982,
      "step": 672
    },
    {
      "epoch": 0.005384,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009933572994206992,
      "loss": 2.3622,
      "step": 673
    },
    {
      "epoch": 0.005392,
      "grad_norm": 2.765625,
      "learning_rate": 0.000993336827132205,
      "loss": 2.5144,
      "step": 674
    },
    {
      "epoch": 0.0054,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009933163235568369,
      "loss": 2.4361,
      "step": 675
    },
    {
      "epoch": 0.005408,
      "grad_norm": 3.125,
      "learning_rate": 0.000993295788695895,
      "loss": 2.3388,
      "step": 676
    },
    {
      "epoch": 0.005416,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009932752225506822,
      "loss": 1.6465,
      "step": 677
    },
    {
      "epoch": 0.005424,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009932546251225022,
      "loss": 2.2241,
      "step": 678
    },
    {
      "epoch": 0.005432,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009932339964126618,
      "loss": 2.1922,
      "step": 679
    },
    {
      "epoch": 0.00544,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009932133364224688,
      "loss": 2.9018,
      "step": 680
    },
    {
      "epoch": 0.005448,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009931926451532337,
      "loss": 1.8611,
      "step": 681
    },
    {
      "epoch": 0.005456,
      "grad_norm": 3.609375,
      "learning_rate": 0.0009931719226062687,
      "loss": 3.0274,
      "step": 682
    },
    {
      "epoch": 0.005464,
      "grad_norm": 2.46875,
      "learning_rate": 0.000993151168782888,
      "loss": 2.7502,
      "step": 683
    },
    {
      "epoch": 0.005472,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009931303836844074,
      "loss": 2.4806,
      "step": 684
    },
    {
      "epoch": 0.00548,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009931095673121454,
      "loss": 2.423,
      "step": 685
    },
    {
      "epoch": 0.005488,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009930887196674225,
      "loss": 2.7128,
      "step": 686
    },
    {
      "epoch": 0.005496,
      "grad_norm": 2.4375,
      "learning_rate": 0.00099306784075156,
      "loss": 2.4644,
      "step": 687
    },
    {
      "epoch": 0.005504,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009930469305658828,
      "loss": 2.2745,
      "step": 688
    },
    {
      "epoch": 0.005512,
      "grad_norm": 3.125,
      "learning_rate": 0.0009930259891117166,
      "loss": 2.1802,
      "step": 689
    },
    {
      "epoch": 0.00552,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009930050163903896,
      "loss": 2.3705,
      "step": 690
    },
    {
      "epoch": 0.005528,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009929840124032318,
      "loss": 2.3073,
      "step": 691
    },
    {
      "epoch": 0.005536,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009929629771515754,
      "loss": 2.7613,
      "step": 692
    },
    {
      "epoch": 0.005544,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009929419106367542,
      "loss": 2.1751,
      "step": 693
    },
    {
      "epoch": 0.005552,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009929208128601043,
      "loss": 2.5635,
      "step": 694
    },
    {
      "epoch": 0.00556,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0009928996838229637,
      "loss": 1.9968,
      "step": 695
    },
    {
      "epoch": 0.005568,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009928785235266725,
      "loss": 2.4433,
      "step": 696
    },
    {
      "epoch": 0.005576,
      "grad_norm": 2.21875,
      "learning_rate": 0.0009928573319725727,
      "loss": 2.2108,
      "step": 697
    },
    {
      "epoch": 0.005584,
      "grad_norm": 2.8125,
      "learning_rate": 0.000992836109162008,
      "loss": 2.6701,
      "step": 698
    },
    {
      "epoch": 0.005592,
      "grad_norm": 2.234375,
      "learning_rate": 0.0009928148550963243,
      "loss": 2.0521,
      "step": 699
    },
    {
      "epoch": 0.0056,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009927935697768698,
      "loss": 2.3745,
      "step": 700
    },
    {
      "epoch": 0.005608,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009927722532049941,
      "loss": 2.604,
      "step": 701
    },
    {
      "epoch": 0.005616,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009927509053820494,
      "loss": 2.3476,
      "step": 702
    },
    {
      "epoch": 0.005624,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009927295263093894,
      "loss": 2.2101,
      "step": 703
    },
    {
      "epoch": 0.005632,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009927081159883696,
      "loss": 2.3007,
      "step": 704
    },
    {
      "epoch": 0.00564,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009926866744203486,
      "loss": 2.4521,
      "step": 705
    },
    {
      "epoch": 0.005648,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009926652016066853,
      "loss": 2.7074,
      "step": 706
    },
    {
      "epoch": 0.005656,
      "grad_norm": 2.78125,
      "learning_rate": 0.0009926436975487424,
      "loss": 2.6884,
      "step": 707
    },
    {
      "epoch": 0.005664,
      "grad_norm": 2.75,
      "learning_rate": 0.0009926221622478825,
      "loss": 2.6637,
      "step": 708
    },
    {
      "epoch": 0.005672,
      "grad_norm": 2.15625,
      "learning_rate": 0.0009926005957054725,
      "loss": 2.3447,
      "step": 709
    },
    {
      "epoch": 0.00568,
      "grad_norm": 3.0625,
      "learning_rate": 0.0009925789979228796,
      "loss": 2.0975,
      "step": 710
    },
    {
      "epoch": 0.005688,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009925573689014734,
      "loss": 2.7824,
      "step": 711
    },
    {
      "epoch": 0.005696,
      "grad_norm": 2.625,
      "learning_rate": 0.0009925357086426258,
      "loss": 2.5509,
      "step": 712
    },
    {
      "epoch": 0.005704,
      "grad_norm": 2.5,
      "learning_rate": 0.0009925140171477104,
      "loss": 2.4388,
      "step": 713
    },
    {
      "epoch": 0.005712,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009924922944181032,
      "loss": 2.5461,
      "step": 714
    },
    {
      "epoch": 0.00572,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009924705404551811,
      "loss": 2.23,
      "step": 715
    },
    {
      "epoch": 0.005728,
      "grad_norm": 2.859375,
      "learning_rate": 0.0009924487552603243,
      "loss": 2.2301,
      "step": 716
    },
    {
      "epoch": 0.005736,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009924269388349141,
      "loss": 2.4589,
      "step": 717
    },
    {
      "epoch": 0.005744,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009924050911803341,
      "loss": 2.4995,
      "step": 718
    },
    {
      "epoch": 0.005752,
      "grad_norm": 2.53125,
      "learning_rate": 0.0009923832122979701,
      "loss": 2.1376,
      "step": 719
    },
    {
      "epoch": 0.00576,
      "grad_norm": 3.546875,
      "learning_rate": 0.0009923613021892092,
      "loss": 2.0861,
      "step": 720
    },
    {
      "epoch": 0.005768,
      "grad_norm": 3.0625,
      "learning_rate": 0.0009923393608554415,
      "loss": 2.2836,
      "step": 721
    },
    {
      "epoch": 0.005776,
      "grad_norm": 2.390625,
      "learning_rate": 0.000992317388298058,
      "loss": 1.9761,
      "step": 722
    },
    {
      "epoch": 0.005784,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009922953845184523,
      "loss": 2.3847,
      "step": 723
    },
    {
      "epoch": 0.005792,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009922733495180199,
      "loss": 2.5171,
      "step": 724
    },
    {
      "epoch": 0.0058,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009922512832981584,
      "loss": 2.0619,
      "step": 725
    },
    {
      "epoch": 0.005808,
      "grad_norm": 2.359375,
      "learning_rate": 0.000992229185860267,
      "loss": 2.1805,
      "step": 726
    },
    {
      "epoch": 0.005816,
      "grad_norm": 3.296875,
      "learning_rate": 0.000992207057205747,
      "loss": 2.9828,
      "step": 727
    },
    {
      "epoch": 0.005824,
      "grad_norm": 2.84375,
      "learning_rate": 0.000992184897336002,
      "loss": 1.9876,
      "step": 728
    },
    {
      "epoch": 0.005832,
      "grad_norm": 2.46875,
      "learning_rate": 0.0009921627062524374,
      "loss": 2.4258,
      "step": 729
    },
    {
      "epoch": 0.00584,
      "grad_norm": 3.203125,
      "learning_rate": 0.0009921404839564603,
      "loss": 2.4812,
      "step": 730
    },
    {
      "epoch": 0.005848,
      "grad_norm": 2.5,
      "learning_rate": 0.00099211823044948,
      "loss": 2.6338,
      "step": 731
    },
    {
      "epoch": 0.005856,
      "grad_norm": 3.109375,
      "learning_rate": 0.0009920959457329081,
      "loss": 2.4888,
      "step": 732
    },
    {
      "epoch": 0.005864,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009920736298081577,
      "loss": 2.637,
      "step": 733
    },
    {
      "epoch": 0.005872,
      "grad_norm": 2.171875,
      "learning_rate": 0.000992051282676644,
      "loss": 2.5485,
      "step": 734
    },
    {
      "epoch": 0.00588,
      "grad_norm": 3.46875,
      "learning_rate": 0.000992028904339784,
      "loss": 2.1388,
      "step": 735
    },
    {
      "epoch": 0.005888,
      "grad_norm": 2.828125,
      "learning_rate": 0.0009920064947989975,
      "loss": 2.6359,
      "step": 736
    },
    {
      "epoch": 0.005896,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009919840540557054,
      "loss": 2.197,
      "step": 737
    },
    {
      "epoch": 0.005904,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009919615821113307,
      "loss": 2.4005,
      "step": 738
    },
    {
      "epoch": 0.005912,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009919390789672987,
      "loss": 2.707,
      "step": 739
    },
    {
      "epoch": 0.00592,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009919165446250365,
      "loss": 2.2453,
      "step": 740
    },
    {
      "epoch": 0.005928,
      "grad_norm": 2.546875,
      "learning_rate": 0.000991893979085973,
      "loss": 2.4185,
      "step": 741
    },
    {
      "epoch": 0.005936,
      "grad_norm": 2.96875,
      "learning_rate": 0.00099187138235154,
      "loss": 2.607,
      "step": 742
    },
    {
      "epoch": 0.005944,
      "grad_norm": 5.09375,
      "learning_rate": 0.0009918487544231696,
      "loss": 2.272,
      "step": 743
    },
    {
      "epoch": 0.005952,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009918260953022974,
      "loss": 2.7594,
      "step": 744
    },
    {
      "epoch": 0.00596,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009918034049903603,
      "loss": 2.2736,
      "step": 745
    },
    {
      "epoch": 0.005968,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009917806834887975,
      "loss": 2.2974,
      "step": 746
    },
    {
      "epoch": 0.005976,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009917579307990496,
      "loss": 2.1313,
      "step": 747
    },
    {
      "epoch": 0.005984,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009917351469225597,
      "loss": 2.4566,
      "step": 748
    },
    {
      "epoch": 0.005992,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009917123318607726,
      "loss": 2.3872,
      "step": 749
    },
    {
      "epoch": 0.006,
      "grad_norm": 3.15625,
      "learning_rate": 0.0009916894856151356,
      "loss": 1.8446,
      "step": 750
    },
    {
      "epoch": 0.006008,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009916666081870972,
      "loss": 2.2886,
      "step": 751
    },
    {
      "epoch": 0.006016,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009916436995781085,
      "loss": 2.0598,
      "step": 752
    },
    {
      "epoch": 0.006024,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009916207597896222,
      "loss": 2.2278,
      "step": 753
    },
    {
      "epoch": 0.006032,
      "grad_norm": 3.453125,
      "learning_rate": 0.000991597788823093,
      "loss": 2.7454,
      "step": 754
    },
    {
      "epoch": 0.00604,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0009915747866799782,
      "loss": 1.76,
      "step": 755
    },
    {
      "epoch": 0.006048,
      "grad_norm": 2.734375,
      "learning_rate": 0.000991551753361736,
      "loss": 2.2332,
      "step": 756
    },
    {
      "epoch": 0.006056,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009915286888698275,
      "loss": 1.7404,
      "step": 757
    },
    {
      "epoch": 0.006064,
      "grad_norm": 2.5,
      "learning_rate": 0.000991505593205715,
      "loss": 2.0035,
      "step": 758
    },
    {
      "epoch": 0.006072,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009914824663708637,
      "loss": 2.7774,
      "step": 759
    },
    {
      "epoch": 0.00608,
      "grad_norm": 6.5625,
      "learning_rate": 0.00099145930836674,
      "loss": 2.4951,
      "step": 760
    },
    {
      "epoch": 0.006088,
      "grad_norm": 2.625,
      "learning_rate": 0.0009914361191948125,
      "loss": 2.3349,
      "step": 761
    },
    {
      "epoch": 0.006096,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009914128988565522,
      "loss": 2.4641,
      "step": 762
    },
    {
      "epoch": 0.006104,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009913896473534313,
      "loss": 2.1047,
      "step": 763
    },
    {
      "epoch": 0.006112,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009913663646869246,
      "loss": 2.6953,
      "step": 764
    },
    {
      "epoch": 0.00612,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009913430508585086,
      "loss": 2.0658,
      "step": 765
    },
    {
      "epoch": 0.006128,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009913197058696618,
      "loss": 2.5393,
      "step": 766
    },
    {
      "epoch": 0.006136,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009912963297218648,
      "loss": 2.6389,
      "step": 767
    },
    {
      "epoch": 0.006144,
      "grad_norm": 4.375,
      "learning_rate": 0.0009912729224166002,
      "loss": 2.2208,
      "step": 768
    },
    {
      "epoch": 0.006152,
      "grad_norm": 3.0,
      "learning_rate": 0.0009912494839553522,
      "loss": 2.653,
      "step": 769
    },
    {
      "epoch": 0.00616,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009912260143396072,
      "loss": 2.2034,
      "step": 770
    },
    {
      "epoch": 0.006168,
      "grad_norm": 3.296875,
      "learning_rate": 0.000991202513570854,
      "loss": 2.1248,
      "step": 771
    },
    {
      "epoch": 0.006176,
      "grad_norm": 3.421875,
      "learning_rate": 0.0009911789816505825,
      "loss": 2.2735,
      "step": 772
    },
    {
      "epoch": 0.006184,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009911554185802855,
      "loss": 2.6349,
      "step": 773
    },
    {
      "epoch": 0.006192,
      "grad_norm": 2.375,
      "learning_rate": 0.0009911318243614573,
      "loss": 2.204,
      "step": 774
    },
    {
      "epoch": 0.0062,
      "grad_norm": 2.203125,
      "learning_rate": 0.000991108198995594,
      "loss": 2.0909,
      "step": 775
    },
    {
      "epoch": 0.006208,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009910845424841938,
      "loss": 2.1299,
      "step": 776
    },
    {
      "epoch": 0.006216,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009910608548287575,
      "loss": 2.5299,
      "step": 777
    },
    {
      "epoch": 0.006224,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009910371360307866,
      "loss": 2.648,
      "step": 778
    },
    {
      "epoch": 0.006232,
      "grad_norm": 3.1875,
      "learning_rate": 0.000991013386091786,
      "loss": 2.7741,
      "step": 779
    },
    {
      "epoch": 0.00624,
      "grad_norm": 3.0,
      "learning_rate": 0.0009909896050132616,
      "loss": 2.3313,
      "step": 780
    },
    {
      "epoch": 0.006248,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009909657927967214,
      "loss": 2.4001,
      "step": 781
    },
    {
      "epoch": 0.006256,
      "grad_norm": 3.15625,
      "learning_rate": 0.000990941949443676,
      "loss": 2.2397,
      "step": 782
    },
    {
      "epoch": 0.006264,
      "grad_norm": 2.703125,
      "learning_rate": 0.000990918074955637,
      "loss": 2.3784,
      "step": 783
    },
    {
      "epoch": 0.006272,
      "grad_norm": 2.71875,
      "learning_rate": 0.000990894169334119,
      "loss": 2.5539,
      "step": 784
    },
    {
      "epoch": 0.00628,
      "grad_norm": 2.625,
      "learning_rate": 0.0009908702325806375,
      "loss": 2.1632,
      "step": 785
    },
    {
      "epoch": 0.006288,
      "grad_norm": 2.796875,
      "learning_rate": 0.000990846264696711,
      "loss": 2.0973,
      "step": 786
    },
    {
      "epoch": 0.006296,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009908222656838596,
      "loss": 2.3819,
      "step": 787
    },
    {
      "epoch": 0.006304,
      "grad_norm": 2.265625,
      "learning_rate": 0.000990798235543605,
      "loss": 2.2243,
      "step": 788
    },
    {
      "epoch": 0.006312,
      "grad_norm": 3.265625,
      "learning_rate": 0.000990774174277471,
      "loss": 2.4756,
      "step": 789
    },
    {
      "epoch": 0.00632,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009907500818869838,
      "loss": 2.1659,
      "step": 790
    },
    {
      "epoch": 0.006328,
      "grad_norm": 2.921875,
      "learning_rate": 0.0009907259583736715,
      "loss": 2.4194,
      "step": 791
    },
    {
      "epoch": 0.006336,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009907018037390634,
      "loss": 2.0159,
      "step": 792
    },
    {
      "epoch": 0.006344,
      "grad_norm": 3.046875,
      "learning_rate": 0.000990677617984692,
      "loss": 2.9039,
      "step": 793
    },
    {
      "epoch": 0.006352,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009906534011120908,
      "loss": 2.3807,
      "step": 794
    },
    {
      "epoch": 0.00636,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009906291531227955,
      "loss": 1.994,
      "step": 795
    },
    {
      "epoch": 0.006368,
      "grad_norm": 1.984375,
      "learning_rate": 0.0009906048740183442,
      "loss": 1.9649,
      "step": 796
    },
    {
      "epoch": 0.006376,
      "grad_norm": 2.96875,
      "learning_rate": 0.0009905805638002767,
      "loss": 2.5107,
      "step": 797
    },
    {
      "epoch": 0.006384,
      "grad_norm": 2.40625,
      "learning_rate": 0.0009905562224701344,
      "loss": 2.3981,
      "step": 798
    },
    {
      "epoch": 0.006392,
      "grad_norm": 2.359375,
      "learning_rate": 0.0009905318500294609,
      "loss": 2.3173,
      "step": 799
    },
    {
      "epoch": 0.0064,
      "grad_norm": 4.375,
      "learning_rate": 0.0009905074464798022,
      "loss": 2.2309,
      "step": 800
    },
    {
      "epoch": 0.006408,
      "grad_norm": 3.140625,
      "learning_rate": 0.000990483011822706,
      "loss": 2.0817,
      "step": 801
    },
    {
      "epoch": 0.006416,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009904585460597216,
      "loss": 2.0485,
      "step": 802
    },
    {
      "epoch": 0.006424,
      "grad_norm": 3.109375,
      "learning_rate": 0.000990434049192401,
      "loss": 2.5482,
      "step": 803
    },
    {
      "epoch": 0.006432,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009904095212222971,
      "loss": 2.4535,
      "step": 804
    },
    {
      "epoch": 0.00644,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009903849621509663,
      "loss": 2.8395,
      "step": 805
    },
    {
      "epoch": 0.006448,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009903603719799655,
      "loss": 1.8746,
      "step": 806
    },
    {
      "epoch": 0.006456,
      "grad_norm": 11.0,
      "learning_rate": 0.0009903357507108543,
      "loss": 2.2944,
      "step": 807
    },
    {
      "epoch": 0.006464,
      "grad_norm": 3.125,
      "learning_rate": 0.0009903110983451942,
      "loss": 2.2825,
      "step": 808
    },
    {
      "epoch": 0.006472,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009902864148845485,
      "loss": 2.3946,
      "step": 809
    },
    {
      "epoch": 0.00648,
      "grad_norm": 2.765625,
      "learning_rate": 0.000990261700330483,
      "loss": 2.6699,
      "step": 810
    },
    {
      "epoch": 0.006488,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009902369546845647,
      "loss": 1.9501,
      "step": 811
    },
    {
      "epoch": 0.006496,
      "grad_norm": 3.0,
      "learning_rate": 0.000990212177948363,
      "loss": 2.7386,
      "step": 812
    },
    {
      "epoch": 0.006504,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009901873701234493,
      "loss": 2.4213,
      "step": 813
    },
    {
      "epoch": 0.006512,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009901625312113966,
      "loss": 2.0697,
      "step": 814
    },
    {
      "epoch": 0.00652,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009901376612137807,
      "loss": 2.2036,
      "step": 815
    },
    {
      "epoch": 0.006528,
      "grad_norm": 3.125,
      "learning_rate": 0.0009901127601321784,
      "loss": 2.4044,
      "step": 816
    },
    {
      "epoch": 0.006536,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009900878279681688,
      "loss": 2.9324,
      "step": 817
    },
    {
      "epoch": 0.006544,
      "grad_norm": 3.203125,
      "learning_rate": 0.0009900628647233336,
      "loss": 2.3495,
      "step": 818
    },
    {
      "epoch": 0.006552,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009900378703992557,
      "loss": 2.1946,
      "step": 819
    },
    {
      "epoch": 0.00656,
      "grad_norm": 2.296875,
      "learning_rate": 0.0009900128449975198,
      "loss": 2.1172,
      "step": 820
    },
    {
      "epoch": 0.006568,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009899877885197134,
      "loss": 2.2507,
      "step": 821
    },
    {
      "epoch": 0.006576,
      "grad_norm": 7.625,
      "learning_rate": 0.0009899627009674256,
      "loss": 2.2173,
      "step": 822
    },
    {
      "epoch": 0.006584,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009899375823422473,
      "loss": 2.6946,
      "step": 823
    },
    {
      "epoch": 0.006592,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009899124326457715,
      "loss": 2.5422,
      "step": 824
    },
    {
      "epoch": 0.0066,
      "grad_norm": 2.78125,
      "learning_rate": 0.0009898872518795932,
      "loss": 2.3561,
      "step": 825
    },
    {
      "epoch": 0.006608,
      "grad_norm": 2.6875,
      "learning_rate": 0.000989862040045309,
      "loss": 2.589,
      "step": 826
    },
    {
      "epoch": 0.006616,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009898367971445186,
      "loss": 1.9889,
      "step": 827
    },
    {
      "epoch": 0.006624,
      "grad_norm": 3.15625,
      "learning_rate": 0.000989811523178822,
      "loss": 2.9384,
      "step": 828
    },
    {
      "epoch": 0.006632,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009897862181498225,
      "loss": 2.5604,
      "step": 829
    },
    {
      "epoch": 0.00664,
      "grad_norm": 3.234375,
      "learning_rate": 0.0009897608820591249,
      "loss": 2.525,
      "step": 830
    },
    {
      "epoch": 0.006648,
      "grad_norm": 4.8125,
      "learning_rate": 0.000989735514908336,
      "loss": 2.7396,
      "step": 831
    },
    {
      "epoch": 0.006656,
      "grad_norm": 3.25,
      "learning_rate": 0.0009897101166990645,
      "loss": 2.2525,
      "step": 832
    },
    {
      "epoch": 0.006664,
      "grad_norm": 3.578125,
      "learning_rate": 0.000989684687432921,
      "loss": 2.2011,
      "step": 833
    },
    {
      "epoch": 0.006672,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009896592271115183,
      "loss": 2.6506,
      "step": 834
    },
    {
      "epoch": 0.00668,
      "grad_norm": 4.5,
      "learning_rate": 0.000989633735736471,
      "loss": 2.3357,
      "step": 835
    },
    {
      "epoch": 0.006688,
      "grad_norm": 9.3125,
      "learning_rate": 0.000989608213309396,
      "loss": 2.696,
      "step": 836
    },
    {
      "epoch": 0.006696,
      "grad_norm": 3.34375,
      "learning_rate": 0.0009895826598319117,
      "loss": 2.1965,
      "step": 837
    },
    {
      "epoch": 0.006704,
      "grad_norm": 2.859375,
      "learning_rate": 0.0009895570753056387,
      "loss": 2.3956,
      "step": 838
    },
    {
      "epoch": 0.006712,
      "grad_norm": 4.28125,
      "learning_rate": 0.0009895314597321994,
      "loss": 2.0995,
      "step": 839
    },
    {
      "epoch": 0.00672,
      "grad_norm": 3.6875,
      "learning_rate": 0.0009895058131132186,
      "loss": 2.1329,
      "step": 840
    },
    {
      "epoch": 0.006728,
      "grad_norm": 3.953125,
      "learning_rate": 0.0009894801354503224,
      "loss": 2.1301,
      "step": 841
    },
    {
      "epoch": 0.006736,
      "grad_norm": 2.265625,
      "learning_rate": 0.0009894544267451397,
      "loss": 2.1504,
      "step": 842
    },
    {
      "epoch": 0.006744,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009894286869993006,
      "loss": 2.8615,
      "step": 843
    },
    {
      "epoch": 0.006752,
      "grad_norm": 3.96875,
      "learning_rate": 0.0009894029162144376,
      "loss": 2.2885,
      "step": 844
    },
    {
      "epoch": 0.00676,
      "grad_norm": 3.203125,
      "learning_rate": 0.000989377114392185,
      "loss": 2.819,
      "step": 845
    },
    {
      "epoch": 0.006768,
      "grad_norm": 6.25,
      "learning_rate": 0.000989351281534179,
      "loss": 2.2356,
      "step": 846
    },
    {
      "epoch": 0.006776,
      "grad_norm": 3.65625,
      "learning_rate": 0.0009893254176420584,
      "loss": 1.9986,
      "step": 847
    },
    {
      "epoch": 0.006784,
      "grad_norm": 8.3125,
      "learning_rate": 0.0009892995227174628,
      "loss": 2.5724,
      "step": 848
    },
    {
      "epoch": 0.006792,
      "grad_norm": 57.0,
      "learning_rate": 0.0009892735967620347,
      "loss": 2.5662,
      "step": 849
    },
    {
      "epoch": 0.0068,
      "grad_norm": 3.375,
      "learning_rate": 0.0009892476397774185,
      "loss": 2.1833,
      "step": 850
    },
    {
      "epoch": 0.006808,
      "grad_norm": 2.859375,
      "learning_rate": 0.0009892216517652602,
      "loss": 2.8256,
      "step": 851
    },
    {
      "epoch": 0.006816,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009891956327272079,
      "loss": 2.5291,
      "step": 852
    },
    {
      "epoch": 0.006824,
      "grad_norm": 2.875,
      "learning_rate": 0.0009891695826649116,
      "loss": 2.7706,
      "step": 853
    },
    {
      "epoch": 0.006832,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009891435015800237,
      "loss": 2.3584,
      "step": 854
    },
    {
      "epoch": 0.00684,
      "grad_norm": 3.84375,
      "learning_rate": 0.0009891173894741978,
      "loss": 2.554,
      "step": 855
    },
    {
      "epoch": 0.006848,
      "grad_norm": 5.875,
      "learning_rate": 0.00098909124634909,
      "loss": 2.249,
      "step": 856
    },
    {
      "epoch": 0.006856,
      "grad_norm": 3.296875,
      "learning_rate": 0.0009890650722063589,
      "loss": 2.6463,
      "step": 857
    },
    {
      "epoch": 0.006864,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009890388670476635,
      "loss": 2.6148,
      "step": 858
    },
    {
      "epoch": 0.006872,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009890126308746664,
      "loss": 2.073,
      "step": 859
    },
    {
      "epoch": 0.00688,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009889863636890308,
      "loss": 2.6188,
      "step": 860
    },
    {
      "epoch": 0.006888,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009889600654924232,
      "loss": 2.0904,
      "step": 861
    },
    {
      "epoch": 0.006896,
      "grad_norm": 5.78125,
      "learning_rate": 0.0009889337362865111,
      "loss": 2.8617,
      "step": 862
    },
    {
      "epoch": 0.006904,
      "grad_norm": 17.0,
      "learning_rate": 0.0009889073760729642,
      "loss": 2.3585,
      "step": 863
    },
    {
      "epoch": 0.006912,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009888809848534546,
      "loss": 2.3237,
      "step": 864
    },
    {
      "epoch": 0.00692,
      "grad_norm": 3.53125,
      "learning_rate": 0.0009888545626296555,
      "loss": 2.143,
      "step": 865
    },
    {
      "epoch": 0.006928,
      "grad_norm": 3.296875,
      "learning_rate": 0.000988828109403243,
      "loss": 2.3857,
      "step": 866
    },
    {
      "epoch": 0.006936,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009888016251758942,
      "loss": 2.1326,
      "step": 867
    },
    {
      "epoch": 0.006944,
      "grad_norm": 5.96875,
      "learning_rate": 0.0009887751099492894,
      "loss": 2.2711,
      "step": 868
    },
    {
      "epoch": 0.006952,
      "grad_norm": 3.65625,
      "learning_rate": 0.0009887485637251096,
      "loss": 2.386,
      "step": 869
    },
    {
      "epoch": 0.00696,
      "grad_norm": 3.0625,
      "learning_rate": 0.0009887219865050386,
      "loss": 2.1976,
      "step": 870
    },
    {
      "epoch": 0.006968,
      "grad_norm": 25.375,
      "learning_rate": 0.0009886953782907617,
      "loss": 1.8421,
      "step": 871
    },
    {
      "epoch": 0.006976,
      "grad_norm": 3.859375,
      "learning_rate": 0.0009886687390839666,
      "loss": 2.7707,
      "step": 872
    },
    {
      "epoch": 0.006984,
      "grad_norm": 3.625,
      "learning_rate": 0.0009886420688863427,
      "loss": 2.4722,
      "step": 873
    },
    {
      "epoch": 0.006992,
      "grad_norm": 4.78125,
      "learning_rate": 0.0009886153676995813,
      "loss": 2.7474,
      "step": 874
    },
    {
      "epoch": 0.007,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009885886355253757,
      "loss": 1.9611,
      "step": 875
    },
    {
      "epoch": 0.007008,
      "grad_norm": 4.46875,
      "learning_rate": 0.0009885618723654214,
      "loss": 2.5036,
      "step": 876
    },
    {
      "epoch": 0.007016,
      "grad_norm": 3.375,
      "learning_rate": 0.0009885350782214155,
      "loss": 2.4786,
      "step": 877
    },
    {
      "epoch": 0.007024,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009885082530950576,
      "loss": 2.5318,
      "step": 878
    },
    {
      "epoch": 0.007032,
      "grad_norm": 15.75,
      "learning_rate": 0.0009884813969880484,
      "loss": 2.124,
      "step": 879
    },
    {
      "epoch": 0.00704,
      "grad_norm": 4.90625,
      "learning_rate": 0.0009884545099020914,
      "loss": 2.3292,
      "step": 880
    },
    {
      "epoch": 0.007048,
      "grad_norm": 2.828125,
      "learning_rate": 0.0009884275918388918,
      "loss": 2.3692,
      "step": 881
    },
    {
      "epoch": 0.007056,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009884006428001566,
      "loss": 2.5219,
      "step": 882
    },
    {
      "epoch": 0.007064,
      "grad_norm": 3.296875,
      "learning_rate": 0.0009883736627875948,
      "loss": 2.6393,
      "step": 883
    },
    {
      "epoch": 0.007072,
      "grad_norm": 3.78125,
      "learning_rate": 0.0009883466518029176,
      "loss": 2.6673,
      "step": 884
    },
    {
      "epoch": 0.00708,
      "grad_norm": 4.8125,
      "learning_rate": 0.000988319609847838,
      "loss": 2.2877,
      "step": 885
    },
    {
      "epoch": 0.007088,
      "grad_norm": 3.671875,
      "learning_rate": 0.0009882925369240707,
      "loss": 2.2581,
      "step": 886
    },
    {
      "epoch": 0.007096,
      "grad_norm": 2.625,
      "learning_rate": 0.0009882654330333333,
      "loss": 2.2624,
      "step": 887
    },
    {
      "epoch": 0.007104,
      "grad_norm": 3.015625,
      "learning_rate": 0.0009882382981773437,
      "loss": 2.6791,
      "step": 888
    },
    {
      "epoch": 0.007112,
      "grad_norm": 3.109375,
      "learning_rate": 0.0009882111323578236,
      "loss": 2.6151,
      "step": 889
    },
    {
      "epoch": 0.00712,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009881839355764955,
      "loss": 2.3607,
      "step": 890
    },
    {
      "epoch": 0.007128,
      "grad_norm": 3.203125,
      "learning_rate": 0.0009881567078350844,
      "loss": 1.8735,
      "step": 891
    },
    {
      "epoch": 0.007136,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009881294491353165,
      "loss": 2.7525,
      "step": 892
    },
    {
      "epoch": 0.007144,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009881021594789212,
      "loss": 1.9622,
      "step": 893
    },
    {
      "epoch": 0.007152,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009880748388676288,
      "loss": 2.5409,
      "step": 894
    },
    {
      "epoch": 0.00716,
      "grad_norm": 4.3125,
      "learning_rate": 0.0009880474873031721,
      "loss": 2.1424,
      "step": 895
    },
    {
      "epoch": 0.007168,
      "grad_norm": 3.6875,
      "learning_rate": 0.0009880201047872854,
      "loss": 2.0962,
      "step": 896
    },
    {
      "epoch": 0.007176,
      "grad_norm": 2.75,
      "learning_rate": 0.0009879926913217059,
      "loss": 2.0898,
      "step": 897
    },
    {
      "epoch": 0.007184,
      "grad_norm": 4.875,
      "learning_rate": 0.0009879652469081711,
      "loss": 2.622,
      "step": 898
    },
    {
      "epoch": 0.007192,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009879377715484227,
      "loss": 2.7861,
      "step": 899
    },
    {
      "epoch": 0.0072,
      "grad_norm": 5.34375,
      "learning_rate": 0.0009879102652442023,
      "loss": 2.4208,
      "step": 900
    },
    {
      "epoch": 0.007208,
      "grad_norm": 3.671875,
      "learning_rate": 0.0009878827279972547,
      "loss": 2.6055,
      "step": 901
    },
    {
      "epoch": 0.007216,
      "grad_norm": 3.5,
      "learning_rate": 0.000987855159809326,
      "loss": 2.6596,
      "step": 902
    },
    {
      "epoch": 0.007224,
      "grad_norm": 3.4375,
      "learning_rate": 0.0009878275606821651,
      "loss": 2.2369,
      "step": 903
    },
    {
      "epoch": 0.007232,
      "grad_norm": 3.5,
      "learning_rate": 0.0009877999306175216,
      "loss": 3.3349,
      "step": 904
    },
    {
      "epoch": 0.00724,
      "grad_norm": 4.59375,
      "learning_rate": 0.0009877722696171483,
      "loss": 2.139,
      "step": 905
    },
    {
      "epoch": 0.007248,
      "grad_norm": 2.734375,
      "learning_rate": 0.000987744577682799,
      "loss": 2.015,
      "step": 906
    },
    {
      "epoch": 0.007256,
      "grad_norm": 8.625,
      "learning_rate": 0.0009877168548162302,
      "loss": 2.4227,
      "step": 907
    },
    {
      "epoch": 0.007264,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009876891010192,
      "loss": 2.8337,
      "step": 908
    },
    {
      "epoch": 0.007272,
      "grad_norm": 3.859375,
      "learning_rate": 0.0009876613162934686,
      "loss": 2.8596,
      "step": 909
    },
    {
      "epoch": 0.00728,
      "grad_norm": 4.1875,
      "learning_rate": 0.0009876335006407978,
      "loss": 2.976,
      "step": 910
    },
    {
      "epoch": 0.007288,
      "grad_norm": 3.84375,
      "learning_rate": 0.000987605654062952,
      "loss": 2.627,
      "step": 911
    },
    {
      "epoch": 0.007296,
      "grad_norm": 3.34375,
      "learning_rate": 0.0009875777765616966,
      "loss": 2.153,
      "step": 912
    },
    {
      "epoch": 0.007304,
      "grad_norm": 2.78125,
      "learning_rate": 0.0009875498681388005,
      "loss": 2.6163,
      "step": 913
    },
    {
      "epoch": 0.007312,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009875219287960326,
      "loss": 2.5867,
      "step": 914
    },
    {
      "epoch": 0.00732,
      "grad_norm": 2.75,
      "learning_rate": 0.0009874939585351654,
      "loss": 2.4676,
      "step": 915
    },
    {
      "epoch": 0.007328,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009874659573579727,
      "loss": 2.7059,
      "step": 916
    },
    {
      "epoch": 0.007336,
      "grad_norm": 3.15625,
      "learning_rate": 0.0009874379252662303,
      "loss": 1.7653,
      "step": 917
    },
    {
      "epoch": 0.007344,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009874098622617157,
      "loss": 2.6246,
      "step": 918
    },
    {
      "epoch": 0.007352,
      "grad_norm": 2.859375,
      "learning_rate": 0.0009873817683462088,
      "loss": 2.3481,
      "step": 919
    },
    {
      "epoch": 0.00736,
      "grad_norm": 3.640625,
      "learning_rate": 0.0009873536435214914,
      "loss": 2.3222,
      "step": 920
    },
    {
      "epoch": 0.007368,
      "grad_norm": 3.25,
      "learning_rate": 0.000987325487789347,
      "loss": 2.9621,
      "step": 921
    },
    {
      "epoch": 0.007376,
      "grad_norm": 3.40625,
      "learning_rate": 0.0009872973011515612,
      "loss": 2.2863,
      "step": 922
    },
    {
      "epoch": 0.007384,
      "grad_norm": 2.921875,
      "learning_rate": 0.0009872690836099215,
      "loss": 2.1246,
      "step": 923
    },
    {
      "epoch": 0.007392,
      "grad_norm": 30.125,
      "learning_rate": 0.0009872408351662176,
      "loss": 2.492,
      "step": 924
    },
    {
      "epoch": 0.0074,
      "grad_norm": 12.5625,
      "learning_rate": 0.000987212555822241,
      "loss": 2.5275,
      "step": 925
    },
    {
      "epoch": 0.007408,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009871842455797851,
      "loss": 2.0341,
      "step": 926
    },
    {
      "epoch": 0.007416,
      "grad_norm": 5.59375,
      "learning_rate": 0.000987155904440645,
      "loss": 2.759,
      "step": 927
    },
    {
      "epoch": 0.007424,
      "grad_norm": 4.625,
      "learning_rate": 0.0009871275324066185,
      "loss": 2.2378,
      "step": 928
    },
    {
      "epoch": 0.007432,
      "grad_norm": 4.84375,
      "learning_rate": 0.0009870991294795046,
      "loss": 2.557,
      "step": 929
    },
    {
      "epoch": 0.00744,
      "grad_norm": 3.5,
      "learning_rate": 0.000987070695661105,
      "loss": 2.2237,
      "step": 930
    },
    {
      "epoch": 0.007448,
      "grad_norm": 4.5,
      "learning_rate": 0.0009870422309532223,
      "loss": 2.4358,
      "step": 931
    },
    {
      "epoch": 0.007456,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009870137353576625,
      "loss": 2.4295,
      "step": 932
    },
    {
      "epoch": 0.007464,
      "grad_norm": 2.8125,
      "learning_rate": 0.000986985208876232,
      "loss": 2.2917,
      "step": 933
    },
    {
      "epoch": 0.007472,
      "grad_norm": 4.0625,
      "learning_rate": 0.0009869566515107403,
      "loss": 2.484,
      "step": 934
    },
    {
      "epoch": 0.00748,
      "grad_norm": 4.34375,
      "learning_rate": 0.0009869280632629983,
      "loss": 2.1957,
      "step": 935
    },
    {
      "epoch": 0.007488,
      "grad_norm": 2.859375,
      "learning_rate": 0.0009868994441348195,
      "loss": 2.3244,
      "step": 936
    },
    {
      "epoch": 0.007496,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009868707941280185,
      "loss": 2.3941,
      "step": 937
    },
    {
      "epoch": 0.007504,
      "grad_norm": 3.4375,
      "learning_rate": 0.000986842113244412,
      "loss": 2.503,
      "step": 938
    },
    {
      "epoch": 0.007512,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009868134014858194,
      "loss": 1.8781,
      "step": 939
    },
    {
      "epoch": 0.00752,
      "grad_norm": 3.5,
      "learning_rate": 0.0009867846588540613,
      "loss": 2.0957,
      "step": 940
    },
    {
      "epoch": 0.007528,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009867558853509607,
      "loss": 1.8741,
      "step": 941
    },
    {
      "epoch": 0.007536,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009867270809783423,
      "loss": 2.2781,
      "step": 942
    },
    {
      "epoch": 0.007544,
      "grad_norm": 3.46875,
      "learning_rate": 0.0009866982457380326,
      "loss": 2.3588,
      "step": 943
    },
    {
      "epoch": 0.007552,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009866693796318608,
      "loss": 2.4459,
      "step": 944
    },
    {
      "epoch": 0.00756,
      "grad_norm": 2.96875,
      "learning_rate": 0.0009866404826616574,
      "loss": 2.7947,
      "step": 945
    },
    {
      "epoch": 0.007568,
      "grad_norm": 3.546875,
      "learning_rate": 0.0009866115548292545,
      "loss": 2.1662,
      "step": 946
    },
    {
      "epoch": 0.007576,
      "grad_norm": 3.015625,
      "learning_rate": 0.0009865825961364874,
      "loss": 2.4436,
      "step": 947
    },
    {
      "epoch": 0.007584,
      "grad_norm": 2.5625,
      "learning_rate": 0.000986553606585192,
      "loss": 2.0949,
      "step": 948
    },
    {
      "epoch": 0.007592,
      "grad_norm": 3.109375,
      "learning_rate": 0.0009865245861772072,
      "loss": 2.4112,
      "step": 949
    },
    {
      "epoch": 0.0076,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009864955349143734,
      "loss": 1.5918,
      "step": 950
    },
    {
      "epoch": 0.007608,
      "grad_norm": 2.453125,
      "learning_rate": 0.000986466452798533,
      "loss": 2.0212,
      "step": 951
    },
    {
      "epoch": 0.007616,
      "grad_norm": 2.953125,
      "learning_rate": 0.00098643733983153,
      "loss": 2.292,
      "step": 952
    },
    {
      "epoch": 0.007624,
      "grad_norm": 3.890625,
      "learning_rate": 0.0009864081960152113,
      "loss": 2.5087,
      "step": 953
    },
    {
      "epoch": 0.007632,
      "grad_norm": 3.234375,
      "learning_rate": 0.0009863790213514247,
      "loss": 2.0412,
      "step": 954
    },
    {
      "epoch": 0.00764,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009863498158420205,
      "loss": 2.513,
      "step": 955
    },
    {
      "epoch": 0.007648,
      "grad_norm": 3.40625,
      "learning_rate": 0.0009863205794888512,
      "loss": 2.4871,
      "step": 956
    },
    {
      "epoch": 0.007656,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009862913122937705,
      "loss": 2.5282,
      "step": 957
    },
    {
      "epoch": 0.007664,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009862620142586348,
      "loss": 2.3203,
      "step": 958
    },
    {
      "epoch": 0.007672,
      "grad_norm": 2.328125,
      "learning_rate": 0.000986232685385302,
      "loss": 2.1069,
      "step": 959
    },
    {
      "epoch": 0.00768,
      "grad_norm": 2.625,
      "learning_rate": 0.0009862033256756322,
      "loss": 2.6188,
      "step": 960
    },
    {
      "epoch": 0.007688,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009861739351314871,
      "loss": 1.9506,
      "step": 961
    },
    {
      "epoch": 0.007696,
      "grad_norm": 2.609375,
      "learning_rate": 0.000986144513754731,
      "loss": 2.6677,
      "step": 962
    },
    {
      "epoch": 0.007704,
      "grad_norm": 3.53125,
      "learning_rate": 0.0009861150615472296,
      "loss": 2.7367,
      "step": 963
    },
    {
      "epoch": 0.007712,
      "grad_norm": 2.625,
      "learning_rate": 0.0009860855785108508,
      "loss": 2.5445,
      "step": 964
    },
    {
      "epoch": 0.00772,
      "grad_norm": 2.78125,
      "learning_rate": 0.0009860560646474642,
      "loss": 2.1327,
      "step": 965
    },
    {
      "epoch": 0.007728,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009860265199589417,
      "loss": 2.6643,
      "step": 966
    },
    {
      "epoch": 0.007736,
      "grad_norm": 3.015625,
      "learning_rate": 0.0009859969444471568,
      "loss": 2.1202,
      "step": 967
    },
    {
      "epoch": 0.007744,
      "grad_norm": 3.40625,
      "learning_rate": 0.0009859673381139855,
      "loss": 2.252,
      "step": 968
    },
    {
      "epoch": 0.007752,
      "grad_norm": 3.03125,
      "learning_rate": 0.000985937700961305,
      "loss": 2.1694,
      "step": 969
    },
    {
      "epoch": 0.00776,
      "grad_norm": 3.46875,
      "learning_rate": 0.000985908032990995,
      "loss": 2.596,
      "step": 970
    },
    {
      "epoch": 0.007768,
      "grad_norm": 2.78125,
      "learning_rate": 0.000985878334204937,
      "loss": 2.4959,
      "step": 971
    },
    {
      "epoch": 0.007776,
      "grad_norm": 2.75,
      "learning_rate": 0.0009858486046050148,
      "loss": 1.8638,
      "step": 972
    },
    {
      "epoch": 0.007784,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009858188441931133,
      "loss": 2.1945,
      "step": 973
    },
    {
      "epoch": 0.007792,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009857890529711203,
      "loss": 1.9514,
      "step": 974
    },
    {
      "epoch": 0.0078,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009857592309409247,
      "loss": 2.4108,
      "step": 975
    },
    {
      "epoch": 0.007808,
      "grad_norm": 3.171875,
      "learning_rate": 0.000985729378104418,
      "loss": 2.3765,
      "step": 976
    },
    {
      "epoch": 0.007816,
      "grad_norm": 3.140625,
      "learning_rate": 0.0009856994944634936,
      "loss": 2.7232,
      "step": 977
    },
    {
      "epoch": 0.007824,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009856695800200463,
      "loss": 2.0792,
      "step": 978
    },
    {
      "epoch": 0.007832,
      "grad_norm": 3.15625,
      "learning_rate": 0.0009856396347759737,
      "loss": 2.1162,
      "step": 979
    },
    {
      "epoch": 0.00784,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009856096587331747,
      "loss": 2.4491,
      "step": 980
    },
    {
      "epoch": 0.007848,
      "grad_norm": 2.53125,
      "learning_rate": 0.00098557965189355,
      "loss": 2.4132,
      "step": 981
    },
    {
      "epoch": 0.007856,
      "grad_norm": 3.1875,
      "learning_rate": 0.000985549614259003,
      "loss": 2.3162,
      "step": 982
    },
    {
      "epoch": 0.007864,
      "grad_norm": 2.625,
      "learning_rate": 0.0009855195458314387,
      "loss": 1.9616,
      "step": 983
    },
    {
      "epoch": 0.007872,
      "grad_norm": 2.859375,
      "learning_rate": 0.000985489446612764,
      "loss": 2.4057,
      "step": 984
    },
    {
      "epoch": 0.00788,
      "grad_norm": 2.875,
      "learning_rate": 0.0009854593166048872,
      "loss": 2.29,
      "step": 985
    },
    {
      "epoch": 0.007888,
      "grad_norm": 3.0625,
      "learning_rate": 0.0009854291558097198,
      "loss": 2.5381,
      "step": 986
    },
    {
      "epoch": 0.007896,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009853989642291743,
      "loss": 1.8944,
      "step": 987
    },
    {
      "epoch": 0.007904,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009853687418651656,
      "loss": 2.2163,
      "step": 988
    },
    {
      "epoch": 0.007912,
      "grad_norm": 2.625,
      "learning_rate": 0.00098533848871961,
      "loss": 2.0262,
      "step": 989
    },
    {
      "epoch": 0.00792,
      "grad_norm": 2.1875,
      "learning_rate": 0.0009853082047944263,
      "loss": 1.9078,
      "step": 990
    },
    {
      "epoch": 0.007928,
      "grad_norm": 3.609375,
      "learning_rate": 0.0009852778900915349,
      "loss": 2.2277,
      "step": 991
    },
    {
      "epoch": 0.007936,
      "grad_norm": 16.625,
      "learning_rate": 0.0009852475446128588,
      "loss": 2.1486,
      "step": 992
    },
    {
      "epoch": 0.007944,
      "grad_norm": 3.015625,
      "learning_rate": 0.000985217168360322,
      "loss": 2.4675,
      "step": 993
    },
    {
      "epoch": 0.007952,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009851867613358512,
      "loss": 2.3384,
      "step": 994
    },
    {
      "epoch": 0.00796,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009851563235413747,
      "loss": 2.0297,
      "step": 995
    },
    {
      "epoch": 0.007968,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009851258549788229,
      "loss": 2.691,
      "step": 996
    },
    {
      "epoch": 0.007976,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009850953556501277,
      "loss": 2.3908,
      "step": 997
    },
    {
      "epoch": 0.007984,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009850648255572238,
      "loss": 2.518,
      "step": 998
    },
    {
      "epoch": 0.007992,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009850342647020475,
      "loss": 1.9422,
      "step": 999
    },
    {
      "epoch": 0.008,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009850036730865363,
      "loss": 1.8628,
      "step": 1000
    },
    {
      "epoch": 0.008008,
      "grad_norm": 4.34375,
      "learning_rate": 0.0009849730507126306,
      "loss": 2.666,
      "step": 1001
    },
    {
      "epoch": 0.008016,
      "grad_norm": 3.625,
      "learning_rate": 0.0009849423975822728,
      "loss": 2.7322,
      "step": 1002
    },
    {
      "epoch": 0.008024,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009849117136974065,
      "loss": 3.0342,
      "step": 1003
    },
    {
      "epoch": 0.008032,
      "grad_norm": 4.5625,
      "learning_rate": 0.0009848809990599776,
      "loss": 2.0444,
      "step": 1004
    },
    {
      "epoch": 0.00804,
      "grad_norm": 2.390625,
      "learning_rate": 0.000984850253671934,
      "loss": 1.9374,
      "step": 1005
    },
    {
      "epoch": 0.008048,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009848194775352256,
      "loss": 2.7859,
      "step": 1006
    },
    {
      "epoch": 0.008056,
      "grad_norm": 3.375,
      "learning_rate": 0.0009847886706518046,
      "loss": 2.4402,
      "step": 1007
    },
    {
      "epoch": 0.008064,
      "grad_norm": 3.0,
      "learning_rate": 0.000984757833023624,
      "loss": 1.9096,
      "step": 1008
    },
    {
      "epoch": 0.008072,
      "grad_norm": 3.203125,
      "learning_rate": 0.00098472696465264,
      "loss": 2.5763,
      "step": 1009
    },
    {
      "epoch": 0.00808,
      "grad_norm": 3.546875,
      "learning_rate": 0.0009846960655408102,
      "loss": 2.4549,
      "step": 1010
    },
    {
      "epoch": 0.008088,
      "grad_norm": 4.8125,
      "learning_rate": 0.0009846651356900937,
      "loss": 2.7397,
      "step": 1011
    },
    {
      "epoch": 0.008096,
      "grad_norm": 3.84375,
      "learning_rate": 0.0009846341751024528,
      "loss": 2.7059,
      "step": 1012
    },
    {
      "epoch": 0.008104,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009846031837798504,
      "loss": 2.5009,
      "step": 1013
    },
    {
      "epoch": 0.008112,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009845721617242522,
      "loss": 2.4021,
      "step": 1014
    },
    {
      "epoch": 0.00812,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009845411089376254,
      "loss": 2.3926,
      "step": 1015
    },
    {
      "epoch": 0.008128,
      "grad_norm": 3.34375,
      "learning_rate": 0.0009845100254219396,
      "loss": 2.559,
      "step": 1016
    },
    {
      "epoch": 0.008136,
      "grad_norm": 2.140625,
      "learning_rate": 0.000984478911179166,
      "loss": 1.9164,
      "step": 1017
    },
    {
      "epoch": 0.008144,
      "grad_norm": 3.625,
      "learning_rate": 0.0009844477662112774,
      "loss": 3.1538,
      "step": 1018
    },
    {
      "epoch": 0.008152,
      "grad_norm": 3.203125,
      "learning_rate": 0.0009844165905202498,
      "loss": 2.6124,
      "step": 1019
    },
    {
      "epoch": 0.00816,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009843853841080595,
      "loss": 2.2735,
      "step": 1020
    },
    {
      "epoch": 0.008168,
      "grad_norm": 3.234375,
      "learning_rate": 0.000984354146976686,
      "loss": 2.3599,
      "step": 1021
    },
    {
      "epoch": 0.008176,
      "grad_norm": 3.828125,
      "learning_rate": 0.0009843228791281105,
      "loss": 2.5344,
      "step": 1022
    },
    {
      "epoch": 0.008184,
      "grad_norm": 2.859375,
      "learning_rate": 0.0009842915805643156,
      "loss": 2.8577,
      "step": 1023
    },
    {
      "epoch": 0.008192,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009842602512872864,
      "loss": 2.1852,
      "step": 1024
    },
    {
      "epoch": 0.0082,
      "grad_norm": 3.9375,
      "learning_rate": 0.0009842288912990096,
      "loss": 2.283,
      "step": 1025
    },
    {
      "epoch": 0.008208,
      "grad_norm": 2.625,
      "learning_rate": 0.0009841975006014744,
      "loss": 2.307,
      "step": 1026
    },
    {
      "epoch": 0.008216,
      "grad_norm": 2.640625,
      "learning_rate": 0.000984166079196671,
      "loss": 2.3357,
      "step": 1027
    },
    {
      "epoch": 0.008224,
      "grad_norm": 2.859375,
      "learning_rate": 0.0009841346270865925,
      "loss": 2.0502,
      "step": 1028
    },
    {
      "epoch": 0.008232,
      "grad_norm": 4.40625,
      "learning_rate": 0.0009841031442732336,
      "loss": 2.5893,
      "step": 1029
    },
    {
      "epoch": 0.00824,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009840716307585907,
      "loss": 1.9426,
      "step": 1030
    },
    {
      "epoch": 0.008248,
      "grad_norm": 3.5625,
      "learning_rate": 0.0009840400865446625,
      "loss": 2.2325,
      "step": 1031
    },
    {
      "epoch": 0.008256,
      "grad_norm": 3.359375,
      "learning_rate": 0.0009840085116334493,
      "loss": 2.4399,
      "step": 1032
    },
    {
      "epoch": 0.008264,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009839769060269539,
      "loss": 2.4184,
      "step": 1033
    },
    {
      "epoch": 0.008272,
      "grad_norm": 3.84375,
      "learning_rate": 0.0009839452697271802,
      "loss": 3.0402,
      "step": 1034
    },
    {
      "epoch": 0.00828,
      "grad_norm": 3.140625,
      "learning_rate": 0.000983913602736135,
      "loss": 2.3744,
      "step": 1035
    },
    {
      "epoch": 0.008288,
      "grad_norm": 2.171875,
      "learning_rate": 0.0009838819050558263,
      "loss": 1.7321,
      "step": 1036
    },
    {
      "epoch": 0.008296,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009838501766882646,
      "loss": 2.4289,
      "step": 1037
    },
    {
      "epoch": 0.008304,
      "grad_norm": 3.1875,
      "learning_rate": 0.0009838184176354616,
      "loss": 2.2972,
      "step": 1038
    },
    {
      "epoch": 0.008312,
      "grad_norm": 3.3125,
      "learning_rate": 0.000983786627899432,
      "loss": 2.4402,
      "step": 1039
    },
    {
      "epoch": 0.00832,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009837548074821917,
      "loss": 2.4661,
      "step": 1040
    },
    {
      "epoch": 0.008328,
      "grad_norm": 3.3125,
      "learning_rate": 0.0009837229563857582,
      "loss": 2.7305,
      "step": 1041
    },
    {
      "epoch": 0.008336,
      "grad_norm": 5.15625,
      "learning_rate": 0.000983691074612152,
      "loss": 2.2036,
      "step": 1042
    },
    {
      "epoch": 0.008344,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009836591621633951,
      "loss": 2.0279,
      "step": 1043
    },
    {
      "epoch": 0.008352,
      "grad_norm": 3.375,
      "learning_rate": 0.000983627219041511,
      "loss": 2.46,
      "step": 1044
    },
    {
      "epoch": 0.00836,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009835952452485258,
      "loss": 1.5747,
      "step": 1045
    },
    {
      "epoch": 0.008368,
      "grad_norm": 3.015625,
      "learning_rate": 0.0009835632407864667,
      "loss": 2.6689,
      "step": 1046
    },
    {
      "epoch": 0.008376,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009835312056573641,
      "loss": 2.4178,
      "step": 1047
    },
    {
      "epoch": 0.008384,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009834991398632491,
      "loss": 2.3673,
      "step": 1048
    },
    {
      "epoch": 0.008392,
      "grad_norm": 3.203125,
      "learning_rate": 0.0009834670434061556,
      "loss": 3.0684,
      "step": 1049
    },
    {
      "epoch": 0.0084,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009834349162881188,
      "loss": 2.1458,
      "step": 1050
    },
    {
      "epoch": 0.008408,
      "grad_norm": 2.5,
      "learning_rate": 0.0009834027585111766,
      "loss": 1.8609,
      "step": 1051
    },
    {
      "epoch": 0.008416,
      "grad_norm": 2.734375,
      "learning_rate": 0.000983370570077368,
      "loss": 2.0292,
      "step": 1052
    },
    {
      "epoch": 0.008424,
      "grad_norm": 3.515625,
      "learning_rate": 0.0009833383509887347,
      "loss": 2.4118,
      "step": 1053
    },
    {
      "epoch": 0.008432,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009833061012473198,
      "loss": 2.2706,
      "step": 1054
    },
    {
      "epoch": 0.00844,
      "grad_norm": 2.25,
      "learning_rate": 0.0009832738208551683,
      "loss": 1.7236,
      "step": 1055
    },
    {
      "epoch": 0.008448,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009832415098143278,
      "loss": 1.9321,
      "step": 1056
    },
    {
      "epoch": 0.008456,
      "grad_norm": 3.640625,
      "learning_rate": 0.0009832091681268474,
      "loss": 2.4944,
      "step": 1057
    },
    {
      "epoch": 0.008464,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009831767957947778,
      "loss": 2.3575,
      "step": 1058
    },
    {
      "epoch": 0.008472,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009831443928201727,
      "loss": 2.5896,
      "step": 1059
    },
    {
      "epoch": 0.00848,
      "grad_norm": 6.09375,
      "learning_rate": 0.0009831119592050863,
      "loss": 1.95,
      "step": 1060
    },
    {
      "epoch": 0.008488,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009830794949515758,
      "loss": 2.4578,
      "step": 1061
    },
    {
      "epoch": 0.008496,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009830470000617003,
      "loss": 2.2318,
      "step": 1062
    },
    {
      "epoch": 0.008504,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009830144745375203,
      "loss": 2.041,
      "step": 1063
    },
    {
      "epoch": 0.008512,
      "grad_norm": 5.5,
      "learning_rate": 0.0009829819183810986,
      "loss": 1.8644,
      "step": 1064
    },
    {
      "epoch": 0.00852,
      "grad_norm": 3.609375,
      "learning_rate": 0.0009829493315944998,
      "loss": 2.6589,
      "step": 1065
    },
    {
      "epoch": 0.008528,
      "grad_norm": 3.375,
      "learning_rate": 0.0009829167141797907,
      "loss": 1.5465,
      "step": 1066
    },
    {
      "epoch": 0.008536,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009828840661390396,
      "loss": 1.8352,
      "step": 1067
    },
    {
      "epoch": 0.008544,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009828513874743175,
      "loss": 2.7023,
      "step": 1068
    },
    {
      "epoch": 0.008552,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009828186781876962,
      "loss": 2.3296,
      "step": 1069
    },
    {
      "epoch": 0.00856,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009827859382812506,
      "loss": 2.8014,
      "step": 1070
    },
    {
      "epoch": 0.008568,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009827531677570568,
      "loss": 2.1331,
      "step": 1071
    },
    {
      "epoch": 0.008576,
      "grad_norm": 3.296875,
      "learning_rate": 0.0009827203666171928,
      "loss": 2.1787,
      "step": 1072
    },
    {
      "epoch": 0.008584,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009826875348637393,
      "loss": 1.9714,
      "step": 1073
    },
    {
      "epoch": 0.008592,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009826546724987784,
      "loss": 2.1727,
      "step": 1074
    },
    {
      "epoch": 0.0086,
      "grad_norm": 9.625,
      "learning_rate": 0.000982621779524394,
      "loss": 2.3304,
      "step": 1075
    },
    {
      "epoch": 0.008608,
      "grad_norm": 2.75,
      "learning_rate": 0.0009825888559426722,
      "loss": 2.0995,
      "step": 1076
    },
    {
      "epoch": 0.008616,
      "grad_norm": 3.125,
      "learning_rate": 0.0009825559017557007,
      "loss": 2.3621,
      "step": 1077
    },
    {
      "epoch": 0.008624,
      "grad_norm": 3.25,
      "learning_rate": 0.0009825229169655699,
      "loss": 2.5175,
      "step": 1078
    },
    {
      "epoch": 0.008632,
      "grad_norm": 5.25,
      "learning_rate": 0.0009824899015743714,
      "loss": 2.5876,
      "step": 1079
    },
    {
      "epoch": 0.00864,
      "grad_norm": 3.078125,
      "learning_rate": 0.000982456855584199,
      "loss": 2.6256,
      "step": 1080
    },
    {
      "epoch": 0.008648,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009824237789971485,
      "loss": 2.2485,
      "step": 1081
    },
    {
      "epoch": 0.008656,
      "grad_norm": 2.625,
      "learning_rate": 0.0009823906718153177,
      "loss": 2.2748,
      "step": 1082
    },
    {
      "epoch": 0.008664,
      "grad_norm": 2.96875,
      "learning_rate": 0.000982357534040806,
      "loss": 2.1666,
      "step": 1083
    },
    {
      "epoch": 0.008672,
      "grad_norm": 2.75,
      "learning_rate": 0.000982324365675715,
      "loss": 2.3536,
      "step": 1084
    },
    {
      "epoch": 0.00868,
      "grad_norm": 4.65625,
      "learning_rate": 0.0009822911667221484,
      "loss": 1.9162,
      "step": 1085
    },
    {
      "epoch": 0.008688,
      "grad_norm": 4.1875,
      "learning_rate": 0.0009822579371822114,
      "loss": 2.2568,
      "step": 1086
    },
    {
      "epoch": 0.008696,
      "grad_norm": 3.234375,
      "learning_rate": 0.0009822246770580113,
      "loss": 2.4055,
      "step": 1087
    },
    {
      "epoch": 0.008704,
      "grad_norm": 6.1875,
      "learning_rate": 0.0009821913863516578,
      "loss": 2.5932,
      "step": 1088
    },
    {
      "epoch": 0.008712,
      "grad_norm": 3.609375,
      "learning_rate": 0.000982158065065262,
      "loss": 2.4764,
      "step": 1089
    },
    {
      "epoch": 0.00872,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009821247132009367,
      "loss": 1.9833,
      "step": 1090
    },
    {
      "epoch": 0.008728,
      "grad_norm": 16.25,
      "learning_rate": 0.0009820913307607977,
      "loss": 2.1663,
      "step": 1091
    },
    {
      "epoch": 0.008736,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009820579177469615,
      "loss": 3.1541,
      "step": 1092
    },
    {
      "epoch": 0.008744,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009820244741615477,
      "loss": 2.456,
      "step": 1093
    },
    {
      "epoch": 0.008752,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009819910000066767,
      "loss": 2.4852,
      "step": 1094
    },
    {
      "epoch": 0.00876,
      "grad_norm": 6.625,
      "learning_rate": 0.0009819574952844717,
      "loss": 1.9264,
      "step": 1095
    },
    {
      "epoch": 0.008768,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009819239599970573,
      "loss": 2.2882,
      "step": 1096
    },
    {
      "epoch": 0.008776,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009818903941465607,
      "loss": 2.4658,
      "step": 1097
    },
    {
      "epoch": 0.008784,
      "grad_norm": 3.375,
      "learning_rate": 0.0009818567977351103,
      "loss": 2.1053,
      "step": 1098
    },
    {
      "epoch": 0.008792,
      "grad_norm": 3.203125,
      "learning_rate": 0.0009818231707648367,
      "loss": 2.5472,
      "step": 1099
    },
    {
      "epoch": 0.0088,
      "grad_norm": 3.828125,
      "learning_rate": 0.0009817895132378724,
      "loss": 2.774,
      "step": 1100
    },
    {
      "epoch": 0.008808,
      "grad_norm": 3.4375,
      "learning_rate": 0.0009817558251563525,
      "loss": 2.5896,
      "step": 1101
    },
    {
      "epoch": 0.008816,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009817221065224126,
      "loss": 2.4196,
      "step": 1102
    },
    {
      "epoch": 0.008824,
      "grad_norm": 3.4375,
      "learning_rate": 0.000981688357338192,
      "loss": 2.2919,
      "step": 1103
    },
    {
      "epoch": 0.008832,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009816545776058303,
      "loss": 2.5547,
      "step": 1104
    },
    {
      "epoch": 0.00884,
      "grad_norm": 3.234375,
      "learning_rate": 0.0009816207673274701,
      "loss": 2.509,
      "step": 1105
    },
    {
      "epoch": 0.008848,
      "grad_norm": 3.375,
      "learning_rate": 0.0009815869265052555,
      "loss": 2.7982,
      "step": 1106
    },
    {
      "epoch": 0.008856,
      "grad_norm": 3.078125,
      "learning_rate": 0.000981553055141333,
      "loss": 2.2748,
      "step": 1107
    },
    {
      "epoch": 0.008864,
      "grad_norm": 3.703125,
      "learning_rate": 0.00098151915323785,
      "loss": 2.4906,
      "step": 1108
    },
    {
      "epoch": 0.008872,
      "grad_norm": 3.0,
      "learning_rate": 0.0009814852207969571,
      "loss": 2.9099,
      "step": 1109
    },
    {
      "epoch": 0.00888,
      "grad_norm": 3.0625,
      "learning_rate": 0.0009814512578208063,
      "loss": 2.236,
      "step": 1110
    },
    {
      "epoch": 0.008888,
      "grad_norm": 3.328125,
      "learning_rate": 0.000981417264311551,
      "loss": 2.43,
      "step": 1111
    },
    {
      "epoch": 0.008896,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009813832402713475,
      "loss": 2.3832,
      "step": 1112
    },
    {
      "epoch": 0.008904,
      "grad_norm": 3.921875,
      "learning_rate": 0.0009813491857023533,
      "loss": 2.4108,
      "step": 1113
    },
    {
      "epoch": 0.008912,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009813151006067282,
      "loss": 2.6897,
      "step": 1114
    },
    {
      "epoch": 0.00892,
      "grad_norm": 4.125,
      "learning_rate": 0.0009812809849866337,
      "loss": 2.2368,
      "step": 1115
    },
    {
      "epoch": 0.008928,
      "grad_norm": 3.53125,
      "learning_rate": 0.0009812468388442335,
      "loss": 2.3624,
      "step": 1116
    },
    {
      "epoch": 0.008936,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009812126621816933,
      "loss": 2.479,
      "step": 1117
    },
    {
      "epoch": 0.008944,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009811784550011803,
      "loss": 2.3829,
      "step": 1118
    },
    {
      "epoch": 0.008952,
      "grad_norm": 22.75,
      "learning_rate": 0.0009811442173048638,
      "loss": 2.4674,
      "step": 1119
    },
    {
      "epoch": 0.00896,
      "grad_norm": 3.625,
      "learning_rate": 0.000981109949094915,
      "loss": 3.1102,
      "step": 1120
    },
    {
      "epoch": 0.008968,
      "grad_norm": 2.921875,
      "learning_rate": 0.000981075650373508,
      "loss": 2.306,
      "step": 1121
    },
    {
      "epoch": 0.008976,
      "grad_norm": 5.03125,
      "learning_rate": 0.000981041321142817,
      "loss": 2.4479,
      "step": 1122
    },
    {
      "epoch": 0.008984,
      "grad_norm": 3.796875,
      "learning_rate": 0.0009810069614050196,
      "loss": 2.273,
      "step": 1123
    },
    {
      "epoch": 0.008992,
      "grad_norm": 4.53125,
      "learning_rate": 0.0009809725711622946,
      "loss": 2.3488,
      "step": 1124
    },
    {
      "epoch": 0.009,
      "grad_norm": 70.5,
      "learning_rate": 0.0009809381504168233,
      "loss": 2.1604,
      "step": 1125
    },
    {
      "epoch": 0.009008,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009809036991707885,
      "loss": 2.2615,
      "step": 1126
    },
    {
      "epoch": 0.009016,
      "grad_norm": 3.578125,
      "learning_rate": 0.000980869217426375,
      "loss": 2.6198,
      "step": 1127
    },
    {
      "epoch": 0.009024,
      "grad_norm": 4.875,
      "learning_rate": 0.0009808347051857697,
      "loss": 2.0359,
      "step": 1128
    },
    {
      "epoch": 0.009032,
      "grad_norm": 32.0,
      "learning_rate": 0.000980800162451161,
      "loss": 2.1418,
      "step": 1129
    },
    {
      "epoch": 0.00904,
      "grad_norm": 4.46875,
      "learning_rate": 0.0009807655892247402,
      "loss": 2.6599,
      "step": 1130
    },
    {
      "epoch": 0.009048,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009807309855086994,
      "loss": 2.5127,
      "step": 1131
    },
    {
      "epoch": 0.009056,
      "grad_norm": 3.15625,
      "learning_rate": 0.0009806963513052333,
      "loss": 3.3395,
      "step": 1132
    },
    {
      "epoch": 0.009064,
      "grad_norm": 3.234375,
      "learning_rate": 0.000980661686616538,
      "loss": 2.9129,
      "step": 1133
    },
    {
      "epoch": 0.009072,
      "grad_norm": 3.0625,
      "learning_rate": 0.0009806269914448124,
      "loss": 2.4544,
      "step": 1134
    },
    {
      "epoch": 0.00908,
      "grad_norm": 3.1875,
      "learning_rate": 0.0009805922657922567,
      "loss": 2.5791,
      "step": 1135
    },
    {
      "epoch": 0.009088,
      "grad_norm": 2.375,
      "learning_rate": 0.0009805575096610728,
      "loss": 2.002,
      "step": 1136
    },
    {
      "epoch": 0.009096,
      "grad_norm": 3.34375,
      "learning_rate": 0.0009805227230534656,
      "loss": 3.0077,
      "step": 1137
    },
    {
      "epoch": 0.009104,
      "grad_norm": 2.375,
      "learning_rate": 0.0009804879059716405,
      "loss": 1.9527,
      "step": 1138
    },
    {
      "epoch": 0.009112,
      "grad_norm": 3.875,
      "learning_rate": 0.0009804530584178062,
      "loss": 2.329,
      "step": 1139
    },
    {
      "epoch": 0.00912,
      "grad_norm": 2.453125,
      "learning_rate": 0.000980418180394172,
      "loss": 2.2127,
      "step": 1140
    },
    {
      "epoch": 0.009128,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009803832719029507,
      "loss": 2.6774,
      "step": 1141
    },
    {
      "epoch": 0.009136,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009803483329463552,
      "loss": 2.5821,
      "step": 1142
    },
    {
      "epoch": 0.009144,
      "grad_norm": 40.75,
      "learning_rate": 0.0009803133635266019,
      "loss": 2.631,
      "step": 1143
    },
    {
      "epoch": 0.009152,
      "grad_norm": 2.328125,
      "learning_rate": 0.0009802783636459083,
      "loss": 1.8657,
      "step": 1144
    },
    {
      "epoch": 0.00916,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009802433333064944,
      "loss": 2.5771,
      "step": 1145
    },
    {
      "epoch": 0.009168,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009802082725105811,
      "loss": 2.3805,
      "step": 1146
    },
    {
      "epoch": 0.009176,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009801731812603927,
      "loss": 2.475,
      "step": 1147
    },
    {
      "epoch": 0.009184,
      "grad_norm": 4.03125,
      "learning_rate": 0.000980138059558154,
      "loss": 2.2304,
      "step": 1148
    },
    {
      "epoch": 0.009192,
      "grad_norm": 3.421875,
      "learning_rate": 0.0009801029074060928,
      "loss": 2.2151,
      "step": 1149
    },
    {
      "epoch": 0.0092,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009800677248064382,
      "loss": 2.3945,
      "step": 1150
    },
    {
      "epoch": 0.009208,
      "grad_norm": 3.515625,
      "learning_rate": 0.0009800325117614215,
      "loss": 2.3751,
      "step": 1151
    },
    {
      "epoch": 0.009216,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009799972682732758,
      "loss": 2.4393,
      "step": 1152
    },
    {
      "epoch": 0.009224,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009799619943442364,
      "loss": 2.4102,
      "step": 1153
    },
    {
      "epoch": 0.009232,
      "grad_norm": 4.40625,
      "learning_rate": 0.0009799266899765402,
      "loss": 2.4199,
      "step": 1154
    },
    {
      "epoch": 0.00924,
      "grad_norm": 3.46875,
      "learning_rate": 0.0009798913551724263,
      "loss": 2.2432,
      "step": 1155
    },
    {
      "epoch": 0.009248,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009798559899341353,
      "loss": 2.8464,
      "step": 1156
    },
    {
      "epoch": 0.009256,
      "grad_norm": 3.5625,
      "learning_rate": 0.0009798205942639102,
      "loss": 1.9444,
      "step": 1157
    },
    {
      "epoch": 0.009264,
      "grad_norm": 3.390625,
      "learning_rate": 0.0009797851681639957,
      "loss": 2.3227,
      "step": 1158
    },
    {
      "epoch": 0.009272,
      "grad_norm": 3.25,
      "learning_rate": 0.0009797497116366388,
      "loss": 2.255,
      "step": 1159
    },
    {
      "epoch": 0.00928,
      "grad_norm": 5.96875,
      "learning_rate": 0.0009797142246840875,
      "loss": 3.0305,
      "step": 1160
    },
    {
      "epoch": 0.009288,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009796787073085929,
      "loss": 2.1952,
      "step": 1161
    },
    {
      "epoch": 0.009296,
      "grad_norm": 2.53125,
      "learning_rate": 0.000979643159512407,
      "loss": 2.4221,
      "step": 1162
    },
    {
      "epoch": 0.009304,
      "grad_norm": 31.0,
      "learning_rate": 0.0009796075812977848,
      "loss": 2.0702,
      "step": 1163
    },
    {
      "epoch": 0.009312,
      "grad_norm": 326.0,
      "learning_rate": 0.0009795719726669824,
      "loss": 2.4645,
      "step": 1164
    },
    {
      "epoch": 0.00932,
      "grad_norm": 2.921875,
      "learning_rate": 0.0009795363336222577,
      "loss": 2.3498,
      "step": 1165
    },
    {
      "epoch": 0.009328,
      "grad_norm": 3.53125,
      "learning_rate": 0.000979500664165871,
      "loss": 2.2098,
      "step": 1166
    },
    {
      "epoch": 0.009336,
      "grad_norm": 3.421875,
      "learning_rate": 0.0009794649643000846,
      "loss": 2.3361,
      "step": 1167
    },
    {
      "epoch": 0.009344,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009794292340271626,
      "loss": 2.2592,
      "step": 1168
    },
    {
      "epoch": 0.009352,
      "grad_norm": 3.09375,
      "learning_rate": 0.000979393473349371,
      "loss": 2.4585,
      "step": 1169
    },
    {
      "epoch": 0.00936,
      "grad_norm": 3.640625,
      "learning_rate": 0.0009793576822689776,
      "loss": 2.2381,
      "step": 1170
    },
    {
      "epoch": 0.009368,
      "grad_norm": 3.203125,
      "learning_rate": 0.000979321860788252,
      "loss": 2.4181,
      "step": 1171
    },
    {
      "epoch": 0.009376,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009792860089094665,
      "loss": 2.2664,
      "step": 1172
    },
    {
      "epoch": 0.009384,
      "grad_norm": 4.28125,
      "learning_rate": 0.000979250126634894,
      "loss": 2.2472,
      "step": 1173
    },
    {
      "epoch": 0.009392,
      "grad_norm": 2.453125,
      "learning_rate": 0.0009792142139668108,
      "loss": 1.8541,
      "step": 1174
    },
    {
      "epoch": 0.0094,
      "grad_norm": 4.1875,
      "learning_rate": 0.0009791782709074944,
      "loss": 2.5951,
      "step": 1175
    },
    {
      "epoch": 0.009408,
      "grad_norm": 2.5,
      "learning_rate": 0.0009791422974592238,
      "loss": 1.8411,
      "step": 1176
    },
    {
      "epoch": 0.009416,
      "grad_norm": 3.25,
      "learning_rate": 0.0009791062936242807,
      "loss": 2.7685,
      "step": 1177
    },
    {
      "epoch": 0.009424,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009790702594049485,
      "loss": 2.0299,
      "step": 1178
    },
    {
      "epoch": 0.009432,
      "grad_norm": 2.875,
      "learning_rate": 0.0009790341948035125,
      "loss": 2.6598,
      "step": 1179
    },
    {
      "epoch": 0.00944,
      "grad_norm": 3.34375,
      "learning_rate": 0.0009789980998222594,
      "loss": 2.0269,
      "step": 1180
    },
    {
      "epoch": 0.009448,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009789619744634789,
      "loss": 1.6833,
      "step": 1181
    },
    {
      "epoch": 0.009456,
      "grad_norm": 3.15625,
      "learning_rate": 0.0009789258187294617,
      "loss": 2.2923,
      "step": 1182
    },
    {
      "epoch": 0.009464,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009788896326225008,
      "loss": 2.1057,
      "step": 1183
    },
    {
      "epoch": 0.009472,
      "grad_norm": 4.125,
      "learning_rate": 0.000978853416144891,
      "loss": 2.3777,
      "step": 1184
    },
    {
      "epoch": 0.00948,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009788171692989294,
      "loss": 3.0205,
      "step": 1185
    },
    {
      "epoch": 0.009488,
      "grad_norm": 3.375,
      "learning_rate": 0.0009787808920869144,
      "loss": 2.5063,
      "step": 1186
    },
    {
      "epoch": 0.009496,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009787445845111469,
      "loss": 2.7476,
      "step": 1187
    },
    {
      "epoch": 0.009504,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009787082465739293,
      "loss": 2.0509,
      "step": 1188
    },
    {
      "epoch": 0.009512,
      "grad_norm": 3.828125,
      "learning_rate": 0.0009786718782775663,
      "loss": 2.532,
      "step": 1189
    },
    {
      "epoch": 0.00952,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009786354796243641,
      "loss": 2.4438,
      "step": 1190
    },
    {
      "epoch": 0.009528,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009785990506166312,
      "loss": 2.7103,
      "step": 1191
    },
    {
      "epoch": 0.009536,
      "grad_norm": 2.734375,
      "learning_rate": 0.000978562591256678,
      "loss": 2.3712,
      "step": 1192
    },
    {
      "epoch": 0.009544,
      "grad_norm": 2.921875,
      "learning_rate": 0.0009785261015468167,
      "loss": 2.2404,
      "step": 1193
    },
    {
      "epoch": 0.009552,
      "grad_norm": 3.421875,
      "learning_rate": 0.0009784895814893611,
      "loss": 3.0869,
      "step": 1194
    },
    {
      "epoch": 0.00956,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009784530310866279,
      "loss": 1.9674,
      "step": 1195
    },
    {
      "epoch": 0.009568,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009784164503409342,
      "loss": 2.2707,
      "step": 1196
    },
    {
      "epoch": 0.009576,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009783798392546006,
      "loss": 1.7093,
      "step": 1197
    },
    {
      "epoch": 0.009584,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009783431978299488,
      "loss": 2.3292,
      "step": 1198
    },
    {
      "epoch": 0.009592,
      "grad_norm": 3.828125,
      "learning_rate": 0.0009783065260693025,
      "loss": 1.7621,
      "step": 1199
    },
    {
      "epoch": 0.0096,
      "grad_norm": 4.34375,
      "learning_rate": 0.0009782698239749871,
      "loss": 2.5254,
      "step": 1200
    },
    {
      "epoch": 0.009608,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009782330915493306,
      "loss": 2.3107,
      "step": 1201
    },
    {
      "epoch": 0.009616,
      "grad_norm": 3.640625,
      "learning_rate": 0.0009781963287946627,
      "loss": 2.3161,
      "step": 1202
    },
    {
      "epoch": 0.009624,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009781595357133144,
      "loss": 2.3591,
      "step": 1203
    },
    {
      "epoch": 0.009632,
      "grad_norm": 3.296875,
      "learning_rate": 0.000978122712307619,
      "loss": 1.8024,
      "step": 1204
    },
    {
      "epoch": 0.00964,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009780858585799123,
      "loss": 1.6226,
      "step": 1205
    },
    {
      "epoch": 0.009648,
      "grad_norm": 4.0,
      "learning_rate": 0.000978048974532531,
      "loss": 2.1137,
      "step": 1206
    },
    {
      "epoch": 0.009656,
      "grad_norm": 3.375,
      "learning_rate": 0.0009780120601678148,
      "loss": 2.2401,
      "step": 1207
    },
    {
      "epoch": 0.009664,
      "grad_norm": 3.578125,
      "learning_rate": 0.0009779751154881044,
      "loss": 2.5201,
      "step": 1208
    },
    {
      "epoch": 0.009672,
      "grad_norm": 4.0625,
      "learning_rate": 0.0009779381404957429,
      "loss": 2.7197,
      "step": 1209
    },
    {
      "epoch": 0.00968,
      "grad_norm": 3.78125,
      "learning_rate": 0.000977901135193075,
      "loss": 2.5489,
      "step": 1210
    },
    {
      "epoch": 0.009688,
      "grad_norm": 2.390625,
      "learning_rate": 0.000977864099582448,
      "loss": 2.0973,
      "step": 1211
    },
    {
      "epoch": 0.009696,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009778270336662101,
      "loss": 2.1735,
      "step": 1212
    },
    {
      "epoch": 0.009704,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009777899374467122,
      "loss": 2.7415,
      "step": 1213
    },
    {
      "epoch": 0.009712,
      "grad_norm": 2.828125,
      "learning_rate": 0.0009777528109263071,
      "loss": 2.037,
      "step": 1214
    },
    {
      "epoch": 0.00972,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009777156541073493,
      "loss": 2.0933,
      "step": 1215
    },
    {
      "epoch": 0.009728,
      "grad_norm": 2.8125,
      "learning_rate": 0.000977678466992195,
      "loss": 2.4089,
      "step": 1216
    },
    {
      "epoch": 0.009736,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009776412495832026,
      "loss": 2.8237,
      "step": 1217
    },
    {
      "epoch": 0.009744,
      "grad_norm": 3.3125,
      "learning_rate": 0.0009776040018827327,
      "loss": 2.8598,
      "step": 1218
    },
    {
      "epoch": 0.009752,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009775667238931468,
      "loss": 2.3332,
      "step": 1219
    },
    {
      "epoch": 0.00976,
      "grad_norm": 167.0,
      "learning_rate": 0.00097752941561681,
      "loss": 2.5717,
      "step": 1220
    },
    {
      "epoch": 0.009768,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009774920770560876,
      "loss": 2.3958,
      "step": 1221
    },
    {
      "epoch": 0.009776,
      "grad_norm": 4.3125,
      "learning_rate": 0.0009774547082133478,
      "loss": 2.3754,
      "step": 1222
    },
    {
      "epoch": 0.009784,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009774173090909607,
      "loss": 2.4088,
      "step": 1223
    },
    {
      "epoch": 0.009792,
      "grad_norm": 3.46875,
      "learning_rate": 0.000977379879691298,
      "loss": 2.5859,
      "step": 1224
    },
    {
      "epoch": 0.0098,
      "grad_norm": 3.515625,
      "learning_rate": 0.000977342420016733,
      "loss": 2.4757,
      "step": 1225
    },
    {
      "epoch": 0.009808,
      "grad_norm": 3.46875,
      "learning_rate": 0.000977304930069642,
      "loss": 2.3152,
      "step": 1226
    },
    {
      "epoch": 0.009816,
      "grad_norm": 3.421875,
      "learning_rate": 0.000977267409852402,
      "loss": 2.0647,
      "step": 1227
    },
    {
      "epoch": 0.009824,
      "grad_norm": 3.78125,
      "learning_rate": 0.000977229859367393,
      "loss": 2.7476,
      "step": 1228
    },
    {
      "epoch": 0.009832,
      "grad_norm": 4.53125,
      "learning_rate": 0.0009771922786169962,
      "loss": 3.6881,
      "step": 1229
    },
    {
      "epoch": 0.00984,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009771546676035946,
      "loss": 2.3479,
      "step": 1230
    },
    {
      "epoch": 0.009848,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009771170263295742,
      "loss": 2.4004,
      "step": 1231
    },
    {
      "epoch": 0.009856,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009770793547973214,
      "loss": 2.3488,
      "step": 1232
    },
    {
      "epoch": 0.009864,
      "grad_norm": 3.734375,
      "learning_rate": 0.0009770416530092258,
      "loss": 2.6904,
      "step": 1233
    },
    {
      "epoch": 0.009872,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009770039209676781,
      "loss": 2.4115,
      "step": 1234
    },
    {
      "epoch": 0.00988,
      "grad_norm": 3.15625,
      "learning_rate": 0.0009769661586750712,
      "loss": 2.6776,
      "step": 1235
    },
    {
      "epoch": 0.009888,
      "grad_norm": 2.875,
      "learning_rate": 0.0009769283661338005,
      "loss": 2.1255,
      "step": 1236
    },
    {
      "epoch": 0.009896,
      "grad_norm": 2.640625,
      "learning_rate": 0.000976890543346262,
      "loss": 2.2408,
      "step": 1237
    },
    {
      "epoch": 0.009904,
      "grad_norm": 2.96875,
      "learning_rate": 0.0009768526903148548,
      "loss": 2.4357,
      "step": 1238
    },
    {
      "epoch": 0.009912,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009768148070419796,
      "loss": 2.313,
      "step": 1239
    },
    {
      "epoch": 0.00992,
      "grad_norm": 3.3125,
      "learning_rate": 0.0009767768935300385,
      "loss": 2.1613,
      "step": 1240
    },
    {
      "epoch": 0.009928,
      "grad_norm": 3.25,
      "learning_rate": 0.0009767389497814362,
      "loss": 2.5997,
      "step": 1241
    },
    {
      "epoch": 0.009936,
      "grad_norm": 3.359375,
      "learning_rate": 0.0009767009757985791,
      "loss": 2.523,
      "step": 1242
    },
    {
      "epoch": 0.009944,
      "grad_norm": 3.546875,
      "learning_rate": 0.0009766629715838755,
      "loss": 2.6024,
      "step": 1243
    },
    {
      "epoch": 0.009952,
      "grad_norm": 2.515625,
      "learning_rate": 0.0009766249371397353,
      "loss": 2.0717,
      "step": 1244
    },
    {
      "epoch": 0.00996,
      "grad_norm": 2.953125,
      "learning_rate": 0.000976586872468571,
      "loss": 2.6215,
      "step": 1245
    },
    {
      "epoch": 0.009968,
      "grad_norm": 2.84375,
      "learning_rate": 0.000976548777572796,
      "loss": 2.7619,
      "step": 1246
    },
    {
      "epoch": 0.009976,
      "grad_norm": 2.75,
      "learning_rate": 0.0009765106524548268,
      "loss": 2.6658,
      "step": 1247
    },
    {
      "epoch": 0.009984,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009764724971170812,
      "loss": 3.0304,
      "step": 1248
    },
    {
      "epoch": 0.009992,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009764343115619787,
      "loss": 2.0588,
      "step": 1249
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.34375,
      "learning_rate": 0.0009763960957919414,
      "loss": 2.0798,
      "step": 1250
    },
    {
      "epoch": 0.010008,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009763578498093923,
      "loss": 1.9572,
      "step": 1251
    },
    {
      "epoch": 0.010016,
      "grad_norm": 2.109375,
      "learning_rate": 0.0009763195736167575,
      "loss": 1.8039,
      "step": 1252
    },
    {
      "epoch": 0.010024,
      "grad_norm": 3.15625,
      "learning_rate": 0.0009762812672164641,
      "loss": 2.2611,
      "step": 1253
    },
    {
      "epoch": 0.010032,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009762429306109415,
      "loss": 2.6415,
      "step": 1254
    },
    {
      "epoch": 0.01004,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009762045638026212,
      "loss": 2.5496,
      "step": 1255
    },
    {
      "epoch": 0.010048,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009761661667939359,
      "loss": 2.5062,
      "step": 1256
    },
    {
      "epoch": 0.010056,
      "grad_norm": 2.3125,
      "learning_rate": 0.0009761277395873211,
      "loss": 1.9854,
      "step": 1257
    },
    {
      "epoch": 0.010064,
      "grad_norm": 2.875,
      "learning_rate": 0.0009760892821852138,
      "loss": 2.1809,
      "step": 1258
    },
    {
      "epoch": 0.010072,
      "grad_norm": 2.875,
      "learning_rate": 0.0009760507945900527,
      "loss": 2.7308,
      "step": 1259
    },
    {
      "epoch": 0.01008,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009760122768042788,
      "loss": 2.1488,
      "step": 1260
    },
    {
      "epoch": 0.010088,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009759737288303347,
      "loss": 2.2404,
      "step": 1261
    },
    {
      "epoch": 0.010096,
      "grad_norm": 3.515625,
      "learning_rate": 0.0009759351506706653,
      "loss": 2.3416,
      "step": 1262
    },
    {
      "epoch": 0.010104,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009758965423277171,
      "loss": 2.0936,
      "step": 1263
    },
    {
      "epoch": 0.010112,
      "grad_norm": 2.921875,
      "learning_rate": 0.0009758579038039386,
      "loss": 3.0488,
      "step": 1264
    },
    {
      "epoch": 0.01012,
      "grad_norm": 2.625,
      "learning_rate": 0.00097581923510178,
      "loss": 2.0543,
      "step": 1265
    },
    {
      "epoch": 0.010128,
      "grad_norm": 3.328125,
      "learning_rate": 0.000975780536223694,
      "loss": 2.3157,
      "step": 1266
    },
    {
      "epoch": 0.010136,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009757418071721345,
      "loss": 2.7794,
      "step": 1267
    },
    {
      "epoch": 0.010144,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009757030479495579,
      "loss": 1.9007,
      "step": 1268
    },
    {
      "epoch": 0.010152,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009756642585584221,
      "loss": 2.449,
      "step": 1269
    },
    {
      "epoch": 0.01016,
      "grad_norm": 3.671875,
      "learning_rate": 0.0009756254390011871,
      "loss": 2.4513,
      "step": 1270
    },
    {
      "epoch": 0.010168,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009755865892803148,
      "loss": 2.2878,
      "step": 1271
    },
    {
      "epoch": 0.010176,
      "grad_norm": 3.203125,
      "learning_rate": 0.0009755477093982692,
      "loss": 2.0278,
      "step": 1272
    },
    {
      "epoch": 0.010184,
      "grad_norm": 3.0,
      "learning_rate": 0.0009755087993575158,
      "loss": 2.4454,
      "step": 1273
    },
    {
      "epoch": 0.010192,
      "grad_norm": 3.53125,
      "learning_rate": 0.0009754698591605223,
      "loss": 3.1321,
      "step": 1274
    },
    {
      "epoch": 0.0102,
      "grad_norm": 3.734375,
      "learning_rate": 0.0009754308888097583,
      "loss": 2.6913,
      "step": 1275
    },
    {
      "epoch": 0.010208,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009753918883076951,
      "loss": 2.3973,
      "step": 1276
    },
    {
      "epoch": 0.010216,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009753528576568061,
      "loss": 2.1189,
      "step": 1277
    },
    {
      "epoch": 0.010224,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009753137968595668,
      "loss": 2.0419,
      "step": 1278
    },
    {
      "epoch": 0.010232,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009752747059184542,
      "loss": 1.7639,
      "step": 1279
    },
    {
      "epoch": 0.01024,
      "grad_norm": 3.65625,
      "learning_rate": 0.0009752355848359473,
      "loss": 2.4456,
      "step": 1280
    },
    {
      "epoch": 0.010248,
      "grad_norm": 3.25,
      "learning_rate": 0.0009751964336145273,
      "loss": 3.0387,
      "step": 1281
    },
    {
      "epoch": 0.010256,
      "grad_norm": 3.390625,
      "learning_rate": 0.0009751572522566773,
      "loss": 2.731,
      "step": 1282
    },
    {
      "epoch": 0.010264,
      "grad_norm": 3.5,
      "learning_rate": 0.0009751180407648818,
      "loss": 2.1537,
      "step": 1283
    },
    {
      "epoch": 0.010272,
      "grad_norm": 2.625,
      "learning_rate": 0.0009750787991416275,
      "loss": 2.2179,
      "step": 1284
    },
    {
      "epoch": 0.01028,
      "grad_norm": 3.828125,
      "learning_rate": 0.0009750395273894034,
      "loss": 2.7144,
      "step": 1285
    },
    {
      "epoch": 0.010288,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009750002255106999,
      "loss": 3.1902,
      "step": 1286
    },
    {
      "epoch": 0.010296,
      "grad_norm": 3.234375,
      "learning_rate": 0.0009749608935080094,
      "loss": 2.7522,
      "step": 1287
    },
    {
      "epoch": 0.010304,
      "grad_norm": 3.421875,
      "learning_rate": 0.0009749215313838265,
      "loss": 2.2962,
      "step": 1288
    },
    {
      "epoch": 0.010312,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009748821391406473,
      "loss": 2.6833,
      "step": 1289
    },
    {
      "epoch": 0.01032,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009748427167809702,
      "loss": 2.3747,
      "step": 1290
    },
    {
      "epoch": 0.010328,
      "grad_norm": 3.484375,
      "learning_rate": 0.000974803264307295,
      "loss": 3.1765,
      "step": 1291
    },
    {
      "epoch": 0.010336,
      "grad_norm": 3.0625,
      "learning_rate": 0.000974763781722124,
      "loss": 2.2542,
      "step": 1292
    },
    {
      "epoch": 0.010344,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009747242690279612,
      "loss": 2.5648,
      "step": 1293
    },
    {
      "epoch": 0.010352,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009746847262273122,
      "loss": 1.8832,
      "step": 1294
    },
    {
      "epoch": 0.01036,
      "grad_norm": 3.828125,
      "learning_rate": 0.000974645153322685,
      "loss": 1.9625,
      "step": 1295
    },
    {
      "epoch": 0.010368,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009746055503165892,
      "loss": 2.4567,
      "step": 1296
    },
    {
      "epoch": 0.010376,
      "grad_norm": 3.25,
      "learning_rate": 0.0009745659172115362,
      "loss": 2.6221,
      "step": 1297
    },
    {
      "epoch": 0.010384,
      "grad_norm": 3.390625,
      "learning_rate": 0.0009745262540100396,
      "loss": 1.7217,
      "step": 1298
    },
    {
      "epoch": 0.010392,
      "grad_norm": 3.109375,
      "learning_rate": 0.0009744865607146151,
      "loss": 2.215,
      "step": 1299
    },
    {
      "epoch": 0.0104,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009744468373277796,
      "loss": 2.2563,
      "step": 1300
    },
    {
      "epoch": 0.010408,
      "grad_norm": 3.109375,
      "learning_rate": 0.0009744070838520524,
      "loss": 2.4379,
      "step": 1301
    },
    {
      "epoch": 0.010416,
      "grad_norm": 3.0625,
      "learning_rate": 0.0009743673002899546,
      "loss": 2.4221,
      "step": 1302
    },
    {
      "epoch": 0.010424,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009743274866440095,
      "loss": 1.9019,
      "step": 1303
    },
    {
      "epoch": 0.010432,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009742876429167418,
      "loss": 2.2665,
      "step": 1304
    },
    {
      "epoch": 0.01044,
      "grad_norm": 3.609375,
      "learning_rate": 0.0009742477691106783,
      "loss": 2.4283,
      "step": 1305
    },
    {
      "epoch": 0.010448,
      "grad_norm": 20.625,
      "learning_rate": 0.0009742078652283478,
      "loss": 1.7838,
      "step": 1306
    },
    {
      "epoch": 0.010456,
      "grad_norm": 3.015625,
      "learning_rate": 0.0009741679312722811,
      "loss": 2.345,
      "step": 1307
    },
    {
      "epoch": 0.010464,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009741279672450105,
      "loss": 2.2458,
      "step": 1308
    },
    {
      "epoch": 0.010472,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009740879731490708,
      "loss": 2.9518,
      "step": 1309
    },
    {
      "epoch": 0.01048,
      "grad_norm": 2.9375,
      "learning_rate": 0.000974047948986998,
      "loss": 2.2773,
      "step": 1310
    },
    {
      "epoch": 0.010488,
      "grad_norm": 2.875,
      "learning_rate": 0.0009740078947613309,
      "loss": 2.2202,
      "step": 1311
    },
    {
      "epoch": 0.010496,
      "grad_norm": 2.875,
      "learning_rate": 0.0009739678104746091,
      "loss": 2.2916,
      "step": 1312
    },
    {
      "epoch": 0.010504,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009739276961293753,
      "loss": 2.0818,
      "step": 1313
    },
    {
      "epoch": 0.010512,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009738875517281729,
      "loss": 2.0408,
      "step": 1314
    },
    {
      "epoch": 0.01052,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009738473772735484,
      "loss": 2.2331,
      "step": 1315
    },
    {
      "epoch": 0.010528,
      "grad_norm": 3.734375,
      "learning_rate": 0.000973807172768049,
      "loss": 2.3156,
      "step": 1316
    },
    {
      "epoch": 0.010536,
      "grad_norm": 3.859375,
      "learning_rate": 0.0009737669382142251,
      "loss": 2.915,
      "step": 1317
    },
    {
      "epoch": 0.010544,
      "grad_norm": 2.484375,
      "learning_rate": 0.0009737266736146279,
      "loss": 1.9253,
      "step": 1318
    },
    {
      "epoch": 0.010552,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009736863789718109,
      "loss": 2.4065,
      "step": 1319
    },
    {
      "epoch": 0.01056,
      "grad_norm": 3.796875,
      "learning_rate": 0.0009736460542883299,
      "loss": 2.616,
      "step": 1320
    },
    {
      "epoch": 0.010568,
      "grad_norm": 3.421875,
      "learning_rate": 0.0009736056995667417,
      "loss": 2.9226,
      "step": 1321
    },
    {
      "epoch": 0.010576,
      "grad_norm": 2.625,
      "learning_rate": 0.0009735653148096062,
      "loss": 1.6124,
      "step": 1322
    },
    {
      "epoch": 0.010584,
      "grad_norm": 2.390625,
      "learning_rate": 0.000973524900019484,
      "loss": 1.9177,
      "step": 1323
    },
    {
      "epoch": 0.010592,
      "grad_norm": 3.5,
      "learning_rate": 0.0009734844551989385,
      "loss": 1.9827,
      "step": 1324
    },
    {
      "epoch": 0.0106,
      "grad_norm": 6.8125,
      "learning_rate": 0.0009734439803505345,
      "loss": 1.9332,
      "step": 1325
    },
    {
      "epoch": 0.010608,
      "grad_norm": 2.828125,
      "learning_rate": 0.000973403475476839,
      "loss": 2.6164,
      "step": 1326
    },
    {
      "epoch": 0.010616,
      "grad_norm": 4.40625,
      "learning_rate": 0.0009733629405804205,
      "loss": 3.4817,
      "step": 1327
    },
    {
      "epoch": 0.010624,
      "grad_norm": 3.25,
      "learning_rate": 0.0009733223756638501,
      "loss": 2.6044,
      "step": 1328
    },
    {
      "epoch": 0.010632,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009732817807297001,
      "loss": 2.2979,
      "step": 1329
    },
    {
      "epoch": 0.01064,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009732411557805449,
      "loss": 1.9315,
      "step": 1330
    },
    {
      "epoch": 0.010648,
      "grad_norm": 2.671875,
      "learning_rate": 0.0009732005008189613,
      "loss": 1.9644,
      "step": 1331
    },
    {
      "epoch": 0.010656,
      "grad_norm": 3.375,
      "learning_rate": 0.0009731598158475271,
      "loss": 2.1789,
      "step": 1332
    },
    {
      "epoch": 0.010664,
      "grad_norm": 2.390625,
      "learning_rate": 0.0009731191008688227,
      "loss": 2.1048,
      "step": 1333
    },
    {
      "epoch": 0.010672,
      "grad_norm": 4.28125,
      "learning_rate": 0.0009730783558854303,
      "loss": 2.1208,
      "step": 1334
    },
    {
      "epoch": 0.01068,
      "grad_norm": 3.296875,
      "learning_rate": 0.0009730375808999339,
      "loss": 2.0629,
      "step": 1335
    },
    {
      "epoch": 0.010688,
      "grad_norm": 5.28125,
      "learning_rate": 0.0009729967759149193,
      "loss": 2.6076,
      "step": 1336
    },
    {
      "epoch": 0.010696,
      "grad_norm": 3.125,
      "learning_rate": 0.0009729559409329743,
      "loss": 2.4443,
      "step": 1337
    },
    {
      "epoch": 0.010704,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009729150759566886,
      "loss": 2.2592,
      "step": 1338
    },
    {
      "epoch": 0.010712,
      "grad_norm": 3.109375,
      "learning_rate": 0.0009728741809886538,
      "loss": 2.538,
      "step": 1339
    },
    {
      "epoch": 0.01072,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009728332560314637,
      "loss": 1.7543,
      "step": 1340
    },
    {
      "epoch": 0.010728,
      "grad_norm": 3.34375,
      "learning_rate": 0.0009727923010877133,
      "loss": 2.6589,
      "step": 1341
    },
    {
      "epoch": 0.010736,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009727513161600001,
      "loss": 1.8743,
      "step": 1342
    },
    {
      "epoch": 0.010744,
      "grad_norm": 3.5625,
      "learning_rate": 0.0009727103012509233,
      "loss": 2.3077,
      "step": 1343
    },
    {
      "epoch": 0.010752,
      "grad_norm": 3.84375,
      "learning_rate": 0.0009726692563630841,
      "loss": 2.082,
      "step": 1344
    },
    {
      "epoch": 0.01076,
      "grad_norm": 3.9375,
      "learning_rate": 0.0009726281814990855,
      "loss": 2.5107,
      "step": 1345
    },
    {
      "epoch": 0.010768,
      "grad_norm": 3.484375,
      "learning_rate": 0.0009725870766615324,
      "loss": 2.7255,
      "step": 1346
    },
    {
      "epoch": 0.010776,
      "grad_norm": 3.5,
      "learning_rate": 0.0009725459418530313,
      "loss": 2.3053,
      "step": 1347
    },
    {
      "epoch": 0.010784,
      "grad_norm": 2.75,
      "learning_rate": 0.0009725047770761916,
      "loss": 2.4367,
      "step": 1348
    },
    {
      "epoch": 0.010792,
      "grad_norm": 3.125,
      "learning_rate": 0.0009724635823336234,
      "loss": 2.4892,
      "step": 1349
    },
    {
      "epoch": 0.0108,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009724223576279394,
      "loss": 2.1743,
      "step": 1350
    },
    {
      "epoch": 0.010808,
      "grad_norm": 3.046875,
      "learning_rate": 0.000972381102961754,
      "loss": 2.345,
      "step": 1351
    },
    {
      "epoch": 0.010816,
      "grad_norm": 4.21875,
      "learning_rate": 0.0009723398183376834,
      "loss": 2.2578,
      "step": 1352
    },
    {
      "epoch": 0.010824,
      "grad_norm": 3.09375,
      "learning_rate": 0.000972298503758346,
      "loss": 2.1018,
      "step": 1353
    },
    {
      "epoch": 0.010832,
      "grad_norm": 3.40625,
      "learning_rate": 0.000972257159226362,
      "loss": 2.6635,
      "step": 1354
    },
    {
      "epoch": 0.01084,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009722157847443534,
      "loss": 2.9847,
      "step": 1355
    },
    {
      "epoch": 0.010848,
      "grad_norm": 3.515625,
      "learning_rate": 0.0009721743803149437,
      "loss": 2.6603,
      "step": 1356
    },
    {
      "epoch": 0.010856,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009721329459407593,
      "loss": 2.2404,
      "step": 1357
    },
    {
      "epoch": 0.010864,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009720914816244275,
      "loss": 2.0596,
      "step": 1358
    },
    {
      "epoch": 0.010872,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009720499873685782,
      "loss": 2.4521,
      "step": 1359
    },
    {
      "epoch": 0.01088,
      "grad_norm": 3.015625,
      "learning_rate": 0.0009720084631758427,
      "loss": 3.1029,
      "step": 1360
    },
    {
      "epoch": 0.010888,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009719669090488547,
      "loss": 2.3492,
      "step": 1361
    },
    {
      "epoch": 0.010896,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009719253249902493,
      "loss": 1.8842,
      "step": 1362
    },
    {
      "epoch": 0.010904,
      "grad_norm": 3.671875,
      "learning_rate": 0.0009718837110026637,
      "loss": 2.8783,
      "step": 1363
    },
    {
      "epoch": 0.010912,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009718420670887371,
      "loss": 2.245,
      "step": 1364
    },
    {
      "epoch": 0.01092,
      "grad_norm": 3.015625,
      "learning_rate": 0.0009718003932511106,
      "loss": 2.969,
      "step": 1365
    },
    {
      "epoch": 0.010928,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009717586894924269,
      "loss": 2.2403,
      "step": 1366
    },
    {
      "epoch": 0.010936,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009717169558153307,
      "loss": 2.7348,
      "step": 1367
    },
    {
      "epoch": 0.010944,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009716751922224691,
      "loss": 2.3191,
      "step": 1368
    },
    {
      "epoch": 0.010952,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009716333987164904,
      "loss": 2.2927,
      "step": 1369
    },
    {
      "epoch": 0.01096,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009715915753000454,
      "loss": 2.2298,
      "step": 1370
    },
    {
      "epoch": 0.010968,
      "grad_norm": 2.375,
      "learning_rate": 0.0009715497219757861,
      "loss": 2.0357,
      "step": 1371
    },
    {
      "epoch": 0.010976,
      "grad_norm": 3.0,
      "learning_rate": 0.0009715078387463672,
      "loss": 2.0499,
      "step": 1372
    },
    {
      "epoch": 0.010984,
      "grad_norm": 3.109375,
      "learning_rate": 0.0009714659256144445,
      "loss": 2.4872,
      "step": 1373
    },
    {
      "epoch": 0.010992,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009714239825826762,
      "loss": 2.6449,
      "step": 1374
    },
    {
      "epoch": 0.011,
      "grad_norm": 3.359375,
      "learning_rate": 0.0009713820096537225,
      "loss": 2.655,
      "step": 1375
    },
    {
      "epoch": 0.011008,
      "grad_norm": 28.125,
      "learning_rate": 0.0009713400068302449,
      "loss": 2.7228,
      "step": 1376
    },
    {
      "epoch": 0.011016,
      "grad_norm": 3.484375,
      "learning_rate": 0.0009712979741149076,
      "loss": 2.1789,
      "step": 1377
    },
    {
      "epoch": 0.011024,
      "grad_norm": 3.890625,
      "learning_rate": 0.0009712559115103759,
      "loss": 2.5858,
      "step": 1378
    },
    {
      "epoch": 0.011032,
      "grad_norm": 3.671875,
      "learning_rate": 0.0009712138190193176,
      "loss": 2.5736,
      "step": 1379
    },
    {
      "epoch": 0.01104,
      "grad_norm": 6.34375,
      "learning_rate": 0.0009711716966444022,
      "loss": 2.1157,
      "step": 1380
    },
    {
      "epoch": 0.011048,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009711295443883008,
      "loss": 2.4231,
      "step": 1381
    },
    {
      "epoch": 0.011056,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009710873622536867,
      "loss": 2.661,
      "step": 1382
    },
    {
      "epoch": 0.011064,
      "grad_norm": 3.3125,
      "learning_rate": 0.0009710451502432354,
      "loss": 2.5302,
      "step": 1383
    },
    {
      "epoch": 0.011072,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009710029083596234,
      "loss": 2.0646,
      "step": 1384
    },
    {
      "epoch": 0.01108,
      "grad_norm": 3.3125,
      "learning_rate": 0.0009709606366055299,
      "loss": 2.8672,
      "step": 1385
    },
    {
      "epoch": 0.011088,
      "grad_norm": 3.359375,
      "learning_rate": 0.0009709183349836359,
      "loss": 2.591,
      "step": 1386
    },
    {
      "epoch": 0.011096,
      "grad_norm": 2.875,
      "learning_rate": 0.0009708760034966239,
      "loss": 2.6294,
      "step": 1387
    },
    {
      "epoch": 0.011104,
      "grad_norm": 3.78125,
      "learning_rate": 0.0009708336421471784,
      "loss": 2.6453,
      "step": 1388
    },
    {
      "epoch": 0.011112,
      "grad_norm": 2.828125,
      "learning_rate": 0.0009707912509379862,
      "loss": 2.0457,
      "step": 1389
    },
    {
      "epoch": 0.01112,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009707488298717356,
      "loss": 2.4065,
      "step": 1390
    },
    {
      "epoch": 0.011128,
      "grad_norm": 5.03125,
      "learning_rate": 0.0009707063789511169,
      "loss": 2.6241,
      "step": 1391
    },
    {
      "epoch": 0.011136,
      "grad_norm": 4.59375,
      "learning_rate": 0.0009706638981788223,
      "loss": 1.7711,
      "step": 1392
    },
    {
      "epoch": 0.011144,
      "grad_norm": 3.578125,
      "learning_rate": 0.0009706213875575457,
      "loss": 2.4802,
      "step": 1393
    },
    {
      "epoch": 0.011152,
      "grad_norm": 3.875,
      "learning_rate": 0.0009705788470899833,
      "loss": 2.4616,
      "step": 1394
    },
    {
      "epoch": 0.01116,
      "grad_norm": 3.40625,
      "learning_rate": 0.0009705362767788329,
      "loss": 2.3647,
      "step": 1395
    },
    {
      "epoch": 0.011168,
      "grad_norm": 4.40625,
      "learning_rate": 0.0009704936766267942,
      "loss": 2.2076,
      "step": 1396
    },
    {
      "epoch": 0.011176,
      "grad_norm": 3.34375,
      "learning_rate": 0.0009704510466365691,
      "loss": 2.6617,
      "step": 1397
    },
    {
      "epoch": 0.011184,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009704083868108607,
      "loss": 2.2743,
      "step": 1398
    },
    {
      "epoch": 0.011192,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009703656971523747,
      "loss": 2.8659,
      "step": 1399
    },
    {
      "epoch": 0.0112,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009703229776638185,
      "loss": 2.3952,
      "step": 1400
    },
    {
      "epoch": 0.011208,
      "grad_norm": 3.125,
      "learning_rate": 0.0009702802283479013,
      "loss": 2.8623,
      "step": 1401
    },
    {
      "epoch": 0.011216,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009702374492073343,
      "loss": 2.4936,
      "step": 1402
    },
    {
      "epoch": 0.011224,
      "grad_norm": 2.828125,
      "learning_rate": 0.0009701946402448301,
      "loss": 2.4413,
      "step": 1403
    },
    {
      "epoch": 0.011232,
      "grad_norm": 3.203125,
      "learning_rate": 0.0009701518014631039,
      "loss": 2.9878,
      "step": 1404
    },
    {
      "epoch": 0.01124,
      "grad_norm": 3.1875,
      "learning_rate": 0.0009701089328648726,
      "loss": 2.225,
      "step": 1405
    },
    {
      "epoch": 0.011248,
      "grad_norm": 3.1875,
      "learning_rate": 0.0009700660344528546,
      "loss": 2.1321,
      "step": 1406
    },
    {
      "epoch": 0.011256,
      "grad_norm": 4.09375,
      "learning_rate": 0.0009700231062297706,
      "loss": 2.5387,
      "step": 1407
    },
    {
      "epoch": 0.011264,
      "grad_norm": 3.109375,
      "learning_rate": 0.0009699801481983432,
      "loss": 2.5182,
      "step": 1408
    },
    {
      "epoch": 0.011272,
      "grad_norm": 3.90625,
      "learning_rate": 0.0009699371603612965,
      "loss": 2.4518,
      "step": 1409
    },
    {
      "epoch": 0.01128,
      "grad_norm": 4.6875,
      "learning_rate": 0.0009698941427213569,
      "loss": 2.6494,
      "step": 1410
    },
    {
      "epoch": 0.011288,
      "grad_norm": 4.1875,
      "learning_rate": 0.0009698510952812525,
      "loss": 2.4309,
      "step": 1411
    },
    {
      "epoch": 0.011296,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009698080180437132,
      "loss": 2.2569,
      "step": 1412
    },
    {
      "epoch": 0.011304,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009697649110114711,
      "loss": 2.5427,
      "step": 1413
    },
    {
      "epoch": 0.011312,
      "grad_norm": 3.484375,
      "learning_rate": 0.00096972177418726,
      "loss": 2.7221,
      "step": 1414
    },
    {
      "epoch": 0.01132,
      "grad_norm": 5.625,
      "learning_rate": 0.0009696786075738153,
      "loss": 2.4023,
      "step": 1415
    },
    {
      "epoch": 0.011328,
      "grad_norm": 3.96875,
      "learning_rate": 0.0009696354111738749,
      "loss": 2.2853,
      "step": 1416
    },
    {
      "epoch": 0.011336,
      "grad_norm": 3.015625,
      "learning_rate": 0.0009695921849901783,
      "loss": 2.3587,
      "step": 1417
    },
    {
      "epoch": 0.011344,
      "grad_norm": 3.234375,
      "learning_rate": 0.0009695489290254665,
      "loss": 2.6046,
      "step": 1418
    },
    {
      "epoch": 0.011352,
      "grad_norm": 4768.0,
      "learning_rate": 0.0009695056432824829,
      "loss": 2.2041,
      "step": 1419
    },
    {
      "epoch": 0.01136,
      "grad_norm": 3.4375,
      "learning_rate": 0.0009694623277639729,
      "loss": 2.5062,
      "step": 1420
    },
    {
      "epoch": 0.011368,
      "grad_norm": 4.75,
      "learning_rate": 0.0009694189824726832,
      "loss": 2.3521,
      "step": 1421
    },
    {
      "epoch": 0.011376,
      "grad_norm": 164.0,
      "learning_rate": 0.0009693756074113628,
      "loss": 2.1865,
      "step": 1422
    },
    {
      "epoch": 0.011384,
      "grad_norm": 5.125,
      "learning_rate": 0.0009693322025827625,
      "loss": 2.3867,
      "step": 1423
    },
    {
      "epoch": 0.011392,
      "grad_norm": 151.0,
      "learning_rate": 0.0009692887679896351,
      "loss": 2.478,
      "step": 1424
    },
    {
      "epoch": 0.0114,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009692453036347351,
      "loss": 2.8059,
      "step": 1425
    },
    {
      "epoch": 0.011408,
      "grad_norm": 3.546875,
      "learning_rate": 0.0009692018095208188,
      "loss": 2.5559,
      "step": 1426
    },
    {
      "epoch": 0.011416,
      "grad_norm": 23.875,
      "learning_rate": 0.0009691582856506447,
      "loss": 2.2487,
      "step": 1427
    },
    {
      "epoch": 0.011424,
      "grad_norm": 3.796875,
      "learning_rate": 0.0009691147320269731,
      "loss": 2.2725,
      "step": 1428
    },
    {
      "epoch": 0.011432,
      "grad_norm": 5.40625,
      "learning_rate": 0.0009690711486525661,
      "loss": 3.0237,
      "step": 1429
    },
    {
      "epoch": 0.01144,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009690275355301874,
      "loss": 2.2087,
      "step": 1430
    },
    {
      "epoch": 0.011448,
      "grad_norm": 6.0625,
      "learning_rate": 0.0009689838926626033,
      "loss": 2.6657,
      "step": 1431
    },
    {
      "epoch": 0.011456,
      "grad_norm": 5.375,
      "learning_rate": 0.0009689402200525816,
      "loss": 2.4717,
      "step": 1432
    },
    {
      "epoch": 0.011464,
      "grad_norm": 4.40625,
      "learning_rate": 0.0009688965177028916,
      "loss": 2.1569,
      "step": 1433
    },
    {
      "epoch": 0.011472,
      "grad_norm": 3.9375,
      "learning_rate": 0.0009688527856163052,
      "loss": 2.0296,
      "step": 1434
    },
    {
      "epoch": 0.01148,
      "grad_norm": 4.4375,
      "learning_rate": 0.0009688090237955955,
      "loss": 3.027,
      "step": 1435
    },
    {
      "epoch": 0.011488,
      "grad_norm": 5.8125,
      "learning_rate": 0.0009687652322435382,
      "loss": 2.37,
      "step": 1436
    },
    {
      "epoch": 0.011496,
      "grad_norm": 4.71875,
      "learning_rate": 0.0009687214109629103,
      "loss": 2.8291,
      "step": 1437
    },
    {
      "epoch": 0.011504,
      "grad_norm": 5.28125,
      "learning_rate": 0.0009686775599564907,
      "loss": 2.6955,
      "step": 1438
    },
    {
      "epoch": 0.011512,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009686336792270609,
      "loss": 2.3192,
      "step": 1439
    },
    {
      "epoch": 0.01152,
      "grad_norm": 4.15625,
      "learning_rate": 0.0009685897687774035,
      "loss": 2.5774,
      "step": 1440
    },
    {
      "epoch": 0.011528,
      "grad_norm": 3.546875,
      "learning_rate": 0.000968545828610303,
      "loss": 2.3518,
      "step": 1441
    },
    {
      "epoch": 0.011536,
      "grad_norm": 4.1875,
      "learning_rate": 0.0009685018587285464,
      "loss": 2.0937,
      "step": 1442
    },
    {
      "epoch": 0.011544,
      "grad_norm": 3.609375,
      "learning_rate": 0.000968457859134922,
      "loss": 2.8268,
      "step": 1443
    },
    {
      "epoch": 0.011552,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009684138298322205,
      "loss": 2.1214,
      "step": 1444
    },
    {
      "epoch": 0.01156,
      "grad_norm": 3.5,
      "learning_rate": 0.0009683697708232338,
      "loss": 2.8602,
      "step": 1445
    },
    {
      "epoch": 0.011568,
      "grad_norm": 4.09375,
      "learning_rate": 0.0009683256821107564,
      "loss": 2.3292,
      "step": 1446
    },
    {
      "epoch": 0.011576,
      "grad_norm": 3.546875,
      "learning_rate": 0.000968281563697584,
      "loss": 3.2053,
      "step": 1447
    },
    {
      "epoch": 0.011584,
      "grad_norm": 3.703125,
      "learning_rate": 0.0009682374155865149,
      "loss": 2.4514,
      "step": 1448
    },
    {
      "epoch": 0.011592,
      "grad_norm": 4.125,
      "learning_rate": 0.0009681932377803488,
      "loss": 3.1021,
      "step": 1449
    },
    {
      "epoch": 0.0116,
      "grad_norm": 4.34375,
      "learning_rate": 0.0009681490302818874,
      "loss": 2.4014,
      "step": 1450
    },
    {
      "epoch": 0.011608,
      "grad_norm": 3.1875,
      "learning_rate": 0.0009681047930939342,
      "loss": 2.3098,
      "step": 1451
    },
    {
      "epoch": 0.011616,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009680605262192948,
      "loss": 2.2594,
      "step": 1452
    },
    {
      "epoch": 0.011624,
      "grad_norm": 5.59375,
      "learning_rate": 0.0009680162296607764,
      "loss": 2.2659,
      "step": 1453
    },
    {
      "epoch": 0.011632,
      "grad_norm": 3.15625,
      "learning_rate": 0.0009679719034211884,
      "loss": 2.1848,
      "step": 1454
    },
    {
      "epoch": 0.01164,
      "grad_norm": 23.125,
      "learning_rate": 0.000967927547503342,
      "loss": 1.9713,
      "step": 1455
    },
    {
      "epoch": 0.011648,
      "grad_norm": 3.8125,
      "learning_rate": 0.0009678831619100499,
      "loss": 2.4353,
      "step": 1456
    },
    {
      "epoch": 0.011656,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009678387466441272,
      "loss": 2.3124,
      "step": 1457
    },
    {
      "epoch": 0.011664,
      "grad_norm": 4.34375,
      "learning_rate": 0.0009677943017083907,
      "loss": 2.3855,
      "step": 1458
    },
    {
      "epoch": 0.011672,
      "grad_norm": 4.25,
      "learning_rate": 0.0009677498271056589,
      "loss": 2.5789,
      "step": 1459
    },
    {
      "epoch": 0.01168,
      "grad_norm": 3.0,
      "learning_rate": 0.0009677053228387525,
      "loss": 2.1612,
      "step": 1460
    },
    {
      "epoch": 0.011688,
      "grad_norm": 5.125,
      "learning_rate": 0.0009676607889104935,
      "loss": 2.7722,
      "step": 1461
    },
    {
      "epoch": 0.011696,
      "grad_norm": 3.375,
      "learning_rate": 0.0009676162253237069,
      "loss": 2.2508,
      "step": 1462
    },
    {
      "epoch": 0.011704,
      "grad_norm": 3.515625,
      "learning_rate": 0.0009675716320812183,
      "loss": 2.2007,
      "step": 1463
    },
    {
      "epoch": 0.011712,
      "grad_norm": 6.75,
      "learning_rate": 0.0009675270091858559,
      "loss": 2.2657,
      "step": 1464
    },
    {
      "epoch": 0.01172,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009674823566404497,
      "loss": 2.4579,
      "step": 1465
    },
    {
      "epoch": 0.011728,
      "grad_norm": 4.8125,
      "learning_rate": 0.0009674376744478315,
      "loss": 2.4689,
      "step": 1466
    },
    {
      "epoch": 0.011736,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009673929626108349,
      "loss": 2.4332,
      "step": 1467
    },
    {
      "epoch": 0.011744,
      "grad_norm": 3.296875,
      "learning_rate": 0.0009673482211322956,
      "loss": 2.7257,
      "step": 1468
    },
    {
      "epoch": 0.011752,
      "grad_norm": 2.8125,
      "learning_rate": 0.000967303450015051,
      "loss": 1.9153,
      "step": 1469
    },
    {
      "epoch": 0.01176,
      "grad_norm": 3.890625,
      "learning_rate": 0.0009672586492619404,
      "loss": 2.6094,
      "step": 1470
    },
    {
      "epoch": 0.011768,
      "grad_norm": 2.859375,
      "learning_rate": 0.0009672138188758051,
      "loss": 2.0719,
      "step": 1471
    },
    {
      "epoch": 0.011776,
      "grad_norm": 4.5,
      "learning_rate": 0.0009671689588594882,
      "loss": 2.6258,
      "step": 1472
    },
    {
      "epoch": 0.011784,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009671240692158345,
      "loss": 1.9683,
      "step": 1473
    },
    {
      "epoch": 0.011792,
      "grad_norm": 3.90625,
      "learning_rate": 0.0009670791499476912,
      "loss": 2.8312,
      "step": 1474
    },
    {
      "epoch": 0.0118,
      "grad_norm": 3.484375,
      "learning_rate": 0.0009670342010579065,
      "loss": 2.5007,
      "step": 1475
    },
    {
      "epoch": 0.011808,
      "grad_norm": 3.671875,
      "learning_rate": 0.0009669892225493315,
      "loss": 2.4419,
      "step": 1476
    },
    {
      "epoch": 0.011816,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009669442144248184,
      "loss": 2.0885,
      "step": 1477
    },
    {
      "epoch": 0.011824,
      "grad_norm": 3.484375,
      "learning_rate": 0.0009668991766872217,
      "loss": 2.617,
      "step": 1478
    },
    {
      "epoch": 0.011832,
      "grad_norm": 3.875,
      "learning_rate": 0.0009668541093393976,
      "loss": 2.4154,
      "step": 1479
    },
    {
      "epoch": 0.01184,
      "grad_norm": 3.40625,
      "learning_rate": 0.0009668090123842043,
      "loss": 2.4021,
      "step": 1480
    },
    {
      "epoch": 0.011848,
      "grad_norm": 1.875,
      "learning_rate": 0.0009667638858245016,
      "loss": 1.3015,
      "step": 1481
    },
    {
      "epoch": 0.011856,
      "grad_norm": 2.859375,
      "learning_rate": 0.0009667187296631517,
      "loss": 2.5513,
      "step": 1482
    },
    {
      "epoch": 0.011864,
      "grad_norm": 4.625,
      "learning_rate": 0.0009666735439030178,
      "loss": 2.1646,
      "step": 1483
    },
    {
      "epoch": 0.011872,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009666283285469663,
      "loss": 2.4925,
      "step": 1484
    },
    {
      "epoch": 0.01188,
      "grad_norm": 2.875,
      "learning_rate": 0.0009665830835978641,
      "loss": 1.9096,
      "step": 1485
    },
    {
      "epoch": 0.011888,
      "grad_norm": 3.421875,
      "learning_rate": 0.0009665378090585807,
      "loss": 2.7434,
      "step": 1486
    },
    {
      "epoch": 0.011896,
      "grad_norm": 3.65625,
      "learning_rate": 0.0009664925049319875,
      "loss": 2.9048,
      "step": 1487
    },
    {
      "epoch": 0.011904,
      "grad_norm": 4.78125,
      "learning_rate": 0.0009664471712209576,
      "loss": 2.5093,
      "step": 1488
    },
    {
      "epoch": 0.011912,
      "grad_norm": 3.84375,
      "learning_rate": 0.0009664018079283661,
      "loss": 3.0774,
      "step": 1489
    },
    {
      "epoch": 0.01192,
      "grad_norm": 3.3125,
      "learning_rate": 0.0009663564150570896,
      "loss": 2.4026,
      "step": 1490
    },
    {
      "epoch": 0.011928,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009663109926100071,
      "loss": 2.2442,
      "step": 1491
    },
    {
      "epoch": 0.011936,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009662655405899992,
      "loss": 2.4311,
      "step": 1492
    },
    {
      "epoch": 0.011944,
      "grad_norm": 3.796875,
      "learning_rate": 0.0009662200589999483,
      "loss": 2.6784,
      "step": 1493
    },
    {
      "epoch": 0.011952,
      "grad_norm": 3.625,
      "learning_rate": 0.000966174547842739,
      "loss": 2.944,
      "step": 1494
    },
    {
      "epoch": 0.01196,
      "grad_norm": 4.375,
      "learning_rate": 0.0009661290071212574,
      "loss": 2.32,
      "step": 1495
    },
    {
      "epoch": 0.011968,
      "grad_norm": 3.234375,
      "learning_rate": 0.0009660834368383918,
      "loss": 2.2396,
      "step": 1496
    },
    {
      "epoch": 0.011976,
      "grad_norm": 4.34375,
      "learning_rate": 0.0009660378369970321,
      "loss": 1.9949,
      "step": 1497
    },
    {
      "epoch": 0.011984,
      "grad_norm": 3.15625,
      "learning_rate": 0.0009659922076000702,
      "loss": 2.2384,
      "step": 1498
    },
    {
      "epoch": 0.011992,
      "grad_norm": 3.34375,
      "learning_rate": 0.0009659465486503998,
      "loss": 2.3727,
      "step": 1499
    },
    {
      "epoch": 0.012,
      "grad_norm": 3.390625,
      "learning_rate": 0.0009659008601509168,
      "loss": 2.2633,
      "step": 1500
    },
    {
      "epoch": 0.012008,
      "grad_norm": 3.125,
      "learning_rate": 0.0009658551421045183,
      "loss": 1.8743,
      "step": 1501
    },
    {
      "epoch": 0.012016,
      "grad_norm": 4.46875,
      "learning_rate": 0.000965809394514104,
      "loss": 2.533,
      "step": 1502
    },
    {
      "epoch": 0.012024,
      "grad_norm": 3.796875,
      "learning_rate": 0.0009657636173825752,
      "loss": 2.6252,
      "step": 1503
    },
    {
      "epoch": 0.012032,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009657178107128348,
      "loss": 2.2181,
      "step": 1504
    },
    {
      "epoch": 0.01204,
      "grad_norm": 3.796875,
      "learning_rate": 0.000965671974507788,
      "loss": 2.5224,
      "step": 1505
    },
    {
      "epoch": 0.012048,
      "grad_norm": 3.15625,
      "learning_rate": 0.0009656261087703415,
      "loss": 2.2692,
      "step": 1506
    },
    {
      "epoch": 0.012056,
      "grad_norm": 3.859375,
      "learning_rate": 0.0009655802135034041,
      "loss": 2.9406,
      "step": 1507
    },
    {
      "epoch": 0.012064,
      "grad_norm": 4.78125,
      "learning_rate": 0.0009655342887098867,
      "loss": 2.3665,
      "step": 1508
    },
    {
      "epoch": 0.012072,
      "grad_norm": 2.609375,
      "learning_rate": 0.0009654883343927015,
      "loss": 2.428,
      "step": 1509
    },
    {
      "epoch": 0.01208,
      "grad_norm": 3.40625,
      "learning_rate": 0.0009654423505547629,
      "loss": 2.3257,
      "step": 1510
    },
    {
      "epoch": 0.012088,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009653963371989874,
      "loss": 1.8533,
      "step": 1511
    },
    {
      "epoch": 0.012096,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009653502943282926,
      "loss": 2.4212,
      "step": 1512
    },
    {
      "epoch": 0.012104,
      "grad_norm": 3.203125,
      "learning_rate": 0.000965304221945599,
      "loss": 2.5108,
      "step": 1513
    },
    {
      "epoch": 0.012112,
      "grad_norm": 2.375,
      "learning_rate": 0.0009652581200538282,
      "loss": 2.1769,
      "step": 1514
    },
    {
      "epoch": 0.01212,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009652119886559041,
      "loss": 2.2921,
      "step": 1515
    },
    {
      "epoch": 0.012128,
      "grad_norm": 4.09375,
      "learning_rate": 0.000965165827754752,
      "loss": 2.4479,
      "step": 1516
    },
    {
      "epoch": 0.012136,
      "grad_norm": 3.125,
      "learning_rate": 0.0009651196373532998,
      "loss": 2.5655,
      "step": 1517
    },
    {
      "epoch": 0.012144,
      "grad_norm": 4.65625,
      "learning_rate": 0.0009650734174544764,
      "loss": 2.9544,
      "step": 1518
    },
    {
      "epoch": 0.012152,
      "grad_norm": 16.25,
      "learning_rate": 0.0009650271680612134,
      "loss": 2.0145,
      "step": 1519
    },
    {
      "epoch": 0.01216,
      "grad_norm": 8.1875,
      "learning_rate": 0.0009649808891764436,
      "loss": 1.4618,
      "step": 1520
    },
    {
      "epoch": 0.012168,
      "grad_norm": 3.0,
      "learning_rate": 0.000964934580803102,
      "loss": 2.1696,
      "step": 1521
    },
    {
      "epoch": 0.012176,
      "grad_norm": 3.765625,
      "learning_rate": 0.0009648882429441257,
      "loss": 2.308,
      "step": 1522
    },
    {
      "epoch": 0.012184,
      "grad_norm": 4.15625,
      "learning_rate": 0.0009648418756024531,
      "loss": 2.3938,
      "step": 1523
    },
    {
      "epoch": 0.012192,
      "grad_norm": 3.5,
      "learning_rate": 0.0009647954787810249,
      "loss": 1.7808,
      "step": 1524
    },
    {
      "epoch": 0.0122,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009647490524827833,
      "loss": 2.0332,
      "step": 1525
    },
    {
      "epoch": 0.012208,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009647025967106731,
      "loss": 2.0282,
      "step": 1526
    },
    {
      "epoch": 0.012216,
      "grad_norm": 4.90625,
      "learning_rate": 0.00096465611146764,
      "loss": 2.1542,
      "step": 1527
    },
    {
      "epoch": 0.012224,
      "grad_norm": 3.921875,
      "learning_rate": 0.0009646095967566321,
      "loss": 2.4009,
      "step": 1528
    },
    {
      "epoch": 0.012232,
      "grad_norm": 3.734375,
      "learning_rate": 0.0009645630525805996,
      "loss": 2.2622,
      "step": 1529
    },
    {
      "epoch": 0.01224,
      "grad_norm": 8.625,
      "learning_rate": 0.0009645164789424941,
      "loss": 2.4965,
      "step": 1530
    },
    {
      "epoch": 0.012248,
      "grad_norm": 4.84375,
      "learning_rate": 0.0009644698758452692,
      "loss": 2.6203,
      "step": 1531
    },
    {
      "epoch": 0.012256,
      "grad_norm": 3.46875,
      "learning_rate": 0.0009644232432918805,
      "loss": 2.4983,
      "step": 1532
    },
    {
      "epoch": 0.012264,
      "grad_norm": 4.71875,
      "learning_rate": 0.0009643765812852852,
      "loss": 2.5532,
      "step": 1533
    },
    {
      "epoch": 0.012272,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009643298898284428,
      "loss": 2.1031,
      "step": 1534
    },
    {
      "epoch": 0.01228,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009642831689243143,
      "loss": 2.4462,
      "step": 1535
    },
    {
      "epoch": 0.012288,
      "grad_norm": 2.875,
      "learning_rate": 0.0009642364185758627,
      "loss": 1.8899,
      "step": 1536
    },
    {
      "epoch": 0.012296,
      "grad_norm": 3.15625,
      "learning_rate": 0.0009641896387860531,
      "loss": 2.1491,
      "step": 1537
    },
    {
      "epoch": 0.012304,
      "grad_norm": 3.515625,
      "learning_rate": 0.0009641428295578519,
      "loss": 2.3769,
      "step": 1538
    },
    {
      "epoch": 0.012312,
      "grad_norm": 4.34375,
      "learning_rate": 0.0009640959908942276,
      "loss": 2.6275,
      "step": 1539
    },
    {
      "epoch": 0.01232,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009640491227981509,
      "loss": 1.9994,
      "step": 1540
    },
    {
      "epoch": 0.012328,
      "grad_norm": 2.546875,
      "learning_rate": 0.0009640022252725939,
      "loss": 2.1096,
      "step": 1541
    },
    {
      "epoch": 0.012336,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009639552983205313,
      "loss": 2.6443,
      "step": 1542
    },
    {
      "epoch": 0.012344,
      "grad_norm": 3.296875,
      "learning_rate": 0.0009639083419449386,
      "loss": 2.3,
      "step": 1543
    },
    {
      "epoch": 0.012352,
      "grad_norm": 3.6875,
      "learning_rate": 0.0009638613561487939,
      "loss": 2.9142,
      "step": 1544
    },
    {
      "epoch": 0.01236,
      "grad_norm": 3.359375,
      "learning_rate": 0.0009638143409350772,
      "loss": 2.4846,
      "step": 1545
    },
    {
      "epoch": 0.012368,
      "grad_norm": 2.859375,
      "learning_rate": 0.0009637672963067697,
      "loss": 2.0504,
      "step": 1546
    },
    {
      "epoch": 0.012376,
      "grad_norm": 2.421875,
      "learning_rate": 0.0009637202222668554,
      "loss": 2.1635,
      "step": 1547
    },
    {
      "epoch": 0.012384,
      "grad_norm": 3.0,
      "learning_rate": 0.0009636731188183194,
      "loss": 2.4472,
      "step": 1548
    },
    {
      "epoch": 0.012392,
      "grad_norm": 3.609375,
      "learning_rate": 0.0009636259859641489,
      "loss": 2.6912,
      "step": 1549
    },
    {
      "epoch": 0.0124,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009635788237073333,
      "loss": 2.5965,
      "step": 1550
    },
    {
      "epoch": 0.012408,
      "grad_norm": 6.0625,
      "learning_rate": 0.0009635316320508634,
      "loss": 1.9129,
      "step": 1551
    },
    {
      "epoch": 0.012416,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009634844109977319,
      "loss": 2.3093,
      "step": 1552
    },
    {
      "epoch": 0.012424,
      "grad_norm": 3.0,
      "learning_rate": 0.0009634371605509337,
      "loss": 2.238,
      "step": 1553
    },
    {
      "epoch": 0.012432,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009633898807134653,
      "loss": 2.0332,
      "step": 1554
    },
    {
      "epoch": 0.01244,
      "grad_norm": 3.765625,
      "learning_rate": 0.0009633425714883253,
      "loss": 2.4543,
      "step": 1555
    },
    {
      "epoch": 0.012448,
      "grad_norm": 3.140625,
      "learning_rate": 0.0009632952328785136,
      "loss": 2.4122,
      "step": 1556
    },
    {
      "epoch": 0.012456,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009632478648870328,
      "loss": 2.0024,
      "step": 1557
    },
    {
      "epoch": 0.012464,
      "grad_norm": 4.40625,
      "learning_rate": 0.0009632004675168869,
      "loss": 2.5701,
      "step": 1558
    },
    {
      "epoch": 0.012472,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009631530407710813,
      "loss": 2.5814,
      "step": 1559
    },
    {
      "epoch": 0.01248,
      "grad_norm": 3.875,
      "learning_rate": 0.0009631055846526243,
      "loss": 2.7172,
      "step": 1560
    },
    {
      "epoch": 0.012488,
      "grad_norm": 3.0,
      "learning_rate": 0.0009630580991645253,
      "loss": 2.5776,
      "step": 1561
    },
    {
      "epoch": 0.012496,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009630105843097957,
      "loss": 2.6966,
      "step": 1562
    },
    {
      "epoch": 0.012504,
      "grad_norm": 3.421875,
      "learning_rate": 0.0009629630400914489,
      "loss": 2.8895,
      "step": 1563
    },
    {
      "epoch": 0.012512,
      "grad_norm": 3.5,
      "learning_rate": 0.0009629154665125001,
      "loss": 2.3854,
      "step": 1564
    },
    {
      "epoch": 0.01252,
      "grad_norm": 60.75,
      "learning_rate": 0.0009628678635759664,
      "loss": 2.6387,
      "step": 1565
    },
    {
      "epoch": 0.012528,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009628202312848669,
      "loss": 2.0645,
      "step": 1566
    },
    {
      "epoch": 0.012536,
      "grad_norm": 3.03125,
      "learning_rate": 0.000962772569642222,
      "loss": 2.5091,
      "step": 1567
    },
    {
      "epoch": 0.012544,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009627248786510546,
      "loss": 2.292,
      "step": 1568
    },
    {
      "epoch": 0.012552,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009626771583143892,
      "loss": 2.6159,
      "step": 1569
    },
    {
      "epoch": 0.01256,
      "grad_norm": 3.359375,
      "learning_rate": 0.0009626294086352519,
      "loss": 2.0226,
      "step": 1570
    },
    {
      "epoch": 0.012568,
      "grad_norm": 3.46875,
      "learning_rate": 0.0009625816296166714,
      "loss": 2.6413,
      "step": 1571
    },
    {
      "epoch": 0.012576,
      "grad_norm": 3.765625,
      "learning_rate": 0.0009625338212616775,
      "loss": 2.5841,
      "step": 1572
    },
    {
      "epoch": 0.012584,
      "grad_norm": 3.890625,
      "learning_rate": 0.0009624859835733021,
      "loss": 2.3383,
      "step": 1573
    },
    {
      "epoch": 0.012592,
      "grad_norm": 3.890625,
      "learning_rate": 0.0009624381165545793,
      "loss": 1.9732,
      "step": 1574
    },
    {
      "epoch": 0.0126,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009623902202085444,
      "loss": 2.3935,
      "step": 1575
    },
    {
      "epoch": 0.012608,
      "grad_norm": 3.953125,
      "learning_rate": 0.0009623422945382351,
      "loss": 2.6001,
      "step": 1576
    },
    {
      "epoch": 0.012616,
      "grad_norm": 3.25,
      "learning_rate": 0.0009622943395466908,
      "loss": 2.4149,
      "step": 1577
    },
    {
      "epoch": 0.012624,
      "grad_norm": 3.84375,
      "learning_rate": 0.000962246355236953,
      "loss": 2.4343,
      "step": 1578
    },
    {
      "epoch": 0.012632,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009621983416120643,
      "loss": 2.5685,
      "step": 1579
    },
    {
      "epoch": 0.01264,
      "grad_norm": 4.28125,
      "learning_rate": 0.0009621502986750701,
      "loss": 2.8761,
      "step": 1580
    },
    {
      "epoch": 0.012648,
      "grad_norm": 3.84375,
      "learning_rate": 0.0009621022264290171,
      "loss": 2.389,
      "step": 1581
    },
    {
      "epoch": 0.012656,
      "grad_norm": 4.0,
      "learning_rate": 0.0009620541248769539,
      "loss": 2.2099,
      "step": 1582
    },
    {
      "epoch": 0.012664,
      "grad_norm": 125.0,
      "learning_rate": 0.0009620059940219311,
      "loss": 2.7344,
      "step": 1583
    },
    {
      "epoch": 0.012672,
      "grad_norm": 66.0,
      "learning_rate": 0.0009619578338670012,
      "loss": 2.1077,
      "step": 1584
    },
    {
      "epoch": 0.01268,
      "grad_norm": 3.515625,
      "learning_rate": 0.0009619096444152182,
      "loss": 2.6258,
      "step": 1585
    },
    {
      "epoch": 0.012688,
      "grad_norm": 3.8125,
      "learning_rate": 0.0009618614256696385,
      "loss": 2.7372,
      "step": 1586
    },
    {
      "epoch": 0.012696,
      "grad_norm": 3.984375,
      "learning_rate": 0.0009618131776333201,
      "loss": 2.7063,
      "step": 1587
    },
    {
      "epoch": 0.012704,
      "grad_norm": 3.90625,
      "learning_rate": 0.0009617649003093226,
      "loss": 2.5683,
      "step": 1588
    },
    {
      "epoch": 0.012712,
      "grad_norm": 3.65625,
      "learning_rate": 0.0009617165937007079,
      "loss": 2.7038,
      "step": 1589
    },
    {
      "epoch": 0.01272,
      "grad_norm": 3.75,
      "learning_rate": 0.0009616682578105394,
      "loss": 2.0722,
      "step": 1590
    },
    {
      "epoch": 0.012728,
      "grad_norm": 3.6875,
      "learning_rate": 0.0009616198926418824,
      "loss": 2.3366,
      "step": 1591
    },
    {
      "epoch": 0.012736,
      "grad_norm": 4.4375,
      "learning_rate": 0.0009615714981978046,
      "loss": 2.9602,
      "step": 1592
    },
    {
      "epoch": 0.012744,
      "grad_norm": 7.40625,
      "learning_rate": 0.0009615230744813748,
      "loss": 2.5518,
      "step": 1593
    },
    {
      "epoch": 0.012752,
      "grad_norm": 11.1875,
      "learning_rate": 0.000961474621495664,
      "loss": 2.9394,
      "step": 1594
    },
    {
      "epoch": 0.01276,
      "grad_norm": 28.375,
      "learning_rate": 0.0009614261392437451,
      "loss": 2.4314,
      "step": 1595
    },
    {
      "epoch": 0.012768,
      "grad_norm": 9.3125,
      "learning_rate": 0.0009613776277286927,
      "loss": 2.5537,
      "step": 1596
    },
    {
      "epoch": 0.012776,
      "grad_norm": 15.75,
      "learning_rate": 0.0009613290869535834,
      "loss": 2.5965,
      "step": 1597
    },
    {
      "epoch": 0.012784,
      "grad_norm": 157.0,
      "learning_rate": 0.0009612805169214957,
      "loss": 2.5689,
      "step": 1598
    },
    {
      "epoch": 0.012792,
      "grad_norm": 66.0,
      "learning_rate": 0.0009612319176355098,
      "loss": 2.8868,
      "step": 1599
    },
    {
      "epoch": 0.0128,
      "grad_norm": 17536.0,
      "learning_rate": 0.0009611832890987075,
      "loss": 2.3734,
      "step": 1600
    },
    {
      "epoch": 0.012808,
      "grad_norm": 1000.0,
      "learning_rate": 0.0009611346313141733,
      "loss": 3.016,
      "step": 1601
    },
    {
      "epoch": 0.012816,
      "grad_norm": 6272.0,
      "learning_rate": 0.0009610859442849928,
      "loss": 3.0115,
      "step": 1602
    },
    {
      "epoch": 0.012824,
      "grad_norm": 300.0,
      "learning_rate": 0.0009610372280142536,
      "loss": 2.7349,
      "step": 1603
    },
    {
      "epoch": 0.012832,
      "grad_norm": 12.0,
      "learning_rate": 0.0009609884825050452,
      "loss": 3.1794,
      "step": 1604
    },
    {
      "epoch": 0.01284,
      "grad_norm": 7.1875,
      "learning_rate": 0.0009609397077604591,
      "loss": 3.1212,
      "step": 1605
    },
    {
      "epoch": 0.012848,
      "grad_norm": 4.5,
      "learning_rate": 0.0009608909037835887,
      "loss": 2.4597,
      "step": 1606
    },
    {
      "epoch": 0.012856,
      "grad_norm": 32.75,
      "learning_rate": 0.0009608420705775287,
      "loss": 2.4046,
      "step": 1607
    },
    {
      "epoch": 0.012864,
      "grad_norm": 6.65625,
      "learning_rate": 0.0009607932081453763,
      "loss": 2.415,
      "step": 1608
    },
    {
      "epoch": 0.012872,
      "grad_norm": 5.875,
      "learning_rate": 0.0009607443164902302,
      "loss": 2.8109,
      "step": 1609
    },
    {
      "epoch": 0.01288,
      "grad_norm": 19.375,
      "learning_rate": 0.0009606953956151911,
      "loss": 2.9357,
      "step": 1610
    },
    {
      "epoch": 0.012888,
      "grad_norm": 3.84375,
      "learning_rate": 0.0009606464455233617,
      "loss": 2.1549,
      "step": 1611
    },
    {
      "epoch": 0.012896,
      "grad_norm": 4.0625,
      "learning_rate": 0.0009605974662178459,
      "loss": 2.8744,
      "step": 1612
    },
    {
      "epoch": 0.012904,
      "grad_norm": 4.5625,
      "learning_rate": 0.0009605484577017504,
      "loss": 2.746,
      "step": 1613
    },
    {
      "epoch": 0.012912,
      "grad_norm": 5.4375,
      "learning_rate": 0.000960499419978183,
      "loss": 2.4468,
      "step": 1614
    },
    {
      "epoch": 0.01292,
      "grad_norm": 6.40625,
      "learning_rate": 0.0009604503530502535,
      "loss": 2.7514,
      "step": 1615
    },
    {
      "epoch": 0.012928,
      "grad_norm": 3.84375,
      "learning_rate": 0.0009604012569210741,
      "loss": 2.6343,
      "step": 1616
    },
    {
      "epoch": 0.012936,
      "grad_norm": 3.953125,
      "learning_rate": 0.000960352131593758,
      "loss": 2.6126,
      "step": 1617
    },
    {
      "epoch": 0.012944,
      "grad_norm": 4.65625,
      "learning_rate": 0.0009603029770714209,
      "loss": 2.3892,
      "step": 1618
    },
    {
      "epoch": 0.012952,
      "grad_norm": 3.5,
      "learning_rate": 0.00096025379335718,
      "loss": 2.4088,
      "step": 1619
    },
    {
      "epoch": 0.01296,
      "grad_norm": 4.09375,
      "learning_rate": 0.0009602045804541546,
      "loss": 2.6671,
      "step": 1620
    },
    {
      "epoch": 0.012968,
      "grad_norm": 5.40625,
      "learning_rate": 0.0009601553383654655,
      "loss": 2.5954,
      "step": 1621
    },
    {
      "epoch": 0.012976,
      "grad_norm": 6.5625,
      "learning_rate": 0.0009601060670942359,
      "loss": 2.7659,
      "step": 1622
    },
    {
      "epoch": 0.012984,
      "grad_norm": 180.0,
      "learning_rate": 0.0009600567666435904,
      "loss": 2.3535,
      "step": 1623
    },
    {
      "epoch": 0.012992,
      "grad_norm": 2640.0,
      "learning_rate": 0.0009600074370166556,
      "loss": 2.2217,
      "step": 1624
    },
    {
      "epoch": 0.013,
      "grad_norm": 91.0,
      "learning_rate": 0.0009599580782165598,
      "loss": 2.0964,
      "step": 1625
    },
    {
      "epoch": 0.013008,
      "grad_norm": 82.5,
      "learning_rate": 0.0009599086902464333,
      "loss": 2.2146,
      "step": 1626
    },
    {
      "epoch": 0.013016,
      "grad_norm": 9216.0,
      "learning_rate": 0.0009598592731094083,
      "loss": 2.2224,
      "step": 1627
    },
    {
      "epoch": 0.013024,
      "grad_norm": 33280.0,
      "learning_rate": 0.0009598098268086188,
      "loss": 2.3503,
      "step": 1628
    },
    {
      "epoch": 0.013032,
      "grad_norm": 117248.0,
      "learning_rate": 0.0009597603513472006,
      "loss": 2.1838,
      "step": 1629
    },
    {
      "epoch": 0.01304,
      "grad_norm": 11072.0,
      "learning_rate": 0.0009597108467282916,
      "loss": 2.2981,
      "step": 1630
    },
    {
      "epoch": 0.013048,
      "grad_norm": 1296.0,
      "learning_rate": 0.0009596613129550309,
      "loss": 2.8694,
      "step": 1631
    },
    {
      "epoch": 0.013056,
      "grad_norm": 8576.0,
      "learning_rate": 0.00095961175003056,
      "loss": 2.5409,
      "step": 1632
    },
    {
      "epoch": 0.013064,
      "grad_norm": 8.625,
      "learning_rate": 0.0009595621579580223,
      "loss": 2.8079,
      "step": 1633
    },
    {
      "epoch": 0.013072,
      "grad_norm": 21504.0,
      "learning_rate": 0.0009595125367405628,
      "loss": 1.9427,
      "step": 1634
    },
    {
      "epoch": 0.01308,
      "grad_norm": 335872.0,
      "learning_rate": 0.0009594628863813284,
      "loss": 2.1921,
      "step": 1635
    },
    {
      "epoch": 0.013088,
      "grad_norm": 47104.0,
      "learning_rate": 0.000959413206883468,
      "loss": 2.8641,
      "step": 1636
    },
    {
      "epoch": 0.013096,
      "grad_norm": 48896.0,
      "learning_rate": 0.000959363498250132,
      "loss": 2.9558,
      "step": 1637
    },
    {
      "epoch": 0.013104,
      "grad_norm": 5376.0,
      "learning_rate": 0.0009593137604844729,
      "loss": 3.3235,
      "step": 1638
    },
    {
      "epoch": 0.013112,
      "grad_norm": 129024.0,
      "learning_rate": 0.0009592639935896452,
      "loss": 3.1197,
      "step": 1639
    },
    {
      "epoch": 0.01312,
      "grad_norm": 25088.0,
      "learning_rate": 0.0009592141975688047,
      "loss": 3.3212,
      "step": 1640
    },
    {
      "epoch": 0.013128,
      "grad_norm": 30464.0,
      "learning_rate": 0.0009591643724251099,
      "loss": 3.2661,
      "step": 1641
    },
    {
      "epoch": 0.013136,
      "grad_norm": 14528.0,
      "learning_rate": 0.0009591145181617203,
      "loss": 3.2571,
      "step": 1642
    },
    {
      "epoch": 0.013144,
      "grad_norm": 5728.0,
      "learning_rate": 0.0009590646347817977,
      "loss": 4.1171,
      "step": 1643
    },
    {
      "epoch": 0.013152,
      "grad_norm": 137.0,
      "learning_rate": 0.0009590147222885057,
      "loss": 4.5435,
      "step": 1644
    },
    {
      "epoch": 0.01316,
      "grad_norm": 748.0,
      "learning_rate": 0.0009589647806850097,
      "loss": 4.9383,
      "step": 1645
    },
    {
      "epoch": 0.013168,
      "grad_norm": 175.0,
      "learning_rate": 0.0009589148099744769,
      "loss": 5.5403,
      "step": 1646
    },
    {
      "epoch": 0.013176,
      "grad_norm": 132.0,
      "learning_rate": 0.0009588648101600762,
      "loss": 4.5043,
      "step": 1647
    },
    {
      "epoch": 0.013184,
      "grad_norm": 25.125,
      "learning_rate": 0.0009588147812449788,
      "loss": 4.6111,
      "step": 1648
    },
    {
      "epoch": 0.013192,
      "grad_norm": 9.375,
      "learning_rate": 0.0009587647232323574,
      "loss": 4.0855,
      "step": 1649
    },
    {
      "epoch": 0.0132,
      "grad_norm": 9.0625,
      "learning_rate": 0.0009587146361253867,
      "loss": 3.891,
      "step": 1650
    },
    {
      "epoch": 0.013208,
      "grad_norm": 6.90625,
      "learning_rate": 0.000958664519927243,
      "loss": 3.5506,
      "step": 1651
    },
    {
      "epoch": 0.013216,
      "grad_norm": 6.9375,
      "learning_rate": 0.0009586143746411047,
      "loss": 3.6185,
      "step": 1652
    },
    {
      "epoch": 0.013224,
      "grad_norm": 10.125,
      "learning_rate": 0.0009585642002701518,
      "loss": 3.4196,
      "step": 1653
    },
    {
      "epoch": 0.013232,
      "grad_norm": 7.84375,
      "learning_rate": 0.0009585139968175667,
      "loss": 3.5538,
      "step": 1654
    },
    {
      "epoch": 0.01324,
      "grad_norm": 5.125,
      "learning_rate": 0.0009584637642865328,
      "loss": 3.2,
      "step": 1655
    },
    {
      "epoch": 0.013248,
      "grad_norm": 5.46875,
      "learning_rate": 0.0009584135026802361,
      "loss": 2.8272,
      "step": 1656
    },
    {
      "epoch": 0.013256,
      "grad_norm": 5.59375,
      "learning_rate": 0.000958363212001864,
      "loss": 3.1665,
      "step": 1657
    },
    {
      "epoch": 0.013264,
      "grad_norm": 5.21875,
      "learning_rate": 0.0009583128922546059,
      "loss": 2.9226,
      "step": 1658
    },
    {
      "epoch": 0.013272,
      "grad_norm": 4.375,
      "learning_rate": 0.0009582625434416532,
      "loss": 3.2731,
      "step": 1659
    },
    {
      "epoch": 0.01328,
      "grad_norm": 3.921875,
      "learning_rate": 0.0009582121655661986,
      "loss": 3.0717,
      "step": 1660
    },
    {
      "epoch": 0.013288,
      "grad_norm": 3.953125,
      "learning_rate": 0.0009581617586314373,
      "loss": 2.8663,
      "step": 1661
    },
    {
      "epoch": 0.013296,
      "grad_norm": 4.09375,
      "learning_rate": 0.0009581113226405658,
      "loss": 3.3486,
      "step": 1662
    },
    {
      "epoch": 0.013304,
      "grad_norm": 3.953125,
      "learning_rate": 0.000958060857596783,
      "loss": 2.6151,
      "step": 1663
    },
    {
      "epoch": 0.013312,
      "grad_norm": 5.3125,
      "learning_rate": 0.0009580103635032892,
      "loss": 2.2261,
      "step": 1664
    },
    {
      "epoch": 0.01332,
      "grad_norm": 9.25,
      "learning_rate": 0.0009579598403632865,
      "loss": 3.035,
      "step": 1665
    },
    {
      "epoch": 0.013328,
      "grad_norm": 4.5625,
      "learning_rate": 0.0009579092881799794,
      "loss": 3.2414,
      "step": 1666
    },
    {
      "epoch": 0.013336,
      "grad_norm": 3.515625,
      "learning_rate": 0.0009578587069565736,
      "loss": 3.5588,
      "step": 1667
    },
    {
      "epoch": 0.013344,
      "grad_norm": 4.21875,
      "learning_rate": 0.0009578080966962768,
      "loss": 2.8918,
      "step": 1668
    },
    {
      "epoch": 0.013352,
      "grad_norm": 31.25,
      "learning_rate": 0.000957757457402299,
      "loss": 2.4318,
      "step": 1669
    },
    {
      "epoch": 0.01336,
      "grad_norm": 14.5,
      "learning_rate": 0.0009577067890778515,
      "loss": 3.0043,
      "step": 1670
    },
    {
      "epoch": 0.013368,
      "grad_norm": 5.25,
      "learning_rate": 0.0009576560917261474,
      "loss": 2.7823,
      "step": 1671
    },
    {
      "epoch": 0.013376,
      "grad_norm": 7.34375,
      "learning_rate": 0.0009576053653504022,
      "loss": 3.8058,
      "step": 1672
    },
    {
      "epoch": 0.013384,
      "grad_norm": 4.84375,
      "learning_rate": 0.0009575546099538329,
      "loss": 2.8067,
      "step": 1673
    },
    {
      "epoch": 0.013392,
      "grad_norm": 4.78125,
      "learning_rate": 0.0009575038255396582,
      "loss": 3.2137,
      "step": 1674
    },
    {
      "epoch": 0.0134,
      "grad_norm": 5.25,
      "learning_rate": 0.0009574530121110989,
      "loss": 3.2006,
      "step": 1675
    },
    {
      "epoch": 0.013408,
      "grad_norm": 5.25,
      "learning_rate": 0.0009574021696713773,
      "loss": 2.7244,
      "step": 1676
    },
    {
      "epoch": 0.013416,
      "grad_norm": 6.96875,
      "learning_rate": 0.000957351298223718,
      "loss": 3.1275,
      "step": 1677
    },
    {
      "epoch": 0.013424,
      "grad_norm": 12.5,
      "learning_rate": 0.0009573003977713472,
      "loss": 3.3999,
      "step": 1678
    },
    {
      "epoch": 0.013432,
      "grad_norm": 8.875,
      "learning_rate": 0.000957249468317493,
      "loss": 2.846,
      "step": 1679
    },
    {
      "epoch": 0.01344,
      "grad_norm": 6.40625,
      "learning_rate": 0.0009571985098653851,
      "loss": 2.546,
      "step": 1680
    },
    {
      "epoch": 0.013448,
      "grad_norm": 6.09375,
      "learning_rate": 0.0009571475224182553,
      "loss": 2.8758,
      "step": 1681
    },
    {
      "epoch": 0.013456,
      "grad_norm": 374.0,
      "learning_rate": 0.0009570965059793373,
      "loss": 2.8854,
      "step": 1682
    },
    {
      "epoch": 0.013464,
      "grad_norm": 3.9375,
      "learning_rate": 0.0009570454605518663,
      "loss": 2.9951,
      "step": 1683
    },
    {
      "epoch": 0.013472,
      "grad_norm": 239.0,
      "learning_rate": 0.0009569943861390797,
      "loss": 2.2862,
      "step": 1684
    },
    {
      "epoch": 0.01348,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009569432827442164,
      "loss": 2.5532,
      "step": 1685
    },
    {
      "epoch": 0.013488,
      "grad_norm": 3.40625,
      "learning_rate": 0.0009568921503705176,
      "loss": 2.8841,
      "step": 1686
    },
    {
      "epoch": 0.013496,
      "grad_norm": 4.5,
      "learning_rate": 0.0009568409890212257,
      "loss": 3.2982,
      "step": 1687
    },
    {
      "epoch": 0.013504,
      "grad_norm": 3.578125,
      "learning_rate": 0.0009567897986995857,
      "loss": 2.9228,
      "step": 1688
    },
    {
      "epoch": 0.013512,
      "grad_norm": 3.578125,
      "learning_rate": 0.0009567385794088438,
      "loss": 2.3325,
      "step": 1689
    },
    {
      "epoch": 0.01352,
      "grad_norm": 4.5,
      "learning_rate": 0.0009566873311522482,
      "loss": 3.0614,
      "step": 1690
    },
    {
      "epoch": 0.013528,
      "grad_norm": 452.0,
      "learning_rate": 0.0009566360539330492,
      "loss": 2.5966,
      "step": 1691
    },
    {
      "epoch": 0.013536,
      "grad_norm": 4.75,
      "learning_rate": 0.0009565847477544986,
      "loss": 2.7207,
      "step": 1692
    },
    {
      "epoch": 0.013544,
      "grad_norm": 4.90625,
      "learning_rate": 0.0009565334126198504,
      "loss": 2.7129,
      "step": 1693
    },
    {
      "epoch": 0.013552,
      "grad_norm": 4.28125,
      "learning_rate": 0.0009564820485323599,
      "loss": 2.7854,
      "step": 1694
    },
    {
      "epoch": 0.01356,
      "grad_norm": 6.21875,
      "learning_rate": 0.0009564306554952847,
      "loss": 3.0738,
      "step": 1695
    },
    {
      "epoch": 0.013568,
      "grad_norm": 3.84375,
      "learning_rate": 0.0009563792335118841,
      "loss": 2.6817,
      "step": 1696
    },
    {
      "epoch": 0.013576,
      "grad_norm": 5.71875,
      "learning_rate": 0.0009563277825854193,
      "loss": 3.2775,
      "step": 1697
    },
    {
      "epoch": 0.013584,
      "grad_norm": 5.125,
      "learning_rate": 0.0009562763027191531,
      "loss": 2.6701,
      "step": 1698
    },
    {
      "epoch": 0.013592,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009562247939163504,
      "loss": 2.4586,
      "step": 1699
    },
    {
      "epoch": 0.0136,
      "grad_norm": 5.5625,
      "learning_rate": 0.0009561732561802779,
      "loss": 3.0457,
      "step": 1700
    },
    {
      "epoch": 0.013608,
      "grad_norm": 152.0,
      "learning_rate": 0.0009561216895142038,
      "loss": 2.3603,
      "step": 1701
    },
    {
      "epoch": 0.013616,
      "grad_norm": 197.0,
      "learning_rate": 0.0009560700939213987,
      "loss": 2.5521,
      "step": 1702
    },
    {
      "epoch": 0.013624,
      "grad_norm": 10.75,
      "learning_rate": 0.0009560184694051345,
      "loss": 2.902,
      "step": 1703
    },
    {
      "epoch": 0.013632,
      "grad_norm": 2624.0,
      "learning_rate": 0.0009559668159686853,
      "loss": 2.7219,
      "step": 1704
    },
    {
      "epoch": 0.01364,
      "grad_norm": 568.0,
      "learning_rate": 0.0009559151336153271,
      "loss": 2.7122,
      "step": 1705
    },
    {
      "epoch": 0.013648,
      "grad_norm": 4352.0,
      "learning_rate": 0.0009558634223483371,
      "loss": 3.0346,
      "step": 1706
    },
    {
      "epoch": 0.013656,
      "grad_norm": 191.0,
      "learning_rate": 0.000955811682170995,
      "loss": 3.8762,
      "step": 1707
    },
    {
      "epoch": 0.013664,
      "grad_norm": 1104.0,
      "learning_rate": 0.0009557599130865822,
      "loss": 3.4818,
      "step": 1708
    },
    {
      "epoch": 0.013672,
      "grad_norm": 4800.0,
      "learning_rate": 0.0009557081150983818,
      "loss": 4.3088,
      "step": 1709
    },
    {
      "epoch": 0.01368,
      "grad_norm": 15360.0,
      "learning_rate": 0.0009556562882096784,
      "loss": 5.2712,
      "step": 1710
    },
    {
      "epoch": 0.013688,
      "grad_norm": 334.0,
      "learning_rate": 0.0009556044324237594,
      "loss": 6.1616,
      "step": 1711
    },
    {
      "epoch": 0.013696,
      "grad_norm": 144.0,
      "learning_rate": 0.0009555525477439131,
      "loss": 4.8233,
      "step": 1712
    },
    {
      "epoch": 0.013704,
      "grad_norm": 21.375,
      "learning_rate": 0.0009555006341734299,
      "loss": 4.2935,
      "step": 1713
    },
    {
      "epoch": 0.013712,
      "grad_norm": 12.0625,
      "learning_rate": 0.0009554486917156022,
      "loss": 4.2703,
      "step": 1714
    },
    {
      "epoch": 0.01372,
      "grad_norm": 11.375,
      "learning_rate": 0.0009553967203737243,
      "loss": 3.4716,
      "step": 1715
    },
    {
      "epoch": 0.013728,
      "grad_norm": 125.0,
      "learning_rate": 0.0009553447201510919,
      "loss": 2.4771,
      "step": 1716
    },
    {
      "epoch": 0.013736,
      "grad_norm": 68.5,
      "learning_rate": 0.0009552926910510029,
      "loss": 2.9879,
      "step": 1717
    },
    {
      "epoch": 0.013744,
      "grad_norm": 7.71875,
      "learning_rate": 0.0009552406330767571,
      "loss": 3.2319,
      "step": 1718
    },
    {
      "epoch": 0.013752,
      "grad_norm": 7.78125,
      "learning_rate": 0.0009551885462316557,
      "loss": 2.8692,
      "step": 1719
    },
    {
      "epoch": 0.01376,
      "grad_norm": 6.96875,
      "learning_rate": 0.000955136430519002,
      "loss": 3.608,
      "step": 1720
    },
    {
      "epoch": 0.013768,
      "grad_norm": 5.78125,
      "learning_rate": 0.0009550842859421013,
      "loss": 2.7555,
      "step": 1721
    },
    {
      "epoch": 0.013776,
      "grad_norm": 5.65625,
      "learning_rate": 0.0009550321125042604,
      "loss": 2.9747,
      "step": 1722
    },
    {
      "epoch": 0.013784,
      "grad_norm": 6.03125,
      "learning_rate": 0.0009549799102087881,
      "loss": 3.2478,
      "step": 1723
    },
    {
      "epoch": 0.013792,
      "grad_norm": 4.96875,
      "learning_rate": 0.0009549276790589953,
      "loss": 2.627,
      "step": 1724
    },
    {
      "epoch": 0.0138,
      "grad_norm": 13.3125,
      "learning_rate": 0.0009548754190581938,
      "loss": 3.0301,
      "step": 1725
    },
    {
      "epoch": 0.013808,
      "grad_norm": 4.34375,
      "learning_rate": 0.0009548231302096985,
      "loss": 3.0802,
      "step": 1726
    },
    {
      "epoch": 0.013816,
      "grad_norm": 4.84375,
      "learning_rate": 0.0009547708125168253,
      "loss": 3.3961,
      "step": 1727
    },
    {
      "epoch": 0.013824,
      "grad_norm": 4.65625,
      "learning_rate": 0.000954718465982892,
      "loss": 2.964,
      "step": 1728
    },
    {
      "epoch": 0.013832,
      "grad_norm": 7.96875,
      "learning_rate": 0.0009546660906112186,
      "loss": 2.8581,
      "step": 1729
    },
    {
      "epoch": 0.01384,
      "grad_norm": 548.0,
      "learning_rate": 0.0009546136864051265,
      "loss": 2.7159,
      "step": 1730
    },
    {
      "epoch": 0.013848,
      "grad_norm": 15.1875,
      "learning_rate": 0.0009545612533679391,
      "loss": 2.8315,
      "step": 1731
    },
    {
      "epoch": 0.013856,
      "grad_norm": 12.125,
      "learning_rate": 0.0009545087915029816,
      "loss": 3.3987,
      "step": 1732
    },
    {
      "epoch": 0.013864,
      "grad_norm": 980.0,
      "learning_rate": 0.0009544563008135814,
      "loss": 2.8346,
      "step": 1733
    },
    {
      "epoch": 0.013872,
      "grad_norm": 15.375,
      "learning_rate": 0.0009544037813030669,
      "loss": 3.2528,
      "step": 1734
    },
    {
      "epoch": 0.01388,
      "grad_norm": 25.5,
      "learning_rate": 0.0009543512329747693,
      "loss": 3.391,
      "step": 1735
    },
    {
      "epoch": 0.013888,
      "grad_norm": 91.5,
      "learning_rate": 0.0009542986558320209,
      "loss": 4.3246,
      "step": 1736
    },
    {
      "epoch": 0.013896,
      "grad_norm": 11.8125,
      "learning_rate": 0.0009542460498781562,
      "loss": 3.719,
      "step": 1737
    },
    {
      "epoch": 0.013904,
      "grad_norm": 7.09375,
      "learning_rate": 0.0009541934151165113,
      "loss": 2.8787,
      "step": 1738
    },
    {
      "epoch": 0.013912,
      "grad_norm": 11.1875,
      "learning_rate": 0.0009541407515504243,
      "loss": 3.1767,
      "step": 1739
    },
    {
      "epoch": 0.01392,
      "grad_norm": 11.5625,
      "learning_rate": 0.0009540880591832352,
      "loss": 2.5473,
      "step": 1740
    },
    {
      "epoch": 0.013928,
      "grad_norm": 9.125,
      "learning_rate": 0.0009540353380182853,
      "loss": 2.4701,
      "step": 1741
    },
    {
      "epoch": 0.013936,
      "grad_norm": 12.5,
      "learning_rate": 0.0009539825880589185,
      "loss": 3.2072,
      "step": 1742
    },
    {
      "epoch": 0.013944,
      "grad_norm": 7.5625,
      "learning_rate": 0.0009539298093084801,
      "loss": 3.0691,
      "step": 1743
    },
    {
      "epoch": 0.013952,
      "grad_norm": 5.0625,
      "learning_rate": 0.0009538770017703171,
      "loss": 2.9854,
      "step": 1744
    },
    {
      "epoch": 0.01396,
      "grad_norm": 6.875,
      "learning_rate": 0.0009538241654477787,
      "loss": 3.0524,
      "step": 1745
    },
    {
      "epoch": 0.013968,
      "grad_norm": 6.75,
      "learning_rate": 0.0009537713003442155,
      "loss": 2.8033,
      "step": 1746
    },
    {
      "epoch": 0.013976,
      "grad_norm": 6.0625,
      "learning_rate": 0.0009537184064629801,
      "loss": 3.0128,
      "step": 1747
    },
    {
      "epoch": 0.013984,
      "grad_norm": 7.875,
      "learning_rate": 0.0009536654838074272,
      "loss": 2.5575,
      "step": 1748
    },
    {
      "epoch": 0.013992,
      "grad_norm": 6.9375,
      "learning_rate": 0.0009536125323809132,
      "loss": 2.6684,
      "step": 1749
    },
    {
      "epoch": 0.014,
      "grad_norm": 10.0,
      "learning_rate": 0.0009535595521867959,
      "loss": 2.8853,
      "step": 1750
    },
    {
      "epoch": 0.014008,
      "grad_norm": 4.40625,
      "learning_rate": 0.0009535065432284353,
      "loss": 3.4151,
      "step": 1751
    },
    {
      "epoch": 0.014016,
      "grad_norm": 3.765625,
      "learning_rate": 0.0009534535055091934,
      "loss": 2.0483,
      "step": 1752
    },
    {
      "epoch": 0.014024,
      "grad_norm": 20.875,
      "learning_rate": 0.0009534004390324335,
      "loss": 2.6147,
      "step": 1753
    },
    {
      "epoch": 0.014032,
      "grad_norm": 4.5,
      "learning_rate": 0.0009533473438015213,
      "loss": 2.7359,
      "step": 1754
    },
    {
      "epoch": 0.01404,
      "grad_norm": 5.34375,
      "learning_rate": 0.0009532942198198237,
      "loss": 2.7111,
      "step": 1755
    },
    {
      "epoch": 0.014048,
      "grad_norm": 4.28125,
      "learning_rate": 0.0009532410670907102,
      "loss": 3.01,
      "step": 1756
    },
    {
      "epoch": 0.014056,
      "grad_norm": 4.59375,
      "learning_rate": 0.0009531878856175513,
      "loss": 3.3613,
      "step": 1757
    },
    {
      "epoch": 0.014064,
      "grad_norm": 4.25,
      "learning_rate": 0.00095313467540372,
      "loss": 3.2192,
      "step": 1758
    },
    {
      "epoch": 0.014072,
      "grad_norm": 4.25,
      "learning_rate": 0.0009530814364525906,
      "loss": 2.6988,
      "step": 1759
    },
    {
      "epoch": 0.01408,
      "grad_norm": 3.96875,
      "learning_rate": 0.0009530281687675395,
      "loss": 2.5759,
      "step": 1760
    },
    {
      "epoch": 0.014088,
      "grad_norm": 125.5,
      "learning_rate": 0.0009529748723519451,
      "loss": 2.5701,
      "step": 1761
    },
    {
      "epoch": 0.014096,
      "grad_norm": 3.578125,
      "learning_rate": 0.000952921547209187,
      "loss": 2.4377,
      "step": 1762
    },
    {
      "epoch": 0.014104,
      "grad_norm": 4.25,
      "learning_rate": 0.0009528681933426474,
      "loss": 2.9383,
      "step": 1763
    },
    {
      "epoch": 0.014112,
      "grad_norm": 4.5,
      "learning_rate": 0.0009528148107557097,
      "loss": 2.5384,
      "step": 1764
    },
    {
      "epoch": 0.01412,
      "grad_norm": 4.09375,
      "learning_rate": 0.0009527613994517595,
      "loss": 2.9215,
      "step": 1765
    },
    {
      "epoch": 0.014128,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009527079594341841,
      "loss": 2.1204,
      "step": 1766
    },
    {
      "epoch": 0.014136,
      "grad_norm": 3.125,
      "learning_rate": 0.0009526544907063726,
      "loss": 2.2696,
      "step": 1767
    },
    {
      "epoch": 0.014144,
      "grad_norm": 6.375,
      "learning_rate": 0.0009526009932717157,
      "loss": 3.41,
      "step": 1768
    },
    {
      "epoch": 0.014152,
      "grad_norm": 3.796875,
      "learning_rate": 0.0009525474671336063,
      "loss": 2.3253,
      "step": 1769
    },
    {
      "epoch": 0.01416,
      "grad_norm": 4.96875,
      "learning_rate": 0.0009524939122954391,
      "loss": 2.945,
      "step": 1770
    },
    {
      "epoch": 0.014168,
      "grad_norm": 5.25,
      "learning_rate": 0.0009524403287606104,
      "loss": 3.0538,
      "step": 1771
    },
    {
      "epoch": 0.014176,
      "grad_norm": 3.546875,
      "learning_rate": 0.0009523867165325183,
      "loss": 3.1133,
      "step": 1772
    },
    {
      "epoch": 0.014184,
      "grad_norm": 3.734375,
      "learning_rate": 0.000952333075614563,
      "loss": 3.1749,
      "step": 1773
    },
    {
      "epoch": 0.014192,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009522794060101462,
      "loss": 2.6868,
      "step": 1774
    },
    {
      "epoch": 0.0142,
      "grad_norm": 3.25,
      "learning_rate": 0.0009522257077226717,
      "loss": 3.1258,
      "step": 1775
    },
    {
      "epoch": 0.014208,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009521719807555448,
      "loss": 3.0222,
      "step": 1776
    },
    {
      "epoch": 0.014216,
      "grad_norm": 3.671875,
      "learning_rate": 0.000952118225112173,
      "loss": 3.0513,
      "step": 1777
    },
    {
      "epoch": 0.014224,
      "grad_norm": 4.40625,
      "learning_rate": 0.0009520644407959653,
      "loss": 2.5004,
      "step": 1778
    },
    {
      "epoch": 0.014232,
      "grad_norm": 3.53125,
      "learning_rate": 0.0009520106278103327,
      "loss": 2.8099,
      "step": 1779
    },
    {
      "epoch": 0.01424,
      "grad_norm": 4.21875,
      "learning_rate": 0.000951956786158688,
      "loss": 2.6422,
      "step": 1780
    },
    {
      "epoch": 0.014248,
      "grad_norm": 3.640625,
      "learning_rate": 0.0009519029158444457,
      "loss": 3.2086,
      "step": 1781
    },
    {
      "epoch": 0.014256,
      "grad_norm": 4.59375,
      "learning_rate": 0.0009518490168710221,
      "loss": 2.8074,
      "step": 1782
    },
    {
      "epoch": 0.014264,
      "grad_norm": 4.6875,
      "learning_rate": 0.0009517950892418357,
      "loss": 3.0088,
      "step": 1783
    },
    {
      "epoch": 0.014272,
      "grad_norm": 5.0625,
      "learning_rate": 0.0009517411329603063,
      "loss": 2.9066,
      "step": 1784
    },
    {
      "epoch": 0.01428,
      "grad_norm": 13.0,
      "learning_rate": 0.0009516871480298556,
      "loss": 3.0485,
      "step": 1785
    },
    {
      "epoch": 0.014288,
      "grad_norm": 7.3125,
      "learning_rate": 0.0009516331344539078,
      "loss": 2.0322,
      "step": 1786
    },
    {
      "epoch": 0.014296,
      "grad_norm": 4.40625,
      "learning_rate": 0.0009515790922358879,
      "loss": 2.5755,
      "step": 1787
    },
    {
      "epoch": 0.014304,
      "grad_norm": 4.59375,
      "learning_rate": 0.0009515250213792234,
      "loss": 2.5647,
      "step": 1788
    },
    {
      "epoch": 0.014312,
      "grad_norm": 4.6875,
      "learning_rate": 0.0009514709218873433,
      "loss": 2.9292,
      "step": 1789
    },
    {
      "epoch": 0.01432,
      "grad_norm": 5.75,
      "learning_rate": 0.0009514167937636787,
      "loss": 2.8275,
      "step": 1790
    },
    {
      "epoch": 0.014328,
      "grad_norm": 6.03125,
      "learning_rate": 0.0009513626370116621,
      "loss": 1.9728,
      "step": 1791
    },
    {
      "epoch": 0.014336,
      "grad_norm": 6.9375,
      "learning_rate": 0.0009513084516347283,
      "loss": 3.1421,
      "step": 1792
    },
    {
      "epoch": 0.014344,
      "grad_norm": 11.3125,
      "learning_rate": 0.0009512542376363136,
      "loss": 2.3259,
      "step": 1793
    },
    {
      "epoch": 0.014352,
      "grad_norm": 11.875,
      "learning_rate": 0.000951199995019856,
      "loss": 2.666,
      "step": 1794
    },
    {
      "epoch": 0.01436,
      "grad_norm": 4.71875,
      "learning_rate": 0.0009511457237887959,
      "loss": 2.7335,
      "step": 1795
    },
    {
      "epoch": 0.014368,
      "grad_norm": 3.875,
      "learning_rate": 0.0009510914239465746,
      "loss": 2.6984,
      "step": 1796
    },
    {
      "epoch": 0.014376,
      "grad_norm": 3.8125,
      "learning_rate": 0.0009510370954966363,
      "loss": 2.3031,
      "step": 1797
    },
    {
      "epoch": 0.014384,
      "grad_norm": 3.875,
      "learning_rate": 0.0009509827384424261,
      "loss": 3.1488,
      "step": 1798
    },
    {
      "epoch": 0.014392,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009509283527873914,
      "loss": 3.1044,
      "step": 1799
    },
    {
      "epoch": 0.0144,
      "grad_norm": 4.15625,
      "learning_rate": 0.0009508739385349812,
      "loss": 2.1658,
      "step": 1800
    },
    {
      "epoch": 0.014408,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009508194956886465,
      "loss": 3.1576,
      "step": 1801
    },
    {
      "epoch": 0.014416,
      "grad_norm": 3.578125,
      "learning_rate": 0.0009507650242518398,
      "loss": 2.7158,
      "step": 1802
    },
    {
      "epoch": 0.014424,
      "grad_norm": 2.828125,
      "learning_rate": 0.0009507105242280156,
      "loss": 2.0661,
      "step": 1803
    },
    {
      "epoch": 0.014432,
      "grad_norm": 3.609375,
      "learning_rate": 0.0009506559956206307,
      "loss": 2.4327,
      "step": 1804
    },
    {
      "epoch": 0.01444,
      "grad_norm": 4.09375,
      "learning_rate": 0.0009506014384331426,
      "loss": 2.8335,
      "step": 1805
    },
    {
      "epoch": 0.014448,
      "grad_norm": 3.65625,
      "learning_rate": 0.0009505468526690117,
      "loss": 2.0364,
      "step": 1806
    },
    {
      "epoch": 0.014456,
      "grad_norm": 20.375,
      "learning_rate": 0.0009504922383316996,
      "loss": 2.6127,
      "step": 1807
    },
    {
      "epoch": 0.014464,
      "grad_norm": 4.625,
      "learning_rate": 0.00095043759542467,
      "loss": 2.8881,
      "step": 1808
    },
    {
      "epoch": 0.014472,
      "grad_norm": 4.15625,
      "learning_rate": 0.0009503829239513881,
      "loss": 2.675,
      "step": 1809
    },
    {
      "epoch": 0.01448,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009503282239153215,
      "loss": 2.9855,
      "step": 1810
    },
    {
      "epoch": 0.014488,
      "grad_norm": 4.4375,
      "learning_rate": 0.0009502734953199386,
      "loss": 1.93,
      "step": 1811
    },
    {
      "epoch": 0.014496,
      "grad_norm": 3.609375,
      "learning_rate": 0.0009502187381687106,
      "loss": 2.6029,
      "step": 1812
    },
    {
      "epoch": 0.014504,
      "grad_norm": 3.796875,
      "learning_rate": 0.0009501639524651102,
      "loss": 2.5635,
      "step": 1813
    },
    {
      "epoch": 0.014512,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009501091382126118,
      "loss": 2.6514,
      "step": 1814
    },
    {
      "epoch": 0.01452,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009500542954146916,
      "loss": 2.3883,
      "step": 1815
    },
    {
      "epoch": 0.014528,
      "grad_norm": 3.125,
      "learning_rate": 0.0009499994240748275,
      "loss": 2.8491,
      "step": 1816
    },
    {
      "epoch": 0.014536,
      "grad_norm": 4.59375,
      "learning_rate": 0.0009499445241964997,
      "loss": 2.9087,
      "step": 1817
    },
    {
      "epoch": 0.014544,
      "grad_norm": 27.75,
      "learning_rate": 0.0009498895957831897,
      "loss": 2.0625,
      "step": 1818
    },
    {
      "epoch": 0.014552,
      "grad_norm": 7.21875,
      "learning_rate": 0.0009498346388383809,
      "loss": 2.8301,
      "step": 1819
    },
    {
      "epoch": 0.01456,
      "grad_norm": 5.1875,
      "learning_rate": 0.0009497796533655589,
      "loss": 2.8156,
      "step": 1820
    },
    {
      "epoch": 0.014568,
      "grad_norm": 5.15625,
      "learning_rate": 0.0009497246393682107,
      "loss": 2.0179,
      "step": 1821
    },
    {
      "epoch": 0.014576,
      "grad_norm": 3.609375,
      "learning_rate": 0.0009496695968498252,
      "loss": 2.9041,
      "step": 1822
    },
    {
      "epoch": 0.014584,
      "grad_norm": 4.5,
      "learning_rate": 0.0009496145258138931,
      "loss": 2.6081,
      "step": 1823
    },
    {
      "epoch": 0.014592,
      "grad_norm": 4.3125,
      "learning_rate": 0.0009495594262639069,
      "loss": 2.853,
      "step": 1824
    },
    {
      "epoch": 0.0146,
      "grad_norm": 5.875,
      "learning_rate": 0.0009495042982033611,
      "loss": 2.4539,
      "step": 1825
    },
    {
      "epoch": 0.014608,
      "grad_norm": 5.0625,
      "learning_rate": 0.0009494491416357517,
      "loss": 2.9195,
      "step": 1826
    },
    {
      "epoch": 0.014616,
      "grad_norm": 6.78125,
      "learning_rate": 0.0009493939565645769,
      "loss": 2.3782,
      "step": 1827
    },
    {
      "epoch": 0.014624,
      "grad_norm": 5.5,
      "learning_rate": 0.0009493387429933362,
      "loss": 2.2934,
      "step": 1828
    },
    {
      "epoch": 0.014632,
      "grad_norm": 5.625,
      "learning_rate": 0.0009492835009255313,
      "loss": 2.9106,
      "step": 1829
    },
    {
      "epoch": 0.01464,
      "grad_norm": 4.75,
      "learning_rate": 0.0009492282303646657,
      "loss": 2.8992,
      "step": 1830
    },
    {
      "epoch": 0.014648,
      "grad_norm": 9.5,
      "learning_rate": 0.0009491729313142443,
      "loss": 2.7906,
      "step": 1831
    },
    {
      "epoch": 0.014656,
      "grad_norm": 4.90625,
      "learning_rate": 0.0009491176037777745,
      "loss": 2.7732,
      "step": 1832
    },
    {
      "epoch": 0.014664,
      "grad_norm": 4.0625,
      "learning_rate": 0.000949062247758765,
      "loss": 2.6595,
      "step": 1833
    },
    {
      "epoch": 0.014672,
      "grad_norm": 4.1875,
      "learning_rate": 0.0009490068632607261,
      "loss": 2.7472,
      "step": 1834
    },
    {
      "epoch": 0.01468,
      "grad_norm": 3.6875,
      "learning_rate": 0.0009489514502871706,
      "loss": 2.928,
      "step": 1835
    },
    {
      "epoch": 0.014688,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009488960088416125,
      "loss": 2.9612,
      "step": 1836
    },
    {
      "epoch": 0.014696,
      "grad_norm": 4.0,
      "learning_rate": 0.0009488405389275681,
      "loss": 3.5982,
      "step": 1837
    },
    {
      "epoch": 0.014704,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009487850405485547,
      "loss": 2.7505,
      "step": 1838
    },
    {
      "epoch": 0.014712,
      "grad_norm": 3.0625,
      "learning_rate": 0.0009487295137080927,
      "loss": 2.1785,
      "step": 1839
    },
    {
      "epoch": 0.01472,
      "grad_norm": 3.734375,
      "learning_rate": 0.0009486739584097031,
      "loss": 3.3794,
      "step": 1840
    },
    {
      "epoch": 0.014728,
      "grad_norm": 2.875,
      "learning_rate": 0.000948618374656909,
      "loss": 2.3741,
      "step": 1841
    },
    {
      "epoch": 0.014736,
      "grad_norm": 3.625,
      "learning_rate": 0.0009485627624532358,
      "loss": 2.9139,
      "step": 1842
    },
    {
      "epoch": 0.014744,
      "grad_norm": 4.125,
      "learning_rate": 0.0009485071218022102,
      "loss": 2.7489,
      "step": 1843
    },
    {
      "epoch": 0.014752,
      "grad_norm": 4.4375,
      "learning_rate": 0.0009484514527073609,
      "loss": 2.7216,
      "step": 1844
    },
    {
      "epoch": 0.01476,
      "grad_norm": 5.71875,
      "learning_rate": 0.0009483957551722185,
      "loss": 2.7165,
      "step": 1845
    },
    {
      "epoch": 0.014768,
      "grad_norm": 6.0625,
      "learning_rate": 0.000948340029200315,
      "loss": 2.7619,
      "step": 1846
    },
    {
      "epoch": 0.014776,
      "grad_norm": 5.03125,
      "learning_rate": 0.0009482842747951846,
      "loss": 2.9371,
      "step": 1847
    },
    {
      "epoch": 0.014784,
      "grad_norm": 19.75,
      "learning_rate": 0.0009482284919603635,
      "loss": 1.722,
      "step": 1848
    },
    {
      "epoch": 0.014792,
      "grad_norm": 6.6875,
      "learning_rate": 0.0009481726806993887,
      "loss": 3.0757,
      "step": 1849
    },
    {
      "epoch": 0.0148,
      "grad_norm": 6.09375,
      "learning_rate": 0.0009481168410158003,
      "loss": 2.7552,
      "step": 1850
    },
    {
      "epoch": 0.014808,
      "grad_norm": 12.875,
      "learning_rate": 0.0009480609729131394,
      "loss": 2.3559,
      "step": 1851
    },
    {
      "epoch": 0.014816,
      "grad_norm": 3.875,
      "learning_rate": 0.000948005076394949,
      "loss": 2.3399,
      "step": 1852
    },
    {
      "epoch": 0.014824,
      "grad_norm": 5.65625,
      "learning_rate": 0.0009479491514647742,
      "loss": 2.485,
      "step": 1853
    },
    {
      "epoch": 0.014832,
      "grad_norm": 4.53125,
      "learning_rate": 0.0009478931981261615,
      "loss": 2.9368,
      "step": 1854
    },
    {
      "epoch": 0.01484,
      "grad_norm": 109.5,
      "learning_rate": 0.0009478372163826594,
      "loss": 2.159,
      "step": 1855
    },
    {
      "epoch": 0.014848,
      "grad_norm": 3.875,
      "learning_rate": 0.0009477812062378184,
      "loss": 2.2016,
      "step": 1856
    },
    {
      "epoch": 0.014856,
      "grad_norm": 3.921875,
      "learning_rate": 0.0009477251676951904,
      "loss": 2.9366,
      "step": 1857
    },
    {
      "epoch": 0.014864,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009476691007583292,
      "loss": 2.2999,
      "step": 1858
    },
    {
      "epoch": 0.014872,
      "grad_norm": 3.46875,
      "learning_rate": 0.0009476130054307907,
      "loss": 2.4486,
      "step": 1859
    },
    {
      "epoch": 0.01488,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009475568817161326,
      "loss": 2.0589,
      "step": 1860
    },
    {
      "epoch": 0.014888,
      "grad_norm": 4.34375,
      "learning_rate": 0.0009475007296179139,
      "loss": 2.6214,
      "step": 1861
    },
    {
      "epoch": 0.014896,
      "grad_norm": 3.875,
      "learning_rate": 0.0009474445491396957,
      "loss": 2.5102,
      "step": 1862
    },
    {
      "epoch": 0.014904,
      "grad_norm": 5.03125,
      "learning_rate": 0.000947388340285041,
      "loss": 3.3666,
      "step": 1863
    },
    {
      "epoch": 0.014912,
      "grad_norm": 4.0,
      "learning_rate": 0.0009473321030575145,
      "loss": 2.5488,
      "step": 1864
    },
    {
      "epoch": 0.01492,
      "grad_norm": 3.78125,
      "learning_rate": 0.0009472758374606826,
      "loss": 2.0991,
      "step": 1865
    },
    {
      "epoch": 0.014928,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009472195434981139,
      "loss": 2.0958,
      "step": 1866
    },
    {
      "epoch": 0.014936,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009471632211733781,
      "loss": 2.6042,
      "step": 1867
    },
    {
      "epoch": 0.014944,
      "grad_norm": 3.796875,
      "learning_rate": 0.0009471068704900474,
      "loss": 2.5553,
      "step": 1868
    },
    {
      "epoch": 0.014952,
      "grad_norm": 3.25,
      "learning_rate": 0.0009470504914516953,
      "loss": 2.8116,
      "step": 1869
    },
    {
      "epoch": 0.01496,
      "grad_norm": 3.609375,
      "learning_rate": 0.0009469940840618976,
      "loss": 2.2402,
      "step": 1870
    },
    {
      "epoch": 0.014968,
      "grad_norm": 4.3125,
      "learning_rate": 0.0009469376483242311,
      "loss": 2.8208,
      "step": 1871
    },
    {
      "epoch": 0.014976,
      "grad_norm": 4.4375,
      "learning_rate": 0.0009468811842422753,
      "loss": 3.0299,
      "step": 1872
    },
    {
      "epoch": 0.014984,
      "grad_norm": 3.171875,
      "learning_rate": 0.000946824691819611,
      "loss": 2.1514,
      "step": 1873
    },
    {
      "epoch": 0.014992,
      "grad_norm": 4.1875,
      "learning_rate": 0.0009467681710598208,
      "loss": 2.2212,
      "step": 1874
    },
    {
      "epoch": 0.015,
      "grad_norm": 4.6875,
      "learning_rate": 0.0009467116219664893,
      "loss": 2.4741,
      "step": 1875
    },
    {
      "epoch": 0.015008,
      "grad_norm": 3.390625,
      "learning_rate": 0.0009466550445432026,
      "loss": 2.4659,
      "step": 1876
    },
    {
      "epoch": 0.015016,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009465984387935489,
      "loss": 1.9163,
      "step": 1877
    },
    {
      "epoch": 0.015024,
      "grad_norm": 4.9375,
      "learning_rate": 0.000946541804721118,
      "loss": 3.0359,
      "step": 1878
    },
    {
      "epoch": 0.015032,
      "grad_norm": 3.3125,
      "learning_rate": 0.0009464851423295018,
      "loss": 2.0166,
      "step": 1879
    },
    {
      "epoch": 0.01504,
      "grad_norm": 4.6875,
      "learning_rate": 0.0009464284516222936,
      "loss": 3.2692,
      "step": 1880
    },
    {
      "epoch": 0.015048,
      "grad_norm": 4.6875,
      "learning_rate": 0.0009463717326030885,
      "loss": 2.4765,
      "step": 1881
    },
    {
      "epoch": 0.015056,
      "grad_norm": 4.96875,
      "learning_rate": 0.0009463149852754838,
      "loss": 3.1633,
      "step": 1882
    },
    {
      "epoch": 0.015064,
      "grad_norm": 9.4375,
      "learning_rate": 0.0009462582096430783,
      "loss": 2.2447,
      "step": 1883
    },
    {
      "epoch": 0.015072,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009462014057094726,
      "loss": 2.3417,
      "step": 1884
    },
    {
      "epoch": 0.01508,
      "grad_norm": 5.71875,
      "learning_rate": 0.0009461445734782692,
      "loss": 2.87,
      "step": 1885
    },
    {
      "epoch": 0.015088,
      "grad_norm": 3.515625,
      "learning_rate": 0.0009460877129530723,
      "loss": 2.294,
      "step": 1886
    },
    {
      "epoch": 0.015096,
      "grad_norm": 4.8125,
      "learning_rate": 0.0009460308241374878,
      "loss": 3.237,
      "step": 1887
    },
    {
      "epoch": 0.015104,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009459739070351237,
      "loss": 1.9501,
      "step": 1888
    },
    {
      "epoch": 0.015112,
      "grad_norm": 4.40625,
      "learning_rate": 0.0009459169616495896,
      "loss": 2.4074,
      "step": 1889
    },
    {
      "epoch": 0.01512,
      "grad_norm": 3.578125,
      "learning_rate": 0.0009458599879844969,
      "loss": 2.6216,
      "step": 1890
    },
    {
      "epoch": 0.015128,
      "grad_norm": 5.0625,
      "learning_rate": 0.0009458029860434587,
      "loss": 2.3515,
      "step": 1891
    },
    {
      "epoch": 0.015136,
      "grad_norm": 3.828125,
      "learning_rate": 0.0009457459558300902,
      "loss": 2.659,
      "step": 1892
    },
    {
      "epoch": 0.015144,
      "grad_norm": 4.5625,
      "learning_rate": 0.0009456888973480082,
      "loss": 2.8854,
      "step": 1893
    },
    {
      "epoch": 0.015152,
      "grad_norm": 4.78125,
      "learning_rate": 0.0009456318106008309,
      "loss": 3.4461,
      "step": 1894
    },
    {
      "epoch": 0.01516,
      "grad_norm": 3.734375,
      "learning_rate": 0.0009455746955921791,
      "loss": 2.6252,
      "step": 1895
    },
    {
      "epoch": 0.015168,
      "grad_norm": 6.96875,
      "learning_rate": 0.0009455175523256748,
      "loss": 2.7538,
      "step": 1896
    },
    {
      "epoch": 0.015176,
      "grad_norm": 4.5625,
      "learning_rate": 0.000945460380804942,
      "loss": 3.2448,
      "step": 1897
    },
    {
      "epoch": 0.015184,
      "grad_norm": 3.640625,
      "learning_rate": 0.0009454031810336064,
      "loss": 2.4004,
      "step": 1898
    },
    {
      "epoch": 0.015192,
      "grad_norm": 8.5625,
      "learning_rate": 0.0009453459530152957,
      "loss": 2.3201,
      "step": 1899
    },
    {
      "epoch": 0.0152,
      "grad_norm": 3.0625,
      "learning_rate": 0.0009452886967536389,
      "loss": 2.2991,
      "step": 1900
    },
    {
      "epoch": 0.015208,
      "grad_norm": 3.40625,
      "learning_rate": 0.0009452314122522676,
      "loss": 2.6102,
      "step": 1901
    },
    {
      "epoch": 0.015216,
      "grad_norm": 5.09375,
      "learning_rate": 0.0009451740995148143,
      "loss": 2.4964,
      "step": 1902
    },
    {
      "epoch": 0.015224,
      "grad_norm": 4.09375,
      "learning_rate": 0.0009451167585449138,
      "loss": 3.0048,
      "step": 1903
    },
    {
      "epoch": 0.015232,
      "grad_norm": 5.21875,
      "learning_rate": 0.0009450593893462029,
      "loss": 2.6576,
      "step": 1904
    },
    {
      "epoch": 0.01524,
      "grad_norm": 6.6875,
      "learning_rate": 0.0009450019919223196,
      "loss": 2.1955,
      "step": 1905
    },
    {
      "epoch": 0.015248,
      "grad_norm": 5.53125,
      "learning_rate": 0.0009449445662769039,
      "loss": 2.3768,
      "step": 1906
    },
    {
      "epoch": 0.015256,
      "grad_norm": 6.75,
      "learning_rate": 0.000944887112413598,
      "loss": 2.6027,
      "step": 1907
    },
    {
      "epoch": 0.015264,
      "grad_norm": 3.578125,
      "learning_rate": 0.0009448296303360453,
      "loss": 2.3413,
      "step": 1908
    },
    {
      "epoch": 0.015272,
      "grad_norm": 5.53125,
      "learning_rate": 0.0009447721200478914,
      "loss": 2.1708,
      "step": 1909
    },
    {
      "epoch": 0.01528,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009447145815527833,
      "loss": 2.1096,
      "step": 1910
    },
    {
      "epoch": 0.015288,
      "grad_norm": 4.4375,
      "learning_rate": 0.0009446570148543703,
      "loss": 2.1626,
      "step": 1911
    },
    {
      "epoch": 0.015296,
      "grad_norm": 3.828125,
      "learning_rate": 0.0009445994199563033,
      "loss": 2.6211,
      "step": 1912
    },
    {
      "epoch": 0.015304,
      "grad_norm": 3.125,
      "learning_rate": 0.0009445417968622345,
      "loss": 2.2542,
      "step": 1913
    },
    {
      "epoch": 0.015312,
      "grad_norm": 3.734375,
      "learning_rate": 0.0009444841455758186,
      "loss": 2.3958,
      "step": 1914
    },
    {
      "epoch": 0.01532,
      "grad_norm": 4.125,
      "learning_rate": 0.0009444264661007116,
      "loss": 2.5735,
      "step": 1915
    },
    {
      "epoch": 0.015328,
      "grad_norm": 4.28125,
      "learning_rate": 0.0009443687584405715,
      "loss": 2.9886,
      "step": 1916
    },
    {
      "epoch": 0.015336,
      "grad_norm": 4.0,
      "learning_rate": 0.0009443110225990583,
      "loss": 2.9884,
      "step": 1917
    },
    {
      "epoch": 0.015344,
      "grad_norm": 4.0,
      "learning_rate": 0.0009442532585798334,
      "loss": 2.5562,
      "step": 1918
    },
    {
      "epoch": 0.015352,
      "grad_norm": 4.15625,
      "learning_rate": 0.0009441954663865599,
      "loss": 2.6029,
      "step": 1919
    },
    {
      "epoch": 0.01536,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009441376460229032,
      "loss": 2.6729,
      "step": 1920
    },
    {
      "epoch": 0.015368,
      "grad_norm": 3.140625,
      "learning_rate": 0.0009440797974925301,
      "loss": 2.5426,
      "step": 1921
    },
    {
      "epoch": 0.015376,
      "grad_norm": 3.75,
      "learning_rate": 0.0009440219207991092,
      "loss": 2.4386,
      "step": 1922
    },
    {
      "epoch": 0.015384,
      "grad_norm": 4.34375,
      "learning_rate": 0.0009439640159463112,
      "loss": 3.1575,
      "step": 1923
    },
    {
      "epoch": 0.015392,
      "grad_norm": 5.4375,
      "learning_rate": 0.0009439060829378081,
      "loss": 2.6503,
      "step": 1924
    },
    {
      "epoch": 0.0154,
      "grad_norm": 3.75,
      "learning_rate": 0.0009438481217772743,
      "loss": 2.4362,
      "step": 1925
    },
    {
      "epoch": 0.015408,
      "grad_norm": 4.1875,
      "learning_rate": 0.0009437901324683853,
      "loss": 2.4339,
      "step": 1926
    },
    {
      "epoch": 0.015416,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009437321150148188,
      "loss": 2.7774,
      "step": 1927
    },
    {
      "epoch": 0.015424,
      "grad_norm": 4.96875,
      "learning_rate": 0.0009436740694202542,
      "loss": 2.4869,
      "step": 1928
    },
    {
      "epoch": 0.015432,
      "grad_norm": 6.15625,
      "learning_rate": 0.0009436159956883728,
      "loss": 2.8365,
      "step": 1929
    },
    {
      "epoch": 0.01544,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009435578938228576,
      "loss": 2.1284,
      "step": 1930
    },
    {
      "epoch": 0.015448,
      "grad_norm": 3.125,
      "learning_rate": 0.000943499763827393,
      "loss": 2.349,
      "step": 1931
    },
    {
      "epoch": 0.015456,
      "grad_norm": 3.890625,
      "learning_rate": 0.0009434416057056658,
      "loss": 2.7074,
      "step": 1932
    },
    {
      "epoch": 0.015464,
      "grad_norm": 55.5,
      "learning_rate": 0.0009433834194613644,
      "loss": 3.0703,
      "step": 1933
    },
    {
      "epoch": 0.015472,
      "grad_norm": 4.40625,
      "learning_rate": 0.0009433252050981788,
      "loss": 2.7756,
      "step": 1934
    },
    {
      "epoch": 0.01548,
      "grad_norm": 4.53125,
      "learning_rate": 0.000943266962619801,
      "loss": 2.1847,
      "step": 1935
    },
    {
      "epoch": 0.015488,
      "grad_norm": 4.0625,
      "learning_rate": 0.0009432086920299245,
      "loss": 2.7324,
      "step": 1936
    },
    {
      "epoch": 0.015496,
      "grad_norm": 4.9375,
      "learning_rate": 0.0009431503933322449,
      "loss": 2.3152,
      "step": 1937
    },
    {
      "epoch": 0.015504,
      "grad_norm": 4.375,
      "learning_rate": 0.0009430920665304594,
      "loss": 2.6246,
      "step": 1938
    },
    {
      "epoch": 0.015512,
      "grad_norm": 4.3125,
      "learning_rate": 0.0009430337116282669,
      "loss": 2.8588,
      "step": 1939
    },
    {
      "epoch": 0.01552,
      "grad_norm": 3.78125,
      "learning_rate": 0.0009429753286293683,
      "loss": 2.1823,
      "step": 1940
    },
    {
      "epoch": 0.015528,
      "grad_norm": 4.375,
      "learning_rate": 0.0009429169175374664,
      "loss": 2.5238,
      "step": 1941
    },
    {
      "epoch": 0.015536,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009428584783562651,
      "loss": 2.3751,
      "step": 1942
    },
    {
      "epoch": 0.015544,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009428000110894707,
      "loss": 2.0947,
      "step": 1943
    },
    {
      "epoch": 0.015552,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009427415157407916,
      "loss": 2.404,
      "step": 1944
    },
    {
      "epoch": 0.01556,
      "grad_norm": 3.234375,
      "learning_rate": 0.000942682992313937,
      "loss": 2.2722,
      "step": 1945
    },
    {
      "epoch": 0.015568,
      "grad_norm": 3.890625,
      "learning_rate": 0.0009426244408126184,
      "loss": 2.6087,
      "step": 1946
    },
    {
      "epoch": 0.015576,
      "grad_norm": 4.8125,
      "learning_rate": 0.0009425658612405492,
      "loss": 2.4504,
      "step": 1947
    },
    {
      "epoch": 0.015584,
      "grad_norm": 3.4375,
      "learning_rate": 0.0009425072536014446,
      "loss": 2.6989,
      "step": 1948
    },
    {
      "epoch": 0.015592,
      "grad_norm": 3.203125,
      "learning_rate": 0.0009424486178990213,
      "loss": 2.3022,
      "step": 1949
    },
    {
      "epoch": 0.0156,
      "grad_norm": 3.578125,
      "learning_rate": 0.0009423899541369978,
      "loss": 2.7904,
      "step": 1950
    },
    {
      "epoch": 0.015608,
      "grad_norm": 4.5,
      "learning_rate": 0.0009423312623190946,
      "loss": 2.4093,
      "step": 1951
    },
    {
      "epoch": 0.015616,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009422725424490339,
      "loss": 3.3801,
      "step": 1952
    },
    {
      "epoch": 0.015624,
      "grad_norm": 3.46875,
      "learning_rate": 0.0009422137945305396,
      "loss": 2.4903,
      "step": 1953
    },
    {
      "epoch": 0.015632,
      "grad_norm": 3.1875,
      "learning_rate": 0.0009421550185673373,
      "loss": 1.9227,
      "step": 1954
    },
    {
      "epoch": 0.01564,
      "grad_norm": 3.53125,
      "learning_rate": 0.0009420962145631549,
      "loss": 2.6215,
      "step": 1955
    },
    {
      "epoch": 0.015648,
      "grad_norm": 5.09375,
      "learning_rate": 0.0009420373825217212,
      "loss": 3.2512,
      "step": 1956
    },
    {
      "epoch": 0.015656,
      "grad_norm": 3.4375,
      "learning_rate": 0.0009419785224467675,
      "loss": 2.5304,
      "step": 1957
    },
    {
      "epoch": 0.015664,
      "grad_norm": 5.40625,
      "learning_rate": 0.0009419196343420267,
      "loss": 2.6813,
      "step": 1958
    },
    {
      "epoch": 0.015672,
      "grad_norm": 3.890625,
      "learning_rate": 0.0009418607182112333,
      "loss": 2.2973,
      "step": 1959
    },
    {
      "epoch": 0.01568,
      "grad_norm": 4.25,
      "learning_rate": 0.0009418017740581237,
      "loss": 2.4879,
      "step": 1960
    },
    {
      "epoch": 0.015688,
      "grad_norm": 6.34375,
      "learning_rate": 0.0009417428018864362,
      "loss": 2.7289,
      "step": 1961
    },
    {
      "epoch": 0.015696,
      "grad_norm": 5.8125,
      "learning_rate": 0.0009416838016999108,
      "loss": 2.4296,
      "step": 1962
    },
    {
      "epoch": 0.015704,
      "grad_norm": 4.28125,
      "learning_rate": 0.000941624773502289,
      "loss": 1.7515,
      "step": 1963
    },
    {
      "epoch": 0.015712,
      "grad_norm": 5.5625,
      "learning_rate": 0.0009415657172973142,
      "loss": 2.8563,
      "step": 1964
    },
    {
      "epoch": 0.01572,
      "grad_norm": 4.59375,
      "learning_rate": 0.0009415066330887321,
      "loss": 2.7107,
      "step": 1965
    },
    {
      "epoch": 0.015728,
      "grad_norm": 4.75,
      "learning_rate": 0.0009414475208802893,
      "loss": 2.7495,
      "step": 1966
    },
    {
      "epoch": 0.015736,
      "grad_norm": 3.9375,
      "learning_rate": 0.0009413883806757351,
      "loss": 3.8129,
      "step": 1967
    },
    {
      "epoch": 0.015744,
      "grad_norm": 4.75,
      "learning_rate": 0.0009413292124788198,
      "loss": 2.8651,
      "step": 1968
    },
    {
      "epoch": 0.015752,
      "grad_norm": 3.578125,
      "learning_rate": 0.0009412700162932957,
      "loss": 2.2376,
      "step": 1969
    },
    {
      "epoch": 0.01576,
      "grad_norm": 3.546875,
      "learning_rate": 0.0009412107921229171,
      "loss": 2.7573,
      "step": 1970
    },
    {
      "epoch": 0.015768,
      "grad_norm": 3.234375,
      "learning_rate": 0.00094115153997144,
      "loss": 2.3501,
      "step": 1971
    },
    {
      "epoch": 0.015776,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009410922598426222,
      "loss": 2.5103,
      "step": 1972
    },
    {
      "epoch": 0.015784,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009410329517402229,
      "loss": 2.7505,
      "step": 1973
    },
    {
      "epoch": 0.015792,
      "grad_norm": 3.875,
      "learning_rate": 0.0009409736156680035,
      "loss": 2.6499,
      "step": 1974
    },
    {
      "epoch": 0.0158,
      "grad_norm": 185.0,
      "learning_rate": 0.0009409142516297269,
      "loss": 2.3101,
      "step": 1975
    },
    {
      "epoch": 0.015808,
      "grad_norm": 3.90625,
      "learning_rate": 0.0009408548596291582,
      "loss": 2.3463,
      "step": 1976
    },
    {
      "epoch": 0.015816,
      "grad_norm": 4.0,
      "learning_rate": 0.0009407954396700635,
      "loss": 2.494,
      "step": 1977
    },
    {
      "epoch": 0.015824,
      "grad_norm": 3.765625,
      "learning_rate": 0.0009407359917562116,
      "loss": 3.3045,
      "step": 1978
    },
    {
      "epoch": 0.015832,
      "grad_norm": 3.34375,
      "learning_rate": 0.0009406765158913724,
      "loss": 2.7237,
      "step": 1979
    },
    {
      "epoch": 0.01584,
      "grad_norm": 4.0,
      "learning_rate": 0.0009406170120793178,
      "loss": 2.596,
      "step": 1980
    },
    {
      "epoch": 0.015848,
      "grad_norm": 5.65625,
      "learning_rate": 0.0009405574803238216,
      "loss": 2.4166,
      "step": 1981
    },
    {
      "epoch": 0.015856,
      "grad_norm": 3.765625,
      "learning_rate": 0.0009404979206286591,
      "loss": 2.8139,
      "step": 1982
    },
    {
      "epoch": 0.015864,
      "grad_norm": 4.75,
      "learning_rate": 0.0009404383329976076,
      "loss": 2.7056,
      "step": 1983
    },
    {
      "epoch": 0.015872,
      "grad_norm": 5.0625,
      "learning_rate": 0.0009403787174344459,
      "loss": 2.1955,
      "step": 1984
    },
    {
      "epoch": 0.01588,
      "grad_norm": 3.421875,
      "learning_rate": 0.0009403190739429551,
      "loss": 2.5193,
      "step": 1985
    },
    {
      "epoch": 0.015888,
      "grad_norm": 4.53125,
      "learning_rate": 0.0009402594025269173,
      "loss": 3.0063,
      "step": 1986
    },
    {
      "epoch": 0.015896,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009401997031901171,
      "loss": 2.7363,
      "step": 1987
    },
    {
      "epoch": 0.015904,
      "grad_norm": 2.96875,
      "learning_rate": 0.0009401399759363404,
      "loss": 2.2394,
      "step": 1988
    },
    {
      "epoch": 0.015912,
      "grad_norm": 2.78125,
      "learning_rate": 0.0009400802207693752,
      "loss": 2.1539,
      "step": 1989
    },
    {
      "epoch": 0.01592,
      "grad_norm": 3.25,
      "learning_rate": 0.0009400204376930109,
      "loss": 2.4843,
      "step": 1990
    },
    {
      "epoch": 0.015928,
      "grad_norm": 3.625,
      "learning_rate": 0.000939960626711039,
      "loss": 2.5537,
      "step": 1991
    },
    {
      "epoch": 0.015936,
      "grad_norm": 3.5,
      "learning_rate": 0.0009399007878272528,
      "loss": 2.4474,
      "step": 1992
    },
    {
      "epoch": 0.015944,
      "grad_norm": 3.40625,
      "learning_rate": 0.0009398409210454468,
      "loss": 2.4076,
      "step": 1993
    },
    {
      "epoch": 0.015952,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009397810263694183,
      "loss": 2.7296,
      "step": 1994
    },
    {
      "epoch": 0.01596,
      "grad_norm": 4.5625,
      "learning_rate": 0.000939721103802965,
      "loss": 3.219,
      "step": 1995
    },
    {
      "epoch": 0.015968,
      "grad_norm": 2.96875,
      "learning_rate": 0.0009396611533498877,
      "loss": 2.0824,
      "step": 1996
    },
    {
      "epoch": 0.015976,
      "grad_norm": 3.46875,
      "learning_rate": 0.0009396011750139881,
      "loss": 2.8568,
      "step": 1997
    },
    {
      "epoch": 0.015984,
      "grad_norm": 3.734375,
      "learning_rate": 0.0009395411687990702,
      "loss": 2.6632,
      "step": 1998
    },
    {
      "epoch": 0.015992,
      "grad_norm": 4.5,
      "learning_rate": 0.0009394811347089394,
      "loss": 2.4228,
      "step": 1999
    },
    {
      "epoch": 0.016,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009394210727474029,
      "loss": 2.2591,
      "step": 2000
    },
    {
      "epoch": 0.016008,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009393609829182699,
      "loss": 2.4385,
      "step": 2001
    },
    {
      "epoch": 0.016016,
      "grad_norm": 3.390625,
      "learning_rate": 0.0009393008652253511,
      "loss": 2.314,
      "step": 2002
    },
    {
      "epoch": 0.016024,
      "grad_norm": 3.484375,
      "learning_rate": 0.0009392407196724593,
      "loss": 3.0159,
      "step": 2003
    },
    {
      "epoch": 0.016032,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009391805462634086,
      "loss": 2.6126,
      "step": 2004
    },
    {
      "epoch": 0.01604,
      "grad_norm": 4.0,
      "learning_rate": 0.0009391203450020153,
      "loss": 2.8481,
      "step": 2005
    },
    {
      "epoch": 0.016048,
      "grad_norm": 4.5,
      "learning_rate": 0.0009390601158920973,
      "loss": 1.9919,
      "step": 2006
    },
    {
      "epoch": 0.016056,
      "grad_norm": 3.859375,
      "learning_rate": 0.0009389998589374742,
      "loss": 2.755,
      "step": 2007
    },
    {
      "epoch": 0.016064,
      "grad_norm": 3.546875,
      "learning_rate": 0.0009389395741419675,
      "loss": 2.3468,
      "step": 2008
    },
    {
      "epoch": 0.016072,
      "grad_norm": 3.4375,
      "learning_rate": 0.0009388792615094004,
      "loss": 2.9755,
      "step": 2009
    },
    {
      "epoch": 0.01608,
      "grad_norm": 3.125,
      "learning_rate": 0.0009388189210435977,
      "loss": 2.8788,
      "step": 2010
    },
    {
      "epoch": 0.016088,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009387585527483864,
      "loss": 2.5872,
      "step": 2011
    },
    {
      "epoch": 0.016096,
      "grad_norm": 2.6875,
      "learning_rate": 0.0009386981566275948,
      "loss": 2.3506,
      "step": 2012
    },
    {
      "epoch": 0.016104,
      "grad_norm": 3.546875,
      "learning_rate": 0.000938637732685053,
      "loss": 2.3691,
      "step": 2013
    },
    {
      "epoch": 0.016112,
      "grad_norm": 21.0,
      "learning_rate": 0.0009385772809245934,
      "loss": 2.22,
      "step": 2014
    },
    {
      "epoch": 0.01612,
      "grad_norm": 2.640625,
      "learning_rate": 0.0009385168013500495,
      "loss": 1.8839,
      "step": 2015
    },
    {
      "epoch": 0.016128,
      "grad_norm": 3.625,
      "learning_rate": 0.0009384562939652569,
      "loss": 2.3058,
      "step": 2016
    },
    {
      "epoch": 0.016136,
      "grad_norm": 3.171875,
      "learning_rate": 0.000938395758774053,
      "loss": 2.3271,
      "step": 2017
    },
    {
      "epoch": 0.016144,
      "grad_norm": 3.90625,
      "learning_rate": 0.0009383351957802769,
      "loss": 2.3144,
      "step": 2018
    },
    {
      "epoch": 0.016152,
      "grad_norm": 3.703125,
      "learning_rate": 0.0009382746049877693,
      "loss": 3.0411,
      "step": 2019
    },
    {
      "epoch": 0.01616,
      "grad_norm": 4.90625,
      "learning_rate": 0.0009382139864003726,
      "loss": 3.1926,
      "step": 2020
    },
    {
      "epoch": 0.016168,
      "grad_norm": 4.0625,
      "learning_rate": 0.0009381533400219318,
      "loss": 2.7928,
      "step": 2021
    },
    {
      "epoch": 0.016176,
      "grad_norm": 4.84375,
      "learning_rate": 0.0009380926658562925,
      "loss": 2.625,
      "step": 2022
    },
    {
      "epoch": 0.016184,
      "grad_norm": 5.125,
      "learning_rate": 0.0009380319639073027,
      "loss": 3.3456,
      "step": 2023
    },
    {
      "epoch": 0.016192,
      "grad_norm": 3.578125,
      "learning_rate": 0.0009379712341788122,
      "loss": 2.1623,
      "step": 2024
    },
    {
      "epoch": 0.0162,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009379104766746722,
      "loss": 2.4597,
      "step": 2025
    },
    {
      "epoch": 0.016208,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009378496913987362,
      "loss": 2.5024,
      "step": 2026
    },
    {
      "epoch": 0.016216,
      "grad_norm": 3.5,
      "learning_rate": 0.0009377888783548586,
      "loss": 2.5561,
      "step": 2027
    },
    {
      "epoch": 0.016224,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009377280375468967,
      "loss": 1.9854,
      "step": 2028
    },
    {
      "epoch": 0.016232,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009376671689787084,
      "loss": 2.6797,
      "step": 2029
    },
    {
      "epoch": 0.01624,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009376062726541543,
      "loss": 3.1037,
      "step": 2030
    },
    {
      "epoch": 0.016248,
      "grad_norm": 3.65625,
      "learning_rate": 0.0009375453485770961,
      "loss": 2.4399,
      "step": 2031
    },
    {
      "epoch": 0.016256,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009374843967513978,
      "loss": 2.4585,
      "step": 2032
    },
    {
      "epoch": 0.016264,
      "grad_norm": 16.125,
      "learning_rate": 0.0009374234171809248,
      "loss": 2.795,
      "step": 2033
    },
    {
      "epoch": 0.016272,
      "grad_norm": 3.375,
      "learning_rate": 0.0009373624098695443,
      "loss": 2.4325,
      "step": 2034
    },
    {
      "epoch": 0.01628,
      "grad_norm": 3.65625,
      "learning_rate": 0.0009373013748211255,
      "loss": 2.3227,
      "step": 2035
    },
    {
      "epoch": 0.016288,
      "grad_norm": 3.390625,
      "learning_rate": 0.000937240312039539,
      "loss": 2.2662,
      "step": 2036
    },
    {
      "epoch": 0.016296,
      "grad_norm": 3.46875,
      "learning_rate": 0.0009371792215286574,
      "loss": 3.3621,
      "step": 2037
    },
    {
      "epoch": 0.016304,
      "grad_norm": 10.6875,
      "learning_rate": 0.0009371181032923548,
      "loss": 2.4241,
      "step": 2038
    },
    {
      "epoch": 0.016312,
      "grad_norm": 3.25,
      "learning_rate": 0.0009370569573345076,
      "loss": 2.3139,
      "step": 2039
    },
    {
      "epoch": 0.01632,
      "grad_norm": 3.96875,
      "learning_rate": 0.0009369957836589933,
      "loss": 2.658,
      "step": 2040
    },
    {
      "epoch": 0.016328,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009369345822696915,
      "loss": 1.9932,
      "step": 2041
    },
    {
      "epoch": 0.016336,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009368733531704838,
      "loss": 2.025,
      "step": 2042
    },
    {
      "epoch": 0.016344,
      "grad_norm": 4.0625,
      "learning_rate": 0.000936812096365253,
      "loss": 2.7226,
      "step": 2043
    },
    {
      "epoch": 0.016352,
      "grad_norm": 14.625,
      "learning_rate": 0.0009367508118578841,
      "loss": 2.6992,
      "step": 2044
    },
    {
      "epoch": 0.01636,
      "grad_norm": 4.125,
      "learning_rate": 0.0009366894996522636,
      "loss": 3.2481,
      "step": 2045
    },
    {
      "epoch": 0.016368,
      "grad_norm": 3.671875,
      "learning_rate": 0.0009366281597522799,
      "loss": 2.5633,
      "step": 2046
    },
    {
      "epoch": 0.016376,
      "grad_norm": 3.921875,
      "learning_rate": 0.000936566792161823,
      "loss": 2.5563,
      "step": 2047
    },
    {
      "epoch": 0.016384,
      "grad_norm": 3.703125,
      "learning_rate": 0.0009365053968847849,
      "loss": 2.4693,
      "step": 2048
    },
    {
      "epoch": 0.016392,
      "grad_norm": 4.0,
      "learning_rate": 0.0009364439739250592,
      "loss": 2.3194,
      "step": 2049
    },
    {
      "epoch": 0.0164,
      "grad_norm": 3.671875,
      "learning_rate": 0.0009363825232865413,
      "loss": 2.9378,
      "step": 2050
    },
    {
      "epoch": 0.016408,
      "grad_norm": 3.859375,
      "learning_rate": 0.0009363210449731281,
      "loss": 2.453,
      "step": 2051
    },
    {
      "epoch": 0.016416,
      "grad_norm": 4.6875,
      "learning_rate": 0.0009362595389887188,
      "loss": 3.031,
      "step": 2052
    },
    {
      "epoch": 0.016424,
      "grad_norm": 4.0625,
      "learning_rate": 0.0009361980053372139,
      "loss": 2.5704,
      "step": 2053
    },
    {
      "epoch": 0.016432,
      "grad_norm": 4.84375,
      "learning_rate": 0.0009361364440225159,
      "loss": 3.0753,
      "step": 2054
    },
    {
      "epoch": 0.01644,
      "grad_norm": 3.3125,
      "learning_rate": 0.0009360748550485288,
      "loss": 2.5644,
      "step": 2055
    },
    {
      "epoch": 0.016448,
      "grad_norm": 3.703125,
      "learning_rate": 0.0009360132384191584,
      "loss": 2.4295,
      "step": 2056
    },
    {
      "epoch": 0.016456,
      "grad_norm": 12.625,
      "learning_rate": 0.0009359515941383126,
      "loss": 2.4892,
      "step": 2057
    },
    {
      "epoch": 0.016464,
      "grad_norm": 3.84375,
      "learning_rate": 0.0009358899222099009,
      "loss": 1.9385,
      "step": 2058
    },
    {
      "epoch": 0.016472,
      "grad_norm": 3.375,
      "learning_rate": 0.000935828222637834,
      "loss": 2.1664,
      "step": 2059
    },
    {
      "epoch": 0.01648,
      "grad_norm": 4.09375,
      "learning_rate": 0.0009357664954260253,
      "loss": 2.3635,
      "step": 2060
    },
    {
      "epoch": 0.016488,
      "grad_norm": 3.578125,
      "learning_rate": 0.0009357047405783893,
      "loss": 2.5212,
      "step": 2061
    },
    {
      "epoch": 0.016496,
      "grad_norm": 3.46875,
      "learning_rate": 0.0009356429580988423,
      "loss": 2.2528,
      "step": 2062
    },
    {
      "epoch": 0.016504,
      "grad_norm": 3.359375,
      "learning_rate": 0.0009355811479913028,
      "loss": 2.3958,
      "step": 2063
    },
    {
      "epoch": 0.016512,
      "grad_norm": 4.78125,
      "learning_rate": 0.0009355193102596903,
      "loss": 2.2866,
      "step": 2064
    },
    {
      "epoch": 0.01652,
      "grad_norm": 4.40625,
      "learning_rate": 0.0009354574449079267,
      "loss": 2.8604,
      "step": 2065
    },
    {
      "epoch": 0.016528,
      "grad_norm": 4.625,
      "learning_rate": 0.0009353955519399355,
      "loss": 2.4946,
      "step": 2066
    },
    {
      "epoch": 0.016536,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009353336313596419,
      "loss": 2.1779,
      "step": 2067
    },
    {
      "epoch": 0.016544,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009352716831709726,
      "loss": 2.2651,
      "step": 2068
    },
    {
      "epoch": 0.016552,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009352097073778564,
      "loss": 2.3193,
      "step": 2069
    },
    {
      "epoch": 0.01656,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009351477039842238,
      "loss": 2.5496,
      "step": 2070
    },
    {
      "epoch": 0.016568,
      "grad_norm": 5.15625,
      "learning_rate": 0.000935085672994007,
      "loss": 3.2691,
      "step": 2071
    },
    {
      "epoch": 0.016576,
      "grad_norm": 5.25,
      "learning_rate": 0.0009350236144111399,
      "loss": 2.8406,
      "step": 2072
    },
    {
      "epoch": 0.016584,
      "grad_norm": 4.34375,
      "learning_rate": 0.000934961528239558,
      "loss": 2.7381,
      "step": 2073
    },
    {
      "epoch": 0.016592,
      "grad_norm": 4.84375,
      "learning_rate": 0.0009348994144831992,
      "loss": 2.4841,
      "step": 2074
    },
    {
      "epoch": 0.0166,
      "grad_norm": 4.09375,
      "learning_rate": 0.0009348372731460022,
      "loss": 1.8333,
      "step": 2075
    },
    {
      "epoch": 0.016608,
      "grad_norm": 4.125,
      "learning_rate": 0.0009347751042319081,
      "loss": 2.6566,
      "step": 2076
    },
    {
      "epoch": 0.016616,
      "grad_norm": 4.3125,
      "learning_rate": 0.0009347129077448597,
      "loss": 3.2214,
      "step": 2077
    },
    {
      "epoch": 0.016624,
      "grad_norm": 10.9375,
      "learning_rate": 0.0009346506836888013,
      "loss": 2.6956,
      "step": 2078
    },
    {
      "epoch": 0.016632,
      "grad_norm": 3.484375,
      "learning_rate": 0.0009345884320676791,
      "loss": 2.8038,
      "step": 2079
    },
    {
      "epoch": 0.01664,
      "grad_norm": 3.53125,
      "learning_rate": 0.0009345261528854412,
      "loss": 2.38,
      "step": 2080
    },
    {
      "epoch": 0.016648,
      "grad_norm": 3.296875,
      "learning_rate": 0.0009344638461460371,
      "loss": 2.988,
      "step": 2081
    },
    {
      "epoch": 0.016656,
      "grad_norm": 3.5,
      "learning_rate": 0.0009344015118534181,
      "loss": 3.2494,
      "step": 2082
    },
    {
      "epoch": 0.016664,
      "grad_norm": 3.671875,
      "learning_rate": 0.0009343391500115377,
      "loss": 2.9925,
      "step": 2083
    },
    {
      "epoch": 0.016672,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009342767606243506,
      "loss": 2.048,
      "step": 2084
    },
    {
      "epoch": 0.01668,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009342143436958135,
      "loss": 2.3595,
      "step": 2085
    },
    {
      "epoch": 0.016688,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009341518992298847,
      "loss": 2.7399,
      "step": 2086
    },
    {
      "epoch": 0.016696,
      "grad_norm": 3.140625,
      "learning_rate": 0.0009340894272305245,
      "loss": 2.9021,
      "step": 2087
    },
    {
      "epoch": 0.016704,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009340269277016951,
      "loss": 2.2788,
      "step": 2088
    },
    {
      "epoch": 0.016712,
      "grad_norm": 3.90625,
      "learning_rate": 0.0009339644006473596,
      "loss": 2.41,
      "step": 2089
    },
    {
      "epoch": 0.01672,
      "grad_norm": 4.25,
      "learning_rate": 0.0009339018460714835,
      "loss": 3.3685,
      "step": 2090
    },
    {
      "epoch": 0.016728,
      "grad_norm": 3.40625,
      "learning_rate": 0.0009338392639780342,
      "loss": 2.0549,
      "step": 2091
    },
    {
      "epoch": 0.016736,
      "grad_norm": 2.65625,
      "learning_rate": 0.0009337766543709806,
      "loss": 2.131,
      "step": 2092
    },
    {
      "epoch": 0.016744,
      "grad_norm": 5.46875,
      "learning_rate": 0.000933714017254293,
      "loss": 3.6759,
      "step": 2093
    },
    {
      "epoch": 0.016752,
      "grad_norm": 3.65625,
      "learning_rate": 0.0009336513526319441,
      "loss": 2.4974,
      "step": 2094
    },
    {
      "epoch": 0.01676,
      "grad_norm": 3.015625,
      "learning_rate": 0.0009335886605079078,
      "loss": 2.2813,
      "step": 2095
    },
    {
      "epoch": 0.016768,
      "grad_norm": 4.09375,
      "learning_rate": 0.00093352594088616,
      "loss": 2.1836,
      "step": 2096
    },
    {
      "epoch": 0.016776,
      "grad_norm": 3.1875,
      "learning_rate": 0.0009334631937706786,
      "loss": 2.1425,
      "step": 2097
    },
    {
      "epoch": 0.016784,
      "grad_norm": 3.65625,
      "learning_rate": 0.0009334004191654426,
      "loss": 2.3551,
      "step": 2098
    },
    {
      "epoch": 0.016792,
      "grad_norm": 4.4375,
      "learning_rate": 0.0009333376170744332,
      "loss": 2.4509,
      "step": 2099
    },
    {
      "epoch": 0.0168,
      "grad_norm": 6.125,
      "learning_rate": 0.0009332747875016332,
      "loss": 2.8729,
      "step": 2100
    },
    {
      "epoch": 0.016808,
      "grad_norm": 9.1875,
      "learning_rate": 0.0009332119304510274,
      "loss": 2.4937,
      "step": 2101
    },
    {
      "epoch": 0.016816,
      "grad_norm": 9.375,
      "learning_rate": 0.0009331490459266018,
      "loss": 2.3177,
      "step": 2102
    },
    {
      "epoch": 0.016824,
      "grad_norm": 6.96875,
      "learning_rate": 0.0009330861339323447,
      "loss": 1.857,
      "step": 2103
    },
    {
      "epoch": 0.016832,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009330231944722459,
      "loss": 3.2133,
      "step": 2104
    },
    {
      "epoch": 0.01684,
      "grad_norm": 7.0625,
      "learning_rate": 0.0009329602275502967,
      "loss": 2.6366,
      "step": 2105
    },
    {
      "epoch": 0.016848,
      "grad_norm": 16.5,
      "learning_rate": 0.0009328972331704905,
      "loss": 2.7524,
      "step": 2106
    },
    {
      "epoch": 0.016856,
      "grad_norm": 226.0,
      "learning_rate": 0.0009328342113368226,
      "loss": 2.3661,
      "step": 2107
    },
    {
      "epoch": 0.016864,
      "grad_norm": 246.0,
      "learning_rate": 0.0009327711620532897,
      "loss": 2.4512,
      "step": 2108
    },
    {
      "epoch": 0.016872,
      "grad_norm": 8256.0,
      "learning_rate": 0.0009327080853238901,
      "loss": 2.2009,
      "step": 2109
    },
    {
      "epoch": 0.01688,
      "grad_norm": 2992.0,
      "learning_rate": 0.000932644981152624,
      "loss": 2.8088,
      "step": 2110
    },
    {
      "epoch": 0.016888,
      "grad_norm": 9600.0,
      "learning_rate": 0.0009325818495434936,
      "loss": 2.5808,
      "step": 2111
    },
    {
      "epoch": 0.016896,
      "grad_norm": 2336.0,
      "learning_rate": 0.0009325186905005027,
      "loss": 3.0399,
      "step": 2112
    },
    {
      "epoch": 0.016904,
      "grad_norm": 71.5,
      "learning_rate": 0.0009324555040276567,
      "loss": 3.1409,
      "step": 2113
    },
    {
      "epoch": 0.016912,
      "grad_norm": 4672.0,
      "learning_rate": 0.0009323922901289627,
      "loss": 3.2622,
      "step": 2114
    },
    {
      "epoch": 0.01692,
      "grad_norm": 19.125,
      "learning_rate": 0.0009323290488084297,
      "loss": 4.2641,
      "step": 2115
    },
    {
      "epoch": 0.016928,
      "grad_norm": 10752.0,
      "learning_rate": 0.0009322657800700687,
      "loss": 3.3647,
      "step": 2116
    },
    {
      "epoch": 0.016936,
      "grad_norm": 15.75,
      "learning_rate": 0.0009322024839178916,
      "loss": 3.8015,
      "step": 2117
    },
    {
      "epoch": 0.016944,
      "grad_norm": 118.0,
      "learning_rate": 0.0009321391603559128,
      "loss": 2.028,
      "step": 2118
    },
    {
      "epoch": 0.016952,
      "grad_norm": 1688.0,
      "learning_rate": 0.0009320758093881483,
      "loss": 3.1056,
      "step": 2119
    },
    {
      "epoch": 0.01696,
      "grad_norm": 171.0,
      "learning_rate": 0.0009320124310186157,
      "loss": 3.0985,
      "step": 2120
    },
    {
      "epoch": 0.016968,
      "grad_norm": 2096.0,
      "learning_rate": 0.0009319490252513345,
      "loss": 3.2333,
      "step": 2121
    },
    {
      "epoch": 0.016976,
      "grad_norm": 904.0,
      "learning_rate": 0.0009318855920903257,
      "loss": 3.5808,
      "step": 2122
    },
    {
      "epoch": 0.016984,
      "grad_norm": 199.0,
      "learning_rate": 0.000931822131539612,
      "loss": 2.9975,
      "step": 2123
    },
    {
      "epoch": 0.016992,
      "grad_norm": 29.0,
      "learning_rate": 0.0009317586436032182,
      "loss": 3.1559,
      "step": 2124
    },
    {
      "epoch": 0.017,
      "grad_norm": 179.0,
      "learning_rate": 0.0009316951282851706,
      "loss": 3.1939,
      "step": 2125
    },
    {
      "epoch": 0.017008,
      "grad_norm": 7.0,
      "learning_rate": 0.0009316315855894975,
      "loss": 3.109,
      "step": 2126
    },
    {
      "epoch": 0.017016,
      "grad_norm": 40.25,
      "learning_rate": 0.0009315680155202281,
      "loss": 3.2259,
      "step": 2127
    },
    {
      "epoch": 0.017024,
      "grad_norm": 7.59375,
      "learning_rate": 0.0009315044180813946,
      "loss": 3.9242,
      "step": 2128
    },
    {
      "epoch": 0.017032,
      "grad_norm": 5.375,
      "learning_rate": 0.00093144079327703,
      "loss": 3.4375,
      "step": 2129
    },
    {
      "epoch": 0.01704,
      "grad_norm": 4.40625,
      "learning_rate": 0.0009313771411111691,
      "loss": 2.8263,
      "step": 2130
    },
    {
      "epoch": 0.017048,
      "grad_norm": 3.765625,
      "learning_rate": 0.000931313461587849,
      "loss": 2.468,
      "step": 2131
    },
    {
      "epoch": 0.017056,
      "grad_norm": 3.59375,
      "learning_rate": 0.000931249754711108,
      "loss": 2.2882,
      "step": 2132
    },
    {
      "epoch": 0.017064,
      "grad_norm": 16.875,
      "learning_rate": 0.0009311860204849865,
      "loss": 2.4905,
      "step": 2133
    },
    {
      "epoch": 0.017072,
      "grad_norm": 18.625,
      "learning_rate": 0.0009311222589135262,
      "loss": 2.5819,
      "step": 2134
    },
    {
      "epoch": 0.01708,
      "grad_norm": 5.03125,
      "learning_rate": 0.0009310584700007708,
      "loss": 2.3318,
      "step": 2135
    },
    {
      "epoch": 0.017088,
      "grad_norm": 7.1875,
      "learning_rate": 0.000930994653750766,
      "loss": 3.006,
      "step": 2136
    },
    {
      "epoch": 0.017096,
      "grad_norm": 10.5625,
      "learning_rate": 0.0009309308101675586,
      "loss": 2.8428,
      "step": 2137
    },
    {
      "epoch": 0.017104,
      "grad_norm": 5.03125,
      "learning_rate": 0.000930866939255198,
      "loss": 2.503,
      "step": 2138
    },
    {
      "epoch": 0.017112,
      "grad_norm": 22.875,
      "learning_rate": 0.0009308030410177341,
      "loss": 1.8549,
      "step": 2139
    },
    {
      "epoch": 0.01712,
      "grad_norm": 6.21875,
      "learning_rate": 0.00093073911545922,
      "loss": 3.0851,
      "step": 2140
    },
    {
      "epoch": 0.017128,
      "grad_norm": 5.1875,
      "learning_rate": 0.0009306751625837094,
      "loss": 3.2023,
      "step": 2141
    },
    {
      "epoch": 0.017136,
      "grad_norm": 5.96875,
      "learning_rate": 0.0009306111823952579,
      "loss": 2.5284,
      "step": 2142
    },
    {
      "epoch": 0.017144,
      "grad_norm": 5.8125,
      "learning_rate": 0.0009305471748979235,
      "loss": 2.128,
      "step": 2143
    },
    {
      "epoch": 0.017152,
      "grad_norm": 6.25,
      "learning_rate": 0.0009304831400957652,
      "loss": 2.203,
      "step": 2144
    },
    {
      "epoch": 0.01716,
      "grad_norm": 4.625,
      "learning_rate": 0.0009304190779928443,
      "loss": 2.815,
      "step": 2145
    },
    {
      "epoch": 0.017168,
      "grad_norm": 3.671875,
      "learning_rate": 0.0009303549885932232,
      "loss": 3.0454,
      "step": 2146
    },
    {
      "epoch": 0.017176,
      "grad_norm": 10.125,
      "learning_rate": 0.0009302908719009666,
      "loss": 3.0005,
      "step": 2147
    },
    {
      "epoch": 0.017184,
      "grad_norm": 3.1875,
      "learning_rate": 0.0009302267279201406,
      "loss": 3.1673,
      "step": 2148
    },
    {
      "epoch": 0.017192,
      "grad_norm": 5.28125,
      "learning_rate": 0.0009301625566548132,
      "loss": 2.68,
      "step": 2149
    },
    {
      "epoch": 0.0172,
      "grad_norm": 83.5,
      "learning_rate": 0.0009300983581090541,
      "loss": 2.8563,
      "step": 2150
    },
    {
      "epoch": 0.017208,
      "grad_norm": 73.0,
      "learning_rate": 0.0009300341322869346,
      "loss": 2.9367,
      "step": 2151
    },
    {
      "epoch": 0.017216,
      "grad_norm": 3.75,
      "learning_rate": 0.0009299698791925279,
      "loss": 2.3506,
      "step": 2152
    },
    {
      "epoch": 0.017224,
      "grad_norm": 43.0,
      "learning_rate": 0.000929905598829909,
      "loss": 2.4594,
      "step": 2153
    },
    {
      "epoch": 0.017232,
      "grad_norm": 4.46875,
      "learning_rate": 0.0009298412912031542,
      "loss": 3.2487,
      "step": 2154
    },
    {
      "epoch": 0.01724,
      "grad_norm": 12.875,
      "learning_rate": 0.000929776956316342,
      "loss": 2.5846,
      "step": 2155
    },
    {
      "epoch": 0.017248,
      "grad_norm": 3.921875,
      "learning_rate": 0.0009297125941735522,
      "loss": 2.3342,
      "step": 2156
    },
    {
      "epoch": 0.017256,
      "grad_norm": 18.625,
      "learning_rate": 0.000929648204778867,
      "loss": 2.3348,
      "step": 2157
    },
    {
      "epoch": 0.017264,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009295837881363696,
      "loss": 2.5302,
      "step": 2158
    },
    {
      "epoch": 0.017272,
      "grad_norm": 3.890625,
      "learning_rate": 0.0009295193442501455,
      "loss": 2.1917,
      "step": 2159
    },
    {
      "epoch": 0.01728,
      "grad_norm": 3.828125,
      "learning_rate": 0.0009294548731242813,
      "loss": 3.1456,
      "step": 2160
    },
    {
      "epoch": 0.017288,
      "grad_norm": 4.09375,
      "learning_rate": 0.0009293903747628659,
      "loss": 2.2481,
      "step": 2161
    },
    {
      "epoch": 0.017296,
      "grad_norm": 4.0625,
      "learning_rate": 0.0009293258491699896,
      "loss": 2.4385,
      "step": 2162
    },
    {
      "epoch": 0.017304,
      "grad_norm": 3.4375,
      "learning_rate": 0.0009292612963497448,
      "loss": 3.0649,
      "step": 2163
    },
    {
      "epoch": 0.017312,
      "grad_norm": 4.21875,
      "learning_rate": 0.000929196716306225,
      "loss": 2.5103,
      "step": 2164
    },
    {
      "epoch": 0.01732,
      "grad_norm": 3.296875,
      "learning_rate": 0.0009291321090435261,
      "loss": 2.5761,
      "step": 2165
    },
    {
      "epoch": 0.017328,
      "grad_norm": 3.8125,
      "learning_rate": 0.0009290674745657452,
      "loss": 3.079,
      "step": 2166
    },
    {
      "epoch": 0.017336,
      "grad_norm": 3.640625,
      "learning_rate": 0.0009290028128769816,
      "loss": 2.2416,
      "step": 2167
    },
    {
      "epoch": 0.017344,
      "grad_norm": 2.953125,
      "learning_rate": 0.000928938123981336,
      "loss": 2.9053,
      "step": 2168
    },
    {
      "epoch": 0.017352,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009288734078829105,
      "loss": 2.6221,
      "step": 2169
    },
    {
      "epoch": 0.01736,
      "grad_norm": 3.4375,
      "learning_rate": 0.0009288086645858098,
      "loss": 2.5606,
      "step": 2170
    },
    {
      "epoch": 0.017368,
      "grad_norm": 3.234375,
      "learning_rate": 0.0009287438940941398,
      "loss": 2.3412,
      "step": 2171
    },
    {
      "epoch": 0.017376,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009286790964120079,
      "loss": 2.2654,
      "step": 2172
    },
    {
      "epoch": 0.017384,
      "grad_norm": 3.578125,
      "learning_rate": 0.0009286142715435237,
      "loss": 2.6602,
      "step": 2173
    },
    {
      "epoch": 0.017392,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009285494194927983,
      "loss": 2.6552,
      "step": 2174
    },
    {
      "epoch": 0.0174,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009284845402639446,
      "loss": 2.1998,
      "step": 2175
    },
    {
      "epoch": 0.017408,
      "grad_norm": 3.421875,
      "learning_rate": 0.0009284196338610769,
      "loss": 2.8249,
      "step": 2176
    },
    {
      "epoch": 0.017416,
      "grad_norm": 27.5,
      "learning_rate": 0.0009283547002883119,
      "loss": 2.8347,
      "step": 2177
    },
    {
      "epoch": 0.017424,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009282897395497672,
      "loss": 2.2427,
      "step": 2178
    },
    {
      "epoch": 0.017432,
      "grad_norm": 4.4375,
      "learning_rate": 0.000928224751649563,
      "loss": 2.7544,
      "step": 2179
    },
    {
      "epoch": 0.01744,
      "grad_norm": 3.515625,
      "learning_rate": 0.0009281597365918202,
      "loss": 2.2296,
      "step": 2180
    },
    {
      "epoch": 0.017448,
      "grad_norm": 11.8125,
      "learning_rate": 0.0009280946943806625,
      "loss": 3.0469,
      "step": 2181
    },
    {
      "epoch": 0.017456,
      "grad_norm": 3.234375,
      "learning_rate": 0.0009280296250202144,
      "loss": 2.1292,
      "step": 2182
    },
    {
      "epoch": 0.017464,
      "grad_norm": 4.375,
      "learning_rate": 0.000927964528514603,
      "loss": 2.8419,
      "step": 2183
    },
    {
      "epoch": 0.017472,
      "grad_norm": 4.1875,
      "learning_rate": 0.0009278994048679563,
      "loss": 2.2281,
      "step": 2184
    },
    {
      "epoch": 0.01748,
      "grad_norm": 3.890625,
      "learning_rate": 0.0009278342540844042,
      "loss": 2.8262,
      "step": 2185
    },
    {
      "epoch": 0.017488,
      "grad_norm": 4.9375,
      "learning_rate": 0.0009277690761680788,
      "loss": 2.6041,
      "step": 2186
    },
    {
      "epoch": 0.017496,
      "grad_norm": 5.75,
      "learning_rate": 0.0009277038711231137,
      "loss": 2.9476,
      "step": 2187
    },
    {
      "epoch": 0.017504,
      "grad_norm": 4.625,
      "learning_rate": 0.0009276386389536438,
      "loss": 2.6236,
      "step": 2188
    },
    {
      "epoch": 0.017512,
      "grad_norm": 3.8125,
      "learning_rate": 0.0009275733796638064,
      "loss": 3.3677,
      "step": 2189
    },
    {
      "epoch": 0.01752,
      "grad_norm": 4.21875,
      "learning_rate": 0.0009275080932577398,
      "loss": 3.0498,
      "step": 2190
    },
    {
      "epoch": 0.017528,
      "grad_norm": 3.609375,
      "learning_rate": 0.0009274427797395846,
      "loss": 2.908,
      "step": 2191
    },
    {
      "epoch": 0.017536,
      "grad_norm": 3.359375,
      "learning_rate": 0.000927377439113483,
      "loss": 2.0476,
      "step": 2192
    },
    {
      "epoch": 0.017544,
      "grad_norm": 3.78125,
      "learning_rate": 0.0009273120713835786,
      "loss": 2.6328,
      "step": 2193
    },
    {
      "epoch": 0.017552,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009272466765540172,
      "loss": 3.1633,
      "step": 2194
    },
    {
      "epoch": 0.01756,
      "grad_norm": 3.125,
      "learning_rate": 0.0009271812546289459,
      "loss": 2.7775,
      "step": 2195
    },
    {
      "epoch": 0.017568,
      "grad_norm": 3.015625,
      "learning_rate": 0.0009271158056125137,
      "loss": 2.6394,
      "step": 2196
    },
    {
      "epoch": 0.017576,
      "grad_norm": 3.390625,
      "learning_rate": 0.0009270503295088713,
      "loss": 3.4154,
      "step": 2197
    },
    {
      "epoch": 0.017584,
      "grad_norm": 3.34375,
      "learning_rate": 0.0009269848263221712,
      "loss": 2.6796,
      "step": 2198
    },
    {
      "epoch": 0.017592,
      "grad_norm": 3.34375,
      "learning_rate": 0.0009269192960565674,
      "loss": 2.4133,
      "step": 2199
    },
    {
      "epoch": 0.0176,
      "grad_norm": 2.90625,
      "learning_rate": 0.000926853738716216,
      "loss": 2.6611,
      "step": 2200
    },
    {
      "epoch": 0.017608,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009267881543052743,
      "loss": 1.8857,
      "step": 2201
    },
    {
      "epoch": 0.017616,
      "grad_norm": 4.125,
      "learning_rate": 0.0009267225428279018,
      "loss": 2.8207,
      "step": 2202
    },
    {
      "epoch": 0.017624,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009266569042882593,
      "loss": 2.5235,
      "step": 2203
    },
    {
      "epoch": 0.017632,
      "grad_norm": 3.78125,
      "learning_rate": 0.0009265912386905097,
      "loss": 3.9373,
      "step": 2204
    },
    {
      "epoch": 0.01764,
      "grad_norm": 3.765625,
      "learning_rate": 0.0009265255460388173,
      "loss": 2.4131,
      "step": 2205
    },
    {
      "epoch": 0.017648,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009264598263373485,
      "loss": 2.4032,
      "step": 2206
    },
    {
      "epoch": 0.017656,
      "grad_norm": 4.59375,
      "learning_rate": 0.0009263940795902708,
      "loss": 2.7719,
      "step": 2207
    },
    {
      "epoch": 0.017664,
      "grad_norm": 3.859375,
      "learning_rate": 0.0009263283058017542,
      "loss": 2.7663,
      "step": 2208
    },
    {
      "epoch": 0.017672,
      "grad_norm": 3.375,
      "learning_rate": 0.0009262625049759696,
      "loss": 2.2653,
      "step": 2209
    },
    {
      "epoch": 0.01768,
      "grad_norm": 3.796875,
      "learning_rate": 0.0009261966771170905,
      "loss": 2.098,
      "step": 2210
    },
    {
      "epoch": 0.017688,
      "grad_norm": 3.71875,
      "learning_rate": 0.000926130822229291,
      "loss": 2.4269,
      "step": 2211
    },
    {
      "epoch": 0.017696,
      "grad_norm": 4.25,
      "learning_rate": 0.0009260649403167481,
      "loss": 3.5167,
      "step": 2212
    },
    {
      "epoch": 0.017704,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009259990313836397,
      "loss": 2.3059,
      "step": 2213
    },
    {
      "epoch": 0.017712,
      "grad_norm": 3.859375,
      "learning_rate": 0.0009259330954341456,
      "loss": 2.6145,
      "step": 2214
    },
    {
      "epoch": 0.01772,
      "grad_norm": 4.6875,
      "learning_rate": 0.0009258671324724474,
      "loss": 3.0465,
      "step": 2215
    },
    {
      "epoch": 0.017728,
      "grad_norm": 3.515625,
      "learning_rate": 0.0009258011425027289,
      "loss": 2.3148,
      "step": 2216
    },
    {
      "epoch": 0.017736,
      "grad_norm": 3.359375,
      "learning_rate": 0.0009257351255291744,
      "loss": 2.6897,
      "step": 2217
    },
    {
      "epoch": 0.017744,
      "grad_norm": 3.234375,
      "learning_rate": 0.0009256690815559709,
      "loss": 2.5784,
      "step": 2218
    },
    {
      "epoch": 0.017752,
      "grad_norm": 3.515625,
      "learning_rate": 0.000925603010587307,
      "loss": 2.4038,
      "step": 2219
    },
    {
      "epoch": 0.01776,
      "grad_norm": 4.0625,
      "learning_rate": 0.0009255369126273726,
      "loss": 2.9252,
      "step": 2220
    },
    {
      "epoch": 0.017768,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009254707876803597,
      "loss": 2.6204,
      "step": 2221
    },
    {
      "epoch": 0.017776,
      "grad_norm": 3.828125,
      "learning_rate": 0.0009254046357504619,
      "loss": 2.4755,
      "step": 2222
    },
    {
      "epoch": 0.017784,
      "grad_norm": 4.15625,
      "learning_rate": 0.0009253384568418744,
      "loss": 2.5767,
      "step": 2223
    },
    {
      "epoch": 0.017792,
      "grad_norm": 4.78125,
      "learning_rate": 0.0009252722509587941,
      "loss": 2.6855,
      "step": 2224
    },
    {
      "epoch": 0.0178,
      "grad_norm": 3.59375,
      "learning_rate": 0.00092520601810542,
      "loss": 2.7832,
      "step": 2225
    },
    {
      "epoch": 0.017808,
      "grad_norm": 3.640625,
      "learning_rate": 0.0009251397582859522,
      "loss": 2.7348,
      "step": 2226
    },
    {
      "epoch": 0.017816,
      "grad_norm": 3.0625,
      "learning_rate": 0.000925073471504593,
      "loss": 2.504,
      "step": 2227
    },
    {
      "epoch": 0.017824,
      "grad_norm": 6.0,
      "learning_rate": 0.0009250071577655461,
      "loss": 3.1314,
      "step": 2228
    },
    {
      "epoch": 0.017832,
      "grad_norm": 4.46875,
      "learning_rate": 0.0009249408170730174,
      "loss": 2.4998,
      "step": 2229
    },
    {
      "epoch": 0.01784,
      "grad_norm": 3.5,
      "learning_rate": 0.0009248744494312136,
      "loss": 2.4687,
      "step": 2230
    },
    {
      "epoch": 0.017848,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009248080548443441,
      "loss": 2.8608,
      "step": 2231
    },
    {
      "epoch": 0.017856,
      "grad_norm": 3.984375,
      "learning_rate": 0.0009247416333166193,
      "loss": 3.0936,
      "step": 2232
    },
    {
      "epoch": 0.017864,
      "grad_norm": 4.28125,
      "learning_rate": 0.0009246751848522518,
      "loss": 2.8581,
      "step": 2233
    },
    {
      "epoch": 0.017872,
      "grad_norm": 3.9375,
      "learning_rate": 0.0009246087094554556,
      "loss": 2.7502,
      "step": 2234
    },
    {
      "epoch": 0.01788,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009245422071304464,
      "loss": 2.353,
      "step": 2235
    },
    {
      "epoch": 0.017888,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009244756778814419,
      "loss": 2.9249,
      "step": 2236
    },
    {
      "epoch": 0.017896,
      "grad_norm": 3.78125,
      "learning_rate": 0.000924409121712661,
      "loss": 2.3732,
      "step": 2237
    },
    {
      "epoch": 0.017904,
      "grad_norm": 5.28125,
      "learning_rate": 0.000924342538628325,
      "loss": 2.9259,
      "step": 2238
    },
    {
      "epoch": 0.017912,
      "grad_norm": 4.25,
      "learning_rate": 0.0009242759286326564,
      "loss": 2.6797,
      "step": 2239
    },
    {
      "epoch": 0.01792,
      "grad_norm": 4.25,
      "learning_rate": 0.0009242092917298792,
      "loss": 3.1398,
      "step": 2240
    },
    {
      "epoch": 0.017928,
      "grad_norm": 4.96875,
      "learning_rate": 0.0009241426279242199,
      "loss": 2.8393,
      "step": 2241
    },
    {
      "epoch": 0.017936,
      "grad_norm": 3.796875,
      "learning_rate": 0.0009240759372199059,
      "loss": 2.8723,
      "step": 2242
    },
    {
      "epoch": 0.017944,
      "grad_norm": 3.25,
      "learning_rate": 0.0009240092196211669,
      "loss": 2.4302,
      "step": 2243
    },
    {
      "epoch": 0.017952,
      "grad_norm": 3.875,
      "learning_rate": 0.0009239424751322341,
      "loss": 2.7938,
      "step": 2244
    },
    {
      "epoch": 0.01796,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009238757037573401,
      "loss": 2.2891,
      "step": 2245
    },
    {
      "epoch": 0.017968,
      "grad_norm": 5.0,
      "learning_rate": 0.0009238089055007195,
      "loss": 2.5019,
      "step": 2246
    },
    {
      "epoch": 0.017976,
      "grad_norm": 3.734375,
      "learning_rate": 0.0009237420803666087,
      "loss": 2.6066,
      "step": 2247
    },
    {
      "epoch": 0.017984,
      "grad_norm": 4.15625,
      "learning_rate": 0.0009236752283592458,
      "loss": 3.5008,
      "step": 2248
    },
    {
      "epoch": 0.017992,
      "grad_norm": 3.515625,
      "learning_rate": 0.0009236083494828702,
      "loss": 1.9688,
      "step": 2249
    },
    {
      "epoch": 0.018,
      "grad_norm": 2.78125,
      "learning_rate": 0.0009235414437417234,
      "loss": 2.742,
      "step": 2250
    },
    {
      "epoch": 0.018008,
      "grad_norm": 3.359375,
      "learning_rate": 0.0009234745111400486,
      "loss": 2.3685,
      "step": 2251
    },
    {
      "epoch": 0.018016,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009234075516820903,
      "loss": 2.7371,
      "step": 2252
    },
    {
      "epoch": 0.018024,
      "grad_norm": 4.25,
      "learning_rate": 0.0009233405653720953,
      "loss": 2.82,
      "step": 2253
    },
    {
      "epoch": 0.018032,
      "grad_norm": 3.40625,
      "learning_rate": 0.0009232735522143117,
      "loss": 2.6189,
      "step": 2254
    },
    {
      "epoch": 0.01804,
      "grad_norm": 3.359375,
      "learning_rate": 0.0009232065122129892,
      "loss": 2.4723,
      "step": 2255
    },
    {
      "epoch": 0.018048,
      "grad_norm": 3.0625,
      "learning_rate": 0.0009231394453723797,
      "loss": 3.1268,
      "step": 2256
    },
    {
      "epoch": 0.018056,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009230723516967363,
      "loss": 2.4574,
      "step": 2257
    },
    {
      "epoch": 0.018064,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009230052311903143,
      "loss": 2.7475,
      "step": 2258
    },
    {
      "epoch": 0.018072,
      "grad_norm": 3.515625,
      "learning_rate": 0.00092293808385737,
      "loss": 2.9969,
      "step": 2259
    },
    {
      "epoch": 0.01808,
      "grad_norm": 4.6875,
      "learning_rate": 0.0009228709097021619,
      "loss": 2.469,
      "step": 2260
    },
    {
      "epoch": 0.018088,
      "grad_norm": 3.4375,
      "learning_rate": 0.0009228037087289505,
      "loss": 1.973,
      "step": 2261
    },
    {
      "epoch": 0.018096,
      "grad_norm": 5.09375,
      "learning_rate": 0.0009227364809419971,
      "loss": 2.7187,
      "step": 2262
    },
    {
      "epoch": 0.018104,
      "grad_norm": 3.84375,
      "learning_rate": 0.0009226692263455656,
      "loss": 2.8224,
      "step": 2263
    },
    {
      "epoch": 0.018112,
      "grad_norm": 5.0,
      "learning_rate": 0.000922601944943921,
      "loss": 2.8971,
      "step": 2264
    },
    {
      "epoch": 0.01812,
      "grad_norm": 4.3125,
      "learning_rate": 0.00092253463674133,
      "loss": 2.2827,
      "step": 2265
    },
    {
      "epoch": 0.018128,
      "grad_norm": 4.125,
      "learning_rate": 0.0009224673017420617,
      "loss": 2.4297,
      "step": 2266
    },
    {
      "epoch": 0.018136,
      "grad_norm": 5.875,
      "learning_rate": 0.0009223999399503863,
      "loss": 2.6724,
      "step": 2267
    },
    {
      "epoch": 0.018144,
      "grad_norm": 3.34375,
      "learning_rate": 0.0009223325513705756,
      "loss": 2.8771,
      "step": 2268
    },
    {
      "epoch": 0.018152,
      "grad_norm": 3.78125,
      "learning_rate": 0.0009222651360069034,
      "loss": 2.1285,
      "step": 2269
    },
    {
      "epoch": 0.01816,
      "grad_norm": 3.515625,
      "learning_rate": 0.000922197693863645,
      "loss": 3.0716,
      "step": 2270
    },
    {
      "epoch": 0.018168,
      "grad_norm": 4.0625,
      "learning_rate": 0.0009221302249450777,
      "loss": 3.024,
      "step": 2271
    },
    {
      "epoch": 0.018176,
      "grad_norm": 2.859375,
      "learning_rate": 0.0009220627292554802,
      "loss": 2.7266,
      "step": 2272
    },
    {
      "epoch": 0.018184,
      "grad_norm": 3.40625,
      "learning_rate": 0.000921995206799133,
      "loss": 2.7752,
      "step": 2273
    },
    {
      "epoch": 0.018192,
      "grad_norm": 2.4375,
      "learning_rate": 0.0009219276575803183,
      "loss": 2.2556,
      "step": 2274
    },
    {
      "epoch": 0.0182,
      "grad_norm": 57.5,
      "learning_rate": 0.0009218600816033201,
      "loss": 2.0034,
      "step": 2275
    },
    {
      "epoch": 0.018208,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009217924788724239,
      "loss": 2.2515,
      "step": 2276
    },
    {
      "epoch": 0.018216,
      "grad_norm": 3.15625,
      "learning_rate": 0.0009217248493919169,
      "loss": 2.2495,
      "step": 2277
    },
    {
      "epoch": 0.018224,
      "grad_norm": 9.25,
      "learning_rate": 0.0009216571931660884,
      "loss": 2.046,
      "step": 2278
    },
    {
      "epoch": 0.018232,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009215895101992287,
      "loss": 3.1049,
      "step": 2279
    },
    {
      "epoch": 0.01824,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009215218004956303,
      "loss": 2.4327,
      "step": 2280
    },
    {
      "epoch": 0.018248,
      "grad_norm": 4.6875,
      "learning_rate": 0.0009214540640595874,
      "loss": 3.3104,
      "step": 2281
    },
    {
      "epoch": 0.018256,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009213863008953957,
      "loss": 2.9406,
      "step": 2282
    },
    {
      "epoch": 0.018264,
      "grad_norm": 3.765625,
      "learning_rate": 0.0009213185110073526,
      "loss": 2.5438,
      "step": 2283
    },
    {
      "epoch": 0.018272,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009212506943997574,
      "loss": 2.4274,
      "step": 2284
    },
    {
      "epoch": 0.01828,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009211828510769108,
      "loss": 2.6615,
      "step": 2285
    },
    {
      "epoch": 0.018288,
      "grad_norm": 2.78125,
      "learning_rate": 0.0009211149810431155,
      "loss": 2.6903,
      "step": 2286
    },
    {
      "epoch": 0.018296,
      "grad_norm": 3.140625,
      "learning_rate": 0.0009210470843026754,
      "loss": 2.6647,
      "step": 2287
    },
    {
      "epoch": 0.018304,
      "grad_norm": 2.90625,
      "learning_rate": 0.000920979160859897,
      "loss": 3.274,
      "step": 2288
    },
    {
      "epoch": 0.018312,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009209112107190875,
      "loss": 2.3362,
      "step": 2289
    },
    {
      "epoch": 0.01832,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009208432338845564,
      "loss": 2.6059,
      "step": 2290
    },
    {
      "epoch": 0.018328,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009207752303606146,
      "loss": 2.3004,
      "step": 2291
    },
    {
      "epoch": 0.018336,
      "grad_norm": 3.15625,
      "learning_rate": 0.000920707200151575,
      "loss": 2.4611,
      "step": 2292
    },
    {
      "epoch": 0.018344,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009206391432617516,
      "loss": 2.9711,
      "step": 2293
    },
    {
      "epoch": 0.018352,
      "grad_norm": 3.125,
      "learning_rate": 0.000920571059695461,
      "loss": 2.2244,
      "step": 2294
    },
    {
      "epoch": 0.01836,
      "grad_norm": 4.34375,
      "learning_rate": 0.0009205029494570207,
      "loss": 2.4426,
      "step": 2295
    },
    {
      "epoch": 0.018368,
      "grad_norm": 4.75,
      "learning_rate": 0.0009204348125507502,
      "loss": 2.4134,
      "step": 2296
    },
    {
      "epoch": 0.018376,
      "grad_norm": 3.546875,
      "learning_rate": 0.0009203666489809708,
      "loss": 2.4262,
      "step": 2297
    },
    {
      "epoch": 0.018384,
      "grad_norm": 2.75,
      "learning_rate": 0.0009202984587520051,
      "loss": 1.8437,
      "step": 2298
    },
    {
      "epoch": 0.018392,
      "grad_norm": 4.0,
      "learning_rate": 0.0009202302418681778,
      "loss": 2.8536,
      "step": 2299
    },
    {
      "epoch": 0.0184,
      "grad_norm": 3.375,
      "learning_rate": 0.0009201619983338152,
      "loss": 2.3073,
      "step": 2300
    },
    {
      "epoch": 0.018408,
      "grad_norm": 3.921875,
      "learning_rate": 0.0009200937281532451,
      "loss": 2.7559,
      "step": 2301
    },
    {
      "epoch": 0.018416,
      "grad_norm": 2.921875,
      "learning_rate": 0.0009200254313307972,
      "loss": 2.5209,
      "step": 2302
    },
    {
      "epoch": 0.018424,
      "grad_norm": 15.6875,
      "learning_rate": 0.0009199571078708029,
      "loss": 2.6217,
      "step": 2303
    },
    {
      "epoch": 0.018432,
      "grad_norm": 3.3125,
      "learning_rate": 0.0009198887577775949,
      "loss": 2.4569,
      "step": 2304
    },
    {
      "epoch": 0.01844,
      "grad_norm": 3.3125,
      "learning_rate": 0.0009198203810555081,
      "loss": 2.2659,
      "step": 2305
    },
    {
      "epoch": 0.018448,
      "grad_norm": 3.25,
      "learning_rate": 0.000919751977708879,
      "loss": 2.7942,
      "step": 2306
    },
    {
      "epoch": 0.018456,
      "grad_norm": 3.5625,
      "learning_rate": 0.0009196835477420455,
      "loss": 2.9777,
      "step": 2307
    },
    {
      "epoch": 0.018464,
      "grad_norm": 3.984375,
      "learning_rate": 0.0009196150911593473,
      "loss": 2.3514,
      "step": 2308
    },
    {
      "epoch": 0.018472,
      "grad_norm": 3.265625,
      "learning_rate": 0.000919546607965126,
      "loss": 2.5987,
      "step": 2309
    },
    {
      "epoch": 0.01848,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009194780981637245,
      "loss": 2.8687,
      "step": 2310
    },
    {
      "epoch": 0.018488,
      "grad_norm": 3.015625,
      "learning_rate": 0.0009194095617594878,
      "loss": 2.7535,
      "step": 2311
    },
    {
      "epoch": 0.018496,
      "grad_norm": 3.625,
      "learning_rate": 0.0009193409987567625,
      "loss": 2.7528,
      "step": 2312
    },
    {
      "epoch": 0.018504,
      "grad_norm": 3.25,
      "learning_rate": 0.0009192724091598966,
      "loss": 2.8159,
      "step": 2313
    },
    {
      "epoch": 0.018512,
      "grad_norm": 4.625,
      "learning_rate": 0.0009192037929732398,
      "loss": 2.9354,
      "step": 2314
    },
    {
      "epoch": 0.01852,
      "grad_norm": 3.828125,
      "learning_rate": 0.000919135150201144,
      "loss": 2.838,
      "step": 2315
    },
    {
      "epoch": 0.018528,
      "grad_norm": 3.65625,
      "learning_rate": 0.0009190664808479624,
      "loss": 2.3953,
      "step": 2316
    },
    {
      "epoch": 0.018536,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009189977849180499,
      "loss": 3.0693,
      "step": 2317
    },
    {
      "epoch": 0.018544,
      "grad_norm": 3.71875,
      "learning_rate": 0.000918929062415763,
      "loss": 2.7846,
      "step": 2318
    },
    {
      "epoch": 0.018552,
      "grad_norm": 2.703125,
      "learning_rate": 0.00091886031334546,
      "loss": 2.3004,
      "step": 2319
    },
    {
      "epoch": 0.01856,
      "grad_norm": 3.703125,
      "learning_rate": 0.0009187915377115013,
      "loss": 3.0836,
      "step": 2320
    },
    {
      "epoch": 0.018568,
      "grad_norm": 2.578125,
      "learning_rate": 0.0009187227355182479,
      "loss": 2.2797,
      "step": 2321
    },
    {
      "epoch": 0.018576,
      "grad_norm": 4.21875,
      "learning_rate": 0.0009186539067700636,
      "loss": 2.4035,
      "step": 2322
    },
    {
      "epoch": 0.018584,
      "grad_norm": 3.375,
      "learning_rate": 0.0009185850514713134,
      "loss": 2.5165,
      "step": 2323
    },
    {
      "epoch": 0.018592,
      "grad_norm": 3.765625,
      "learning_rate": 0.000918516169626364,
      "loss": 2.5383,
      "step": 2324
    },
    {
      "epoch": 0.0186,
      "grad_norm": 4.25,
      "learning_rate": 0.000918447261239584,
      "loss": 2.4932,
      "step": 2325
    },
    {
      "epoch": 0.018608,
      "grad_norm": 3.671875,
      "learning_rate": 0.000918378326315343,
      "loss": 2.2256,
      "step": 2326
    },
    {
      "epoch": 0.018616,
      "grad_norm": 4.71875,
      "learning_rate": 0.0009183093648580132,
      "loss": 2.7815,
      "step": 2327
    },
    {
      "epoch": 0.018624,
      "grad_norm": 4.0625,
      "learning_rate": 0.0009182403768719678,
      "loss": 2.6409,
      "step": 2328
    },
    {
      "epoch": 0.018632,
      "grad_norm": 4.6875,
      "learning_rate": 0.0009181713623615822,
      "loss": 2.7556,
      "step": 2329
    },
    {
      "epoch": 0.01864,
      "grad_norm": 4.1875,
      "learning_rate": 0.000918102321331233,
      "loss": 3.0644,
      "step": 2330
    },
    {
      "epoch": 0.018648,
      "grad_norm": 3.6875,
      "learning_rate": 0.0009180332537852988,
      "loss": 2.9007,
      "step": 2331
    },
    {
      "epoch": 0.018656,
      "grad_norm": 3.5625,
      "learning_rate": 0.0009179641597281598,
      "loss": 2.737,
      "step": 2332
    },
    {
      "epoch": 0.018664,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009178950391641978,
      "loss": 2.3259,
      "step": 2333
    },
    {
      "epoch": 0.018672,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009178258920977964,
      "loss": 2.6193,
      "step": 2334
    },
    {
      "epoch": 0.01868,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009177567185333408,
      "loss": 2.8549,
      "step": 2335
    },
    {
      "epoch": 0.018688,
      "grad_norm": 3.375,
      "learning_rate": 0.0009176875184752181,
      "loss": 2.5857,
      "step": 2336
    },
    {
      "epoch": 0.018696,
      "grad_norm": 2.796875,
      "learning_rate": 0.0009176182919278167,
      "loss": 2.2455,
      "step": 2337
    },
    {
      "epoch": 0.018704,
      "grad_norm": 3.46875,
      "learning_rate": 0.0009175490388955269,
      "loss": 3.0094,
      "step": 2338
    },
    {
      "epoch": 0.018712,
      "grad_norm": 3.4375,
      "learning_rate": 0.0009174797593827405,
      "loss": 2.5632,
      "step": 2339
    },
    {
      "epoch": 0.01872,
      "grad_norm": 3.5,
      "learning_rate": 0.0009174104533938514,
      "loss": 2.5248,
      "step": 2340
    },
    {
      "epoch": 0.018728,
      "grad_norm": 3.765625,
      "learning_rate": 0.0009173411209332547,
      "loss": 2.6971,
      "step": 2341
    },
    {
      "epoch": 0.018736,
      "grad_norm": 3.234375,
      "learning_rate": 0.0009172717620053477,
      "loss": 2.0706,
      "step": 2342
    },
    {
      "epoch": 0.018744,
      "grad_norm": 5.5,
      "learning_rate": 0.0009172023766145285,
      "loss": 2.6736,
      "step": 2343
    },
    {
      "epoch": 0.018752,
      "grad_norm": 3.296875,
      "learning_rate": 0.000917132964765198,
      "loss": 3.2301,
      "step": 2344
    },
    {
      "epoch": 0.01876,
      "grad_norm": 3.359375,
      "learning_rate": 0.0009170635264617581,
      "loss": 2.788,
      "step": 2345
    },
    {
      "epoch": 0.018768,
      "grad_norm": 8.625,
      "learning_rate": 0.0009169940617086124,
      "loss": 2.1928,
      "step": 2346
    },
    {
      "epoch": 0.018776,
      "grad_norm": 3.125,
      "learning_rate": 0.0009169245705101661,
      "loss": 2.2712,
      "step": 2347
    },
    {
      "epoch": 0.018784,
      "grad_norm": 3.515625,
      "learning_rate": 0.0009168550528708265,
      "loss": 3.0987,
      "step": 2348
    },
    {
      "epoch": 0.018792,
      "grad_norm": 3.015625,
      "learning_rate": 0.0009167855087950024,
      "loss": 2.6765,
      "step": 2349
    },
    {
      "epoch": 0.0188,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009167159382871039,
      "loss": 2.7783,
      "step": 2350
    },
    {
      "epoch": 0.018808,
      "grad_norm": 3.9375,
      "learning_rate": 0.0009166463413515433,
      "loss": 2.8879,
      "step": 2351
    },
    {
      "epoch": 0.018816,
      "grad_norm": 2.921875,
      "learning_rate": 0.0009165767179927345,
      "loss": 2.6766,
      "step": 2352
    },
    {
      "epoch": 0.018824,
      "grad_norm": 3.0,
      "learning_rate": 0.0009165070682150924,
      "loss": 2.5336,
      "step": 2353
    },
    {
      "epoch": 0.018832,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009164373920230349,
      "loss": 2.4318,
      "step": 2354
    },
    {
      "epoch": 0.01884,
      "grad_norm": 3.109375,
      "learning_rate": 0.0009163676894209801,
      "loss": 2.9485,
      "step": 2355
    },
    {
      "epoch": 0.018848,
      "grad_norm": 2.5625,
      "learning_rate": 0.0009162979604133488,
      "loss": 2.0691,
      "step": 2356
    },
    {
      "epoch": 0.018856,
      "grad_norm": 2.734375,
      "learning_rate": 0.0009162282050045631,
      "loss": 2.0089,
      "step": 2357
    },
    {
      "epoch": 0.018864,
      "grad_norm": 2.859375,
      "learning_rate": 0.0009161584231990467,
      "loss": 2.5931,
      "step": 2358
    },
    {
      "epoch": 0.018872,
      "grad_norm": 2.828125,
      "learning_rate": 0.0009160886150012252,
      "loss": 2.4036,
      "step": 2359
    },
    {
      "epoch": 0.01888,
      "grad_norm": 2.875,
      "learning_rate": 0.0009160187804155257,
      "loss": 2.5498,
      "step": 2360
    },
    {
      "epoch": 0.018888,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009159489194463771,
      "loss": 3.1026,
      "step": 2361
    },
    {
      "epoch": 0.018896,
      "grad_norm": 2.875,
      "learning_rate": 0.0009158790320982098,
      "loss": 2.179,
      "step": 2362
    },
    {
      "epoch": 0.018904,
      "grad_norm": 3.234375,
      "learning_rate": 0.000915809118375456,
      "loss": 2.4734,
      "step": 2363
    },
    {
      "epoch": 0.018912,
      "grad_norm": 2.75,
      "learning_rate": 0.0009157391782825495,
      "loss": 2.2862,
      "step": 2364
    },
    {
      "epoch": 0.01892,
      "grad_norm": 3.78125,
      "learning_rate": 0.000915669211823926,
      "loss": 2.9255,
      "step": 2365
    },
    {
      "epoch": 0.018928,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009155992190040225,
      "loss": 2.2619,
      "step": 2366
    },
    {
      "epoch": 0.018936,
      "grad_norm": 3.890625,
      "learning_rate": 0.0009155291998272781,
      "loss": 2.5723,
      "step": 2367
    },
    {
      "epoch": 0.018944,
      "grad_norm": 3.390625,
      "learning_rate": 0.0009154591542981332,
      "loss": 2.1764,
      "step": 2368
    },
    {
      "epoch": 0.018952,
      "grad_norm": 3.109375,
      "learning_rate": 0.00091538908242103,
      "loss": 2.2723,
      "step": 2369
    },
    {
      "epoch": 0.01896,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009153189842004124,
      "loss": 2.1569,
      "step": 2370
    },
    {
      "epoch": 0.018968,
      "grad_norm": 4.1875,
      "learning_rate": 0.0009152488596407258,
      "loss": 2.3359,
      "step": 2371
    },
    {
      "epoch": 0.018976,
      "grad_norm": 3.1875,
      "learning_rate": 0.0009151787087464178,
      "loss": 1.9837,
      "step": 2372
    },
    {
      "epoch": 0.018984,
      "grad_norm": 3.515625,
      "learning_rate": 0.000915108531521937,
      "loss": 2.7286,
      "step": 2373
    },
    {
      "epoch": 0.018992,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009150383279717338,
      "loss": 2.2795,
      "step": 2374
    },
    {
      "epoch": 0.019,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009149680981002608,
      "loss": 2.4059,
      "step": 2375
    },
    {
      "epoch": 0.019008,
      "grad_norm": 61.75,
      "learning_rate": 0.0009148978419119718,
      "loss": 2.9623,
      "step": 2376
    },
    {
      "epoch": 0.019016,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009148275594113223,
      "loss": 2.9817,
      "step": 2377
    },
    {
      "epoch": 0.019024,
      "grad_norm": 2.96875,
      "learning_rate": 0.0009147572506027695,
      "loss": 2.7588,
      "step": 2378
    },
    {
      "epoch": 0.019032,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009146869154907724,
      "loss": 2.3699,
      "step": 2379
    },
    {
      "epoch": 0.01904,
      "grad_norm": 2.703125,
      "learning_rate": 0.0009146165540797915,
      "loss": 2.6702,
      "step": 2380
    },
    {
      "epoch": 0.019048,
      "grad_norm": 3.5,
      "learning_rate": 0.0009145461663742892,
      "loss": 3.3261,
      "step": 2381
    },
    {
      "epoch": 0.019056,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009144757523787292,
      "loss": 2.777,
      "step": 2382
    },
    {
      "epoch": 0.019064,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009144053120975771,
      "loss": 2.9181,
      "step": 2383
    },
    {
      "epoch": 0.019072,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009143348455353003,
      "loss": 2.2603,
      "step": 2384
    },
    {
      "epoch": 0.01908,
      "grad_norm": 3.421875,
      "learning_rate": 0.0009142643526963677,
      "loss": 2.3733,
      "step": 2385
    },
    {
      "epoch": 0.019088,
      "grad_norm": 2.890625,
      "learning_rate": 0.0009141938335852498,
      "loss": 2.7216,
      "step": 2386
    },
    {
      "epoch": 0.019096,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009141232882064187,
      "loss": 2.0942,
      "step": 2387
    },
    {
      "epoch": 0.019104,
      "grad_norm": 3.4375,
      "learning_rate": 0.0009140527165643484,
      "loss": 2.1025,
      "step": 2388
    },
    {
      "epoch": 0.019112,
      "grad_norm": 2.5,
      "learning_rate": 0.0009139821186635147,
      "loss": 2.2907,
      "step": 2389
    },
    {
      "epoch": 0.01912,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009139114945083946,
      "loss": 2.6083,
      "step": 2390
    },
    {
      "epoch": 0.019128,
      "grad_norm": 3.375,
      "learning_rate": 0.0009138408441034671,
      "loss": 2.0627,
      "step": 2391
    },
    {
      "epoch": 0.019136,
      "grad_norm": 3.3125,
      "learning_rate": 0.0009137701674532128,
      "loss": 2.4537,
      "step": 2392
    },
    {
      "epoch": 0.019144,
      "grad_norm": 3.3125,
      "learning_rate": 0.0009136994645621138,
      "loss": 2.5263,
      "step": 2393
    },
    {
      "epoch": 0.019152,
      "grad_norm": 3.109375,
      "learning_rate": 0.0009136287354346542,
      "loss": 2.35,
      "step": 2394
    },
    {
      "epoch": 0.01916,
      "grad_norm": 3.46875,
      "learning_rate": 0.0009135579800753194,
      "loss": 3.0208,
      "step": 2395
    },
    {
      "epoch": 0.019168,
      "grad_norm": 3.421875,
      "learning_rate": 0.0009134871984885966,
      "loss": 1.9791,
      "step": 2396
    },
    {
      "epoch": 0.019176,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009134163906789748,
      "loss": 2.6874,
      "step": 2397
    },
    {
      "epoch": 0.019184,
      "grad_norm": 3.0,
      "learning_rate": 0.0009133455566509445,
      "loss": 2.4864,
      "step": 2398
    },
    {
      "epoch": 0.019192,
      "grad_norm": 3.65625,
      "learning_rate": 0.000913274696408998,
      "loss": 3.1304,
      "step": 2399
    },
    {
      "epoch": 0.0192,
      "grad_norm": 3.453125,
      "learning_rate": 0.000913203809957629,
      "loss": 2.8444,
      "step": 2400
    },
    {
      "epoch": 0.019208,
      "grad_norm": 3.1875,
      "learning_rate": 0.000913132897301333,
      "loss": 2.4354,
      "step": 2401
    },
    {
      "epoch": 0.019216,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009130619584446076,
      "loss": 2.4562,
      "step": 2402
    },
    {
      "epoch": 0.019224,
      "grad_norm": 3.125,
      "learning_rate": 0.0009129909933919514,
      "loss": 1.9645,
      "step": 2403
    },
    {
      "epoch": 0.019232,
      "grad_norm": 4.34375,
      "learning_rate": 0.0009129200021478649,
      "loss": 3.1137,
      "step": 2404
    },
    {
      "epoch": 0.01924,
      "grad_norm": 3.5,
      "learning_rate": 0.0009128489847168502,
      "loss": 2.4006,
      "step": 2405
    },
    {
      "epoch": 0.019248,
      "grad_norm": 3.375,
      "learning_rate": 0.0009127779411034114,
      "loss": 2.8204,
      "step": 2406
    },
    {
      "epoch": 0.019256,
      "grad_norm": 3.890625,
      "learning_rate": 0.0009127068713120539,
      "loss": 2.5569,
      "step": 2407
    },
    {
      "epoch": 0.019264,
      "grad_norm": 3.734375,
      "learning_rate": 0.0009126357753472847,
      "loss": 2.8012,
      "step": 2408
    },
    {
      "epoch": 0.019272,
      "grad_norm": 2.625,
      "learning_rate": 0.0009125646532136129,
      "loss": 2.0939,
      "step": 2409
    },
    {
      "epoch": 0.01928,
      "grad_norm": 3.125,
      "learning_rate": 0.0009124935049155488,
      "loss": 2.2295,
      "step": 2410
    },
    {
      "epoch": 0.019288,
      "grad_norm": 3.25,
      "learning_rate": 0.0009124223304576047,
      "loss": 3.007,
      "step": 2411
    },
    {
      "epoch": 0.019296,
      "grad_norm": 3.640625,
      "learning_rate": 0.0009123511298442941,
      "loss": 2.5513,
      "step": 2412
    },
    {
      "epoch": 0.019304,
      "grad_norm": 3.359375,
      "learning_rate": 0.0009122799030801329,
      "loss": 2.3395,
      "step": 2413
    },
    {
      "epoch": 0.019312,
      "grad_norm": 4.15625,
      "learning_rate": 0.0009122086501696378,
      "loss": 2.6891,
      "step": 2414
    },
    {
      "epoch": 0.01932,
      "grad_norm": 3.25,
      "learning_rate": 0.000912137371117328,
      "loss": 2.503,
      "step": 2415
    },
    {
      "epoch": 0.019328,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009120660659277236,
      "loss": 2.6117,
      "step": 2416
    },
    {
      "epoch": 0.019336,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009119947346053468,
      "loss": 2.8145,
      "step": 2417
    },
    {
      "epoch": 0.019344,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009119233771547213,
      "loss": 2.8505,
      "step": 2418
    },
    {
      "epoch": 0.019352,
      "grad_norm": 2.96875,
      "learning_rate": 0.0009118519935803726,
      "loss": 2.4176,
      "step": 2419
    },
    {
      "epoch": 0.01936,
      "grad_norm": 4.09375,
      "learning_rate": 0.0009117805838868277,
      "loss": 2.1643,
      "step": 2420
    },
    {
      "epoch": 0.019368,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009117091480786155,
      "loss": 1.9782,
      "step": 2421
    },
    {
      "epoch": 0.019376,
      "grad_norm": 3.25,
      "learning_rate": 0.0009116376861602662,
      "loss": 3.3748,
      "step": 2422
    },
    {
      "epoch": 0.019384,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009115661981363118,
      "loss": 2.5912,
      "step": 2423
    },
    {
      "epoch": 0.019392,
      "grad_norm": 2.8125,
      "learning_rate": 0.0009114946840112861,
      "loss": 2.462,
      "step": 2424
    },
    {
      "epoch": 0.0194,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009114231437897244,
      "loss": 2.8131,
      "step": 2425
    },
    {
      "epoch": 0.019408,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009113515774761638,
      "loss": 2.7796,
      "step": 2426
    },
    {
      "epoch": 0.019416,
      "grad_norm": 3.5625,
      "learning_rate": 0.0009112799850751427,
      "loss": 2.7514,
      "step": 2427
    },
    {
      "epoch": 0.019424,
      "grad_norm": 3.109375,
      "learning_rate": 0.0009112083665912019,
      "loss": 2.0839,
      "step": 2428
    },
    {
      "epoch": 0.019432,
      "grad_norm": 3.6875,
      "learning_rate": 0.0009111367220288829,
      "loss": 3.2493,
      "step": 2429
    },
    {
      "epoch": 0.01944,
      "grad_norm": 5.84375,
      "learning_rate": 0.0009110650513927293,
      "loss": 2.1617,
      "step": 2430
    },
    {
      "epoch": 0.019448,
      "grad_norm": 3.90625,
      "learning_rate": 0.0009109933546872866,
      "loss": 2.291,
      "step": 2431
    },
    {
      "epoch": 0.019456,
      "grad_norm": 3.0625,
      "learning_rate": 0.0009109216319171017,
      "loss": 2.5659,
      "step": 2432
    },
    {
      "epoch": 0.019464,
      "grad_norm": 3.3125,
      "learning_rate": 0.0009108498830867232,
      "loss": 2.6353,
      "step": 2433
    },
    {
      "epoch": 0.019472,
      "grad_norm": 3.109375,
      "learning_rate": 0.0009107781082007013,
      "loss": 2.3882,
      "step": 2434
    },
    {
      "epoch": 0.01948,
      "grad_norm": 2.875,
      "learning_rate": 0.0009107063072635878,
      "loss": 1.8294,
      "step": 2435
    },
    {
      "epoch": 0.019488,
      "grad_norm": 3.640625,
      "learning_rate": 0.0009106344802799364,
      "loss": 2.1865,
      "step": 2436
    },
    {
      "epoch": 0.019496,
      "grad_norm": 3.890625,
      "learning_rate": 0.000910562627254302,
      "loss": 2.8398,
      "step": 2437
    },
    {
      "epoch": 0.019504,
      "grad_norm": 3.609375,
      "learning_rate": 0.0009104907481912416,
      "loss": 3.1294,
      "step": 2438
    },
    {
      "epoch": 0.019512,
      "grad_norm": 4.0625,
      "learning_rate": 0.000910418843095314,
      "loss": 2.5365,
      "step": 2439
    },
    {
      "epoch": 0.01952,
      "grad_norm": 3.1875,
      "learning_rate": 0.0009103469119710788,
      "loss": 2.2013,
      "step": 2440
    },
    {
      "epoch": 0.019528,
      "grad_norm": 3.4375,
      "learning_rate": 0.0009102749548230981,
      "loss": 2.7818,
      "step": 2441
    },
    {
      "epoch": 0.019536,
      "grad_norm": 4.1875,
      "learning_rate": 0.0009102029716559352,
      "loss": 2.6291,
      "step": 2442
    },
    {
      "epoch": 0.019544,
      "grad_norm": 2.84375,
      "learning_rate": 0.0009101309624741554,
      "loss": 2.5438,
      "step": 2443
    },
    {
      "epoch": 0.019552,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009100589272823253,
      "loss": 2.502,
      "step": 2444
    },
    {
      "epoch": 0.01956,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009099868660850134,
      "loss": 2.8484,
      "step": 2445
    },
    {
      "epoch": 0.019568,
      "grad_norm": 3.25,
      "learning_rate": 0.0009099147788867894,
      "loss": 2.5302,
      "step": 2446
    },
    {
      "epoch": 0.019576,
      "grad_norm": 9.875,
      "learning_rate": 0.0009098426656922255,
      "loss": 2.2132,
      "step": 2447
    },
    {
      "epoch": 0.019584,
      "grad_norm": 78.0,
      "learning_rate": 0.0009097705265058946,
      "loss": 2.1222,
      "step": 2448
    },
    {
      "epoch": 0.019592,
      "grad_norm": 28.0,
      "learning_rate": 0.000909698361332372,
      "loss": 2.807,
      "step": 2449
    },
    {
      "epoch": 0.0196,
      "grad_norm": 119.0,
      "learning_rate": 0.0009096261701762343,
      "loss": 2.3598,
      "step": 2450
    },
    {
      "epoch": 0.019608,
      "grad_norm": 3.890625,
      "learning_rate": 0.0009095539530420595,
      "loss": 2.6326,
      "step": 2451
    },
    {
      "epoch": 0.019616,
      "grad_norm": 96.0,
      "learning_rate": 0.0009094817099344277,
      "loss": 1.5923,
      "step": 2452
    },
    {
      "epoch": 0.019624,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009094094408579205,
      "loss": 2.8329,
      "step": 2453
    },
    {
      "epoch": 0.019632,
      "grad_norm": 4.5625,
      "learning_rate": 0.0009093371458171214,
      "loss": 3.2114,
      "step": 2454
    },
    {
      "epoch": 0.01964,
      "grad_norm": 3.6875,
      "learning_rate": 0.0009092648248166147,
      "loss": 2.3595,
      "step": 2455
    },
    {
      "epoch": 0.019648,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009091924778609873,
      "loss": 2.8072,
      "step": 2456
    },
    {
      "epoch": 0.019656,
      "grad_norm": 2.203125,
      "learning_rate": 0.0009091201049548272,
      "loss": 1.7327,
      "step": 2457
    },
    {
      "epoch": 0.019664,
      "grad_norm": 2.765625,
      "learning_rate": 0.0009090477061027244,
      "loss": 1.9418,
      "step": 2458
    },
    {
      "epoch": 0.019672,
      "grad_norm": 3.6875,
      "learning_rate": 0.0009089752813092702,
      "loss": 2.3995,
      "step": 2459
    },
    {
      "epoch": 0.01968,
      "grad_norm": 3.078125,
      "learning_rate": 0.0009089028305790577,
      "loss": 2.0555,
      "step": 2460
    },
    {
      "epoch": 0.019688,
      "grad_norm": 8.375,
      "learning_rate": 0.0009088303539166818,
      "loss": 2.3978,
      "step": 2461
    },
    {
      "epoch": 0.019696,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009087578513267388,
      "loss": 2.562,
      "step": 2462
    },
    {
      "epoch": 0.019704,
      "grad_norm": 3.875,
      "learning_rate": 0.0009086853228138263,
      "loss": 2.6942,
      "step": 2463
    },
    {
      "epoch": 0.019712,
      "grad_norm": 3.0,
      "learning_rate": 0.0009086127683825447,
      "loss": 2.4669,
      "step": 2464
    },
    {
      "epoch": 0.01972,
      "grad_norm": 3.5,
      "learning_rate": 0.0009085401880374951,
      "loss": 3.2457,
      "step": 2465
    },
    {
      "epoch": 0.019728,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009084675817832801,
      "loss": 2.8118,
      "step": 2466
    },
    {
      "epoch": 0.019736,
      "grad_norm": 3.796875,
      "learning_rate": 0.0009083949496245049,
      "loss": 2.4193,
      "step": 2467
    },
    {
      "epoch": 0.019744,
      "grad_norm": 3.40625,
      "learning_rate": 0.0009083222915657751,
      "loss": 2.6598,
      "step": 2468
    },
    {
      "epoch": 0.019752,
      "grad_norm": 4.21875,
      "learning_rate": 0.0009082496076116992,
      "loss": 2.8943,
      "step": 2469
    },
    {
      "epoch": 0.01976,
      "grad_norm": 3.8125,
      "learning_rate": 0.0009081768977668862,
      "loss": 2.5952,
      "step": 2470
    },
    {
      "epoch": 0.019768,
      "grad_norm": 3.984375,
      "learning_rate": 0.0009081041620359478,
      "loss": 2.6943,
      "step": 2471
    },
    {
      "epoch": 0.019776,
      "grad_norm": 13.9375,
      "learning_rate": 0.0009080314004234965,
      "loss": 2.703,
      "step": 2472
    },
    {
      "epoch": 0.019784,
      "grad_norm": 3.8125,
      "learning_rate": 0.0009079586129341466,
      "loss": 2.5667,
      "step": 2473
    },
    {
      "epoch": 0.019792,
      "grad_norm": 3.90625,
      "learning_rate": 0.0009078857995725146,
      "loss": 2.5246,
      "step": 2474
    },
    {
      "epoch": 0.0198,
      "grad_norm": 3.4375,
      "learning_rate": 0.0009078129603432181,
      "loss": 2.5303,
      "step": 2475
    },
    {
      "epoch": 0.019808,
      "grad_norm": 3.0625,
      "learning_rate": 0.0009077400952508763,
      "loss": 2.6667,
      "step": 2476
    },
    {
      "epoch": 0.019816,
      "grad_norm": 3.9375,
      "learning_rate": 0.0009076672043001104,
      "loss": 2.1974,
      "step": 2477
    },
    {
      "epoch": 0.019824,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009075942874955431,
      "loss": 2.8986,
      "step": 2478
    },
    {
      "epoch": 0.019832,
      "grad_norm": 3.203125,
      "learning_rate": 0.0009075213448417985,
      "loss": 3.0384,
      "step": 2479
    },
    {
      "epoch": 0.01984,
      "grad_norm": 3.984375,
      "learning_rate": 0.0009074483763435028,
      "loss": 2.3786,
      "step": 2480
    },
    {
      "epoch": 0.019848,
      "grad_norm": 2.625,
      "learning_rate": 0.0009073753820052835,
      "loss": 1.6224,
      "step": 2481
    },
    {
      "epoch": 0.019856,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009073023618317696,
      "loss": 2.6112,
      "step": 2482
    },
    {
      "epoch": 0.019864,
      "grad_norm": 2.59375,
      "learning_rate": 0.0009072293158275922,
      "loss": 2.0364,
      "step": 2483
    },
    {
      "epoch": 0.019872,
      "grad_norm": 2.875,
      "learning_rate": 0.0009071562439973838,
      "loss": 2.3109,
      "step": 2484
    },
    {
      "epoch": 0.01988,
      "grad_norm": 2.71875,
      "learning_rate": 0.0009070831463457785,
      "loss": 2.8376,
      "step": 2485
    },
    {
      "epoch": 0.019888,
      "grad_norm": 3.15625,
      "learning_rate": 0.0009070100228774119,
      "loss": 2.1396,
      "step": 2486
    },
    {
      "epoch": 0.019896,
      "grad_norm": 3.34375,
      "learning_rate": 0.0009069368735969217,
      "loss": 2.2146,
      "step": 2487
    },
    {
      "epoch": 0.019904,
      "grad_norm": 3.5625,
      "learning_rate": 0.0009068636985089467,
      "loss": 2.7157,
      "step": 2488
    },
    {
      "epoch": 0.019912,
      "grad_norm": 3.0,
      "learning_rate": 0.0009067904976181276,
      "loss": 2.4751,
      "step": 2489
    },
    {
      "epoch": 0.01992,
      "grad_norm": 3.640625,
      "learning_rate": 0.0009067172709291069,
      "loss": 1.9963,
      "step": 2490
    },
    {
      "epoch": 0.019928,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009066440184465283,
      "loss": 2.1975,
      "step": 2491
    },
    {
      "epoch": 0.019936,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009065707401750377,
      "loss": 2.4545,
      "step": 2492
    },
    {
      "epoch": 0.019944,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009064974361192821,
      "loss": 2.1467,
      "step": 2493
    },
    {
      "epoch": 0.019952,
      "grad_norm": 2.984375,
      "learning_rate": 0.0009064241062839105,
      "loss": 2.5052,
      "step": 2494
    },
    {
      "epoch": 0.01996,
      "grad_norm": 2.859375,
      "learning_rate": 0.0009063507506735733,
      "loss": 2.3151,
      "step": 2495
    },
    {
      "epoch": 0.019968,
      "grad_norm": 3.046875,
      "learning_rate": 0.0009062773692929226,
      "loss": 2.2871,
      "step": 2496
    },
    {
      "epoch": 0.019976,
      "grad_norm": 3.09375,
      "learning_rate": 0.0009062039621466122,
      "loss": 2.1649,
      "step": 2497
    },
    {
      "epoch": 0.019984,
      "grad_norm": 3.46875,
      "learning_rate": 0.0009061305292392976,
      "loss": 2.9647,
      "step": 2498
    },
    {
      "epoch": 0.019992,
      "grad_norm": 3.671875,
      "learning_rate": 0.0009060570705756357,
      "loss": 2.6765,
      "step": 2499
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.453125,
      "learning_rate": 0.0009059835861602853,
      "loss": 2.274,
      "step": 2500
    },
    {
      "epoch": 0.020008,
      "grad_norm": 3.65625,
      "learning_rate": 0.0009059100759979066,
      "loss": 2.9982,
      "step": 2501
    },
    {
      "epoch": 0.020016,
      "grad_norm": 3.734375,
      "learning_rate": 0.0009058365400931615,
      "loss": 2.6198,
      "step": 2502
    },
    {
      "epoch": 0.020024,
      "grad_norm": 3.265625,
      "learning_rate": 0.0009057629784507137,
      "loss": 2.3434,
      "step": 2503
    },
    {
      "epoch": 0.020032,
      "grad_norm": 3.5625,
      "learning_rate": 0.0009056893910752282,
      "loss": 2.6835,
      "step": 2504
    },
    {
      "epoch": 0.02004,
      "grad_norm": 3.40625,
      "learning_rate": 0.000905615777971372,
      "loss": 2.6453,
      "step": 2505
    },
    {
      "epoch": 0.020048,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009055421391438137,
      "loss": 2.8609,
      "step": 2506
    },
    {
      "epoch": 0.020056,
      "grad_norm": 4.15625,
      "learning_rate": 0.000905468474597223,
      "loss": 2.9827,
      "step": 2507
    },
    {
      "epoch": 0.020064,
      "grad_norm": 7.03125,
      "learning_rate": 0.0009053947843362717,
      "loss": 2.3582,
      "step": 2508
    },
    {
      "epoch": 0.020072,
      "grad_norm": 3.609375,
      "learning_rate": 0.0009053210683656335,
      "loss": 2.6425,
      "step": 2509
    },
    {
      "epoch": 0.02008,
      "grad_norm": 4.125,
      "learning_rate": 0.0009052473266899832,
      "loss": 2.3927,
      "step": 2510
    },
    {
      "epoch": 0.020088,
      "grad_norm": 4.40625,
      "learning_rate": 0.0009051735593139972,
      "loss": 2.3312,
      "step": 2511
    },
    {
      "epoch": 0.020096,
      "grad_norm": 4.625,
      "learning_rate": 0.0009050997662423541,
      "loss": 2.748,
      "step": 2512
    },
    {
      "epoch": 0.020104,
      "grad_norm": 5.375,
      "learning_rate": 0.0009050259474797335,
      "loss": 2.901,
      "step": 2513
    },
    {
      "epoch": 0.020112,
      "grad_norm": 5.96875,
      "learning_rate": 0.0009049521030308169,
      "loss": 2.782,
      "step": 2514
    },
    {
      "epoch": 0.02012,
      "grad_norm": 3.875,
      "learning_rate": 0.0009048782329002878,
      "loss": 2.5188,
      "step": 2515
    },
    {
      "epoch": 0.020128,
      "grad_norm": 3.984375,
      "learning_rate": 0.0009048043370928305,
      "loss": 2.5122,
      "step": 2516
    },
    {
      "epoch": 0.020136,
      "grad_norm": 4.625,
      "learning_rate": 0.0009047304156131317,
      "loss": 2.2245,
      "step": 2517
    },
    {
      "epoch": 0.020144,
      "grad_norm": 4.15625,
      "learning_rate": 0.0009046564684658791,
      "loss": 2.5169,
      "step": 2518
    },
    {
      "epoch": 0.020152,
      "grad_norm": 2.78125,
      "learning_rate": 0.0009045824956557628,
      "loss": 2.1314,
      "step": 2519
    },
    {
      "epoch": 0.02016,
      "grad_norm": 3.5,
      "learning_rate": 0.0009045084971874737,
      "loss": 2.7989,
      "step": 2520
    },
    {
      "epoch": 0.020168,
      "grad_norm": 4.5625,
      "learning_rate": 0.000904434473065705,
      "loss": 2.9613,
      "step": 2521
    },
    {
      "epoch": 0.020176,
      "grad_norm": 2.90625,
      "learning_rate": 0.0009043604232951508,
      "loss": 2.3781,
      "step": 2522
    },
    {
      "epoch": 0.020184,
      "grad_norm": 3.421875,
      "learning_rate": 0.0009042863478805076,
      "loss": 2.5135,
      "step": 2523
    },
    {
      "epoch": 0.020192,
      "grad_norm": 3.046875,
      "learning_rate": 0.000904212246826473,
      "loss": 2.5289,
      "step": 2524
    },
    {
      "epoch": 0.0202,
      "grad_norm": 3.125,
      "learning_rate": 0.0009041381201377467,
      "loss": 2.1626,
      "step": 2525
    },
    {
      "epoch": 0.020208,
      "grad_norm": 5.15625,
      "learning_rate": 0.0009040639678190294,
      "loss": 2.8493,
      "step": 2526
    },
    {
      "epoch": 0.020216,
      "grad_norm": 3.828125,
      "learning_rate": 0.000903989789875024,
      "loss": 2.3531,
      "step": 2527
    },
    {
      "epoch": 0.020224,
      "grad_norm": 3.390625,
      "learning_rate": 0.0009039155863104347,
      "loss": 2.5852,
      "step": 2528
    },
    {
      "epoch": 0.020232,
      "grad_norm": 3.328125,
      "learning_rate": 0.0009038413571299672,
      "loss": 2.2532,
      "step": 2529
    },
    {
      "epoch": 0.02024,
      "grad_norm": 4.96875,
      "learning_rate": 0.0009037671023383293,
      "loss": 2.6153,
      "step": 2530
    },
    {
      "epoch": 0.020248,
      "grad_norm": 4.15625,
      "learning_rate": 0.0009036928219402301,
      "loss": 2.5298,
      "step": 2531
    },
    {
      "epoch": 0.020256,
      "grad_norm": 4.90625,
      "learning_rate": 0.0009036185159403803,
      "loss": 3.1042,
      "step": 2532
    },
    {
      "epoch": 0.020264,
      "grad_norm": 5.28125,
      "learning_rate": 0.0009035441843434922,
      "loss": 2.727,
      "step": 2533
    },
    {
      "epoch": 0.020272,
      "grad_norm": 3.390625,
      "learning_rate": 0.0009034698271542801,
      "loss": 2.6284,
      "step": 2534
    },
    {
      "epoch": 0.02028,
      "grad_norm": 3.390625,
      "learning_rate": 0.0009033954443774597,
      "loss": 2.7247,
      "step": 2535
    },
    {
      "epoch": 0.020288,
      "grad_norm": 5.1875,
      "learning_rate": 0.0009033210360177478,
      "loss": 2.2813,
      "step": 2536
    },
    {
      "epoch": 0.020296,
      "grad_norm": 3.171875,
      "learning_rate": 0.0009032466020798637,
      "loss": 2.3192,
      "step": 2537
    },
    {
      "epoch": 0.020304,
      "grad_norm": 3.125,
      "learning_rate": 0.000903172142568528,
      "loss": 2.1757,
      "step": 2538
    },
    {
      "epoch": 0.020312,
      "grad_norm": 2.9375,
      "learning_rate": 0.0009030976574884624,
      "loss": 2.064,
      "step": 2539
    },
    {
      "epoch": 0.02032,
      "grad_norm": 3.296875,
      "learning_rate": 0.0009030231468443911,
      "loss": 2.84,
      "step": 2540
    },
    {
      "epoch": 0.020328,
      "grad_norm": 3.140625,
      "learning_rate": 0.0009029486106410391,
      "loss": 2.0335,
      "step": 2541
    },
    {
      "epoch": 0.020336,
      "grad_norm": 4.0,
      "learning_rate": 0.0009028740488831339,
      "loss": 2.9501,
      "step": 2542
    },
    {
      "epoch": 0.020344,
      "grad_norm": 3.03125,
      "learning_rate": 0.0009027994615754036,
      "loss": 2.6019,
      "step": 2543
    },
    {
      "epoch": 0.020352,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009027248487225788,
      "loss": 2.4497,
      "step": 2544
    },
    {
      "epoch": 0.02036,
      "grad_norm": 3.25,
      "learning_rate": 0.000902650210329391,
      "loss": 2.9991,
      "step": 2545
    },
    {
      "epoch": 0.020368,
      "grad_norm": 3.421875,
      "learning_rate": 0.0009025755464005742,
      "loss": 2.8695,
      "step": 2546
    },
    {
      "epoch": 0.020376,
      "grad_norm": 8.8125,
      "learning_rate": 0.0009025008569408631,
      "loss": 1.9543,
      "step": 2547
    },
    {
      "epoch": 0.020384,
      "grad_norm": 2.921875,
      "learning_rate": 0.0009024261419549944,
      "loss": 2.1235,
      "step": 2548
    },
    {
      "epoch": 0.020392,
      "grad_norm": 3.25,
      "learning_rate": 0.0009023514014477068,
      "loss": 2.1355,
      "step": 2549
    },
    {
      "epoch": 0.0204,
      "grad_norm": 4.21875,
      "learning_rate": 0.0009022766354237399,
      "loss": 2.6541,
      "step": 2550
    },
    {
      "epoch": 0.020408,
      "grad_norm": 3.578125,
      "learning_rate": 0.0009022018438878354,
      "loss": 3.2486,
      "step": 2551
    },
    {
      "epoch": 0.020416,
      "grad_norm": 2.953125,
      "learning_rate": 0.0009021270268447367,
      "loss": 1.9664,
      "step": 2552
    },
    {
      "epoch": 0.020424,
      "grad_norm": 3.703125,
      "learning_rate": 0.0009020521842991883,
      "loss": 2.5749,
      "step": 2553
    },
    {
      "epoch": 0.020432,
      "grad_norm": 3.125,
      "learning_rate": 0.0009019773162559368,
      "loss": 2.7795,
      "step": 2554
    },
    {
      "epoch": 0.02044,
      "grad_norm": 1120.0,
      "learning_rate": 0.0009019024227197303,
      "loss": 2.799,
      "step": 2555
    },
    {
      "epoch": 0.020448,
      "grad_norm": 5.875,
      "learning_rate": 0.0009018275036953183,
      "loss": 3.7429,
      "step": 2556
    },
    {
      "epoch": 0.020456,
      "grad_norm": 3.3125,
      "learning_rate": 0.000901752559187452,
      "loss": 2.4368,
      "step": 2557
    },
    {
      "epoch": 0.020464,
      "grad_norm": 3.3125,
      "learning_rate": 0.0009016775892008846,
      "loss": 2.1775,
      "step": 2558
    },
    {
      "epoch": 0.020472,
      "grad_norm": 3.59375,
      "learning_rate": 0.0009016025937403706,
      "loss": 2.9924,
      "step": 2559
    },
    {
      "epoch": 0.02048,
      "grad_norm": 3.15625,
      "learning_rate": 0.0009015275728106658,
      "loss": 2.3662,
      "step": 2560
    },
    {
      "epoch": 0.020488,
      "grad_norm": 3.84375,
      "learning_rate": 0.0009014525264165282,
      "loss": 2.8616,
      "step": 2561
    },
    {
      "epoch": 0.020496,
      "grad_norm": 3.765625,
      "learning_rate": 0.0009013774545627171,
      "loss": 2.6391,
      "step": 2562
    },
    {
      "epoch": 0.020504,
      "grad_norm": 3.21875,
      "learning_rate": 0.0009013023572539935,
      "loss": 2.2897,
      "step": 2563
    },
    {
      "epoch": 0.020512,
      "grad_norm": 3.234375,
      "learning_rate": 0.00090122723449512,
      "loss": 2.3268,
      "step": 2564
    },
    {
      "epoch": 0.02052,
      "grad_norm": 3.5,
      "learning_rate": 0.0009011520862908606,
      "loss": 2.1998,
      "step": 2565
    },
    {
      "epoch": 0.020528,
      "grad_norm": 3.25,
      "learning_rate": 0.0009010769126459815,
      "loss": 2.5281,
      "step": 2566
    },
    {
      "epoch": 0.020536,
      "grad_norm": 3.828125,
      "learning_rate": 0.0009010017135652498,
      "loss": 2.9825,
      "step": 2567
    },
    {
      "epoch": 0.020544,
      "grad_norm": 3.359375,
      "learning_rate": 0.0009009264890534346,
      "loss": 2.5809,
      "step": 2568
    },
    {
      "epoch": 0.020552,
      "grad_norm": 3.140625,
      "learning_rate": 0.0009008512391153067,
      "loss": 2.3823,
      "step": 2569
    },
    {
      "epoch": 0.02056,
      "grad_norm": 3.734375,
      "learning_rate": 0.0009007759637556382,
      "loss": 1.7502,
      "step": 2570
    },
    {
      "epoch": 0.020568,
      "grad_norm": 3.140625,
      "learning_rate": 0.000900700662979203,
      "loss": 2.667,
      "step": 2571
    },
    {
      "epoch": 0.020576,
      "grad_norm": 3.390625,
      "learning_rate": 0.0009006253367907769,
      "loss": 2.1561,
      "step": 2572
    },
    {
      "epoch": 0.020584,
      "grad_norm": 3.921875,
      "learning_rate": 0.0009005499851951365,
      "loss": 3.1898,
      "step": 2573
    },
    {
      "epoch": 0.020592,
      "grad_norm": 4.125,
      "learning_rate": 0.0009004746081970607,
      "loss": 1.9408,
      "step": 2574
    },
    {
      "epoch": 0.0206,
      "grad_norm": 3.546875,
      "learning_rate": 0.0009003992058013302,
      "loss": 2.3539,
      "step": 2575
    },
    {
      "epoch": 0.020608,
      "grad_norm": 3.765625,
      "learning_rate": 0.0009003237780127265,
      "loss": 2.888,
      "step": 2576
    },
    {
      "epoch": 0.020616,
      "grad_norm": 3.40625,
      "learning_rate": 0.0009002483248360332,
      "loss": 2.4032,
      "step": 2577
    },
    {
      "epoch": 0.020624,
      "grad_norm": 2.96875,
      "learning_rate": 0.0009001728462760354,
      "loss": 1.9875,
      "step": 2578
    },
    {
      "epoch": 0.020632,
      "grad_norm": 3.234375,
      "learning_rate": 0.0009000973423375203,
      "loss": 2.3386,
      "step": 2579
    },
    {
      "epoch": 0.02064,
      "grad_norm": 3.71875,
      "learning_rate": 0.0009000218130252758,
      "loss": 2.2545,
      "step": 2580
    },
    {
      "epoch": 0.020648,
      "grad_norm": 3.203125,
      "learning_rate": 0.0008999462583440921,
      "loss": 2.3824,
      "step": 2581
    },
    {
      "epoch": 0.020656,
      "grad_norm": 3.578125,
      "learning_rate": 0.0008998706782987607,
      "loss": 2.71,
      "step": 2582
    },
    {
      "epoch": 0.020664,
      "grad_norm": 5.0,
      "learning_rate": 0.0008997950728940748,
      "loss": 2.055,
      "step": 2583
    },
    {
      "epoch": 0.020672,
      "grad_norm": 3.46875,
      "learning_rate": 0.0008997194421348294,
      "loss": 2.8624,
      "step": 2584
    },
    {
      "epoch": 0.02068,
      "grad_norm": 3.703125,
      "learning_rate": 0.0008996437860258206,
      "loss": 2.4417,
      "step": 2585
    },
    {
      "epoch": 0.020688,
      "grad_norm": 2.96875,
      "learning_rate": 0.0008995681045718467,
      "loss": 2.1408,
      "step": 2586
    },
    {
      "epoch": 0.020696,
      "grad_norm": 14.125,
      "learning_rate": 0.0008994923977777074,
      "loss": 2.1482,
      "step": 2587
    },
    {
      "epoch": 0.020704,
      "grad_norm": 4.28125,
      "learning_rate": 0.0008994166656482035,
      "loss": 2.7117,
      "step": 2588
    },
    {
      "epoch": 0.020712,
      "grad_norm": 3.40625,
      "learning_rate": 0.0008993409081881384,
      "loss": 2.6751,
      "step": 2589
    },
    {
      "epoch": 0.02072,
      "grad_norm": 16.625,
      "learning_rate": 0.0008992651254023162,
      "loss": 2.5228,
      "step": 2590
    },
    {
      "epoch": 0.020728,
      "grad_norm": 3.046875,
      "learning_rate": 0.000899189317295543,
      "loss": 1.9078,
      "step": 2591
    },
    {
      "epoch": 0.020736,
      "grad_norm": 3.890625,
      "learning_rate": 0.0008991134838726265,
      "loss": 2.3015,
      "step": 2592
    },
    {
      "epoch": 0.020744,
      "grad_norm": 3.96875,
      "learning_rate": 0.0008990376251383761,
      "loss": 2.504,
      "step": 2593
    },
    {
      "epoch": 0.020752,
      "grad_norm": 2.65625,
      "learning_rate": 0.0008989617410976023,
      "loss": 2.1792,
      "step": 2594
    },
    {
      "epoch": 0.02076,
      "grad_norm": 3.671875,
      "learning_rate": 0.0008988858317551181,
      "loss": 2.1618,
      "step": 2595
    },
    {
      "epoch": 0.020768,
      "grad_norm": 3.734375,
      "learning_rate": 0.0008988098971157373,
      "loss": 1.9219,
      "step": 2596
    },
    {
      "epoch": 0.020776,
      "grad_norm": 3.921875,
      "learning_rate": 0.0008987339371842755,
      "loss": 2.2211,
      "step": 2597
    },
    {
      "epoch": 0.020784,
      "grad_norm": 5.21875,
      "learning_rate": 0.0008986579519655501,
      "loss": 2.9678,
      "step": 2598
    },
    {
      "epoch": 0.020792,
      "grad_norm": 4.28125,
      "learning_rate": 0.0008985819414643801,
      "loss": 1.8855,
      "step": 2599
    },
    {
      "epoch": 0.0208,
      "grad_norm": 4.3125,
      "learning_rate": 0.0008985059056855857,
      "loss": 2.4614,
      "step": 2600
    },
    {
      "epoch": 0.020808,
      "grad_norm": 4.71875,
      "learning_rate": 0.0008984298446339896,
      "loss": 3.2572,
      "step": 2601
    },
    {
      "epoch": 0.020816,
      "grad_norm": 4.53125,
      "learning_rate": 0.0008983537583144148,
      "loss": 3.0256,
      "step": 2602
    },
    {
      "epoch": 0.020824,
      "grad_norm": 4.1875,
      "learning_rate": 0.0008982776467316871,
      "loss": 2.5088,
      "step": 2603
    },
    {
      "epoch": 0.020832,
      "grad_norm": 3.875,
      "learning_rate": 0.0008982015098906332,
      "loss": 3.0255,
      "step": 2604
    },
    {
      "epoch": 0.02084,
      "grad_norm": 3.453125,
      "learning_rate": 0.0008981253477960816,
      "loss": 2.7386,
      "step": 2605
    },
    {
      "epoch": 0.020848,
      "grad_norm": 3.3125,
      "learning_rate": 0.0008980491604528625,
      "loss": 2.6299,
      "step": 2606
    },
    {
      "epoch": 0.020856,
      "grad_norm": 3.09375,
      "learning_rate": 0.0008979729478658078,
      "loss": 2.2976,
      "step": 2607
    },
    {
      "epoch": 0.020864,
      "grad_norm": 3.203125,
      "learning_rate": 0.0008978967100397503,
      "loss": 2.9939,
      "step": 2608
    },
    {
      "epoch": 0.020872,
      "grad_norm": 2.953125,
      "learning_rate": 0.0008978204469795254,
      "loss": 2.0225,
      "step": 2609
    },
    {
      "epoch": 0.02088,
      "grad_norm": 3.0,
      "learning_rate": 0.0008977441586899694,
      "loss": 2.1205,
      "step": 2610
    },
    {
      "epoch": 0.020888,
      "grad_norm": 2.984375,
      "learning_rate": 0.0008976678451759205,
      "loss": 2.7568,
      "step": 2611
    },
    {
      "epoch": 0.020896,
      "grad_norm": 3.90625,
      "learning_rate": 0.0008975915064422183,
      "loss": 2.2658,
      "step": 2612
    },
    {
      "epoch": 0.020904,
      "grad_norm": 2.953125,
      "learning_rate": 0.0008975151424937042,
      "loss": 2.3873,
      "step": 2613
    },
    {
      "epoch": 0.020912,
      "grad_norm": 3.578125,
      "learning_rate": 0.0008974387533352213,
      "loss": 2.7211,
      "step": 2614
    },
    {
      "epoch": 0.02092,
      "grad_norm": 3.109375,
      "learning_rate": 0.0008973623389716137,
      "loss": 2.8344,
      "step": 2615
    },
    {
      "epoch": 0.020928,
      "grad_norm": 3.578125,
      "learning_rate": 0.0008972858994077279,
      "loss": 2.3558,
      "step": 2616
    },
    {
      "epoch": 0.020936,
      "grad_norm": 2.828125,
      "learning_rate": 0.0008972094346484114,
      "loss": 2.1594,
      "step": 2617
    },
    {
      "epoch": 0.020944,
      "grad_norm": 3.796875,
      "learning_rate": 0.0008971329446985136,
      "loss": 2.4694,
      "step": 2618
    },
    {
      "epoch": 0.020952,
      "grad_norm": 2.578125,
      "learning_rate": 0.0008970564295628854,
      "loss": 2.1444,
      "step": 2619
    },
    {
      "epoch": 0.02096,
      "grad_norm": 3.328125,
      "learning_rate": 0.0008969798892463792,
      "loss": 2.5511,
      "step": 2620
    },
    {
      "epoch": 0.020968,
      "grad_norm": 2.8125,
      "learning_rate": 0.0008969033237538494,
      "loss": 2.0183,
      "step": 2621
    },
    {
      "epoch": 0.020976,
      "grad_norm": 2.734375,
      "learning_rate": 0.0008968267330901512,
      "loss": 2.4551,
      "step": 2622
    },
    {
      "epoch": 0.020984,
      "grad_norm": 3.75,
      "learning_rate": 0.0008967501172601422,
      "loss": 2.9177,
      "step": 2623
    },
    {
      "epoch": 0.020992,
      "grad_norm": 3.984375,
      "learning_rate": 0.0008966734762686814,
      "loss": 2.6315,
      "step": 2624
    },
    {
      "epoch": 0.021,
      "grad_norm": 3.46875,
      "learning_rate": 0.0008965968101206292,
      "loss": 2.9285,
      "step": 2625
    },
    {
      "epoch": 0.021008,
      "grad_norm": 3.578125,
      "learning_rate": 0.0008965201188208474,
      "loss": 2.1791,
      "step": 2626
    },
    {
      "epoch": 0.021016,
      "grad_norm": 2.875,
      "learning_rate": 0.0008964434023742001,
      "loss": 2.4958,
      "step": 2627
    },
    {
      "epoch": 0.021024,
      "grad_norm": 3.375,
      "learning_rate": 0.0008963666607855523,
      "loss": 2.5476,
      "step": 2628
    },
    {
      "epoch": 0.021032,
      "grad_norm": 4.25,
      "learning_rate": 0.000896289894059771,
      "loss": 2.8554,
      "step": 2629
    },
    {
      "epoch": 0.02104,
      "grad_norm": 3.15625,
      "learning_rate": 0.0008962131022017246,
      "loss": 2.9635,
      "step": 2630
    },
    {
      "epoch": 0.021048,
      "grad_norm": 4.21875,
      "learning_rate": 0.0008961362852162832,
      "loss": 3.0229,
      "step": 2631
    },
    {
      "epoch": 0.021056,
      "grad_norm": 3.484375,
      "learning_rate": 0.0008960594431083184,
      "loss": 2.3788,
      "step": 2632
    },
    {
      "epoch": 0.021064,
      "grad_norm": 3.3125,
      "learning_rate": 0.0008959825758827035,
      "loss": 2.7151,
      "step": 2633
    },
    {
      "epoch": 0.021072,
      "grad_norm": 3.6875,
      "learning_rate": 0.0008959056835443133,
      "loss": 3.3296,
      "step": 2634
    },
    {
      "epoch": 0.02108,
      "grad_norm": 3.765625,
      "learning_rate": 0.000895828766098024,
      "loss": 2.8752,
      "step": 2635
    },
    {
      "epoch": 0.021088,
      "grad_norm": 3.046875,
      "learning_rate": 0.0008957518235487141,
      "loss": 2.3133,
      "step": 2636
    },
    {
      "epoch": 0.021096,
      "grad_norm": 4.21875,
      "learning_rate": 0.0008956748559012629,
      "loss": 2.3479,
      "step": 2637
    },
    {
      "epoch": 0.021104,
      "grad_norm": 3.125,
      "learning_rate": 0.0008955978631605514,
      "loss": 2.9216,
      "step": 2638
    },
    {
      "epoch": 0.021112,
      "grad_norm": 3.0625,
      "learning_rate": 0.0008955208453314629,
      "loss": 2.6983,
      "step": 2639
    },
    {
      "epoch": 0.02112,
      "grad_norm": 3.171875,
      "learning_rate": 0.0008954438024188814,
      "loss": 2.0678,
      "step": 2640
    },
    {
      "epoch": 0.021128,
      "grad_norm": 3.28125,
      "learning_rate": 0.0008953667344276931,
      "loss": 2.7622,
      "step": 2641
    },
    {
      "epoch": 0.021136,
      "grad_norm": 3.4375,
      "learning_rate": 0.0008952896413627855,
      "loss": 2.6003,
      "step": 2642
    },
    {
      "epoch": 0.021144,
      "grad_norm": 3.359375,
      "learning_rate": 0.0008952125232290477,
      "loss": 2.7186,
      "step": 2643
    },
    {
      "epoch": 0.021152,
      "grad_norm": 3.9375,
      "learning_rate": 0.0008951353800313704,
      "loss": 3.3478,
      "step": 2644
    },
    {
      "epoch": 0.02116,
      "grad_norm": 4.28125,
      "learning_rate": 0.0008950582117746461,
      "loss": 2.9125,
      "step": 2645
    },
    {
      "epoch": 0.021168,
      "grad_norm": 3.9375,
      "learning_rate": 0.0008949810184637686,
      "loss": 2.1657,
      "step": 2646
    },
    {
      "epoch": 0.021176,
      "grad_norm": 4.1875,
      "learning_rate": 0.0008949038001036334,
      "loss": 2.6946,
      "step": 2647
    },
    {
      "epoch": 0.021184,
      "grad_norm": 5.03125,
      "learning_rate": 0.0008948265566991377,
      "loss": 3.0319,
      "step": 2648
    },
    {
      "epoch": 0.021192,
      "grad_norm": 6.3125,
      "learning_rate": 0.0008947492882551802,
      "loss": 2.5085,
      "step": 2649
    },
    {
      "epoch": 0.0212,
      "grad_norm": 4.4375,
      "learning_rate": 0.0008946719947766611,
      "loss": 3.0574,
      "step": 2650
    },
    {
      "epoch": 0.021208,
      "grad_norm": 3.4375,
      "learning_rate": 0.0008945946762684823,
      "loss": 2.479,
      "step": 2651
    },
    {
      "epoch": 0.021216,
      "grad_norm": 3.1875,
      "learning_rate": 0.0008945173327355471,
      "loss": 2.4835,
      "step": 2652
    },
    {
      "epoch": 0.021224,
      "grad_norm": 3.34375,
      "learning_rate": 0.0008944399641827608,
      "loss": 1.9927,
      "step": 2653
    },
    {
      "epoch": 0.021232,
      "grad_norm": 5.0625,
      "learning_rate": 0.0008943625706150298,
      "loss": 2.4701,
      "step": 2654
    },
    {
      "epoch": 0.02124,
      "grad_norm": 3.5625,
      "learning_rate": 0.0008942851520372625,
      "loss": 2.5215,
      "step": 2655
    },
    {
      "epoch": 0.021248,
      "grad_norm": 3.453125,
      "learning_rate": 0.0008942077084543687,
      "loss": 2.1113,
      "step": 2656
    },
    {
      "epoch": 0.021256,
      "grad_norm": 3.015625,
      "learning_rate": 0.0008941302398712596,
      "loss": 2.6541,
      "step": 2657
    },
    {
      "epoch": 0.021264,
      "grad_norm": 3.09375,
      "learning_rate": 0.0008940527462928483,
      "loss": 2.0576,
      "step": 2658
    },
    {
      "epoch": 0.021272,
      "grad_norm": 3.90625,
      "learning_rate": 0.0008939752277240493,
      "loss": 2.5583,
      "step": 2659
    },
    {
      "epoch": 0.02128,
      "grad_norm": 3.421875,
      "learning_rate": 0.000893897684169779,
      "loss": 2.6924,
      "step": 2660
    },
    {
      "epoch": 0.021288,
      "grad_norm": 3.109375,
      "learning_rate": 0.0008938201156349548,
      "loss": 2.6125,
      "step": 2661
    },
    {
      "epoch": 0.021296,
      "grad_norm": 3.671875,
      "learning_rate": 0.000893742522124496,
      "loss": 2.0131,
      "step": 2662
    },
    {
      "epoch": 0.021304,
      "grad_norm": 16.875,
      "learning_rate": 0.0008936649036433238,
      "loss": 2.572,
      "step": 2663
    },
    {
      "epoch": 0.021312,
      "grad_norm": 3.65625,
      "learning_rate": 0.0008935872601963604,
      "loss": 2.8629,
      "step": 2664
    },
    {
      "epoch": 0.02132,
      "grad_norm": 3.046875,
      "learning_rate": 0.0008935095917885299,
      "loss": 2.3847,
      "step": 2665
    },
    {
      "epoch": 0.021328,
      "grad_norm": 3.6875,
      "learning_rate": 0.0008934318984247581,
      "loss": 2.8426,
      "step": 2666
    },
    {
      "epoch": 0.021336,
      "grad_norm": 3.671875,
      "learning_rate": 0.0008933541801099719,
      "loss": 2.8297,
      "step": 2667
    },
    {
      "epoch": 0.021344,
      "grad_norm": 3.515625,
      "learning_rate": 0.0008932764368491005,
      "loss": 3.0855,
      "step": 2668
    },
    {
      "epoch": 0.021352,
      "grad_norm": 3.015625,
      "learning_rate": 0.0008931986686470741,
      "loss": 2.8691,
      "step": 2669
    },
    {
      "epoch": 0.02136,
      "grad_norm": 3.15625,
      "learning_rate": 0.0008931208755088247,
      "loss": 2.4085,
      "step": 2670
    },
    {
      "epoch": 0.021368,
      "grad_norm": 3.484375,
      "learning_rate": 0.0008930430574392859,
      "loss": 2.7677,
      "step": 2671
    },
    {
      "epoch": 0.021376,
      "grad_norm": 3.578125,
      "learning_rate": 0.0008929652144433927,
      "loss": 2.3039,
      "step": 2672
    },
    {
      "epoch": 0.021384,
      "grad_norm": 3.359375,
      "learning_rate": 0.000892887346526082,
      "loss": 2.1761,
      "step": 2673
    },
    {
      "epoch": 0.021392,
      "grad_norm": 11.8125,
      "learning_rate": 0.0008928094536922919,
      "loss": 2.535,
      "step": 2674
    },
    {
      "epoch": 0.0214,
      "grad_norm": 3.59375,
      "learning_rate": 0.0008927315359469625,
      "loss": 2.5631,
      "step": 2675
    },
    {
      "epoch": 0.021408,
      "grad_norm": 3.5625,
      "learning_rate": 0.000892653593295035,
      "loss": 2.5955,
      "step": 2676
    },
    {
      "epoch": 0.021416,
      "grad_norm": 3.484375,
      "learning_rate": 0.0008925756257414526,
      "loss": 2.7084,
      "step": 2677
    },
    {
      "epoch": 0.021424,
      "grad_norm": 4.21875,
      "learning_rate": 0.0008924976332911601,
      "loss": 2.4632,
      "step": 2678
    },
    {
      "epoch": 0.021432,
      "grad_norm": 3.921875,
      "learning_rate": 0.0008924196159491031,
      "loss": 2.3484,
      "step": 2679
    },
    {
      "epoch": 0.02144,
      "grad_norm": 4.03125,
      "learning_rate": 0.00089234157372023,
      "loss": 2.7477,
      "step": 2680
    },
    {
      "epoch": 0.021448,
      "grad_norm": 3.484375,
      "learning_rate": 0.00089226350660949,
      "loss": 2.4548,
      "step": 2681
    },
    {
      "epoch": 0.021456,
      "grad_norm": 3.25,
      "learning_rate": 0.0008921854146218338,
      "loss": 2.5051,
      "step": 2682
    },
    {
      "epoch": 0.021464,
      "grad_norm": 4.0625,
      "learning_rate": 0.000892107297762214,
      "loss": 2.7906,
      "step": 2683
    },
    {
      "epoch": 0.021472,
      "grad_norm": 3.359375,
      "learning_rate": 0.0008920291560355846,
      "loss": 2.9013,
      "step": 2684
    },
    {
      "epoch": 0.02148,
      "grad_norm": 3.390625,
      "learning_rate": 0.0008919509894469016,
      "loss": 2.5616,
      "step": 2685
    },
    {
      "epoch": 0.021488,
      "grad_norm": 4.5625,
      "learning_rate": 0.0008918727980011219,
      "loss": 2.4202,
      "step": 2686
    },
    {
      "epoch": 0.021496,
      "grad_norm": 3.546875,
      "learning_rate": 0.0008917945817032047,
      "loss": 1.5655,
      "step": 2687
    },
    {
      "epoch": 0.021504,
      "grad_norm": 4.125,
      "learning_rate": 0.0008917163405581099,
      "loss": 3.0738,
      "step": 2688
    },
    {
      "epoch": 0.021512,
      "grad_norm": 2.78125,
      "learning_rate": 0.0008916380745707997,
      "loss": 3.3932,
      "step": 2689
    },
    {
      "epoch": 0.02152,
      "grad_norm": 2.390625,
      "learning_rate": 0.0008915597837462377,
      "loss": 2.3127,
      "step": 2690
    },
    {
      "epoch": 0.021528,
      "grad_norm": 4.03125,
      "learning_rate": 0.0008914814680893889,
      "loss": 2.5285,
      "step": 2691
    },
    {
      "epoch": 0.021536,
      "grad_norm": 3.015625,
      "learning_rate": 0.0008914031276052201,
      "loss": 2.4783,
      "step": 2692
    },
    {
      "epoch": 0.021544,
      "grad_norm": 2.484375,
      "learning_rate": 0.0008913247622986995,
      "loss": 2.6391,
      "step": 2693
    },
    {
      "epoch": 0.021552,
      "grad_norm": 2.296875,
      "learning_rate": 0.000891246372174797,
      "loss": 1.7518,
      "step": 2694
    },
    {
      "epoch": 0.02156,
      "grad_norm": 3.015625,
      "learning_rate": 0.0008911679572384839,
      "loss": 2.499,
      "step": 2695
    },
    {
      "epoch": 0.021568,
      "grad_norm": 3.203125,
      "learning_rate": 0.0008910895174947333,
      "loss": 2.5082,
      "step": 2696
    },
    {
      "epoch": 0.021576,
      "grad_norm": 3.265625,
      "learning_rate": 0.0008910110529485197,
      "loss": 3.1358,
      "step": 2697
    },
    {
      "epoch": 0.021584,
      "grad_norm": 2.625,
      "learning_rate": 0.0008909325636048193,
      "loss": 2.093,
      "step": 2698
    },
    {
      "epoch": 0.021592,
      "grad_norm": 3.734375,
      "learning_rate": 0.0008908540494686097,
      "loss": 2.8209,
      "step": 2699
    },
    {
      "epoch": 0.0216,
      "grad_norm": 3.1875,
      "learning_rate": 0.0008907755105448703,
      "loss": 2.5655,
      "step": 2700
    },
    {
      "epoch": 0.021608,
      "grad_norm": 3.578125,
      "learning_rate": 0.0008906969468385818,
      "loss": 2.7825,
      "step": 2701
    },
    {
      "epoch": 0.021616,
      "grad_norm": 3.03125,
      "learning_rate": 0.0008906183583547268,
      "loss": 2.4338,
      "step": 2702
    },
    {
      "epoch": 0.021624,
      "grad_norm": 2.96875,
      "learning_rate": 0.0008905397450982891,
      "loss": 2.8684,
      "step": 2703
    },
    {
      "epoch": 0.021632,
      "grad_norm": 3.484375,
      "learning_rate": 0.0008904611070742546,
      "loss": 2.7274,
      "step": 2704
    },
    {
      "epoch": 0.02164,
      "grad_norm": 2.859375,
      "learning_rate": 0.0008903824442876101,
      "loss": 1.9534,
      "step": 2705
    },
    {
      "epoch": 0.021648,
      "grad_norm": 2.890625,
      "learning_rate": 0.0008903037567433443,
      "loss": 2.0491,
      "step": 2706
    },
    {
      "epoch": 0.021656,
      "grad_norm": 2.875,
      "learning_rate": 0.0008902250444464477,
      "loss": 2.2807,
      "step": 2707
    },
    {
      "epoch": 0.021664,
      "grad_norm": 2.9375,
      "learning_rate": 0.0008901463074019121,
      "loss": 2.1918,
      "step": 2708
    },
    {
      "epoch": 0.021672,
      "grad_norm": 3.40625,
      "learning_rate": 0.0008900675456147307,
      "loss": 2.73,
      "step": 2709
    },
    {
      "epoch": 0.02168,
      "grad_norm": 3.1875,
      "learning_rate": 0.0008899887590898988,
      "loss": 2.5409,
      "step": 2710
    },
    {
      "epoch": 0.021688,
      "grad_norm": 3.34375,
      "learning_rate": 0.0008899099478324126,
      "loss": 3.2864,
      "step": 2711
    },
    {
      "epoch": 0.021696,
      "grad_norm": 3.3125,
      "learning_rate": 0.0008898311118472706,
      "loss": 2.4325,
      "step": 2712
    },
    {
      "epoch": 0.021704,
      "grad_norm": 3.671875,
      "learning_rate": 0.0008897522511394723,
      "loss": 2.6474,
      "step": 2713
    },
    {
      "epoch": 0.021712,
      "grad_norm": 2.609375,
      "learning_rate": 0.000889673365714019,
      "loss": 2.2069,
      "step": 2714
    },
    {
      "epoch": 0.02172,
      "grad_norm": 3.3125,
      "learning_rate": 0.0008895944555759134,
      "loss": 2.8683,
      "step": 2715
    },
    {
      "epoch": 0.021728,
      "grad_norm": 2.515625,
      "learning_rate": 0.00088951552073016,
      "loss": 2.6151,
      "step": 2716
    },
    {
      "epoch": 0.021736,
      "grad_norm": 2.875,
      "learning_rate": 0.0008894365611817648,
      "loss": 2.3645,
      "step": 2717
    },
    {
      "epoch": 0.021744,
      "grad_norm": 2.8125,
      "learning_rate": 0.0008893575769357352,
      "loss": 2.3964,
      "step": 2718
    },
    {
      "epoch": 0.021752,
      "grad_norm": 3.25,
      "learning_rate": 0.0008892785679970804,
      "loss": 1.9139,
      "step": 2719
    },
    {
      "epoch": 0.02176,
      "grad_norm": 2.796875,
      "learning_rate": 0.0008891995343708111,
      "loss": 2.5362,
      "step": 2720
    },
    {
      "epoch": 0.021768,
      "grad_norm": 3.09375,
      "learning_rate": 0.0008891204760619392,
      "loss": 2.6862,
      "step": 2721
    },
    {
      "epoch": 0.021776,
      "grad_norm": 3.34375,
      "learning_rate": 0.0008890413930754791,
      "loss": 2.4314,
      "step": 2722
    },
    {
      "epoch": 0.021784,
      "grad_norm": 3.515625,
      "learning_rate": 0.0008889622854164454,
      "loss": 2.8433,
      "step": 2723
    },
    {
      "epoch": 0.021792,
      "grad_norm": 4.34375,
      "learning_rate": 0.0008888831530898558,
      "loss": 3.4026,
      "step": 2724
    },
    {
      "epoch": 0.0218,
      "grad_norm": 2.96875,
      "learning_rate": 0.0008888039961007281,
      "loss": 2.4028,
      "step": 2725
    },
    {
      "epoch": 0.021808,
      "grad_norm": 3.0625,
      "learning_rate": 0.000888724814454083,
      "loss": 2.6242,
      "step": 2726
    },
    {
      "epoch": 0.021816,
      "grad_norm": 3.6875,
      "learning_rate": 0.0008886456081549414,
      "loss": 2.8139,
      "step": 2727
    },
    {
      "epoch": 0.021824,
      "grad_norm": 3.28125,
      "learning_rate": 0.000888566377208327,
      "loss": 2.8317,
      "step": 2728
    },
    {
      "epoch": 0.021832,
      "grad_norm": 3.203125,
      "learning_rate": 0.0008884871216192643,
      "loss": 2.4725,
      "step": 2729
    },
    {
      "epoch": 0.02184,
      "grad_norm": 2.84375,
      "learning_rate": 0.0008884078413927799,
      "loss": 1.8723,
      "step": 2730
    },
    {
      "epoch": 0.021848,
      "grad_norm": 2.890625,
      "learning_rate": 0.0008883285365339013,
      "loss": 2.7564,
      "step": 2731
    },
    {
      "epoch": 0.021856,
      "grad_norm": 2.671875,
      "learning_rate": 0.000888249207047658,
      "loss": 2.455,
      "step": 2732
    },
    {
      "epoch": 0.021864,
      "grad_norm": 2.25,
      "learning_rate": 0.0008881698529390811,
      "loss": 1.7137,
      "step": 2733
    },
    {
      "epoch": 0.021872,
      "grad_norm": 3.125,
      "learning_rate": 0.000888090474213203,
      "loss": 2.4261,
      "step": 2734
    },
    {
      "epoch": 0.02188,
      "grad_norm": 3.28125,
      "learning_rate": 0.0008880110708750581,
      "loss": 2.7401,
      "step": 2735
    },
    {
      "epoch": 0.021888,
      "grad_norm": 2.671875,
      "learning_rate": 0.0008879316429296817,
      "loss": 2.0955,
      "step": 2736
    },
    {
      "epoch": 0.021896,
      "grad_norm": 3.03125,
      "learning_rate": 0.0008878521903821115,
      "loss": 2.4692,
      "step": 2737
    },
    {
      "epoch": 0.021904,
      "grad_norm": 3.140625,
      "learning_rate": 0.0008877727132373858,
      "loss": 2.3211,
      "step": 2738
    },
    {
      "epoch": 0.021912,
      "grad_norm": 4.0,
      "learning_rate": 0.0008876932115005452,
      "loss": 3.1478,
      "step": 2739
    },
    {
      "epoch": 0.02192,
      "grad_norm": 3.4375,
      "learning_rate": 0.0008876136851766316,
      "loss": 2.5793,
      "step": 2740
    },
    {
      "epoch": 0.021928,
      "grad_norm": 3.25,
      "learning_rate": 0.0008875341342706884,
      "loss": 2.8439,
      "step": 2741
    },
    {
      "epoch": 0.021936,
      "grad_norm": 3.546875,
      "learning_rate": 0.0008874545587877607,
      "loss": 2.9575,
      "step": 2742
    },
    {
      "epoch": 0.021944,
      "grad_norm": 2.859375,
      "learning_rate": 0.0008873749587328951,
      "loss": 2.421,
      "step": 2743
    },
    {
      "epoch": 0.021952,
      "grad_norm": 3.171875,
      "learning_rate": 0.0008872953341111397,
      "loss": 2.6495,
      "step": 2744
    },
    {
      "epoch": 0.02196,
      "grad_norm": 3.46875,
      "learning_rate": 0.0008872156849275442,
      "loss": 2.2591,
      "step": 2745
    },
    {
      "epoch": 0.021968,
      "grad_norm": 3.046875,
      "learning_rate": 0.0008871360111871598,
      "loss": 2.775,
      "step": 2746
    },
    {
      "epoch": 0.021976,
      "grad_norm": 3.625,
      "learning_rate": 0.0008870563128950395,
      "loss": 2.8198,
      "step": 2747
    },
    {
      "epoch": 0.021984,
      "grad_norm": 3.09375,
      "learning_rate": 0.0008869765900562376,
      "loss": 2.9607,
      "step": 2748
    },
    {
      "epoch": 0.021992,
      "grad_norm": 2.953125,
      "learning_rate": 0.0008868968426758101,
      "loss": 2.5933,
      "step": 2749
    },
    {
      "epoch": 0.022,
      "grad_norm": 2.5625,
      "learning_rate": 0.0008868170707588142,
      "loss": 3.2335,
      "step": 2750
    },
    {
      "epoch": 0.022008,
      "grad_norm": 2.65625,
      "learning_rate": 0.0008867372743103092,
      "loss": 2.535,
      "step": 2751
    },
    {
      "epoch": 0.022016,
      "grad_norm": 2.796875,
      "learning_rate": 0.0008866574533353557,
      "loss": 3.044,
      "step": 2752
    },
    {
      "epoch": 0.022024,
      "grad_norm": 2.5625,
      "learning_rate": 0.0008865776078390157,
      "loss": 2.2158,
      "step": 2753
    },
    {
      "epoch": 0.022032,
      "grad_norm": 3.40625,
      "learning_rate": 0.0008864977378263533,
      "loss": 2.188,
      "step": 2754
    },
    {
      "epoch": 0.02204,
      "grad_norm": 3.0625,
      "learning_rate": 0.0008864178433024332,
      "loss": 2.7737,
      "step": 2755
    },
    {
      "epoch": 0.022048,
      "grad_norm": 2.859375,
      "learning_rate": 0.0008863379242723226,
      "loss": 2.3352,
      "step": 2756
    },
    {
      "epoch": 0.022056,
      "grad_norm": 2.671875,
      "learning_rate": 0.0008862579807410898,
      "loss": 2.5635,
      "step": 2757
    },
    {
      "epoch": 0.022064,
      "grad_norm": 3.953125,
      "learning_rate": 0.0008861780127138048,
      "loss": 2.6765,
      "step": 2758
    },
    {
      "epoch": 0.022072,
      "grad_norm": 2.984375,
      "learning_rate": 0.0008860980201955389,
      "loss": 1.8557,
      "step": 2759
    },
    {
      "epoch": 0.02208,
      "grad_norm": 3.78125,
      "learning_rate": 0.0008860180031913653,
      "loss": 1.9509,
      "step": 2760
    },
    {
      "epoch": 0.022088,
      "grad_norm": 2.84375,
      "learning_rate": 0.0008859379617063585,
      "loss": 1.9022,
      "step": 2761
    },
    {
      "epoch": 0.022096,
      "grad_norm": 2.921875,
      "learning_rate": 0.0008858578957455947,
      "loss": 2.6258,
      "step": 2762
    },
    {
      "epoch": 0.022104,
      "grad_norm": 3.03125,
      "learning_rate": 0.0008857778053141514,
      "loss": 2.5362,
      "step": 2763
    },
    {
      "epoch": 0.022112,
      "grad_norm": 3.578125,
      "learning_rate": 0.0008856976904171081,
      "loss": 2.2888,
      "step": 2764
    },
    {
      "epoch": 0.02212,
      "grad_norm": 3.3125,
      "learning_rate": 0.0008856175510595455,
      "loss": 2.3145,
      "step": 2765
    },
    {
      "epoch": 0.022128,
      "grad_norm": 3.234375,
      "learning_rate": 0.000885537387246546,
      "loss": 2.5289,
      "step": 2766
    },
    {
      "epoch": 0.022136,
      "grad_norm": 3.015625,
      "learning_rate": 0.0008854571989831935,
      "loss": 2.4304,
      "step": 2767
    },
    {
      "epoch": 0.022144,
      "grad_norm": 2.5625,
      "learning_rate": 0.0008853769862745733,
      "loss": 2.4834,
      "step": 2768
    },
    {
      "epoch": 0.022152,
      "grad_norm": 3.15625,
      "learning_rate": 0.0008852967491257726,
      "loss": 2.8067,
      "step": 2769
    },
    {
      "epoch": 0.02216,
      "grad_norm": 3.3125,
      "learning_rate": 0.0008852164875418798,
      "loss": 2.4413,
      "step": 2770
    },
    {
      "epoch": 0.022168,
      "grad_norm": 2.78125,
      "learning_rate": 0.0008851362015279852,
      "loss": 2.5524,
      "step": 2771
    },
    {
      "epoch": 0.022176,
      "grad_norm": 2.90625,
      "learning_rate": 0.0008850558910891801,
      "loss": 2.6897,
      "step": 2772
    },
    {
      "epoch": 0.022184,
      "grad_norm": 14.75,
      "learning_rate": 0.000884975556230558,
      "loss": 2.6999,
      "step": 2773
    },
    {
      "epoch": 0.022192,
      "grad_norm": 3.125,
      "learning_rate": 0.0008848951969572137,
      "loss": 2.1837,
      "step": 2774
    },
    {
      "epoch": 0.0222,
      "grad_norm": 3.046875,
      "learning_rate": 0.000884814813274243,
      "loss": 2.8914,
      "step": 2775
    },
    {
      "epoch": 0.022208,
      "grad_norm": 2.6875,
      "learning_rate": 0.0008847344051867443,
      "loss": 2.2246,
      "step": 2776
    },
    {
      "epoch": 0.022216,
      "grad_norm": 2.515625,
      "learning_rate": 0.0008846539726998169,
      "loss": 2.52,
      "step": 2777
    },
    {
      "epoch": 0.022224,
      "grad_norm": 3.0,
      "learning_rate": 0.0008845735158185615,
      "loss": 2.7436,
      "step": 2778
    },
    {
      "epoch": 0.022232,
      "grad_norm": 3.109375,
      "learning_rate": 0.0008844930345480807,
      "loss": 2.7163,
      "step": 2779
    },
    {
      "epoch": 0.02224,
      "grad_norm": 3.9375,
      "learning_rate": 0.0008844125288934786,
      "loss": 2.7293,
      "step": 2780
    },
    {
      "epoch": 0.022248,
      "grad_norm": 3.09375,
      "learning_rate": 0.0008843319988598607,
      "loss": 2.5206,
      "step": 2781
    },
    {
      "epoch": 0.022256,
      "grad_norm": 2.859375,
      "learning_rate": 0.0008842514444523342,
      "loss": 2.0235,
      "step": 2782
    },
    {
      "epoch": 0.022264,
      "grad_norm": 2.703125,
      "learning_rate": 0.0008841708656760075,
      "loss": 1.846,
      "step": 2783
    },
    {
      "epoch": 0.022272,
      "grad_norm": 3.140625,
      "learning_rate": 0.0008840902625359912,
      "loss": 2.215,
      "step": 2784
    },
    {
      "epoch": 0.02228,
      "grad_norm": 4.46875,
      "learning_rate": 0.0008840096350373971,
      "loss": 2.5577,
      "step": 2785
    },
    {
      "epoch": 0.022288,
      "grad_norm": 3.484375,
      "learning_rate": 0.0008839289831853381,
      "loss": 2.711,
      "step": 2786
    },
    {
      "epoch": 0.022296,
      "grad_norm": 2.578125,
      "learning_rate": 0.0008838483069849292,
      "loss": 2.0991,
      "step": 2787
    },
    {
      "epoch": 0.022304,
      "grad_norm": 2.84375,
      "learning_rate": 0.000883767606441287,
      "loss": 2.4561,
      "step": 2788
    },
    {
      "epoch": 0.022312,
      "grad_norm": 3.09375,
      "learning_rate": 0.0008836868815595291,
      "loss": 2.5592,
      "step": 2789
    },
    {
      "epoch": 0.02232,
      "grad_norm": 2.984375,
      "learning_rate": 0.0008836061323447752,
      "loss": 2.9781,
      "step": 2790
    },
    {
      "epoch": 0.022328,
      "grad_norm": 2.640625,
      "learning_rate": 0.0008835253588021465,
      "loss": 2.5323,
      "step": 2791
    },
    {
      "epoch": 0.022336,
      "grad_norm": 3.34375,
      "learning_rate": 0.0008834445609367652,
      "loss": 2.1223,
      "step": 2792
    },
    {
      "epoch": 0.022344,
      "grad_norm": 2.796875,
      "learning_rate": 0.0008833637387537556,
      "loss": 2.587,
      "step": 2793
    },
    {
      "epoch": 0.022352,
      "grad_norm": 2.796875,
      "learning_rate": 0.0008832828922582434,
      "loss": 2.4655,
      "step": 2794
    },
    {
      "epoch": 0.02236,
      "grad_norm": 3.265625,
      "learning_rate": 0.0008832020214553558,
      "loss": 2.3633,
      "step": 2795
    },
    {
      "epoch": 0.022368,
      "grad_norm": 2.796875,
      "learning_rate": 0.0008831211263502213,
      "loss": 2.4294,
      "step": 2796
    },
    {
      "epoch": 0.022376,
      "grad_norm": 2.328125,
      "learning_rate": 0.0008830402069479703,
      "loss": 2.1127,
      "step": 2797
    },
    {
      "epoch": 0.022384,
      "grad_norm": 2.734375,
      "learning_rate": 0.0008829592632537349,
      "loss": 2.4517,
      "step": 2798
    },
    {
      "epoch": 0.022392,
      "grad_norm": 3.484375,
      "learning_rate": 0.0008828782952726479,
      "loss": 2.3421,
      "step": 2799
    },
    {
      "epoch": 0.0224,
      "grad_norm": 2.65625,
      "learning_rate": 0.0008827973030098447,
      "loss": 2.0168,
      "step": 2800
    },
    {
      "epoch": 0.022408,
      "grad_norm": 3.6875,
      "learning_rate": 0.0008827162864704616,
      "loss": 2.2745,
      "step": 2801
    },
    {
      "epoch": 0.022416,
      "grad_norm": 3.390625,
      "learning_rate": 0.0008826352456596365,
      "loss": 2.3537,
      "step": 2802
    },
    {
      "epoch": 0.022424,
      "grad_norm": 3.28125,
      "learning_rate": 0.0008825541805825088,
      "loss": 2.5425,
      "step": 2803
    },
    {
      "epoch": 0.022432,
      "grad_norm": 2.859375,
      "learning_rate": 0.0008824730912442198,
      "loss": 2.2649,
      "step": 2804
    },
    {
      "epoch": 0.02244,
      "grad_norm": 3.421875,
      "learning_rate": 0.000882391977649912,
      "loss": 2.762,
      "step": 2805
    },
    {
      "epoch": 0.022448,
      "grad_norm": 3.75,
      "learning_rate": 0.0008823108398047297,
      "loss": 2.5092,
      "step": 2806
    },
    {
      "epoch": 0.022456,
      "grad_norm": 3.984375,
      "learning_rate": 0.0008822296777138183,
      "loss": 3.276,
      "step": 2807
    },
    {
      "epoch": 0.022464,
      "grad_norm": 3.171875,
      "learning_rate": 0.0008821484913823252,
      "loss": 2.6396,
      "step": 2808
    },
    {
      "epoch": 0.022472,
      "grad_norm": 3.421875,
      "learning_rate": 0.0008820672808153989,
      "loss": 2.2788,
      "step": 2809
    },
    {
      "epoch": 0.02248,
      "grad_norm": 2.625,
      "learning_rate": 0.00088198604601819,
      "loss": 2.8166,
      "step": 2810
    },
    {
      "epoch": 0.022488,
      "grad_norm": 2.59375,
      "learning_rate": 0.0008819047869958501,
      "loss": 2.7837,
      "step": 2811
    },
    {
      "epoch": 0.022496,
      "grad_norm": 2.84375,
      "learning_rate": 0.0008818235037535327,
      "loss": 3.1881,
      "step": 2812
    },
    {
      "epoch": 0.022504,
      "grad_norm": 3.015625,
      "learning_rate": 0.0008817421962963925,
      "loss": 3.0649,
      "step": 2813
    },
    {
      "epoch": 0.022512,
      "grad_norm": 9.625,
      "learning_rate": 0.0008816608646295861,
      "loss": 2.7576,
      "step": 2814
    },
    {
      "epoch": 0.02252,
      "grad_norm": 3.09375,
      "learning_rate": 0.0008815795087582715,
      "loss": 2.3427,
      "step": 2815
    },
    {
      "epoch": 0.022528,
      "grad_norm": 2.71875,
      "learning_rate": 0.0008814981286876081,
      "loss": 2.1709,
      "step": 2816
    },
    {
      "epoch": 0.022536,
      "grad_norm": 2.609375,
      "learning_rate": 0.0008814167244227569,
      "loss": 2.2394,
      "step": 2817
    },
    {
      "epoch": 0.022544,
      "grad_norm": 2.296875,
      "learning_rate": 0.0008813352959688805,
      "loss": 1.9512,
      "step": 2818
    },
    {
      "epoch": 0.022552,
      "grad_norm": 3.1875,
      "learning_rate": 0.000881253843331143,
      "loss": 2.6879,
      "step": 2819
    },
    {
      "epoch": 0.02256,
      "grad_norm": 2.859375,
      "learning_rate": 0.0008811723665147101,
      "loss": 2.7899,
      "step": 2820
    },
    {
      "epoch": 0.022568,
      "grad_norm": 5.625,
      "learning_rate": 0.0008810908655247489,
      "loss": 2.3615,
      "step": 2821
    },
    {
      "epoch": 0.022576,
      "grad_norm": 2.78125,
      "learning_rate": 0.0008810093403664281,
      "loss": 2.3863,
      "step": 2822
    },
    {
      "epoch": 0.022584,
      "grad_norm": 4.3125,
      "learning_rate": 0.000880927791044918,
      "loss": 2.8119,
      "step": 2823
    },
    {
      "epoch": 0.022592,
      "grad_norm": 2.640625,
      "learning_rate": 0.0008808462175653904,
      "loss": 2.3276,
      "step": 2824
    },
    {
      "epoch": 0.0226,
      "grad_norm": 2.625,
      "learning_rate": 0.0008807646199330186,
      "loss": 2.5113,
      "step": 2825
    },
    {
      "epoch": 0.022608,
      "grad_norm": 2.65625,
      "learning_rate": 0.0008806829981529772,
      "loss": 2.2518,
      "step": 2826
    },
    {
      "epoch": 0.022616,
      "grad_norm": 2.734375,
      "learning_rate": 0.0008806013522304427,
      "loss": 3.0695,
      "step": 2827
    },
    {
      "epoch": 0.022624,
      "grad_norm": 2.703125,
      "learning_rate": 0.0008805196821705931,
      "loss": 2.6053,
      "step": 2828
    },
    {
      "epoch": 0.022632,
      "grad_norm": 2.40625,
      "learning_rate": 0.0008804379879786078,
      "loss": 2.4383,
      "step": 2829
    },
    {
      "epoch": 0.02264,
      "grad_norm": 2.890625,
      "learning_rate": 0.0008803562696596675,
      "loss": 2.8487,
      "step": 2830
    },
    {
      "epoch": 0.022648,
      "grad_norm": 2.734375,
      "learning_rate": 0.0008802745272189549,
      "loss": 2.8328,
      "step": 2831
    },
    {
      "epoch": 0.022656,
      "grad_norm": 2.71875,
      "learning_rate": 0.0008801927606616542,
      "loss": 3.1136,
      "step": 2832
    },
    {
      "epoch": 0.022664,
      "grad_norm": 2.484375,
      "learning_rate": 0.0008801109699929506,
      "loss": 2.4261,
      "step": 2833
    },
    {
      "epoch": 0.022672,
      "grad_norm": 2.765625,
      "learning_rate": 0.0008800291552180314,
      "loss": 2.5252,
      "step": 2834
    },
    {
      "epoch": 0.02268,
      "grad_norm": 3.34375,
      "learning_rate": 0.0008799473163420851,
      "loss": 2.5203,
      "step": 2835
    },
    {
      "epoch": 0.022688,
      "grad_norm": 3.421875,
      "learning_rate": 0.0008798654533703017,
      "loss": 2.5983,
      "step": 2836
    },
    {
      "epoch": 0.022696,
      "grad_norm": 2.640625,
      "learning_rate": 0.0008797835663078732,
      "loss": 2.285,
      "step": 2837
    },
    {
      "epoch": 0.022704,
      "grad_norm": 2.9375,
      "learning_rate": 0.0008797016551599924,
      "loss": 2.721,
      "step": 2838
    },
    {
      "epoch": 0.022712,
      "grad_norm": 2.8125,
      "learning_rate": 0.0008796197199318544,
      "loss": 2.4805,
      "step": 2839
    },
    {
      "epoch": 0.02272,
      "grad_norm": 3.46875,
      "learning_rate": 0.0008795377606286552,
      "loss": 2.7407,
      "step": 2840
    },
    {
      "epoch": 0.022728,
      "grad_norm": 3.296875,
      "learning_rate": 0.0008794557772555925,
      "loss": 2.7285,
      "step": 2841
    },
    {
      "epoch": 0.022736,
      "grad_norm": 2.828125,
      "learning_rate": 0.0008793737698178658,
      "loss": 2.4165,
      "step": 2842
    },
    {
      "epoch": 0.022744,
      "grad_norm": 3.125,
      "learning_rate": 0.0008792917383206758,
      "loss": 2.4299,
      "step": 2843
    },
    {
      "epoch": 0.022752,
      "grad_norm": 2.828125,
      "learning_rate": 0.0008792096827692249,
      "loss": 2.212,
      "step": 2844
    },
    {
      "epoch": 0.02276,
      "grad_norm": 3.28125,
      "learning_rate": 0.000879127603168717,
      "loss": 2.2449,
      "step": 2845
    },
    {
      "epoch": 0.022768,
      "grad_norm": 3.375,
      "learning_rate": 0.0008790454995243573,
      "loss": 2.6798,
      "step": 2846
    },
    {
      "epoch": 0.022776,
      "grad_norm": 2.53125,
      "learning_rate": 0.000878963371841353,
      "loss": 2.1559,
      "step": 2847
    },
    {
      "epoch": 0.022784,
      "grad_norm": 3.28125,
      "learning_rate": 0.0008788812201249123,
      "loss": 2.6488,
      "step": 2848
    },
    {
      "epoch": 0.022792,
      "grad_norm": 3.140625,
      "learning_rate": 0.0008787990443802453,
      "loss": 2.6832,
      "step": 2849
    },
    {
      "epoch": 0.0228,
      "grad_norm": 2.875,
      "learning_rate": 0.0008787168446125638,
      "loss": 2.8295,
      "step": 2850
    },
    {
      "epoch": 0.022808,
      "grad_norm": 2.90625,
      "learning_rate": 0.0008786346208270802,
      "loss": 2.2445,
      "step": 2851
    },
    {
      "epoch": 0.022816,
      "grad_norm": 3.359375,
      "learning_rate": 0.0008785523730290093,
      "loss": 2.466,
      "step": 2852
    },
    {
      "epoch": 0.022824,
      "grad_norm": 3.0,
      "learning_rate": 0.0008784701012235672,
      "loss": 2.1399,
      "step": 2853
    },
    {
      "epoch": 0.022832,
      "grad_norm": 3.734375,
      "learning_rate": 0.0008783878054159717,
      "loss": 2.5488,
      "step": 2854
    },
    {
      "epoch": 0.02284,
      "grad_norm": 2.515625,
      "learning_rate": 0.0008783054856114414,
      "loss": 2.1396,
      "step": 2855
    },
    {
      "epoch": 0.022848,
      "grad_norm": 3.3125,
      "learning_rate": 0.0008782231418151975,
      "loss": 3.2186,
      "step": 2856
    },
    {
      "epoch": 0.022856,
      "grad_norm": 2.75,
      "learning_rate": 0.0008781407740324617,
      "loss": 2.4643,
      "step": 2857
    },
    {
      "epoch": 0.022864,
      "grad_norm": 3.28125,
      "learning_rate": 0.000878058382268458,
      "loss": 2.2218,
      "step": 2858
    },
    {
      "epoch": 0.022872,
      "grad_norm": 2.8125,
      "learning_rate": 0.0008779759665284114,
      "loss": 2.3279,
      "step": 2859
    },
    {
      "epoch": 0.02288,
      "grad_norm": 7.6875,
      "learning_rate": 0.0008778935268175485,
      "loss": 2.0179,
      "step": 2860
    },
    {
      "epoch": 0.022888,
      "grad_norm": 16.25,
      "learning_rate": 0.000877811063141098,
      "loss": 2.4557,
      "step": 2861
    },
    {
      "epoch": 0.022896,
      "grad_norm": 3.046875,
      "learning_rate": 0.0008777285755042891,
      "loss": 2.957,
      "step": 2862
    },
    {
      "epoch": 0.022904,
      "grad_norm": 3.25,
      "learning_rate": 0.0008776460639123533,
      "loss": 3.0638,
      "step": 2863
    },
    {
      "epoch": 0.022912,
      "grad_norm": 2.953125,
      "learning_rate": 0.0008775635283705237,
      "loss": 2.4461,
      "step": 2864
    },
    {
      "epoch": 0.02292,
      "grad_norm": 2.84375,
      "learning_rate": 0.0008774809688840341,
      "loss": 2.6661,
      "step": 2865
    },
    {
      "epoch": 0.022928,
      "grad_norm": 3.328125,
      "learning_rate": 0.0008773983854581207,
      "loss": 2.219,
      "step": 2866
    },
    {
      "epoch": 0.022936,
      "grad_norm": 2.78125,
      "learning_rate": 0.0008773157780980206,
      "loss": 2.4478,
      "step": 2867
    },
    {
      "epoch": 0.022944,
      "grad_norm": 2.375,
      "learning_rate": 0.0008772331468089729,
      "loss": 2.3522,
      "step": 2868
    },
    {
      "epoch": 0.022952,
      "grad_norm": 2.9375,
      "learning_rate": 0.0008771504915962179,
      "loss": 2.4297,
      "step": 2869
    },
    {
      "epoch": 0.02296,
      "grad_norm": 3.3125,
      "learning_rate": 0.0008770678124649972,
      "loss": 2.7615,
      "step": 2870
    },
    {
      "epoch": 0.022968,
      "grad_norm": 3.484375,
      "learning_rate": 0.0008769851094205548,
      "loss": 3.221,
      "step": 2871
    },
    {
      "epoch": 0.022976,
      "grad_norm": 2.5625,
      "learning_rate": 0.000876902382468135,
      "loss": 2.6064,
      "step": 2872
    },
    {
      "epoch": 0.022984,
      "grad_norm": 3.328125,
      "learning_rate": 0.0008768196316129848,
      "loss": 2.4384,
      "step": 2873
    },
    {
      "epoch": 0.022992,
      "grad_norm": 2.5625,
      "learning_rate": 0.0008767368568603519,
      "loss": 2.0107,
      "step": 2874
    },
    {
      "epoch": 0.023,
      "grad_norm": 2.328125,
      "learning_rate": 0.0008766540582154859,
      "loss": 2.2702,
      "step": 2875
    },
    {
      "epoch": 0.023008,
      "grad_norm": 3.59375,
      "learning_rate": 0.0008765712356836375,
      "loss": 2.5891,
      "step": 2876
    },
    {
      "epoch": 0.023016,
      "grad_norm": 3.171875,
      "learning_rate": 0.0008764883892700596,
      "loss": 2.7782,
      "step": 2877
    },
    {
      "epoch": 0.023024,
      "grad_norm": 2.609375,
      "learning_rate": 0.000876405518980006,
      "loss": 2.3485,
      "step": 2878
    },
    {
      "epoch": 0.023032,
      "grad_norm": 2.84375,
      "learning_rate": 0.0008763226248187324,
      "loss": 2.4231,
      "step": 2879
    },
    {
      "epoch": 0.02304,
      "grad_norm": 2.8125,
      "learning_rate": 0.0008762397067914956,
      "loss": 2.6416,
      "step": 2880
    },
    {
      "epoch": 0.023048,
      "grad_norm": 3.671875,
      "learning_rate": 0.0008761567649035545,
      "loss": 1.9322,
      "step": 2881
    },
    {
      "epoch": 0.023056,
      "grad_norm": 2.78125,
      "learning_rate": 0.0008760737991601688,
      "loss": 2.4219,
      "step": 2882
    },
    {
      "epoch": 0.023064,
      "grad_norm": 2.90625,
      "learning_rate": 0.0008759908095666005,
      "loss": 2.3416,
      "step": 2883
    },
    {
      "epoch": 0.023072,
      "grad_norm": 3.265625,
      "learning_rate": 0.0008759077961281123,
      "loss": 2.8897,
      "step": 2884
    },
    {
      "epoch": 0.02308,
      "grad_norm": 3.6875,
      "learning_rate": 0.0008758247588499693,
      "loss": 2.3112,
      "step": 2885
    },
    {
      "epoch": 0.023088,
      "grad_norm": 4.15625,
      "learning_rate": 0.0008757416977374372,
      "loss": 3.238,
      "step": 2886
    },
    {
      "epoch": 0.023096,
      "grad_norm": 3.0625,
      "learning_rate": 0.0008756586127957838,
      "loss": 2.5813,
      "step": 2887
    },
    {
      "epoch": 0.023104,
      "grad_norm": 3.140625,
      "learning_rate": 0.0008755755040302784,
      "loss": 3.0471,
      "step": 2888
    },
    {
      "epoch": 0.023112,
      "grad_norm": 2.921875,
      "learning_rate": 0.0008754923714461915,
      "loss": 2.4548,
      "step": 2889
    },
    {
      "epoch": 0.02312,
      "grad_norm": 2.828125,
      "learning_rate": 0.0008754092150487951,
      "loss": 2.4525,
      "step": 2890
    },
    {
      "epoch": 0.023128,
      "grad_norm": 2.875,
      "learning_rate": 0.0008753260348433634,
      "loss": 2.4126,
      "step": 2891
    },
    {
      "epoch": 0.023136,
      "grad_norm": 2.984375,
      "learning_rate": 0.000875242830835171,
      "loss": 2.6144,
      "step": 2892
    },
    {
      "epoch": 0.023144,
      "grad_norm": 3.125,
      "learning_rate": 0.0008751596030294951,
      "loss": 2.7061,
      "step": 2893
    },
    {
      "epoch": 0.023152,
      "grad_norm": 2.953125,
      "learning_rate": 0.0008750763514316135,
      "loss": 2.1614,
      "step": 2894
    },
    {
      "epoch": 0.02316,
      "grad_norm": 3.046875,
      "learning_rate": 0.0008749930760468063,
      "loss": 2.3196,
      "step": 2895
    },
    {
      "epoch": 0.023168,
      "grad_norm": 3.078125,
      "learning_rate": 0.0008749097768803546,
      "loss": 2.5121,
      "step": 2896
    },
    {
      "epoch": 0.023176,
      "grad_norm": 3.234375,
      "learning_rate": 0.000874826453937541,
      "loss": 2.7553,
      "step": 2897
    },
    {
      "epoch": 0.023184,
      "grad_norm": 3.09375,
      "learning_rate": 0.0008747431072236499,
      "loss": 2.6713,
      "step": 2898
    },
    {
      "epoch": 0.023192,
      "grad_norm": 3.234375,
      "learning_rate": 0.000874659736743967,
      "loss": 2.6395,
      "step": 2899
    },
    {
      "epoch": 0.0232,
      "grad_norm": 2.890625,
      "learning_rate": 0.0008745763425037796,
      "loss": 2.8244,
      "step": 2900
    },
    {
      "epoch": 0.023208,
      "grad_norm": 2.71875,
      "learning_rate": 0.0008744929245083764,
      "loss": 2.2366,
      "step": 2901
    },
    {
      "epoch": 0.023216,
      "grad_norm": 2.6875,
      "learning_rate": 0.0008744094827630478,
      "loss": 2.4064,
      "step": 2902
    },
    {
      "epoch": 0.023224,
      "grad_norm": 3.28125,
      "learning_rate": 0.0008743260172730853,
      "loss": 2.3792,
      "step": 2903
    },
    {
      "epoch": 0.023232,
      "grad_norm": 2.9375,
      "learning_rate": 0.0008742425280437826,
      "loss": 2.2661,
      "step": 2904
    },
    {
      "epoch": 0.02324,
      "grad_norm": 2.484375,
      "learning_rate": 0.0008741590150804343,
      "loss": 1.8845,
      "step": 2905
    },
    {
      "epoch": 0.023248,
      "grad_norm": 3.296875,
      "learning_rate": 0.0008740754783883365,
      "loss": 2.7587,
      "step": 2906
    },
    {
      "epoch": 0.023256,
      "grad_norm": 2.890625,
      "learning_rate": 0.0008739919179727874,
      "loss": 2.256,
      "step": 2907
    },
    {
      "epoch": 0.023264,
      "grad_norm": 3.34375,
      "learning_rate": 0.0008739083338390861,
      "loss": 2.8376,
      "step": 2908
    },
    {
      "epoch": 0.023272,
      "grad_norm": 3.171875,
      "learning_rate": 0.0008738247259925333,
      "loss": 2.7308,
      "step": 2909
    },
    {
      "epoch": 0.02328,
      "grad_norm": 3.453125,
      "learning_rate": 0.0008737410944384314,
      "loss": 2.9598,
      "step": 2910
    },
    {
      "epoch": 0.023288,
      "grad_norm": 3.0,
      "learning_rate": 0.0008736574391820842,
      "loss": 2.3382,
      "step": 2911
    },
    {
      "epoch": 0.023296,
      "grad_norm": 3.3125,
      "learning_rate": 0.0008735737602287972,
      "loss": 2.5425,
      "step": 2912
    },
    {
      "epoch": 0.023304,
      "grad_norm": 2.625,
      "learning_rate": 0.0008734900575838771,
      "loss": 2.1453,
      "step": 2913
    },
    {
      "epoch": 0.023312,
      "grad_norm": 2.8125,
      "learning_rate": 0.0008734063312526323,
      "loss": 2.6776,
      "step": 2914
    },
    {
      "epoch": 0.02332,
      "grad_norm": 2.375,
      "learning_rate": 0.0008733225812403723,
      "loss": 2.1398,
      "step": 2915
    },
    {
      "epoch": 0.023328,
      "grad_norm": 2.859375,
      "learning_rate": 0.0008732388075524089,
      "loss": 2.4419,
      "step": 2916
    },
    {
      "epoch": 0.023336,
      "grad_norm": 2.984375,
      "learning_rate": 0.0008731550101940546,
      "loss": 2.5157,
      "step": 2917
    },
    {
      "epoch": 0.023344,
      "grad_norm": 4.71875,
      "learning_rate": 0.000873071189170624,
      "loss": 2.119,
      "step": 2918
    },
    {
      "epoch": 0.023352,
      "grad_norm": 2.4375,
      "learning_rate": 0.0008729873444874328,
      "loss": 1.9582,
      "step": 2919
    },
    {
      "epoch": 0.02336,
      "grad_norm": 3.140625,
      "learning_rate": 0.0008729034761497983,
      "loss": 2.6796,
      "step": 2920
    },
    {
      "epoch": 0.023368,
      "grad_norm": 3.625,
      "learning_rate": 0.0008728195841630393,
      "loss": 2.8931,
      "step": 2921
    },
    {
      "epoch": 0.023376,
      "grad_norm": 3.21875,
      "learning_rate": 0.0008727356685324761,
      "loss": 2.4876,
      "step": 2922
    },
    {
      "epoch": 0.023384,
      "grad_norm": 2.8125,
      "learning_rate": 0.0008726517292634308,
      "loss": 2.2251,
      "step": 2923
    },
    {
      "epoch": 0.023392,
      "grad_norm": 3.234375,
      "learning_rate": 0.0008725677663612266,
      "loss": 2.7085,
      "step": 2924
    },
    {
      "epoch": 0.0234,
      "grad_norm": 3.265625,
      "learning_rate": 0.0008724837798311882,
      "loss": 2.6973,
      "step": 2925
    },
    {
      "epoch": 0.023408,
      "grad_norm": 3.203125,
      "learning_rate": 0.0008723997696786423,
      "loss": 3.3758,
      "step": 2926
    },
    {
      "epoch": 0.023416,
      "grad_norm": 2.359375,
      "learning_rate": 0.0008723157359089162,
      "loss": 2.4612,
      "step": 2927
    },
    {
      "epoch": 0.023424,
      "grad_norm": 2.515625,
      "learning_rate": 0.0008722316785273397,
      "loss": 2.1113,
      "step": 2928
    },
    {
      "epoch": 0.023432,
      "grad_norm": 2.71875,
      "learning_rate": 0.0008721475975392432,
      "loss": 2.251,
      "step": 2929
    },
    {
      "epoch": 0.02344,
      "grad_norm": 3.1875,
      "learning_rate": 0.0008720634929499595,
      "loss": 2.5482,
      "step": 2930
    },
    {
      "epoch": 0.023448,
      "grad_norm": 2.515625,
      "learning_rate": 0.000871979364764822,
      "loss": 2.2212,
      "step": 2931
    },
    {
      "epoch": 0.023456,
      "grad_norm": 2.328125,
      "learning_rate": 0.0008718952129891662,
      "loss": 2.2323,
      "step": 2932
    },
    {
      "epoch": 0.023464,
      "grad_norm": 2.515625,
      "learning_rate": 0.0008718110376283291,
      "loss": 2.0643,
      "step": 2933
    },
    {
      "epoch": 0.023472,
      "grad_norm": 3.0625,
      "learning_rate": 0.0008717268386876486,
      "loss": 2.5391,
      "step": 2934
    },
    {
      "epoch": 0.02348,
      "grad_norm": 3.25,
      "learning_rate": 0.0008716426161724648,
      "loss": 2.6047,
      "step": 2935
    },
    {
      "epoch": 0.023488,
      "grad_norm": 2.90625,
      "learning_rate": 0.0008715583700881189,
      "loss": 1.9966,
      "step": 2936
    },
    {
      "epoch": 0.023496,
      "grad_norm": 3.421875,
      "learning_rate": 0.0008714741004399536,
      "loss": 2.1759,
      "step": 2937
    },
    {
      "epoch": 0.023504,
      "grad_norm": 2.78125,
      "learning_rate": 0.0008713898072333134,
      "loss": 2.5757,
      "step": 2938
    },
    {
      "epoch": 0.023512,
      "grad_norm": 3.140625,
      "learning_rate": 0.000871305490473544,
      "loss": 2.6564,
      "step": 2939
    },
    {
      "epoch": 0.02352,
      "grad_norm": 3.0625,
      "learning_rate": 0.0008712211501659925,
      "loss": 2.6173,
      "step": 2940
    },
    {
      "epoch": 0.023528,
      "grad_norm": 2.5,
      "learning_rate": 0.000871136786316008,
      "loss": 2.4641,
      "step": 2941
    },
    {
      "epoch": 0.023536,
      "grad_norm": 3.578125,
      "learning_rate": 0.0008710523989289405,
      "loss": 2.7972,
      "step": 2942
    },
    {
      "epoch": 0.023544,
      "grad_norm": 3.25,
      "learning_rate": 0.0008709679880101417,
      "loss": 2.1223,
      "step": 2943
    },
    {
      "epoch": 0.023552,
      "grad_norm": 2.84375,
      "learning_rate": 0.0008708835535649653,
      "loss": 2.2115,
      "step": 2944
    },
    {
      "epoch": 0.02356,
      "grad_norm": 4.09375,
      "learning_rate": 0.0008707990955987655,
      "loss": 2.1785,
      "step": 2945
    },
    {
      "epoch": 0.023568,
      "grad_norm": 2.890625,
      "learning_rate": 0.0008707146141168987,
      "loss": 2.1864,
      "step": 2946
    },
    {
      "epoch": 0.023576,
      "grad_norm": 3.0,
      "learning_rate": 0.0008706301091247229,
      "loss": 2.8291,
      "step": 2947
    },
    {
      "epoch": 0.023584,
      "grad_norm": 3.25,
      "learning_rate": 0.0008705455806275968,
      "loss": 2.6243,
      "step": 2948
    },
    {
      "epoch": 0.023592,
      "grad_norm": 3.234375,
      "learning_rate": 0.0008704610286308816,
      "loss": 2.6982,
      "step": 2949
    },
    {
      "epoch": 0.0236,
      "grad_norm": 2.5625,
      "learning_rate": 0.0008703764531399392,
      "loss": 2.8177,
      "step": 2950
    },
    {
      "epoch": 0.023608,
      "grad_norm": 3.046875,
      "learning_rate": 0.0008702918541601334,
      "loss": 2.4225,
      "step": 2951
    },
    {
      "epoch": 0.023616,
      "grad_norm": 2.8125,
      "learning_rate": 0.0008702072316968293,
      "loss": 2.59,
      "step": 2952
    },
    {
      "epoch": 0.023624,
      "grad_norm": 3.03125,
      "learning_rate": 0.0008701225857553936,
      "loss": 2.3422,
      "step": 2953
    },
    {
      "epoch": 0.023632,
      "grad_norm": 3.015625,
      "learning_rate": 0.0008700379163411945,
      "loss": 2.5393,
      "step": 2954
    },
    {
      "epoch": 0.02364,
      "grad_norm": 2.625,
      "learning_rate": 0.0008699532234596014,
      "loss": 2.3685,
      "step": 2955
    },
    {
      "epoch": 0.023648,
      "grad_norm": 2.453125,
      "learning_rate": 0.0008698685071159858,
      "loss": 2.064,
      "step": 2956
    },
    {
      "epoch": 0.023656,
      "grad_norm": 3.046875,
      "learning_rate": 0.0008697837673157201,
      "loss": 2.3594,
      "step": 2957
    },
    {
      "epoch": 0.023664,
      "grad_norm": 2.625,
      "learning_rate": 0.0008696990040641786,
      "loss": 2.6877,
      "step": 2958
    },
    {
      "epoch": 0.023672,
      "grad_norm": 2.296875,
      "learning_rate": 0.0008696142173667365,
      "loss": 2.1699,
      "step": 2959
    },
    {
      "epoch": 0.02368,
      "grad_norm": 2.96875,
      "learning_rate": 0.0008695294072287712,
      "loss": 2.4378,
      "step": 2960
    },
    {
      "epoch": 0.023688,
      "grad_norm": 2.78125,
      "learning_rate": 0.0008694445736556611,
      "loss": 2.2859,
      "step": 2961
    },
    {
      "epoch": 0.023696,
      "grad_norm": 3.109375,
      "learning_rate": 0.0008693597166527865,
      "loss": 2.8432,
      "step": 2962
    },
    {
      "epoch": 0.023704,
      "grad_norm": 2.734375,
      "learning_rate": 0.0008692748362255286,
      "loss": 2.652,
      "step": 2963
    },
    {
      "epoch": 0.023712,
      "grad_norm": 3.0625,
      "learning_rate": 0.0008691899323792707,
      "loss": 2.1921,
      "step": 2964
    },
    {
      "epoch": 0.02372,
      "grad_norm": 3.203125,
      "learning_rate": 0.0008691050051193972,
      "loss": 2.4141,
      "step": 2965
    },
    {
      "epoch": 0.023728,
      "grad_norm": 3.0625,
      "learning_rate": 0.0008690200544512939,
      "loss": 2.3574,
      "step": 2966
    },
    {
      "epoch": 0.023736,
      "grad_norm": 2.984375,
      "learning_rate": 0.0008689350803803486,
      "loss": 2.4513,
      "step": 2967
    },
    {
      "epoch": 0.023744,
      "grad_norm": 2.90625,
      "learning_rate": 0.0008688500829119501,
      "loss": 2.9565,
      "step": 2968
    },
    {
      "epoch": 0.023752,
      "grad_norm": 2.984375,
      "learning_rate": 0.000868765062051489,
      "loss": 3.0206,
      "step": 2969
    },
    {
      "epoch": 0.02376,
      "grad_norm": 2.8125,
      "learning_rate": 0.000868680017804357,
      "loss": 1.9162,
      "step": 2970
    },
    {
      "epoch": 0.023768,
      "grad_norm": 3.1875,
      "learning_rate": 0.0008685949501759476,
      "loss": 2.5477,
      "step": 2971
    },
    {
      "epoch": 0.023776,
      "grad_norm": 3.046875,
      "learning_rate": 0.0008685098591716558,
      "loss": 2.6591,
      "step": 2972
    },
    {
      "epoch": 0.023784,
      "grad_norm": 3.171875,
      "learning_rate": 0.0008684247447968776,
      "loss": 2.3986,
      "step": 2973
    },
    {
      "epoch": 0.023792,
      "grad_norm": 3.6875,
      "learning_rate": 0.0008683396070570113,
      "loss": 2.8011,
      "step": 2974
    },
    {
      "epoch": 0.0238,
      "grad_norm": 3.171875,
      "learning_rate": 0.0008682544459574561,
      "loss": 2.8189,
      "step": 2975
    },
    {
      "epoch": 0.023808,
      "grad_norm": 3.234375,
      "learning_rate": 0.0008681692615036129,
      "loss": 2.4223,
      "step": 2976
    },
    {
      "epoch": 0.023816,
      "grad_norm": 3.203125,
      "learning_rate": 0.0008680840537008839,
      "loss": 2.866,
      "step": 2977
    },
    {
      "epoch": 0.023824,
      "grad_norm": 2.671875,
      "learning_rate": 0.0008679988225546726,
      "loss": 2.4255,
      "step": 2978
    },
    {
      "epoch": 0.023832,
      "grad_norm": 2.765625,
      "learning_rate": 0.0008679135680703849,
      "loss": 2.4559,
      "step": 2979
    },
    {
      "epoch": 0.02384,
      "grad_norm": 2.953125,
      "learning_rate": 0.0008678282902534272,
      "loss": 2.7454,
      "step": 2980
    },
    {
      "epoch": 0.023848,
      "grad_norm": 2.90625,
      "learning_rate": 0.0008677429891092074,
      "loss": 2.6687,
      "step": 2981
    },
    {
      "epoch": 0.023856,
      "grad_norm": 2.765625,
      "learning_rate": 0.0008676576646431358,
      "loss": 2.6819,
      "step": 2982
    },
    {
      "epoch": 0.023864,
      "grad_norm": 2.78125,
      "learning_rate": 0.0008675723168606232,
      "loss": 2.0289,
      "step": 2983
    },
    {
      "epoch": 0.023872,
      "grad_norm": 2.875,
      "learning_rate": 0.0008674869457670824,
      "loss": 2.6784,
      "step": 2984
    },
    {
      "epoch": 0.02388,
      "grad_norm": 2.875,
      "learning_rate": 0.0008674015513679276,
      "loss": 2.6302,
      "step": 2985
    },
    {
      "epoch": 0.023888,
      "grad_norm": 4.09375,
      "learning_rate": 0.0008673161336685742,
      "loss": 2.2753,
      "step": 2986
    },
    {
      "epoch": 0.023896,
      "grad_norm": 2.953125,
      "learning_rate": 0.0008672306926744396,
      "loss": 2.2744,
      "step": 2987
    },
    {
      "epoch": 0.023904,
      "grad_norm": 2.9375,
      "learning_rate": 0.0008671452283909421,
      "loss": 2.3297,
      "step": 2988
    },
    {
      "epoch": 0.023912,
      "grad_norm": 2.390625,
      "learning_rate": 0.0008670597408235017,
      "loss": 2.038,
      "step": 2989
    },
    {
      "epoch": 0.02392,
      "grad_norm": 3.171875,
      "learning_rate": 0.0008669742299775402,
      "loss": 3.3445,
      "step": 2990
    },
    {
      "epoch": 0.023928,
      "grad_norm": 2.640625,
      "learning_rate": 0.0008668886958584805,
      "loss": 2.8434,
      "step": 2991
    },
    {
      "epoch": 0.023936,
      "grad_norm": 3.203125,
      "learning_rate": 0.0008668031384717471,
      "loss": 2.5344,
      "step": 2992
    },
    {
      "epoch": 0.023944,
      "grad_norm": 2.984375,
      "learning_rate": 0.0008667175578227659,
      "loss": 2.6811,
      "step": 2993
    },
    {
      "epoch": 0.023952,
      "grad_norm": 2.453125,
      "learning_rate": 0.0008666319539169643,
      "loss": 2.1195,
      "step": 2994
    },
    {
      "epoch": 0.02396,
      "grad_norm": 3.453125,
      "learning_rate": 0.0008665463267597713,
      "loss": 3.0531,
      "step": 2995
    },
    {
      "epoch": 0.023968,
      "grad_norm": 2.765625,
      "learning_rate": 0.0008664606763566172,
      "loss": 2.2616,
      "step": 2996
    },
    {
      "epoch": 0.023976,
      "grad_norm": 3.078125,
      "learning_rate": 0.0008663750027129338,
      "loss": 2.5259,
      "step": 2997
    },
    {
      "epoch": 0.023984,
      "grad_norm": 3.078125,
      "learning_rate": 0.0008662893058341546,
      "loss": 2.7361,
      "step": 2998
    },
    {
      "epoch": 0.023992,
      "grad_norm": 2.921875,
      "learning_rate": 0.0008662035857257142,
      "loss": 2.7473,
      "step": 2999
    },
    {
      "epoch": 0.024,
      "grad_norm": 2.90625,
      "learning_rate": 0.0008661178423930491,
      "loss": 2.4512,
      "step": 3000
    },
    {
      "epoch": 0.024008,
      "grad_norm": 3.21875,
      "learning_rate": 0.000866032075841597,
      "loss": 2.5451,
      "step": 3001
    },
    {
      "epoch": 0.024016,
      "grad_norm": 3.15625,
      "learning_rate": 0.0008659462860767968,
      "loss": 3.1254,
      "step": 3002
    },
    {
      "epoch": 0.024024,
      "grad_norm": 2.453125,
      "learning_rate": 0.0008658604731040896,
      "loss": 1.8937,
      "step": 3003
    },
    {
      "epoch": 0.024032,
      "grad_norm": 2.859375,
      "learning_rate": 0.0008657746369289174,
      "loss": 2.5416,
      "step": 3004
    },
    {
      "epoch": 0.02404,
      "grad_norm": 2.859375,
      "learning_rate": 0.000865688777556724,
      "loss": 2.6178,
      "step": 3005
    },
    {
      "epoch": 0.024048,
      "grad_norm": 2.34375,
      "learning_rate": 0.0008656028949929541,
      "loss": 2.2434,
      "step": 3006
    },
    {
      "epoch": 0.024056,
      "grad_norm": 2.671875,
      "learning_rate": 0.0008655169892430545,
      "loss": 2.0667,
      "step": 3007
    },
    {
      "epoch": 0.024064,
      "grad_norm": 2.5625,
      "learning_rate": 0.0008654310603124734,
      "loss": 2.1541,
      "step": 3008
    },
    {
      "epoch": 0.024072,
      "grad_norm": 2.96875,
      "learning_rate": 0.0008653451082066601,
      "loss": 1.7822,
      "step": 3009
    },
    {
      "epoch": 0.02408,
      "grad_norm": 2.984375,
      "learning_rate": 0.0008652591329310657,
      "loss": 2.8261,
      "step": 3010
    },
    {
      "epoch": 0.024088,
      "grad_norm": 3.0625,
      "learning_rate": 0.0008651731344911427,
      "loss": 2.7768,
      "step": 3011
    },
    {
      "epoch": 0.024096,
      "grad_norm": 3.09375,
      "learning_rate": 0.000865087112892345,
      "loss": 2.8121,
      "step": 3012
    },
    {
      "epoch": 0.024104,
      "grad_norm": 3.515625,
      "learning_rate": 0.0008650010681401279,
      "loss": 3.125,
      "step": 3013
    },
    {
      "epoch": 0.024112,
      "grad_norm": 2.484375,
      "learning_rate": 0.0008649150002399482,
      "loss": 2.8669,
      "step": 3014
    },
    {
      "epoch": 0.02412,
      "grad_norm": 2.71875,
      "learning_rate": 0.0008648289091972646,
      "loss": 2.6622,
      "step": 3015
    },
    {
      "epoch": 0.024128,
      "grad_norm": 2.78125,
      "learning_rate": 0.0008647427950175363,
      "loss": 2.382,
      "step": 3016
    },
    {
      "epoch": 0.024136,
      "grad_norm": 2.4375,
      "learning_rate": 0.0008646566577062252,
      "loss": 2.2256,
      "step": 3017
    },
    {
      "epoch": 0.024144,
      "grad_norm": 3.1875,
      "learning_rate": 0.0008645704972687936,
      "loss": 3.1233,
      "step": 3018
    },
    {
      "epoch": 0.024152,
      "grad_norm": 2.84375,
      "learning_rate": 0.0008644843137107057,
      "loss": 2.6507,
      "step": 3019
    },
    {
      "epoch": 0.02416,
      "grad_norm": 2.765625,
      "learning_rate": 0.0008643981070374276,
      "loss": 3.0515,
      "step": 3020
    },
    {
      "epoch": 0.024168,
      "grad_norm": 3.078125,
      "learning_rate": 0.000864311877254426,
      "loss": 2.9681,
      "step": 3021
    },
    {
      "epoch": 0.024176,
      "grad_norm": 2.9375,
      "learning_rate": 0.0008642256243671696,
      "loss": 2.2872,
      "step": 3022
    },
    {
      "epoch": 0.024184,
      "grad_norm": 2.640625,
      "learning_rate": 0.0008641393483811285,
      "loss": 2.293,
      "step": 3023
    },
    {
      "epoch": 0.024192,
      "grad_norm": 3.109375,
      "learning_rate": 0.0008640530493017742,
      "loss": 1.9885,
      "step": 3024
    },
    {
      "epoch": 0.0242,
      "grad_norm": 3.5625,
      "learning_rate": 0.0008639667271345798,
      "loss": 2.4693,
      "step": 3025
    },
    {
      "epoch": 0.024208,
      "grad_norm": 3.40625,
      "learning_rate": 0.0008638803818850196,
      "loss": 2.7523,
      "step": 3026
    },
    {
      "epoch": 0.024216,
      "grad_norm": 2.796875,
      "learning_rate": 0.0008637940135585697,
      "loss": 2.4176,
      "step": 3027
    },
    {
      "epoch": 0.024224,
      "grad_norm": 2.71875,
      "learning_rate": 0.0008637076221607073,
      "loss": 2.5357,
      "step": 3028
    },
    {
      "epoch": 0.024232,
      "grad_norm": 2.9375,
      "learning_rate": 0.0008636212076969113,
      "loss": 2.6171,
      "step": 3029
    },
    {
      "epoch": 0.02424,
      "grad_norm": 2.828125,
      "learning_rate": 0.0008635347701726621,
      "loss": 3.0676,
      "step": 3030
    },
    {
      "epoch": 0.024248,
      "grad_norm": 3.15625,
      "learning_rate": 0.0008634483095934413,
      "loss": 2.9592,
      "step": 3031
    },
    {
      "epoch": 0.024256,
      "grad_norm": 2.734375,
      "learning_rate": 0.0008633618259647324,
      "loss": 2.5606,
      "step": 3032
    },
    {
      "epoch": 0.024264,
      "grad_norm": 2.953125,
      "learning_rate": 0.0008632753192920197,
      "loss": 2.4941,
      "step": 3033
    },
    {
      "epoch": 0.024272,
      "grad_norm": 161.0,
      "learning_rate": 0.0008631887895807898,
      "loss": 2.1177,
      "step": 3034
    },
    {
      "epoch": 0.02428,
      "grad_norm": 2.84375,
      "learning_rate": 0.00086310223683653,
      "loss": 2.6016,
      "step": 3035
    },
    {
      "epoch": 0.024288,
      "grad_norm": 2.484375,
      "learning_rate": 0.0008630156610647295,
      "loss": 1.5545,
      "step": 3036
    },
    {
      "epoch": 0.024296,
      "grad_norm": 3.5,
      "learning_rate": 0.0008629290622708789,
      "loss": 2.3814,
      "step": 3037
    },
    {
      "epoch": 0.024304,
      "grad_norm": 3.0,
      "learning_rate": 0.00086284244046047,
      "loss": 2.2921,
      "step": 3038
    },
    {
      "epoch": 0.024312,
      "grad_norm": 3.734375,
      "learning_rate": 0.0008627557956389964,
      "loss": 2.4122,
      "step": 3039
    },
    {
      "epoch": 0.02432,
      "grad_norm": 2.609375,
      "learning_rate": 0.0008626691278119531,
      "loss": 2.6033,
      "step": 3040
    },
    {
      "epoch": 0.024328,
      "grad_norm": 3.0,
      "learning_rate": 0.0008625824369848363,
      "loss": 2.2241,
      "step": 3041
    },
    {
      "epoch": 0.024336,
      "grad_norm": 3.046875,
      "learning_rate": 0.0008624957231631441,
      "loss": 2.6945,
      "step": 3042
    },
    {
      "epoch": 0.024344,
      "grad_norm": 2.796875,
      "learning_rate": 0.0008624089863523754,
      "loss": 2.4489,
      "step": 3043
    },
    {
      "epoch": 0.024352,
      "grad_norm": 2.953125,
      "learning_rate": 0.0008623222265580311,
      "loss": 2.8369,
      "step": 3044
    },
    {
      "epoch": 0.02436,
      "grad_norm": 2.890625,
      "learning_rate": 0.0008622354437856137,
      "loss": 2.996,
      "step": 3045
    },
    {
      "epoch": 0.024368,
      "grad_norm": 2.78125,
      "learning_rate": 0.0008621486380406265,
      "loss": 2.6307,
      "step": 3046
    },
    {
      "epoch": 0.024376,
      "grad_norm": 2.625,
      "learning_rate": 0.0008620618093285748,
      "loss": 2.5986,
      "step": 3047
    },
    {
      "epoch": 0.024384,
      "grad_norm": 2.546875,
      "learning_rate": 0.0008619749576549651,
      "loss": 2.5019,
      "step": 3048
    },
    {
      "epoch": 0.024392,
      "grad_norm": 2.921875,
      "learning_rate": 0.0008618880830253055,
      "loss": 2.2345,
      "step": 3049
    },
    {
      "epoch": 0.0244,
      "grad_norm": 2.703125,
      "learning_rate": 0.0008618011854451056,
      "loss": 2.5384,
      "step": 3050
    },
    {
      "epoch": 0.024408,
      "grad_norm": 2.40625,
      "learning_rate": 0.000861714264919876,
      "loss": 2.7357,
      "step": 3051
    },
    {
      "epoch": 0.024416,
      "grad_norm": 2.4375,
      "learning_rate": 0.0008616273214551293,
      "loss": 2.3187,
      "step": 3052
    },
    {
      "epoch": 0.024424,
      "grad_norm": 2.5625,
      "learning_rate": 0.0008615403550563796,
      "loss": 2.2372,
      "step": 3053
    },
    {
      "epoch": 0.024432,
      "grad_norm": 2.734375,
      "learning_rate": 0.0008614533657291419,
      "loss": 2.1544,
      "step": 3054
    },
    {
      "epoch": 0.02444,
      "grad_norm": 2.625,
      "learning_rate": 0.0008613663534789331,
      "loss": 2.7091,
      "step": 3055
    },
    {
      "epoch": 0.024448,
      "grad_norm": 2.625,
      "learning_rate": 0.0008612793183112715,
      "loss": 2.4594,
      "step": 3056
    },
    {
      "epoch": 0.024456,
      "grad_norm": 3.03125,
      "learning_rate": 0.0008611922602316764,
      "loss": 2.7221,
      "step": 3057
    },
    {
      "epoch": 0.024464,
      "grad_norm": 3.09375,
      "learning_rate": 0.0008611051792456692,
      "loss": 2.2097,
      "step": 3058
    },
    {
      "epoch": 0.024472,
      "grad_norm": 3.5625,
      "learning_rate": 0.0008610180753587727,
      "loss": 2.472,
      "step": 3059
    },
    {
      "epoch": 0.02448,
      "grad_norm": 3.09375,
      "learning_rate": 0.0008609309485765104,
      "loss": 2.5445,
      "step": 3060
    },
    {
      "epoch": 0.024488,
      "grad_norm": 2.875,
      "learning_rate": 0.0008608437989044083,
      "loss": 1.9301,
      "step": 3061
    },
    {
      "epoch": 0.024496,
      "grad_norm": 2.71875,
      "learning_rate": 0.000860756626347993,
      "loss": 2.5157,
      "step": 3062
    },
    {
      "epoch": 0.024504,
      "grad_norm": 2.90625,
      "learning_rate": 0.0008606694309127931,
      "loss": 2.3455,
      "step": 3063
    },
    {
      "epoch": 0.024512,
      "grad_norm": 2.734375,
      "learning_rate": 0.0008605822126043382,
      "loss": 2.5209,
      "step": 3064
    },
    {
      "epoch": 0.02452,
      "grad_norm": 3.15625,
      "learning_rate": 0.0008604949714281598,
      "loss": 2.6279,
      "step": 3065
    },
    {
      "epoch": 0.024528,
      "grad_norm": 3.375,
      "learning_rate": 0.0008604077073897905,
      "loss": 3.0494,
      "step": 3066
    },
    {
      "epoch": 0.024536,
      "grad_norm": 2.828125,
      "learning_rate": 0.0008603204204947647,
      "loss": 2.4673,
      "step": 3067
    },
    {
      "epoch": 0.024544,
      "grad_norm": 3.046875,
      "learning_rate": 0.0008602331107486179,
      "loss": 2.2317,
      "step": 3068
    },
    {
      "epoch": 0.024552,
      "grad_norm": 3.359375,
      "learning_rate": 0.000860145778156887,
      "loss": 2.8142,
      "step": 3069
    },
    {
      "epoch": 0.02456,
      "grad_norm": 2.703125,
      "learning_rate": 0.0008600584227251107,
      "loss": 2.3699,
      "step": 3070
    },
    {
      "epoch": 0.024568,
      "grad_norm": 4.03125,
      "learning_rate": 0.0008599710444588292,
      "loss": 2.1053,
      "step": 3071
    },
    {
      "epoch": 0.024576,
      "grad_norm": 4.0,
      "learning_rate": 0.0008598836433635835,
      "loss": 3.4209,
      "step": 3072
    },
    {
      "epoch": 0.024584,
      "grad_norm": 3.46875,
      "learning_rate": 0.0008597962194449169,
      "loss": 2.3035,
      "step": 3073
    },
    {
      "epoch": 0.024592,
      "grad_norm": 2.84375,
      "learning_rate": 0.0008597087727083735,
      "loss": 2.4491,
      "step": 3074
    },
    {
      "epoch": 0.0246,
      "grad_norm": 3.3125,
      "learning_rate": 0.000859621303159499,
      "loss": 2.8485,
      "step": 3075
    },
    {
      "epoch": 0.024608,
      "grad_norm": 3.234375,
      "learning_rate": 0.0008595338108038408,
      "loss": 2.7169,
      "step": 3076
    },
    {
      "epoch": 0.024616,
      "grad_norm": 2.984375,
      "learning_rate": 0.0008594462956469475,
      "loss": 2.6056,
      "step": 3077
    },
    {
      "epoch": 0.024624,
      "grad_norm": 3.53125,
      "learning_rate": 0.0008593587576943693,
      "loss": 2.4043,
      "step": 3078
    },
    {
      "epoch": 0.024632,
      "grad_norm": 3.640625,
      "learning_rate": 0.0008592711969516576,
      "loss": 2.7042,
      "step": 3079
    },
    {
      "epoch": 0.02464,
      "grad_norm": 3.984375,
      "learning_rate": 0.0008591836134243656,
      "loss": 2.3026,
      "step": 3080
    },
    {
      "epoch": 0.024648,
      "grad_norm": 3.109375,
      "learning_rate": 0.0008590960071180473,
      "loss": 2.326,
      "step": 3081
    },
    {
      "epoch": 0.024656,
      "grad_norm": 3.40625,
      "learning_rate": 0.0008590083780382594,
      "loss": 2.6963,
      "step": 3082
    },
    {
      "epoch": 0.024664,
      "grad_norm": 2.703125,
      "learning_rate": 0.0008589207261905584,
      "loss": 2.6535,
      "step": 3083
    },
    {
      "epoch": 0.024672,
      "grad_norm": 3.140625,
      "learning_rate": 0.0008588330515805036,
      "loss": 1.8577,
      "step": 3084
    },
    {
      "epoch": 0.02468,
      "grad_norm": 2.875,
      "learning_rate": 0.000858745354213655,
      "loss": 2.1847,
      "step": 3085
    },
    {
      "epoch": 0.024688,
      "grad_norm": 3.0625,
      "learning_rate": 0.0008586576340955745,
      "loss": 2.4716,
      "step": 3086
    },
    {
      "epoch": 0.024696,
      "grad_norm": 2.484375,
      "learning_rate": 0.000858569891231825,
      "loss": 2.4345,
      "step": 3087
    },
    {
      "epoch": 0.024704,
      "grad_norm": 3.078125,
      "learning_rate": 0.0008584821256279708,
      "loss": 2.4492,
      "step": 3088
    },
    {
      "epoch": 0.024712,
      "grad_norm": 2.734375,
      "learning_rate": 0.0008583943372895786,
      "loss": 2.6639,
      "step": 3089
    },
    {
      "epoch": 0.02472,
      "grad_norm": 2.71875,
      "learning_rate": 0.0008583065262222153,
      "loss": 2.9147,
      "step": 3090
    },
    {
      "epoch": 0.024728,
      "grad_norm": 3.828125,
      "learning_rate": 0.0008582186924314499,
      "loss": 2.7793,
      "step": 3091
    },
    {
      "epoch": 0.024736,
      "grad_norm": 2.65625,
      "learning_rate": 0.0008581308359228528,
      "loss": 1.8461,
      "step": 3092
    },
    {
      "epoch": 0.024744,
      "grad_norm": 2.984375,
      "learning_rate": 0.0008580429567019959,
      "loss": 2.4162,
      "step": 3093
    },
    {
      "epoch": 0.024752,
      "grad_norm": 2.8125,
      "learning_rate": 0.0008579550547744519,
      "loss": 2.999,
      "step": 3094
    },
    {
      "epoch": 0.02476,
      "grad_norm": 2.859375,
      "learning_rate": 0.0008578671301457958,
      "loss": 2.7828,
      "step": 3095
    },
    {
      "epoch": 0.024768,
      "grad_norm": 2.84375,
      "learning_rate": 0.0008577791828216037,
      "loss": 2.6608,
      "step": 3096
    },
    {
      "epoch": 0.024776,
      "grad_norm": 2.65625,
      "learning_rate": 0.0008576912128074532,
      "loss": 2.4075,
      "step": 3097
    },
    {
      "epoch": 0.024784,
      "grad_norm": 2.75,
      "learning_rate": 0.000857603220108923,
      "loss": 2.7511,
      "step": 3098
    },
    {
      "epoch": 0.024792,
      "grad_norm": 2.984375,
      "learning_rate": 0.0008575152047315936,
      "loss": 2.3954,
      "step": 3099
    },
    {
      "epoch": 0.0248,
      "grad_norm": 3.484375,
      "learning_rate": 0.0008574271666810469,
      "loss": 2.9334,
      "step": 3100
    },
    {
      "epoch": 0.024808,
      "grad_norm": 2.6875,
      "learning_rate": 0.0008573391059628661,
      "loss": 2.9789,
      "step": 3101
    },
    {
      "epoch": 0.024816,
      "grad_norm": 2.921875,
      "learning_rate": 0.000857251022582636,
      "loss": 3.0947,
      "step": 3102
    },
    {
      "epoch": 0.024824,
      "grad_norm": 3.140625,
      "learning_rate": 0.0008571629165459427,
      "loss": 2.3362,
      "step": 3103
    },
    {
      "epoch": 0.024832,
      "grad_norm": 2.421875,
      "learning_rate": 0.0008570747878583738,
      "loss": 2.1528,
      "step": 3104
    },
    {
      "epoch": 0.02484,
      "grad_norm": 2.796875,
      "learning_rate": 0.0008569866365255181,
      "loss": 2.3441,
      "step": 3105
    },
    {
      "epoch": 0.024848,
      "grad_norm": 3.265625,
      "learning_rate": 0.0008568984625529666,
      "loss": 2.8677,
      "step": 3106
    },
    {
      "epoch": 0.024856,
      "grad_norm": 3.015625,
      "learning_rate": 0.0008568102659463106,
      "loss": 2.7061,
      "step": 3107
    },
    {
      "epoch": 0.024864,
      "grad_norm": 2.828125,
      "learning_rate": 0.0008567220467111438,
      "loss": 2.4004,
      "step": 3108
    },
    {
      "epoch": 0.024872,
      "grad_norm": 3.296875,
      "learning_rate": 0.0008566338048530608,
      "loss": 3.3409,
      "step": 3109
    },
    {
      "epoch": 0.02488,
      "grad_norm": 2.921875,
      "learning_rate": 0.0008565455403776579,
      "loss": 2.5545,
      "step": 3110
    },
    {
      "epoch": 0.024888,
      "grad_norm": 8.25,
      "learning_rate": 0.0008564572532905328,
      "loss": 2.3574,
      "step": 3111
    },
    {
      "epoch": 0.024896,
      "grad_norm": 3.390625,
      "learning_rate": 0.0008563689435972843,
      "loss": 2.7647,
      "step": 3112
    },
    {
      "epoch": 0.024904,
      "grad_norm": 3.21875,
      "learning_rate": 0.0008562806113035131,
      "loss": 3.1895,
      "step": 3113
    },
    {
      "epoch": 0.024912,
      "grad_norm": 3.3125,
      "learning_rate": 0.0008561922564148212,
      "loss": 2.2934,
      "step": 3114
    },
    {
      "epoch": 0.02492,
      "grad_norm": 3.15625,
      "learning_rate": 0.0008561038789368118,
      "loss": 2.5585,
      "step": 3115
    },
    {
      "epoch": 0.024928,
      "grad_norm": 3.15625,
      "learning_rate": 0.0008560154788750898,
      "loss": 2.5898,
      "step": 3116
    },
    {
      "epoch": 0.024936,
      "grad_norm": 2.71875,
      "learning_rate": 0.0008559270562352615,
      "loss": 2.668,
      "step": 3117
    },
    {
      "epoch": 0.024944,
      "grad_norm": 2.5,
      "learning_rate": 0.0008558386110229342,
      "loss": 2.1405,
      "step": 3118
    },
    {
      "epoch": 0.024952,
      "grad_norm": 2.5,
      "learning_rate": 0.0008557501432437175,
      "loss": 2.4516,
      "step": 3119
    },
    {
      "epoch": 0.02496,
      "grad_norm": 2.9375,
      "learning_rate": 0.0008556616529032215,
      "loss": 2.5183,
      "step": 3120
    },
    {
      "epoch": 0.024968,
      "grad_norm": 2.640625,
      "learning_rate": 0.0008555731400070586,
      "loss": 2.4015,
      "step": 3121
    },
    {
      "epoch": 0.024976,
      "grad_norm": 2.765625,
      "learning_rate": 0.0008554846045608418,
      "loss": 2.1858,
      "step": 3122
    },
    {
      "epoch": 0.024984,
      "grad_norm": 3.0,
      "learning_rate": 0.0008553960465701862,
      "loss": 2.8182,
      "step": 3123
    },
    {
      "epoch": 0.024992,
      "grad_norm": 2.90625,
      "learning_rate": 0.0008553074660407078,
      "loss": 2.8037,
      "step": 3124
    },
    {
      "epoch": 0.025,
      "grad_norm": 2.734375,
      "learning_rate": 0.0008552188629780245,
      "loss": 2.4007,
      "step": 3125
    },
    {
      "epoch": 0.025,
      "eval_loss": 2.4781594276428223,
      "eval_runtime": 2848.7828,
      "eval_samples_per_second": 35.103,
      "eval_steps_per_second": 4.388,
      "step": 3125
    }
  ],
  "logging_steps": 1,
  "max_steps": 12500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 3125,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.836543770231603e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}