diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,61698 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999467841672254,
+  "eval_steps": 500,
+  "global_step": 8808,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00011352710991907646,
+      "grad_norm": 8.75,
+      "learning_rate": 0.00020600000000000002,
+      "loss": 11.9909,
+      "step": 1
+    },
+    {
+      "epoch": 0.00022705421983815292,
+      "grad_norm": 7.71875,
+      "learning_rate": 0.00021200000000000003,
+      "loss": 11.428,
+      "step": 2
+    },
+    {
+      "epoch": 0.00034058132975722935,
+      "grad_norm": 8.25,
+      "learning_rate": 0.00021800000000000004,
+      "loss": 10.9556,
+      "step": 3
+    },
+    {
+      "epoch": 0.00045410843967630583,
+      "grad_norm": 7.71875,
+      "learning_rate": 0.000224,
+      "loss": 10.4866,
+      "step": 4
+    },
+    {
+      "epoch": 0.0005676355495953823,
+      "grad_norm": 6.40625,
+      "learning_rate": 0.00023,
+      "loss": 10.0525,
+      "step": 5
+    },
+    {
+      "epoch": 0.0006811626595144587,
+      "grad_norm": 5.5,
+      "learning_rate": 0.00023600000000000002,
+      "loss": 9.6593,
+      "step": 6
+    },
+    {
+      "epoch": 0.0007946897694335352,
+      "grad_norm": 6.375,
+      "learning_rate": 0.00024200000000000003,
+      "loss": 9.3475,
+      "step": 7
+    },
+    {
+      "epoch": 0.0009082168793526117,
+      "grad_norm": 4.34375,
+      "learning_rate": 0.000248,
+      "loss": 9.0679,
+      "step": 8
+    },
+    {
+      "epoch": 0.001021743989271688,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.000254,
+      "loss": 8.8488,
+      "step": 9
+    },
+    {
+      "epoch": 0.0011352710991907645,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00026000000000000003,
+      "loss": 8.6739,
+      "step": 10
+    },
+    {
+      "epoch": 0.001248798209109841,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.000266,
+      "loss": 8.5514,
+      "step": 11
+    },
+    {
+      "epoch": 0.0013623253190289174,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00027200000000000005,
+      "loss": 8.4544,
+      "step": 12
+    },
+    {
+      "epoch": 0.001475852428947994,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00027800000000000004,
+      "loss": 8.3934,
+      "step": 13
+    },
+    {
+      "epoch": 0.0015893795388670705,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.000284,
+      "loss": 8.371,
+      "step": 14
+    },
+    {
+      "epoch": 0.001702906648786147,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00029000000000000006,
+      "loss": 8.3597,
+      "step": 15
+    },
+    {
+      "epoch": 0.0018164337587052233,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00029600000000000004,
+      "loss": 8.3499,
+      "step": 16
+    },
+    {
+      "epoch": 0.0019299608686242998,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000302,
+      "loss": 8.3017,
+      "step": 17
+    },
+    {
+      "epoch": 0.002043487978543376,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.000308,
+      "loss": 8.3102,
+      "step": 18
+    },
+    {
+      "epoch": 0.0021570150884624526,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00031400000000000004,
+      "loss": 8.3016,
+      "step": 19
+    },
+    {
+      "epoch": 0.002270542198381529,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00032,
+      "loss": 8.2926,
+      "step": 20
+    },
+    {
+      "epoch": 0.0023840693083006055,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00032600000000000006,
+      "loss": 8.2745,
+      "step": 21
+    },
+    {
+      "epoch": 0.002497596418219682,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00033200000000000005,
+      "loss": 8.2567,
+      "step": 22
+    },
+    {
+      "epoch": 0.0026111235281387583,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.000338,
+      "loss": 8.238,
+      "step": 23
+    },
+    {
+      "epoch": 0.0027246506380578348,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.000344,
+      "loss": 8.2471,
+      "step": 24
+    },
+    {
+      "epoch": 0.0028381777479769116,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00035,
+      "loss": 8.2487,
+      "step": 25
+    },
+    {
+      "epoch": 0.002951704857895988,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.000356,
+      "loss": 8.2324,
+      "step": 26
+    },
+    {
+      "epoch": 0.0030652319678150645,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.000362,
+      "loss": 8.2254,
+      "step": 27
+    },
+    {
+      "epoch": 0.003178759077734141,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.000368,
+      "loss": 8.2185,
+      "step": 28
+    },
+    {
+      "epoch": 0.0032922861876532174,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.000374,
+      "loss": 8.2334,
+      "step": 29
+    },
+    {
+      "epoch": 0.003405813297572294,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00038,
+      "loss": 8.2103,
+      "step": 30
+    },
+    {
+      "epoch": 0.00351934040749137,
+      "grad_norm": 2.125,
+      "learning_rate": 0.000386,
+      "loss": 8.2065,
+      "step": 31
+    },
+    {
+      "epoch": 0.0036328675174104466,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00039200000000000004,
+      "loss": 8.2215,
+      "step": 32
+    },
+    {
+      "epoch": 0.003746394627329523,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.000398,
+      "loss": 8.2297,
+      "step": 33
+    },
+    {
+      "epoch": 0.0038599217372485995,
+      "grad_norm": 4.65625,
+      "learning_rate": 0.000404,
+      "loss": 8.2004,
+      "step": 34
+    },
+    {
+      "epoch": 0.003973448847167676,
+      "grad_norm": 3.234375,
+      "learning_rate": 0.00041000000000000005,
+      "loss": 8.1734,
+      "step": 35
+    },
+    {
+      "epoch": 0.004086975957086752,
+      "grad_norm": 3.359375,
+      "learning_rate": 0.00041600000000000003,
+      "loss": 8.1699,
+      "step": 36
+    },
+    {
+      "epoch": 0.004200503067005829,
+      "grad_norm": 3.59375,
+      "learning_rate": 0.00042200000000000007,
+      "loss": 8.1847,
+      "step": 37
+    },
+    {
+      "epoch": 0.004314030176924905,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00042800000000000005,
+      "loss": 8.1821,
+      "step": 38
+    },
+    {
+      "epoch": 0.004427557286843982,
+      "grad_norm": 3.046875,
+      "learning_rate": 0.0004340000000000001,
+      "loss": 8.167,
+      "step": 39
+    },
+    {
+      "epoch": 0.004541084396763058,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00044,
+      "loss": 8.1414,
+      "step": 40
+    },
+    {
+      "epoch": 0.004654611506682135,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.000446,
+      "loss": 8.1489,
+      "step": 41
+    },
+    {
+      "epoch": 0.004768138616601211,
+      "grad_norm": 2.75,
+      "learning_rate": 0.0004520000000000001,
+      "loss": 8.1023,
+      "step": 42
+    },
+    {
+      "epoch": 0.004881665726520288,
+      "grad_norm": 3.265625,
+      "learning_rate": 0.000458,
+      "loss": 8.1471,
+      "step": 43
+    },
+    {
+      "epoch": 0.004995192836439364,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00046400000000000006,
+      "loss": 8.117,
+      "step": 44
+    },
+    {
+      "epoch": 0.005108719946358441,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00047000000000000004,
+      "loss": 8.1023,
+      "step": 45
+    },
+    {
+      "epoch": 0.005222247056277517,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00047599999999999997,
+      "loss": 8.0842,
+      "step": 46
+    },
+    {
+      "epoch": 0.0053357741661965935,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00048200000000000006,
+      "loss": 8.071,
+      "step": 47
+    },
+    {
+      "epoch": 0.0054493012761156695,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00048800000000000004,
+      "loss": 8.0508,
+      "step": 48
+    },
+    {
+      "epoch": 0.005562828386034746,
+      "grad_norm": 2.5,
+      "learning_rate": 0.000494,
+      "loss": 8.0878,
+      "step": 49
+    },
+    {
+      "epoch": 0.005676355495953823,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0005,
+      "loss": 8.0276,
+      "step": 50
+    },
+    {
+      "epoch": 0.005789882605872899,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.000506,
+      "loss": 8.0411,
+      "step": 51
+    },
+    {
+      "epoch": 0.005903409715791976,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.000512,
+      "loss": 8.0092,
+      "step": 52
+    },
+    {
+      "epoch": 0.006016936825711052,
+      "grad_norm": 3.5,
+      "learning_rate": 0.000518,
+      "loss": 8.0271,
+      "step": 53
+    },
+    {
+      "epoch": 0.006130463935630129,
+      "grad_norm": 1.5,
+      "learning_rate": 0.000524,
+      "loss": 8.009,
+      "step": 54
+    },
+    {
+      "epoch": 0.006243991045549205,
+      "grad_norm": 3.171875,
+      "learning_rate": 0.0005300000000000001,
+      "loss": 7.9902,
+      "step": 55
+    },
+    {
+      "epoch": 0.006357518155468282,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.000536,
+      "loss": 7.9482,
+      "step": 56
+    },
+    {
+      "epoch": 0.006471045265387358,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0005420000000000001,
+      "loss": 7.9866,
+      "step": 57
+    },
+    {
+      "epoch": 0.006584572375306435,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0005480000000000001,
+      "loss": 7.9408,
+      "step": 58
+    },
+    {
+      "epoch": 0.006698099485225511,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000554,
+      "loss": 7.9441,
+      "step": 59
+    },
+    {
+      "epoch": 0.006811626595144588,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0005600000000000001,
+      "loss": 7.9147,
+      "step": 60
+    },
+    {
+      "epoch": 0.006925153705063664,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0005660000000000001,
+      "loss": 7.9292,
+      "step": 61
+    },
+    {
+      "epoch": 0.00703868081498274,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.000572,
+      "loss": 7.8685,
+      "step": 62
+    },
+    {
+      "epoch": 0.007152207924901816,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0005780000000000001,
+      "loss": 7.8874,
+      "step": 63
+    },
+    {
+      "epoch": 0.007265735034820893,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0005840000000000001,
+      "loss": 7.8855,
+      "step": 64
+    },
+    {
+      "epoch": 0.007379262144739969,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0005900000000000001,
+      "loss": 7.8879,
+      "step": 65
+    },
+    {
+      "epoch": 0.007492789254659046,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0005960000000000001,
+      "loss": 7.834,
+      "step": 66
+    },
+    {
+      "epoch": 0.007606316364578123,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0006020000000000001,
+      "loss": 7.8308,
+      "step": 67
+    },
+    {
+      "epoch": 0.007719843474497199,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.000608,
+      "loss": 7.8146,
+      "step": 68
+    },
+    {
+      "epoch": 0.007833370584416275,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0006140000000000001,
+      "loss": 7.8115,
+      "step": 69
+    },
+    {
+      "epoch": 0.007946897694335353,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.0006200000000000001,
+      "loss": 7.771,
+      "step": 70
+    },
+    {
+      "epoch": 0.008060424804254429,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.000626,
+      "loss": 7.7673,
+      "step": 71
+    },
+    {
+      "epoch": 0.008173951914173505,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.000632,
+      "loss": 7.769,
+      "step": 72
+    },
+    {
+      "epoch": 0.00828747902409258,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0006380000000000001,
+      "loss": 7.7397,
+      "step": 73
+    },
+    {
+      "epoch": 0.008401006134011658,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.000644,
+      "loss": 7.7263,
+      "step": 74
+    },
+    {
+      "epoch": 0.008514533243930734,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0006500000000000001,
+      "loss": 7.7123,
+      "step": 75
+    },
+    {
+      "epoch": 0.00862806035384981,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0006560000000000001,
+      "loss": 7.7051,
+      "step": 76
+    },
+    {
+      "epoch": 0.008741587463768886,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0006619999999999999,
+      "loss": 7.6825,
+      "step": 77
+    },
+    {
+      "epoch": 0.008855114573687964,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0006680000000000001,
+      "loss": 7.6839,
+      "step": 78
+    },
+    {
+      "epoch": 0.00896864168360704,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0006739999999999999,
+      "loss": 7.6467,
+      "step": 79
+    },
+    {
+      "epoch": 0.009082168793526116,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0006799999999999999,
+      "loss": 7.6222,
+      "step": 80
+    },
+    {
+      "epoch": 0.009195695903445192,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0006860000000000001,
+      "loss": 7.6246,
+      "step": 81
+    },
+    {
+      "epoch": 0.00930922301336427,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.000692,
+      "loss": 7.6004,
+      "step": 82
+    },
+    {
+      "epoch": 0.009422750123283346,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0006979999999999999,
+      "loss": 7.5864,
+      "step": 83
+    },
+    {
+      "epoch": 0.009536277233202422,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0007040000000000002,
+      "loss": 7.5924,
+      "step": 84
+    },
+    {
+      "epoch": 0.0096498043431215,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00071,
+      "loss": 7.5853,
+      "step": 85
+    },
+    {
+      "epoch": 0.009763331453040576,
+      "grad_norm": 1.5,
+      "learning_rate": 0.000716,
+      "loss": 7.5521,
+      "step": 86
+    },
+    {
+      "epoch": 0.009876858562959652,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.000722,
+      "loss": 7.5544,
+      "step": 87
+    },
+    {
+      "epoch": 0.009990385672878728,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.000728,
+      "loss": 7.5235,
+      "step": 88
+    },
+    {
+      "epoch": 0.010103912782797805,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.000734,
+      "loss": 7.5459,
+      "step": 89
+    },
+    {
+      "epoch": 0.010217439892716881,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00074,
+      "loss": 7.5094,
+      "step": 90
+    },
+    {
+      "epoch": 0.010330967002635957,
+      "grad_norm": 1.625,
+      "learning_rate": 0.000746,
+      "loss": 7.491,
+      "step": 91
+    },
+    {
+      "epoch": 0.010444494112555033,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0007520000000000001,
+      "loss": 7.4972,
+      "step": 92
+    },
+    {
+      "epoch": 0.010558021222474111,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.000758,
+      "loss": 7.4828,
+      "step": 93
+    },
+    {
+      "epoch": 0.010671548332393187,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.000764,
+      "loss": 7.4417,
+      "step": 94
+    },
+    {
+      "epoch": 0.010785075442312263,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0007700000000000001,
+      "loss": 7.4512,
+      "step": 95
+    },
+    {
+      "epoch": 0.010898602552231339,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.000776,
+      "loss": 7.4023,
+      "step": 96
+    },
+    {
+      "epoch": 0.011012129662150417,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.000782,
+      "loss": 7.4279,
+      "step": 97
+    },
+    {
+      "epoch": 0.011125656772069493,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0007880000000000001,
+      "loss": 7.4036,
+      "step": 98
+    },
+    {
+      "epoch": 0.011239183881988569,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0007940000000000001,
+      "loss": 7.4104,
+      "step": 99
+    },
+    {
+      "epoch": 0.011352710991907647,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0008,
+      "loss": 7.3965,
+      "step": 100
+    },
+    {
+      "epoch": 0.011466238101826723,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0008060000000000001,
+      "loss": 7.3772,
+      "step": 101
+    },
+    {
+      "epoch": 0.011579765211745799,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0008120000000000001,
+      "loss": 7.3716,
+      "step": 102
+    },
+    {
+      "epoch": 0.011693292321664875,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.000818,
+      "loss": 7.3657,
+      "step": 103
+    },
+    {
+      "epoch": 0.011806819431583952,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0008240000000000001,
+      "loss": 7.3203,
+      "step": 104
+    },
+    {
+      "epoch": 0.011920346541503028,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0008300000000000001,
+      "loss": 7.3213,
+      "step": 105
+    },
+    {
+      "epoch": 0.012033873651422104,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.000836,
+      "loss": 7.3194,
+      "step": 106
+    },
+    {
+      "epoch": 0.01214740076134118,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0008420000000000001,
+      "loss": 7.3077,
+      "step": 107
+    },
+    {
+      "epoch": 0.012260927871260258,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0008480000000000001,
+      "loss": 7.2861,
+      "step": 108
+    },
+    {
+      "epoch": 0.012374454981179334,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0008540000000000002,
+      "loss": 7.2577,
+      "step": 109
+    },
+    {
+      "epoch": 0.01248798209109841,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0008599999999999999,
+      "loss": 7.2872,
+      "step": 110
+    },
+    {
+      "epoch": 0.012601509201017486,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0008660000000000001,
+      "loss": 7.2687,
+      "step": 111
+    },
+    {
+      "epoch": 0.012715036310936564,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0008720000000000002,
+      "loss": 7.2544,
+      "step": 112
+    },
+    {
+      "epoch": 0.01282856342085564,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0008779999999999999,
+      "loss": 7.2292,
+      "step": 113
+    },
+    {
+      "epoch": 0.012942090530774716,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0008840000000000001,
+      "loss": 7.2224,
+      "step": 114
+    },
+    {
+      "epoch": 0.013055617640693792,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0008900000000000002,
+      "loss": 7.2196,
+      "step": 115
+    },
+    {
+      "epoch": 0.01316914475061287,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0008959999999999999,
+      "loss": 7.1945,
+      "step": 116
+    },
+    {
+      "epoch": 0.013282671860531945,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0009020000000000001,
+      "loss": 7.1936,
+      "step": 117
+    },
+    {
+      "epoch": 0.013396198970451021,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.000908,
+      "loss": 7.189,
+      "step": 118
+    },
+    {
+      "epoch": 0.0135097260803701,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.000914,
+      "loss": 7.1527,
+      "step": 119
+    },
+    {
+      "epoch": 0.013623253190289175,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0009200000000000001,
+      "loss": 7.1539,
+      "step": 120
+    },
+    {
+      "epoch": 0.013736780300208251,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.000926,
+      "loss": 7.1505,
+      "step": 121
+    },
+    {
+      "epoch": 0.013850307410127327,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.000932,
+      "loss": 7.1208,
+      "step": 122
+    },
+    {
+      "epoch": 0.013963834520046405,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0009379999999999999,
+      "loss": 7.1369,
+      "step": 123
+    },
+    {
+      "epoch": 0.01407736162996548,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.000944,
+      "loss": 7.111,
+      "step": 124
+    },
+    {
+      "epoch": 0.014190888739884557,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00095,
+      "loss": 7.1177,
+      "step": 125
+    },
+    {
+      "epoch": 0.014304415849803633,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.0009559999999999999,
+      "loss": 7.0898,
+      "step": 126
+    },
+    {
+      "epoch": 0.01441794295972271,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.000962,
+      "loss": 7.1302,
+      "step": 127
+    },
+    {
+      "epoch": 0.014531470069641787,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.000968,
+      "loss": 7.0921,
+      "step": 128
+    },
+    {
+      "epoch": 0.014644997179560863,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.000974,
+      "loss": 7.0639,
+      "step": 129
+    },
+    {
+      "epoch": 0.014758524289479939,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00098,
+      "loss": 7.0577,
+      "step": 130
+    },
+    {
+      "epoch": 0.014872051399399016,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0009860000000000001,
+      "loss": 7.0683,
+      "step": 131
+    },
+    {
+      "epoch": 0.014985578509318092,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.000992,
+      "loss": 7.0393,
+      "step": 132
+    },
+    {
+      "epoch": 0.015099105619237168,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.000998,
+      "loss": 7.0399,
+      "step": 133
+    },
+    {
+      "epoch": 0.015212632729156246,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0010040000000000001,
+      "loss": 7.0315,
+      "step": 134
+    },
+    {
+      "epoch": 0.015326159839075322,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00101,
+      "loss": 7.0166,
+      "step": 135
+    },
+    {
+      "epoch": 0.015439686948994398,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.001016,
+      "loss": 7.0053,
+      "step": 136
+    },
+    {
+      "epoch": 0.015553214058913474,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0010220000000000001,
+      "loss": 6.9946,
+      "step": 137
+    },
+    {
+      "epoch": 0.01566674116883255,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.001028,
+      "loss": 7.0057,
+      "step": 138
+    },
+    {
+      "epoch": 0.015780268278751626,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.001034,
+      "loss": 6.9886,
+      "step": 139
+    },
+    {
+      "epoch": 0.015893795388670705,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0010400000000000001,
+      "loss": 6.9678,
+      "step": 140
+    },
+    {
+      "epoch": 0.01600732249858978,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.001046,
+      "loss": 6.9794,
+      "step": 141
+    },
+    {
+      "epoch": 0.016120849608508857,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.001052,
+      "loss": 6.9735,
+      "step": 142
+    },
+    {
+      "epoch": 0.016234376718427933,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.0010580000000000001,
+      "loss": 6.9456,
+      "step": 143
+    },
+    {
+      "epoch": 0.01634790382834701,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.001064,
+      "loss": 6.9397,
+      "step": 144
+    },
+    {
+      "epoch": 0.016461430938266085,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00107,
+      "loss": 6.9411,
+      "step": 145
+    },
+    {
+      "epoch": 0.01657495804818516,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0010760000000000001,
+      "loss": 6.9274,
+      "step": 146
+    },
+    {
+      "epoch": 0.016688485158104237,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.001082,
+      "loss": 6.921,
+      "step": 147
+    },
+    {
+      "epoch": 0.016802012268023317,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0010880000000000002,
+      "loss": 6.9047,
+      "step": 148
+    },
+    {
+      "epoch": 0.016915539377942393,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0010940000000000001,
+      "loss": 6.9285,
+      "step": 149
+    },
+    {
+      "epoch": 0.01702906648786147,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.0011,
+      "loss": 6.8916,
+      "step": 150
+    },
+    {
+      "epoch": 0.017142593597780545,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.001106,
+      "loss": 6.8917,
+      "step": 151
+    },
+    {
+      "epoch": 0.01725612070769962,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0011120000000000001,
+      "loss": 6.8983,
+      "step": 152
+    },
+    {
+      "epoch": 0.017369647817618697,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.001118,
+      "loss": 6.872,
+      "step": 153
+    },
+    {
+      "epoch": 0.017483174927537773,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.001124,
+      "loss": 6.8641,
+      "step": 154
+    },
+    {
+      "epoch": 0.017596702037456852,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0011300000000000001,
+      "loss": 6.8508,
+      "step": 155
+    },
+    {
+      "epoch": 0.01771022914737593,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.001136,
+      "loss": 6.8742,
+      "step": 156
+    },
+    {
+      "epoch": 0.017823756257295004,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.001142,
+      "loss": 6.8687,
+      "step": 157
+    },
+    {
+      "epoch": 0.01793728336721408,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.001148,
+      "loss": 6.8461,
+      "step": 158
+    },
+    {
+      "epoch": 0.018050810477133156,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.001154,
+      "loss": 6.8546,
+      "step": 159
+    },
+    {
+      "epoch": 0.018164337587052232,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00116,
+      "loss": 6.8339,
+      "step": 160
+    },
+    {
+      "epoch": 0.01827786469697131,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.001166,
+      "loss": 6.8353,
+      "step": 161
+    },
+    {
+      "epoch": 0.018391391806890384,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.001172,
+      "loss": 6.8294,
+      "step": 162
+    },
+    {
+      "epoch": 0.018504918916809464,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.001178,
+      "loss": 6.816,
+      "step": 163
+    },
+    {
+      "epoch": 0.01861844602672854,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.001184,
+      "loss": 6.8221,
+      "step": 164
+    },
+    {
+      "epoch": 0.018731973136647616,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00119,
+      "loss": 6.8141,
+      "step": 165
+    },
+    {
+      "epoch": 0.018845500246566692,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.001196,
+      "loss": 6.7819,
+      "step": 166
+    },
+    {
+      "epoch": 0.018959027356485768,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.001202,
+      "loss": 6.8,
+      "step": 167
+    },
+    {
+      "epoch": 0.019072554466404844,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0012080000000000003,
+      "loss": 6.7826,
+      "step": 168
+    },
+    {
+      "epoch": 0.01918608157632392,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.001214,
+      "loss": 6.7639,
+      "step": 169
+    },
+    {
+      "epoch": 0.019299608686243,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00122,
+      "loss": 6.7705,
+      "step": 170
+    },
+    {
+      "epoch": 0.019413135796162075,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.001226,
+      "loss": 6.7804,
+      "step": 171
+    },
+    {
+      "epoch": 0.01952666290608115,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.001232,
+      "loss": 6.7498,
+      "step": 172
+    },
+    {
+      "epoch": 0.019640190016000227,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.001238,
+      "loss": 6.772,
+      "step": 173
+    },
+    {
+      "epoch": 0.019753717125919303,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.001244,
+      "loss": 6.7528,
+      "step": 174
+    },
+    {
+      "epoch": 0.01986724423583838,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00125,
+      "loss": 6.7417,
+      "step": 175
+    },
+    {
+      "epoch": 0.019980771345757455,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.001256,
+      "loss": 6.7512,
+      "step": 176
+    },
+    {
+      "epoch": 0.02009429845567653,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.001262,
+      "loss": 6.7275,
+      "step": 177
+    },
+    {
+      "epoch": 0.02020782556559561,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.001268,
+      "loss": 6.7165,
+      "step": 178
+    },
+    {
+      "epoch": 0.020321352675514687,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.001274,
+      "loss": 6.7315,
+      "step": 179
+    },
+    {
+      "epoch": 0.020434879785433763,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00128,
+      "loss": 6.716,
+      "step": 180
+    },
+    {
+      "epoch": 0.02054840689535284,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.001286,
+      "loss": 6.7107,
+      "step": 181
+    },
+    {
+      "epoch": 0.020661934005271915,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.001292,
+      "loss": 6.6907,
+      "step": 182
+    },
+    {
+      "epoch": 0.02077546111519099,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0012980000000000001,
+      "loss": 6.677,
+      "step": 183
+    },
+    {
+      "epoch": 0.020888988225110067,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0013039999999999998,
+      "loss": 6.6887,
+      "step": 184
+    },
+    {
+      "epoch": 0.021002515335029146,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0013100000000000002,
+      "loss": 6.699,
+      "step": 185
+    },
+    {
+      "epoch": 0.021116042444948222,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0013160000000000001,
+      "loss": 6.6653,
+      "step": 186
+    },
+    {
+      "epoch": 0.021229569554867298,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0013219999999999998,
+      "loss": 6.6719,
+      "step": 187
+    },
+    {
+      "epoch": 0.021343096664786374,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0013280000000000002,
+      "loss": 6.6611,
+      "step": 188
+    },
+    {
+      "epoch": 0.02145662377470545,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0013340000000000001,
+      "loss": 6.6477,
+      "step": 189
+    },
+    {
+      "epoch": 0.021570150884624526,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0013399999999999998,
+      "loss": 6.6759,
+      "step": 190
+    },
+    {
+      "epoch": 0.021683677994543602,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0013460000000000002,
+      "loss": 6.6546,
+      "step": 191
+    },
+    {
+      "epoch": 0.021797205104462678,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0013520000000000001,
+      "loss": 6.6508,
+      "step": 192
+    },
+    {
+      "epoch": 0.021910732214381758,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0013579999999999998,
+      "loss": 6.6491,
+      "step": 193
+    },
+    {
+      "epoch": 0.022024259324300834,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.001364,
+      "loss": 6.6361,
+      "step": 194
+    },
+    {
+      "epoch": 0.02213778643421991,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0013700000000000001,
+      "loss": 6.6242,
+      "step": 195
+    },
+    {
+      "epoch": 0.022251313544138986,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0013759999999999998,
+      "loss": 6.6195,
+      "step": 196
+    },
+    {
+      "epoch": 0.02236484065405806,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.001382,
+      "loss": 6.6039,
+      "step": 197
+    },
+    {
+      "epoch": 0.022478367763977138,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0013880000000000001,
+      "loss": 6.6206,
+      "step": 198
+    },
+    {
+      "epoch": 0.022591894873896214,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0013939999999999998,
+      "loss": 6.6069,
+      "step": 199
+    },
+    {
+      "epoch": 0.022705421983815293,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0014,
+      "loss": 6.6141,
+      "step": 200
+    },
+    {
+      "epoch": 0.02281894909373437,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0014060000000000001,
+      "loss": 6.5865,
+      "step": 201
+    },
+    {
+      "epoch": 0.022932476203653445,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0014119999999999998,
+      "loss": 6.6011,
+      "step": 202
+    },
+    {
+      "epoch": 0.02304600331357252,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.001418,
+      "loss": 6.5957,
+      "step": 203
+    },
+    {
+      "epoch": 0.023159530423491597,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0014240000000000001,
+      "loss": 6.588,
+      "step": 204
+    },
+    {
+      "epoch": 0.023273057533410673,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00143,
+      "loss": 6.5857,
+      "step": 205
+    },
+    {
+      "epoch": 0.02338658464332975,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.001436,
+      "loss": 6.582,
+      "step": 206
+    },
+    {
+      "epoch": 0.023500111753248825,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.001442,
+      "loss": 6.5843,
+      "step": 207
+    },
+    {
+      "epoch": 0.023613638863167905,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.001448,
+      "loss": 6.582,
+      "step": 208
+    },
+    {
+      "epoch": 0.02372716597308698,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.001454,
+      "loss": 6.5853,
+      "step": 209
+    },
+    {
+      "epoch": 0.023840693083006056,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00146,
+      "loss": 6.5648,
+      "step": 210
+    },
+    {
+      "epoch": 0.023954220192925132,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.001466,
+      "loss": 6.5536,
+      "step": 211
+    },
+    {
+      "epoch": 0.02406774730284421,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.001472,
+      "loss": 6.571,
+      "step": 212
+    },
+    {
+      "epoch": 0.024181274412763284,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.001478,
+      "loss": 6.5288,
+      "step": 213
+    },
+    {
+      "epoch": 0.02429480152268236,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.001484,
+      "loss": 6.5409,
+      "step": 214
+    },
+    {
+      "epoch": 0.02440832863260144,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00149,
+      "loss": 6.5423,
+      "step": 215
+    },
+    {
+      "epoch": 0.024521855742520516,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.001496,
+      "loss": 6.5535,
+      "step": 216
+    },
+    {
+      "epoch": 0.024635382852439592,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.001502,
+      "loss": 6.523,
+      "step": 217
+    },
+    {
+      "epoch": 0.024748909962358668,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.001508,
+      "loss": 6.5229,
+      "step": 218
+    },
+    {
+      "epoch": 0.024862437072277744,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.001514,
+      "loss": 6.5246,
+      "step": 219
+    },
+    {
+      "epoch": 0.02497596418219682,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0015199999999999999,
+      "loss": 6.5105,
+      "step": 220
+    },
+    {
+      "epoch": 0.025089491292115896,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.001526,
+      "loss": 6.5089,
+      "step": 221
+    },
+    {
+      "epoch": 0.025203018402034972,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0015320000000000002,
+      "loss": 6.4997,
+      "step": 222
+    },
+    {
+      "epoch": 0.02531654551195405,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0015379999999999999,
+      "loss": 6.4932,
+      "step": 223
+    },
+    {
+      "epoch": 0.025430072621873127,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.001544,
+      "loss": 6.4813,
+      "step": 224
+    },
+    {
+      "epoch": 0.025543599731792203,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0015500000000000002,
+      "loss": 6.4868,
+      "step": 225
+    },
+    {
+      "epoch": 0.02565712684171128,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0015559999999999999,
+      "loss": 6.495,
+      "step": 226
+    },
+    {
+      "epoch": 0.025770653951630355,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.001562,
+      "loss": 6.484,
+      "step": 227
+    },
+    {
+      "epoch": 0.02588418106154943,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0015680000000000002,
+      "loss": 6.4856,
+      "step": 228
+    },
+    {
+      "epoch": 0.025997708171468507,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0015739999999999999,
+      "loss": 6.4888,
+      "step": 229
+    },
+    {
+      "epoch": 0.026111235281387583,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00158,
+      "loss": 6.4776,
+      "step": 230
+    },
+    {
+      "epoch": 0.026224762391306663,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0015860000000000002,
+      "loss": 6.4745,
+      "step": 231
+    },
+    {
+      "epoch": 0.02633828950122574,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0015919999999999999,
+      "loss": 6.4713,
+      "step": 232
+    },
+    {
+      "epoch": 0.026451816611144815,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0015979999999999998,
+      "loss": 6.4736,
+      "step": 233
+    },
+    {
+      "epoch": 0.02656534372106389,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0016040000000000002,
+      "loss": 6.4385,
+      "step": 234
+    },
+    {
+      "epoch": 0.026678870830982967,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0016099999999999999,
+      "loss": 6.4602,
+      "step": 235
+    },
+    {
+      "epoch": 0.026792397940902043,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0016159999999999998,
+      "loss": 6.4555,
+      "step": 236
+    },
+    {
+      "epoch": 0.02690592505082112,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0016220000000000002,
+      "loss": 6.4601,
+      "step": 237
+    },
+    {
+      "epoch": 0.0270194521607402,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0016279999999999999,
+      "loss": 6.4413,
+      "step": 238
+    },
+    {
+      "epoch": 0.027132979270659274,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.001634,
+      "loss": 6.4384,
+      "step": 239
+    },
+    {
+      "epoch": 0.02724650638057835,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0016400000000000002,
+      "loss": 6.4485,
+      "step": 240
+    },
+    {
+      "epoch": 0.027360033490497426,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.001646,
+      "loss": 6.4372,
+      "step": 241
+    },
+    {
+      "epoch": 0.027473560600416502,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.001652,
+      "loss": 6.4507,
+      "step": 242
+    },
+    {
+      "epoch": 0.02758708771033558,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0016580000000000002,
+      "loss": 6.4199,
+      "step": 243
+    },
+    {
+      "epoch": 0.027700614820254654,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.001664,
+      "loss": 6.4132,
+      "step": 244
+    },
+    {
+      "epoch": 0.02781414193017373,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00167,
+      "loss": 6.4206,
+      "step": 245
+    },
+    {
+      "epoch": 0.02792766904009281,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.001676,
+      "loss": 6.4129,
+      "step": 246
+    },
+    {
+      "epoch": 0.028041196150011886,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.001682,
+      "loss": 6.4176,
+      "step": 247
+    },
+    {
+      "epoch": 0.02815472325993096,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.001688,
+      "loss": 6.4286,
+      "step": 248
+    },
+    {
+      "epoch": 0.028268250369850038,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.001694,
+      "loss": 6.4073,
+      "step": 249
+    },
+    {
+      "epoch": 0.028381777479769114,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0017,
+      "loss": 6.4126,
+      "step": 250
+    },
+    {
+      "epoch": 0.02849530458968819,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.001706,
+      "loss": 6.4056,
+      "step": 251
+    },
+    {
+      "epoch": 0.028608831699607266,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.001712,
+      "loss": 6.3884,
+      "step": 252
+    },
+    {
+      "epoch": 0.028722358809526345,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.001718,
+      "loss": 6.4034,
+      "step": 253
+    },
+    {
+      "epoch": 0.02883588591944542,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.001724,
+      "loss": 6.4039,
+      "step": 254
+    },
+    {
+      "epoch": 0.028949413029364497,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00173,
+      "loss": 6.3756,
+      "step": 255
+    },
+    {
+      "epoch": 0.029062940139283573,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0017360000000000001,
+      "loss": 6.3928,
+      "step": 256
+    },
+    {
+      "epoch": 0.02917646724920265,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.001742,
+      "loss": 6.3748,
+      "step": 257
+    },
+    {
+      "epoch": 0.029289994359121725,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.001748,
+      "loss": 6.3946,
+      "step": 258
+    },
+    {
+      "epoch": 0.0294035214690408,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0017540000000000001,
+      "loss": 6.3819,
+      "step": 259
+    },
+    {
+      "epoch": 0.029517048578959877,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00176,
+      "loss": 6.3675,
+      "step": 260
+    },
+    {
+      "epoch": 0.029630575688878957,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.001766,
+      "loss": 6.3723,
+      "step": 261
+    },
+    {
+      "epoch": 0.029744102798798033,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0017720000000000001,
+      "loss": 6.356,
+      "step": 262
+    },
+    {
+      "epoch": 0.02985762990871711,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.001778,
+      "loss": 6.3453,
+      "step": 263
+    },
+    {
+      "epoch": 0.029971157018636185,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.001784,
+      "loss": 6.3625,
+      "step": 264
+    },
+    {
+      "epoch": 0.03008468412855526,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0017900000000000001,
+      "loss": 6.358,
+      "step": 265
+    },
+    {
+      "epoch": 0.030198211238474337,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.001796,
+      "loss": 6.3492,
+      "step": 266
+    },
+    {
+      "epoch": 0.030311738348393413,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.001802,
+      "loss": 6.3527,
+      "step": 267
+    },
+    {
+      "epoch": 0.030425265458312492,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0018080000000000001,
+      "loss": 6.3575,
+      "step": 268
+    },
+    {
+      "epoch": 0.030538792568231568,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0018139999999999999,
+      "loss": 6.3469,
+      "step": 269
+    },
+    {
+      "epoch": 0.030652319678150644,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00182,
+      "loss": 6.3497,
+      "step": 270
+    },
+    {
+      "epoch": 0.03076584678806972,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0018260000000000001,
+      "loss": 6.3454,
+      "step": 271
+    },
+    {
+      "epoch": 0.030879373897988796,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0018319999999999999,
+      "loss": 6.3361,
+      "step": 272
+    },
+    {
+      "epoch": 0.030992901007907872,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0018380000000000002,
+      "loss": 6.3293,
+      "step": 273
+    },
+    {
+      "epoch": 0.031106428117826948,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0018440000000000002,
+      "loss": 6.3312,
+      "step": 274
+    },
+    {
+      "epoch": 0.031219955227746024,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0018499999999999999,
+      "loss": 6.3324,
+      "step": 275
+    },
+    {
+      "epoch": 0.0313334823376651,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0018560000000000002,
+      "loss": 6.3218,
+      "step": 276
+    },
+    {
+      "epoch": 0.031447009447584176,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0018620000000000002,
+      "loss": 6.3279,
+      "step": 277
+    },
+    {
+      "epoch": 0.03156053655750325,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0018679999999999999,
+      "loss": 6.3369,
+      "step": 278
+    },
+    {
+      "epoch": 0.03167406366742233,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0018740000000000002,
+      "loss": 6.3231,
+      "step": 279
+    },
+    {
+      "epoch": 0.03178759077734141,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0018800000000000002,
+      "loss": 6.3226,
+      "step": 280
+    },
+    {
+      "epoch": 0.03190111788726049,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0018859999999999999,
+      "loss": 6.3117,
+      "step": 281
+    },
+    {
+      "epoch": 0.03201464499717956,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.001892,
+      "loss": 6.2882,
+      "step": 282
+    },
+    {
+      "epoch": 0.03212817210709864,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0018980000000000002,
+      "loss": 6.3303,
+      "step": 283
+    },
+    {
+      "epoch": 0.032241699217017715,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0019039999999999999,
+      "loss": 6.3005,
+      "step": 284
+    },
+    {
+      "epoch": 0.03235522632693679,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00191,
+      "loss": 6.313,
+      "step": 285
+    },
+    {
+      "epoch": 0.03246875343685587,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0019160000000000002,
+      "loss": 6.2736,
+      "step": 286
+    },
+    {
+      "epoch": 0.03258228054677494,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0019219999999999999,
+      "loss": 6.2803,
+      "step": 287
+    },
+    {
+      "epoch": 0.03269580765669402,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.001928,
+      "loss": 6.2968,
+      "step": 288
+    },
+    {
+      "epoch": 0.032809334766613095,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0019340000000000002,
+      "loss": 6.3097,
+      "step": 289
+    },
+    {
+      "epoch": 0.03292286187653217,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0019399999999999999,
+      "loss": 6.2837,
+      "step": 290
+    },
+    {
+      "epoch": 0.03303638898645125,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.001946,
+      "loss": 6.2716,
+      "step": 291
+    },
+    {
+      "epoch": 0.03314991609637032,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0019520000000000002,
+      "loss": 6.2953,
+      "step": 292
+    },
+    {
+      "epoch": 0.0332634432062894,
+      "grad_norm": 0.625,
+      "learning_rate": 0.001958,
+      "loss": 6.2702,
+      "step": 293
+    },
+    {
+      "epoch": 0.033376970316208475,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.001964,
+      "loss": 6.2684,
+      "step": 294
+    },
+    {
+      "epoch": 0.03349049742612756,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00197,
+      "loss": 6.2785,
+      "step": 295
+    },
+    {
+      "epoch": 0.033604024536046634,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.001976,
+      "loss": 6.2718,
+      "step": 296
+    },
+    {
+      "epoch": 0.03371755164596571,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.001982,
+      "loss": 6.2632,
+      "step": 297
+    },
+    {
+      "epoch": 0.033831078755884786,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.001988,
+      "loss": 6.253,
+      "step": 298
+    },
+    {
+      "epoch": 0.03394460586580386,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.001994,
+      "loss": 6.261,
+      "step": 299
+    },
+    {
+      "epoch": 0.03405813297572294,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.002,
+      "loss": 6.2471,
+      "step": 300
+    },
+    {
+      "epoch": 0.034171660085642014,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.002,
+      "loss": 6.2568,
+      "step": 301
+    },
+    {
+      "epoch": 0.03428518719556109,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.002,
+      "loss": 6.2433,
+      "step": 302
+    },
+    {
+      "epoch": 0.034398714305480166,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.002,
+      "loss": 6.2558,
+      "step": 303
+    },
+    {
+      "epoch": 0.03451224141539924,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.002,
+      "loss": 6.2437,
+      "step": 304
+    },
+    {
+      "epoch": 0.03462576852531832,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.002,
+      "loss": 6.2434,
+      "step": 305
+    },
+    {
+      "epoch": 0.034739295635237394,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.002,
+      "loss": 6.2332,
+      "step": 306
+    },
+    {
+      "epoch": 0.03485282274515647,
+      "grad_norm": 0.75,
+      "learning_rate": 0.002,
+      "loss": 6.2421,
+      "step": 307
+    },
+    {
+      "epoch": 0.034966349855075546,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.002,
+      "loss": 6.2396,
+      "step": 308
+    },
+    {
+      "epoch": 0.03507987696499462,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.002,
+      "loss": 6.2284,
+      "step": 309
+    },
+    {
+      "epoch": 0.035193404074913705,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.002,
+      "loss": 6.2347,
+      "step": 310
+    },
+    {
+      "epoch": 0.03530693118483278,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.002,
+      "loss": 6.2424,
+      "step": 311
+    },
+    {
+      "epoch": 0.03542045829475186,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.002,
+      "loss": 6.2347,
+      "step": 312
+    },
+    {
+      "epoch": 0.03553398540467093,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.002,
+      "loss": 6.2269,
+      "step": 313
+    },
+    {
+      "epoch": 0.03564751251459001,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.002,
+      "loss": 6.2191,
+      "step": 314
+    },
+    {
+      "epoch": 0.035761039624509085,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.002,
+      "loss": 6.2313,
+      "step": 315
+    },
+    {
+      "epoch": 0.03587456673442816,
+      "grad_norm": 0.625,
+      "learning_rate": 0.002,
+      "loss": 6.2311,
+      "step": 316
+    },
+    {
+      "epoch": 0.03598809384434724,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.002,
+      "loss": 6.2184,
+      "step": 317
+    },
+    {
+      "epoch": 0.03610162095426631,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.002,
+      "loss": 6.1863,
+      "step": 318
+    },
+    {
+      "epoch": 0.03621514806418539,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.002,
+      "loss": 6.2005,
+      "step": 319
+    },
+    {
+      "epoch": 0.036328675174104465,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.002,
+      "loss": 6.2043,
+      "step": 320
+    },
+    {
+      "epoch": 0.03644220228402354,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.002,
+      "loss": 6.2207,
+      "step": 321
+    },
+    {
+      "epoch": 0.03655572939394262,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.002,
+      "loss": 6.2231,
+      "step": 322
+    },
+    {
+      "epoch": 0.03666925650386169,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.002,
+      "loss": 6.1961,
+      "step": 323
+    },
+    {
+      "epoch": 0.03678278361378077,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.002,
+      "loss": 6.1961,
+      "step": 324
+    },
+    {
+      "epoch": 0.03689631072369985,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.002,
+      "loss": 6.2078,
+      "step": 325
+    },
+    {
+      "epoch": 0.03700983783361893,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.002,
+      "loss": 6.2047,
+      "step": 326
+    },
+    {
+      "epoch": 0.037123364943538004,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.002,
+      "loss": 6.1689,
+      "step": 327
+    },
+    {
+      "epoch": 0.03723689205345708,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.002,
+      "loss": 6.1883,
+      "step": 328
+    },
+    {
+      "epoch": 0.037350419163376156,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.002,
+      "loss": 6.1699,
+      "step": 329
+    },
+    {
+      "epoch": 0.03746394627329523,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.002,
+      "loss": 6.1786,
+      "step": 330
+    },
+    {
+      "epoch": 0.03757747338321431,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.002,
+      "loss": 6.1778,
+      "step": 331
+    },
+    {
+      "epoch": 0.037691000493133384,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.002,
+      "loss": 6.1811,
+      "step": 332
+    },
+    {
+      "epoch": 0.03780452760305246,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.002,
+      "loss": 6.1769,
+      "step": 333
+    },
+    {
+      "epoch": 0.037918054712971536,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.002,
+      "loss": 6.1642,
+      "step": 334
+    },
+    {
+      "epoch": 0.03803158182289061,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.002,
+      "loss": 6.1512,
+      "step": 335
+    },
+    {
+      "epoch": 0.03814510893280969,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.002,
+      "loss": 6.1648,
+      "step": 336
+    },
+    {
+      "epoch": 0.038258636042728764,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.002,
+      "loss": 6.1735,
+      "step": 337
+    },
+    {
+      "epoch": 0.03837216315264784,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.002,
+      "loss": 6.1485,
+      "step": 338
+    },
+    {
+      "epoch": 0.038485690262566916,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.002,
+      "loss": 6.1556,
+      "step": 339
+    },
+    {
+      "epoch": 0.038599217372486,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.002,
+      "loss": 6.1484,
+      "step": 340
+    },
+    {
+      "epoch": 0.038712744482405075,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.002,
+      "loss": 6.1459,
+      "step": 341
+    },
+    {
+      "epoch": 0.03882627159232415,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 6.1456,
+      "step": 342
+    },
+    {
+      "epoch": 0.03893979870224323,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 6.1399,
+      "step": 343
+    },
+    {
+      "epoch": 0.0390533258121623,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.002,
+      "loss": 6.1475,
+      "step": 344
+    },
+    {
+      "epoch": 0.03916685292208138,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.002,
+      "loss": 6.1336,
+      "step": 345
+    },
+    {
+      "epoch": 0.039280380032000455,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.002,
+      "loss": 6.133,
+      "step": 346
+    },
+    {
+      "epoch": 0.03939390714191953,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.002,
+      "loss": 6.1262,
+      "step": 347
+    },
+    {
+      "epoch": 0.039507434251838607,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.002,
+      "loss": 6.1377,
+      "step": 348
+    },
+    {
+      "epoch": 0.03962096136175768,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.002,
+      "loss": 6.1175,
+      "step": 349
+    },
+    {
+      "epoch": 0.03973448847167676,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.002,
+      "loss": 6.1234,
+      "step": 350
+    },
+    {
+      "epoch": 0.039848015581595835,
+      "grad_norm": 0.625,
+      "learning_rate": 0.002,
+      "loss": 6.1384,
+      "step": 351
+    },
+    {
+      "epoch": 0.03996154269151491,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.002,
+      "loss": 6.1445,
+      "step": 352
+    },
+    {
+      "epoch": 0.040075069801433987,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.002,
+      "loss": 6.1153,
+      "step": 353
+    },
+    {
+      "epoch": 0.04018859691135306,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.002,
+      "loss": 6.1028,
+      "step": 354
+    },
+    {
+      "epoch": 0.040302124021272145,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.002,
+      "loss": 6.1172,
+      "step": 355
+    },
+    {
+      "epoch": 0.04041565113119122,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.002,
+      "loss": 6.1185,
+      "step": 356
+    },
+    {
+      "epoch": 0.0405291782411103,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.002,
+      "loss": 6.087,
+      "step": 357
+    },
+    {
+      "epoch": 0.04064270535102937,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.002,
+      "loss": 6.0829,
+      "step": 358
+    },
+    {
+      "epoch": 0.04075623246094845,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.002,
+      "loss": 6.0981,
+      "step": 359
+    },
+    {
+      "epoch": 0.040869759570867525,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.002,
+      "loss": 6.1014,
+      "step": 360
+    },
+    {
+      "epoch": 0.0409832866807866,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.002,
+      "loss": 6.0917,
+      "step": 361
+    },
+    {
+      "epoch": 0.04109681379070568,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.002,
+      "loss": 6.1174,
+      "step": 362
+    },
+    {
+      "epoch": 0.04121034090062475,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.002,
+      "loss": 6.0854,
+      "step": 363
+    },
+    {
+      "epoch": 0.04132386801054383,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.002,
+      "loss": 6.081,
+      "step": 364
+    },
+    {
+      "epoch": 0.041437395120462905,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.002,
+      "loss": 6.1024,
+      "step": 365
+    },
+    {
+      "epoch": 0.04155092223038198,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.002,
+      "loss": 6.0949,
+      "step": 366
+    },
+    {
+      "epoch": 0.04166444934030106,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.002,
+      "loss": 6.079,
+      "step": 367
+    },
+    {
+      "epoch": 0.04177797645022013,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.002,
+      "loss": 6.0801,
+      "step": 368
+    },
+    {
+      "epoch": 0.04189150356013921,
+      "grad_norm": 0.625,
+      "learning_rate": 0.002,
+      "loss": 6.0927,
+      "step": 369
+    },
+    {
+      "epoch": 0.04200503067005829,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.002,
+      "loss": 6.0886,
+      "step": 370
+    },
+    {
+      "epoch": 0.04211855777997737,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.002,
+      "loss": 6.0786,
+      "step": 371
+    },
+    {
+      "epoch": 0.042232084889896444,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.002,
+      "loss": 6.0789,
+      "step": 372
+    },
+    {
+      "epoch": 0.04234561199981552,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.002,
+      "loss": 6.0908,
+      "step": 373
+    },
+    {
+      "epoch": 0.042459139109734596,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.002,
+      "loss": 6.0615,
+      "step": 374
+    },
+    {
+      "epoch": 0.04257266621965367,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.002,
+      "loss": 6.0617,
+      "step": 375
+    },
+    {
+      "epoch": 0.04268619332957275,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.002,
+      "loss": 6.0826,
+      "step": 376
+    },
+    {
+      "epoch": 0.042799720439491824,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.002,
+      "loss": 6.0556,
+      "step": 377
+    },
+    {
+      "epoch": 0.0429132475494109,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.002,
+      "loss": 6.0778,
+      "step": 378
+    },
+    {
+      "epoch": 0.043026774659329976,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.002,
+      "loss": 6.0702,
+      "step": 379
+    },
+    {
+      "epoch": 0.04314030176924905,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.002,
+      "loss": 6.062,
+      "step": 380
+    },
+    {
+      "epoch": 0.04325382887916813,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.002,
+      "loss": 6.0649,
+      "step": 381
+    },
+    {
+      "epoch": 0.043367355989087204,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.002,
+      "loss": 6.0691,
+      "step": 382
+    },
+    {
+      "epoch": 0.04348088309900628,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.002,
+      "loss": 6.0609,
+      "step": 383
+    },
+    {
+      "epoch": 0.043594410208925356,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.002,
+      "loss": 6.0438,
+      "step": 384
+    },
+    {
+      "epoch": 0.04370793731884444,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.002,
+      "loss": 6.059,
+      "step": 385
+    },
+    {
+      "epoch": 0.043821464428763515,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.002,
+      "loss": 6.0472,
+      "step": 386
+    },
+    {
+      "epoch": 0.04393499153868259,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.002,
+      "loss": 6.0531,
+      "step": 387
+    },
+    {
+      "epoch": 0.04404851864860167,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.002,
+      "loss": 6.0568,
+      "step": 388
+    },
+    {
+      "epoch": 0.04416204575852074,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 6.047,
+      "step": 389
+    },
+    {
+      "epoch": 0.04427557286843982,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 6.0349,
+      "step": 390
+    },
+    {
+      "epoch": 0.044389099978358895,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.002,
+      "loss": 6.032,
+      "step": 391
+    },
+    {
+      "epoch": 0.04450262708827797,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.002,
+      "loss": 6.034,
+      "step": 392
+    },
+    {
+      "epoch": 0.04461615419819705,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.002,
+      "loss": 6.0341,
+      "step": 393
+    },
+    {
+      "epoch": 0.04472968130811612,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.002,
+      "loss": 6.0278,
+      "step": 394
+    },
+    {
+      "epoch": 0.0448432084180352,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.002,
+      "loss": 6.0432,
+      "step": 395
+    },
+    {
+      "epoch": 0.044956735527954275,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.002,
+      "loss": 6.0241,
+      "step": 396
+    },
+    {
+      "epoch": 0.04507026263787335,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.002,
+      "loss": 6.0446,
+      "step": 397
+    },
+    {
+      "epoch": 0.04518378974779243,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.002,
+      "loss": 6.0416,
+      "step": 398
+    },
+    {
+      "epoch": 0.0452973168577115,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.002,
+      "loss": 6.0243,
+      "step": 399
+    },
+    {
+      "epoch": 0.045410843967630586,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.002,
+      "loss": 6.0252,
+      "step": 400
+    },
+    {
+      "epoch": 0.04552437107754966,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.002,
+      "loss": 6.0338,
+      "step": 401
+    },
+    {
+      "epoch": 0.04563789818746874,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.002,
+      "loss": 6.023,
+      "step": 402
+    },
+    {
+      "epoch": 0.045751425297387814,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.002,
+      "loss": 6.0366,
+      "step": 403
+    },
+    {
+      "epoch": 0.04586495240730689,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.002,
+      "loss": 6.048,
+      "step": 404
+    },
+    {
+      "epoch": 0.045978479517225966,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.002,
+      "loss": 6.0079,
+      "step": 405
+    },
+    {
+      "epoch": 0.04609200662714504,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.002,
+      "loss": 5.9989,
+      "step": 406
+    },
+    {
+      "epoch": 0.04620553373706412,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.002,
+      "loss": 6.0113,
+      "step": 407
+    },
+    {
+      "epoch": 0.046319060846983194,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.002,
+      "loss": 6.0031,
+      "step": 408
+    },
+    {
+      "epoch": 0.04643258795690227,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.002,
+      "loss": 6.0011,
+      "step": 409
+    },
+    {
+      "epoch": 0.046546115066821346,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.002,
+      "loss": 6.0037,
+      "step": 410
+    },
+    {
+      "epoch": 0.04665964217674042,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.002,
+      "loss": 6.0108,
+      "step": 411
+    },
+    {
+      "epoch": 0.0467731692866595,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.002,
+      "loss": 6.0067,
+      "step": 412
+    },
+    {
+      "epoch": 0.046886696396578574,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.002,
+      "loss": 6.0059,
+      "step": 413
+    },
+    {
+      "epoch": 0.04700022350649765,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.002,
+      "loss": 6.0045,
+      "step": 414
+    },
+    {
+      "epoch": 0.04711375061641673,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.002,
+      "loss": 6.0096,
+      "step": 415
+    },
+    {
+      "epoch": 0.04722727772633581,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.002,
+      "loss": 5.9978,
+      "step": 416
+    },
+    {
+      "epoch": 0.047340804836254885,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.002,
+      "loss": 5.9964,
+      "step": 417
+    },
+    {
+      "epoch": 0.04745433194617396,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.002,
+      "loss": 5.985,
+      "step": 418
+    },
+    {
+      "epoch": 0.04756785905609304,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.002,
+      "loss": 5.9858,
+      "step": 419
+    },
+    {
+      "epoch": 0.04768138616601211,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.002,
+      "loss": 5.9872,
+      "step": 420
+    },
+    {
+      "epoch": 0.04779491327593119,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.002,
+      "loss": 5.9986,
+      "step": 421
+    },
+    {
+      "epoch": 0.047908440385850265,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.002,
+      "loss": 5.9791,
+      "step": 422
+    },
+    {
+      "epoch": 0.04802196749576934,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.002,
+      "loss": 5.9766,
+      "step": 423
+    },
+    {
+      "epoch": 0.04813549460568842,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.002,
+      "loss": 5.9913,
+      "step": 424
+    },
+    {
+      "epoch": 0.04824902171560749,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.002,
+      "loss": 5.9781,
+      "step": 425
+    },
+    {
+      "epoch": 0.04836254882552657,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.002,
+      "loss": 5.9782,
+      "step": 426
+    },
+    {
+      "epoch": 0.048476075935445645,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.9753,
+      "step": 427
+    },
+    {
+      "epoch": 0.04858960304536472,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.002,
+      "loss": 5.968,
+      "step": 428
+    },
+    {
+      "epoch": 0.0487031301552838,
+      "grad_norm": 0.5,
+      "learning_rate": 0.002,
+      "loss": 5.973,
+      "step": 429
+    },
+    {
+      "epoch": 0.04881665726520288,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.002,
+      "loss": 5.9632,
+      "step": 430
+    },
+    {
+      "epoch": 0.048930184375121956,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.002,
+      "loss": 5.969,
+      "step": 431
+    },
+    {
+      "epoch": 0.04904371148504103,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.002,
+      "loss": 5.9568,
+      "step": 432
+    },
+    {
+      "epoch": 0.04915723859496011,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.002,
+      "loss": 5.9873,
+      "step": 433
+    },
+    {
+      "epoch": 0.049270765704879184,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.002,
+      "loss": 5.9671,
+      "step": 434
+    },
+    {
+      "epoch": 0.04938429281479826,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.002,
+      "loss": 5.9707,
+      "step": 435
+    },
+    {
+      "epoch": 0.049497819924717336,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.002,
+      "loss": 5.9794,
+      "step": 436
+    },
+    {
+      "epoch": 0.04961134703463641,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.002,
+      "loss": 5.9613,
+      "step": 437
+    },
+    {
+      "epoch": 0.04972487414455549,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.002,
+      "loss": 5.9623,
+      "step": 438
+    },
+    {
+      "epoch": 0.049838401254474564,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.002,
+      "loss": 5.9661,
+      "step": 439
+    },
+    {
+      "epoch": 0.04995192836439364,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.002,
+      "loss": 5.971,
+      "step": 440
+    },
+    {
+      "epoch": 0.050065455474312716,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.002,
+      "loss": 5.9666,
+      "step": 441
+    },
+    {
+      "epoch": 0.05017898258423179,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.002,
+      "loss": 5.9563,
+      "step": 442
+    },
+    {
+      "epoch": 0.05029250969415087,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.002,
+      "loss": 5.9467,
+      "step": 443
+    },
+    {
+      "epoch": 0.050406036804069944,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.002,
+      "loss": 5.961,
+      "step": 444
+    },
+    {
+      "epoch": 0.05051956391398903,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.002,
+      "loss": 5.9395,
+      "step": 445
+    },
+    {
+      "epoch": 0.0506330910239081,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.002,
+      "loss": 5.9592,
+      "step": 446
+    },
+    {
+      "epoch": 0.05074661813382718,
+      "grad_norm": 0.625,
+      "learning_rate": 0.002,
+      "loss": 5.9602,
+      "step": 447
+    },
+    {
+      "epoch": 0.050860145243746255,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.002,
+      "loss": 5.9538,
+      "step": 448
+    },
+    {
+      "epoch": 0.05097367235366533,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.002,
+      "loss": 5.9509,
+      "step": 449
+    },
+    {
+      "epoch": 0.05108719946358441,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.002,
+      "loss": 5.9407,
+      "step": 450
+    },
+    {
+      "epoch": 0.05120072657350348,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.002,
+      "loss": 5.9565,
+      "step": 451
+    },
+    {
+      "epoch": 0.05131425368342256,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.002,
+      "loss": 5.9379,
+      "step": 452
+    },
+    {
+      "epoch": 0.051427780793341635,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.002,
+      "loss": 5.9503,
+      "step": 453
+    },
+    {
+      "epoch": 0.05154130790326071,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.002,
+      "loss": 5.9299,
+      "step": 454
+    },
+    {
+      "epoch": 0.05165483501317979,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.002,
+      "loss": 5.9241,
+      "step": 455
+    },
+    {
+      "epoch": 0.05176836212309886,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.002,
+      "loss": 5.9515,
+      "step": 456
+    },
+    {
+      "epoch": 0.05188188923301794,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.002,
+      "loss": 5.9157,
+      "step": 457
+    },
+    {
+      "epoch": 0.051995416342937015,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.002,
+      "loss": 5.9354,
+      "step": 458
+    },
+    {
+      "epoch": 0.05210894345285609,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.002,
+      "loss": 5.9348,
+      "step": 459
+    },
+    {
+      "epoch": 0.05222247056277517,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.002,
+      "loss": 5.9315,
+      "step": 460
+    },
+    {
+      "epoch": 0.05233599767269425,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 5.9449,
+      "step": 461
+    },
+    {
+      "epoch": 0.052449524782613326,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.9425,
+      "step": 462
+    },
+    {
+      "epoch": 0.0525630518925324,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.002,
+      "loss": 5.9275,
+      "step": 463
+    },
+    {
+      "epoch": 0.05267657900245148,
+      "grad_norm": 0.5,
+      "learning_rate": 0.002,
+      "loss": 5.9382,
+      "step": 464
+    },
+    {
+      "epoch": 0.052790106112370554,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.002,
+      "loss": 5.9147,
+      "step": 465
+    },
+    {
+      "epoch": 0.05290363322228963,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 5.9249,
+      "step": 466
+    },
+    {
+      "epoch": 0.053017160332208706,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.002,
+      "loss": 5.9074,
+      "step": 467
+    },
+    {
+      "epoch": 0.05313068744212778,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.002,
+      "loss": 5.9267,
+      "step": 468
+    },
+    {
+      "epoch": 0.05324421455204686,
+      "grad_norm": 0.625,
+      "learning_rate": 0.002,
+      "loss": 5.9167,
+      "step": 469
+    },
+    {
+      "epoch": 0.053357741661965934,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.002,
+      "loss": 5.9287,
+      "step": 470
+    },
+    {
+      "epoch": 0.05347126877188501,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.002,
+      "loss": 5.9367,
+      "step": 471
+    },
+    {
+      "epoch": 0.053584795881804086,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.002,
+      "loss": 5.9199,
+      "step": 472
+    },
+    {
+      "epoch": 0.05369832299172316,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.002,
+      "loss": 5.9235,
+      "step": 473
+    },
+    {
+      "epoch": 0.05381185010164224,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.8995,
+      "step": 474
+    },
+    {
+      "epoch": 0.053925377211561314,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.002,
+      "loss": 5.9214,
+      "step": 475
+    },
+    {
+      "epoch": 0.0540389043214804,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.002,
+      "loss": 5.9146,
+      "step": 476
+    },
+    {
+      "epoch": 0.05415243143139947,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.002,
+      "loss": 5.9127,
+      "step": 477
+    },
+    {
+      "epoch": 0.05426595854131855,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.002,
+      "loss": 5.9092,
+      "step": 478
+    },
+    {
+      "epoch": 0.054379485651237625,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.914,
+      "step": 479
+    },
+    {
+      "epoch": 0.0544930127611567,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 5.9071,
+      "step": 480
+    },
+    {
+      "epoch": 0.05460653987107578,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.9036,
+      "step": 481
+    },
+    {
+      "epoch": 0.05472006698099485,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.002,
+      "loss": 5.912,
+      "step": 482
+    },
+    {
+      "epoch": 0.05483359409091393,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.002,
+      "loss": 5.9081,
+      "step": 483
+    },
+    {
+      "epoch": 0.054947121200833005,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.002,
+      "loss": 5.8961,
+      "step": 484
+    },
+    {
+      "epoch": 0.05506064831075208,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.002,
+      "loss": 5.8945,
+      "step": 485
+    },
+    {
+      "epoch": 0.05517417542067116,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.896,
+      "step": 486
+    },
+    {
+      "epoch": 0.05528770253059023,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.002,
+      "loss": 5.8671,
+      "step": 487
+    },
+    {
+      "epoch": 0.05540122964050931,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.002,
+      "loss": 5.9104,
+      "step": 488
+    },
+    {
+      "epoch": 0.055514756750428385,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.908,
+      "step": 489
+    },
+    {
+      "epoch": 0.05562828386034746,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.002,
+      "loss": 5.8806,
+      "step": 490
+    },
+    {
+      "epoch": 0.05574181097026654,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.002,
+      "loss": 5.9009,
+      "step": 491
+    },
+    {
+      "epoch": 0.05585533808018562,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 5.8843,
+      "step": 492
+    },
+    {
+      "epoch": 0.055968865190104695,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.002,
+      "loss": 5.8909,
+      "step": 493
+    },
+    {
+      "epoch": 0.05608239230002377,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.002,
+      "loss": 5.8813,
+      "step": 494
+    },
+    {
+      "epoch": 0.05619591940994285,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.002,
+      "loss": 5.9248,
+      "step": 495
+    },
+    {
+      "epoch": 0.05630944651986192,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.002,
+      "loss": 5.8991,
+      "step": 496
+    },
+    {
+      "epoch": 0.056422973629781,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.002,
+      "loss": 5.8897,
+      "step": 497
+    },
+    {
+      "epoch": 0.056536500739700075,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.002,
+      "loss": 5.9034,
+      "step": 498
+    },
+    {
+      "epoch": 0.05665002784961915,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.002,
+      "loss": 5.8889,
+      "step": 499
+    },
+    {
+      "epoch": 0.05676355495953823,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.002,
+      "loss": 5.891,
+      "step": 500
+    },
+    {
+      "epoch": 0.0568770820694573,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.002,
+      "loss": 5.8765,
+      "step": 501
+    },
+    {
+      "epoch": 0.05699060917937638,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.002,
+      "loss": 5.8805,
+      "step": 502
+    },
+    {
+      "epoch": 0.057104136289295455,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.002,
+      "loss": 5.8843,
+      "step": 503
+    },
+    {
+      "epoch": 0.05721766339921453,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.002,
+      "loss": 5.8774,
+      "step": 504
+    },
+    {
+      "epoch": 0.05733119050913361,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.002,
+      "loss": 5.8876,
+      "step": 505
+    },
+    {
+      "epoch": 0.05744471761905269,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.002,
+      "loss": 5.8917,
+      "step": 506
+    },
+    {
+      "epoch": 0.057558244728971766,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.002,
+      "loss": 5.8807,
+      "step": 507
+    },
+    {
+      "epoch": 0.05767177183889084,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.002,
+      "loss": 5.8686,
+      "step": 508
+    },
+    {
+      "epoch": 0.05778529894880992,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.002,
+      "loss": 5.8777,
+      "step": 509
+    },
+    {
+      "epoch": 0.057898826058728994,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 5.878,
+      "step": 510
+    },
+    {
+      "epoch": 0.05801235316864807,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.8672,
+      "step": 511
+    },
+    {
+      "epoch": 0.058125880278567146,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.8738,
+      "step": 512
+    },
+    {
+      "epoch": 0.05823940738848622,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.8602,
+      "step": 513
+    },
+    {
+      "epoch": 0.0583529344984053,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.002,
+      "loss": 5.858,
+      "step": 514
+    },
+    {
+      "epoch": 0.058466461608324374,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.002,
+      "loss": 5.866,
+      "step": 515
+    },
+    {
+      "epoch": 0.05857998871824345,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.002,
+      "loss": 5.8628,
+      "step": 516
+    },
+    {
+      "epoch": 0.058693515828162526,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.002,
+      "loss": 5.8504,
+      "step": 517
+    },
+    {
+      "epoch": 0.0588070429380816,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.002,
+      "loss": 5.8885,
+      "step": 518
+    },
+    {
+      "epoch": 0.05892057004800068,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.002,
+      "loss": 5.8404,
+      "step": 519
+    },
+    {
+      "epoch": 0.059034097157919754,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.8517,
+      "step": 520
+    },
+    {
+      "epoch": 0.05914762426783884,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.8683,
+      "step": 521
+    },
+    {
+      "epoch": 0.05926115137775791,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.002,
+      "loss": 5.8823,
+      "step": 522
+    },
+    {
+      "epoch": 0.05937467848767699,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.8534,
+      "step": 523
+    },
+    {
+      "epoch": 0.059488205597596065,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.002,
+      "loss": 5.861,
+      "step": 524
+    },
+    {
+      "epoch": 0.05960173270751514,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.002,
+      "loss": 5.8512,
+      "step": 525
+    },
+    {
+      "epoch": 0.05971525981743422,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.002,
+      "loss": 5.8561,
+      "step": 526
+    },
+    {
+      "epoch": 0.05982878692735329,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.002,
+      "loss": 5.8532,
+      "step": 527
+    },
+    {
+      "epoch": 0.05994231403727237,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.002,
+      "loss": 5.854,
+      "step": 528
+    },
+    {
+      "epoch": 0.060055841147191445,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.002,
+      "loss": 5.8502,
+      "step": 529
+    },
+    {
+      "epoch": 0.06016936825711052,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.002,
+      "loss": 5.8568,
+      "step": 530
+    },
+    {
+      "epoch": 0.0602828953670296,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 5.8416,
+      "step": 531
+    },
+    {
+      "epoch": 0.06039642247694867,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.002,
+      "loss": 5.8455,
+      "step": 532
+    },
+    {
+      "epoch": 0.06050994958686775,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.002,
+      "loss": 5.8462,
+      "step": 533
+    },
+    {
+      "epoch": 0.060623476696786825,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.002,
+      "loss": 5.8577,
+      "step": 534
+    },
+    {
+      "epoch": 0.0607370038067059,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.002,
+      "loss": 5.8566,
+      "step": 535
+    },
+    {
+      "epoch": 0.060850530916624984,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 5.8322,
+      "step": 536
+    },
+    {
+      "epoch": 0.06096405802654406,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.002,
+      "loss": 5.8367,
+      "step": 537
+    },
+    {
+      "epoch": 0.061077585136463136,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.002,
+      "loss": 5.8477,
+      "step": 538
+    },
+    {
+      "epoch": 0.06119111224638221,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.002,
+      "loss": 5.8544,
+      "step": 539
+    },
+    {
+      "epoch": 0.06130463935630129,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.002,
+      "loss": 5.8441,
+      "step": 540
+    },
+    {
+      "epoch": 0.061418166466220364,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.002,
+      "loss": 5.8381,
+      "step": 541
+    },
+    {
+      "epoch": 0.06153169357613944,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.002,
+      "loss": 5.834,
+      "step": 542
+    },
+    {
+      "epoch": 0.061645220686058516,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.002,
+      "loss": 5.8289,
+      "step": 543
+    },
+    {
+      "epoch": 0.06175874779597759,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.002,
+      "loss": 5.8362,
+      "step": 544
+    },
+    {
+      "epoch": 0.06187227490589667,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.002,
+      "loss": 5.8187,
+      "step": 545
+    },
+    {
+      "epoch": 0.061985802015815744,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.002,
+      "loss": 5.8164,
+      "step": 546
+    },
+    {
+      "epoch": 0.06209932912573482,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.002,
+      "loss": 5.8342,
+      "step": 547
+    },
+    {
+      "epoch": 0.062212856235653896,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.002,
+      "loss": 5.8349,
+      "step": 548
+    },
+    {
+      "epoch": 0.06232638334557297,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.002,
+      "loss": 5.8105,
+      "step": 549
+    },
+    {
+      "epoch": 0.06243991045549205,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.002,
+      "loss": 5.8258,
+      "step": 550
+    },
+    {
+      "epoch": 0.06255343756541112,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.002,
+      "loss": 5.8211,
+      "step": 551
+    },
+    {
+      "epoch": 0.0626669646753302,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.002,
+      "loss": 5.8138,
+      "step": 552
+    },
+    {
+      "epoch": 0.06278049178524928,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.002,
+      "loss": 5.8251,
+      "step": 553
+    },
+    {
+      "epoch": 0.06289401889516835,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.8248,
+      "step": 554
+    },
+    {
+      "epoch": 0.06300754600508743,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.8315,
+      "step": 555
+    },
+    {
+      "epoch": 0.0631210731150065,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.8225,
+      "step": 556
+    },
+    {
+      "epoch": 0.06323460022492558,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.8259,
+      "step": 557
+    },
+    {
+      "epoch": 0.06334812733484466,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.8122,
+      "step": 558
+    },
+    {
+      "epoch": 0.06346165444476375,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.8048,
+      "step": 559
+    },
+    {
+      "epoch": 0.06357518155468282,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.8143,
+      "step": 560
+    },
+    {
+      "epoch": 0.0636887086646019,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 5.82,
+      "step": 561
+    },
+    {
+      "epoch": 0.06380223577452097,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002,
+      "loss": 5.8199,
+      "step": 562
+    },
+    {
+      "epoch": 0.06391576288444005,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.002,
+      "loss": 5.8225,
+      "step": 563
+    },
+    {
+      "epoch": 0.06402928999435913,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 5.8071,
+      "step": 564
+    },
+    {
+      "epoch": 0.0641428171042782,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.002,
+      "loss": 5.7962,
+      "step": 565
+    },
+    {
+      "epoch": 0.06425634421419728,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.002,
+      "loss": 5.8026,
+      "step": 566
+    },
+    {
+      "epoch": 0.06436987132411635,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.002,
+      "loss": 5.7825,
+      "step": 567
+    },
+    {
+      "epoch": 0.06448339843403543,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.002,
+      "loss": 5.8015,
+      "step": 568
+    },
+    {
+      "epoch": 0.0645969255439545,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.002,
+      "loss": 5.8206,
+      "step": 569
+    },
+    {
+      "epoch": 0.06471045265387358,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.002,
+      "loss": 5.8249,
+      "step": 570
+    },
+    {
+      "epoch": 0.06482397976379266,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.002,
+      "loss": 5.7873,
+      "step": 571
+    },
+    {
+      "epoch": 0.06493750687371173,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.002,
+      "loss": 5.8174,
+      "step": 572
+    },
+    {
+      "epoch": 0.06505103398363081,
+      "grad_norm": 0.5,
+      "learning_rate": 0.002,
+      "loss": 5.8136,
+      "step": 573
+    },
+    {
+      "epoch": 0.06516456109354989,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 5.8094,
+      "step": 574
+    },
+    {
+      "epoch": 0.06527808820346896,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.002,
+      "loss": 5.7936,
+      "step": 575
+    },
+    {
+      "epoch": 0.06539161531338804,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.002,
+      "loss": 5.8102,
+      "step": 576
+    },
+    {
+      "epoch": 0.06550514242330711,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.002,
+      "loss": 5.8202,
+      "step": 577
+    },
+    {
+      "epoch": 0.06561866953322619,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.002,
+      "loss": 5.7878,
+      "step": 578
+    },
+    {
+      "epoch": 0.06573219664314527,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.002,
+      "loss": 5.7993,
+      "step": 579
+    },
+    {
+      "epoch": 0.06584572375306434,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.002,
+      "loss": 5.7928,
+      "step": 580
+    },
+    {
+      "epoch": 0.06595925086298342,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.002,
+      "loss": 5.8023,
+      "step": 581
+    },
+    {
+      "epoch": 0.0660727779729025,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.002,
+      "loss": 5.7986,
+      "step": 582
+    },
+    {
+      "epoch": 0.06618630508282157,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.002,
+      "loss": 5.7899,
+      "step": 583
+    },
+    {
+      "epoch": 0.06629983219274065,
+      "grad_norm": 0.5,
+      "learning_rate": 0.002,
+      "loss": 5.7876,
+      "step": 584
+    },
+    {
+      "epoch": 0.06641335930265972,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.002,
+      "loss": 5.7819,
+      "step": 585
+    },
+    {
+      "epoch": 0.0665268864125788,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 5.79,
+      "step": 586
+    },
+    {
+      "epoch": 0.06664041352249787,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.002,
+      "loss": 5.8001,
+      "step": 587
+    },
+    {
+      "epoch": 0.06675394063241695,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.002,
+      "loss": 5.7964,
+      "step": 588
+    },
+    {
+      "epoch": 0.06686746774233604,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.7779,
+      "step": 589
+    },
+    {
+      "epoch": 0.06698099485225512,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.7976,
+      "step": 590
+    },
+    {
+      "epoch": 0.06709452196217419,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.8005,
+      "step": 591
+    },
+    {
+      "epoch": 0.06720804907209327,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.002,
+      "loss": 5.7917,
+      "step": 592
+    },
+    {
+      "epoch": 0.06732157618201234,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.002,
+      "loss": 5.8071,
+      "step": 593
+    },
+    {
+      "epoch": 0.06743510329193142,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.002,
+      "loss": 5.7799,
+      "step": 594
+    },
+    {
+      "epoch": 0.0675486304018505,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.002,
+      "loss": 5.8027,
+      "step": 595
+    },
+    {
+      "epoch": 0.06766215751176957,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.002,
+      "loss": 5.7886,
+      "step": 596
+    },
+    {
+      "epoch": 0.06777568462168865,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.002,
+      "loss": 5.7979,
+      "step": 597
+    },
+    {
+      "epoch": 0.06788921173160772,
+      "grad_norm": 0.5,
+      "learning_rate": 0.002,
+      "loss": 5.7848,
+      "step": 598
+    },
+    {
+      "epoch": 0.0680027388415268,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.002,
+      "loss": 5.7738,
+      "step": 599
+    },
+    {
+      "epoch": 0.06811626595144588,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.002,
+      "loss": 5.793,
+      "step": 600
+    },
+    {
+      "epoch": 0.06822979306136495,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.002,
+      "loss": 5.769,
+      "step": 601
+    },
+    {
+      "epoch": 0.06834332017128403,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.002,
+      "loss": 5.78,
+      "step": 602
+    },
+    {
+      "epoch": 0.0684568472812031,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.002,
+      "loss": 5.7849,
+      "step": 603
+    },
+    {
+      "epoch": 0.06857037439112218,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.002,
+      "loss": 5.7999,
+      "step": 604
+    },
+    {
+      "epoch": 0.06868390150104126,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.7781,
+      "step": 605
+    },
+    {
+      "epoch": 0.06879742861096033,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.7623,
+      "step": 606
+    },
+    {
+      "epoch": 0.06891095572087941,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.002,
+      "loss": 5.7853,
+      "step": 607
+    },
+    {
+      "epoch": 0.06902448283079848,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.002,
+      "loss": 5.7886,
+      "step": 608
+    },
+    {
+      "epoch": 0.06913800994071756,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.002,
+      "loss": 5.7733,
+      "step": 609
+    },
+    {
+      "epoch": 0.06925153705063664,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.002,
+      "loss": 5.7728,
+      "step": 610
+    },
+    {
+      "epoch": 0.06936506416055571,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.002,
+      "loss": 5.7883,
+      "step": 611
+    },
+    {
+      "epoch": 0.06947859127047479,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.002,
+      "loss": 5.7865,
+      "step": 612
+    },
+    {
+      "epoch": 0.06959211838039386,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 5.7726,
+      "step": 613
+    },
+    {
+      "epoch": 0.06970564549031294,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.7789,
+      "step": 614
+    },
+    {
+      "epoch": 0.06981917260023202,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 5.7612,
+      "step": 615
+    },
+    {
+      "epoch": 0.06993269971015109,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002,
+      "loss": 5.7471,
+      "step": 616
+    },
+    {
+      "epoch": 0.07004622682007017,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.002,
+      "loss": 5.766,
+      "step": 617
+    },
+    {
+      "epoch": 0.07015975392998924,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.002,
+      "loss": 5.765,
+      "step": 618
+    },
+    {
+      "epoch": 0.07027328103990833,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.002,
+      "loss": 5.7685,
+      "step": 619
+    },
+    {
+      "epoch": 0.07038680814982741,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.002,
+      "loss": 5.7545,
+      "step": 620
+    },
+    {
+      "epoch": 0.07050033525974649,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.002,
+      "loss": 5.7547,
+      "step": 621
+    },
+    {
+      "epoch": 0.07061386236966556,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.002,
+      "loss": 5.7689,
+      "step": 622
+    },
+    {
+      "epoch": 0.07072738947958464,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.7749,
+      "step": 623
+    },
+    {
+      "epoch": 0.07084091658950371,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.7507,
+      "step": 624
+    },
+    {
+      "epoch": 0.07095444369942279,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.7766,
+      "step": 625
+    },
+    {
+      "epoch": 0.07106797080934187,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.7641,
+      "step": 626
+    },
+    {
+      "epoch": 0.07118149791926094,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.7773,
+      "step": 627
+    },
+    {
+      "epoch": 0.07129502502918002,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.7531,
+      "step": 628
+    },
+    {
+      "epoch": 0.0714085521390991,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.7536,
+      "step": 629
+    },
+    {
+      "epoch": 0.07152207924901817,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 5.7559,
+      "step": 630
+    },
+    {
+      "epoch": 0.07163560635893725,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.779,
+      "step": 631
+    },
+    {
+      "epoch": 0.07174913346885632,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.755,
+      "step": 632
+    },
+    {
+      "epoch": 0.0718626605787754,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.7559,
+      "step": 633
+    },
+    {
+      "epoch": 0.07197618768869447,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.7532,
+      "step": 634
+    },
+    {
+      "epoch": 0.07208971479861355,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.748,
+      "step": 635
+    },
+    {
+      "epoch": 0.07220324190853263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 5.7481,
+      "step": 636
+    },
+    {
+      "epoch": 0.0723167690184517,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.002,
+      "loss": 5.7658,
+      "step": 637
+    },
+    {
+      "epoch": 0.07243029612837078,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.002,
+      "loss": 5.7519,
+      "step": 638
+    },
+    {
+      "epoch": 0.07254382323828985,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.002,
+      "loss": 5.7332,
+      "step": 639
+    },
+    {
+      "epoch": 0.07265735034820893,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.002,
+      "loss": 5.748,
+      "step": 640
+    },
+    {
+      "epoch": 0.072770877458128,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 5.7558,
+      "step": 641
+    },
+    {
+      "epoch": 0.07288440456804708,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 5.7501,
+      "step": 642
+    },
+    {
+      "epoch": 0.07299793167796616,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.002,
+      "loss": 5.7486,
+      "step": 643
+    },
+    {
+      "epoch": 0.07311145878788523,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.002,
+      "loss": 5.75,
+      "step": 644
+    },
+    {
+      "epoch": 0.07322498589780431,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.002,
+      "loss": 5.7473,
+      "step": 645
+    },
+    {
+      "epoch": 0.07333851300772339,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.002,
+      "loss": 5.7436,
+      "step": 646
+    },
+    {
+      "epoch": 0.07345204011764246,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.002,
+      "loss": 5.7425,
+      "step": 647
+    },
+    {
+      "epoch": 0.07356556722756154,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.002,
+      "loss": 5.7354,
+      "step": 648
+    },
+    {
+      "epoch": 0.07367909433748063,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 5.7463,
+      "step": 649
+    },
+    {
+      "epoch": 0.0737926214473997,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.002,
+      "loss": 5.7663,
+      "step": 650
+    },
+    {
+      "epoch": 0.07390614855731878,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.002,
+      "loss": 5.7306,
+      "step": 651
+    },
+    {
+      "epoch": 0.07401967566723786,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.002,
+      "loss": 5.7533,
+      "step": 652
+    },
+    {
+      "epoch": 0.07413320277715693,
+      "grad_norm": 0.5,
+      "learning_rate": 0.002,
+      "loss": 5.7253,
+      "step": 653
+    },
+    {
+      "epoch": 0.07424672988707601,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 5.7278,
+      "step": 654
+    },
+    {
+      "epoch": 0.07436025699699508,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.002,
+      "loss": 5.738,
+      "step": 655
+    },
+    {
+      "epoch": 0.07447378410691416,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.002,
+      "loss": 5.7384,
+      "step": 656
+    },
+    {
+      "epoch": 0.07458731121683324,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.002,
+      "loss": 5.7493,
+      "step": 657
+    },
+    {
+      "epoch": 0.07470083832675231,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.7397,
+      "step": 658
+    },
+    {
+      "epoch": 0.07481436543667139,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.002,
+      "loss": 5.7394,
+      "step": 659
+    },
+    {
+      "epoch": 0.07492789254659046,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.7142,
+      "step": 660
+    },
+    {
+      "epoch": 0.07504141965650954,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.7464,
+      "step": 661
+    },
+    {
+      "epoch": 0.07515494676642862,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.744,
+      "step": 662
+    },
+    {
+      "epoch": 0.07526847387634769,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.7447,
+      "step": 663
+    },
+    {
+      "epoch": 0.07538200098626677,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.7243,
+      "step": 664
+    },
+    {
+      "epoch": 0.07549552809618584,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.7268,
+      "step": 665
+    },
+    {
+      "epoch": 0.07560905520610492,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.7228,
+      "step": 666
+    },
+    {
+      "epoch": 0.075722582316024,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.002,
+      "loss": 5.76,
+      "step": 667
+    },
+    {
+      "epoch": 0.07583610942594307,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.002,
+      "loss": 5.7159,
+      "step": 668
+    },
+    {
+      "epoch": 0.07594963653586215,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 5.7309,
+      "step": 669
+    },
+    {
+      "epoch": 0.07606316364578122,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.7336,
+      "step": 670
+    },
+    {
+      "epoch": 0.0761766907557003,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.7507,
+      "step": 671
+    },
+    {
+      "epoch": 0.07629021786561938,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.7326,
+      "step": 672
+    },
+    {
+      "epoch": 0.07640374497553845,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.728,
+      "step": 673
+    },
+    {
+      "epoch": 0.07651727208545753,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.7249,
+      "step": 674
+    },
+    {
+      "epoch": 0.0766307991953766,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.002,
+      "loss": 5.7227,
+      "step": 675
+    },
+    {
+      "epoch": 0.07674432630529568,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002,
+      "loss": 5.708,
+      "step": 676
+    },
+    {
+      "epoch": 0.07685785341521476,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.002,
+      "loss": 5.7337,
+      "step": 677
+    },
+    {
+      "epoch": 0.07697138052513383,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002,
+      "loss": 5.7041,
+      "step": 678
+    },
+    {
+      "epoch": 0.07708490763505292,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.7493,
+      "step": 679
+    },
+    {
+      "epoch": 0.077198434744972,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.002,
+      "loss": 5.7239,
+      "step": 680
+    },
+    {
+      "epoch": 0.07731196185489107,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.002,
+      "loss": 5.7318,
+      "step": 681
+    },
+    {
+      "epoch": 0.07742548896481015,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.002,
+      "loss": 5.7337,
+      "step": 682
+    },
+    {
+      "epoch": 0.07753901607472923,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.002,
+      "loss": 5.7239,
+      "step": 683
+    },
+    {
+      "epoch": 0.0776525431846483,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.002,
+      "loss": 5.7422,
+      "step": 684
+    },
+    {
+      "epoch": 0.07776607029456738,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.002,
+      "loss": 5.7314,
+      "step": 685
+    },
+    {
+      "epoch": 0.07787959740448645,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.002,
+      "loss": 5.7197,
+      "step": 686
+    },
+    {
+      "epoch": 0.07799312451440553,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.002,
+      "loss": 5.7187,
+      "step": 687
+    },
+    {
+      "epoch": 0.0781066516243246,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.002,
+      "loss": 5.7262,
+      "step": 688
+    },
+    {
+      "epoch": 0.07822017873424368,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.002,
+      "loss": 5.7167,
+      "step": 689
+    },
+    {
+      "epoch": 0.07833370584416276,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 5.7378,
+      "step": 690
+    },
+    {
+      "epoch": 0.07844723295408183,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.002,
+      "loss": 5.7229,
+      "step": 691
+    },
+    {
+      "epoch": 0.07856076006400091,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.002,
+      "loss": 5.7405,
+      "step": 692
+    },
+    {
+      "epoch": 0.07867428717391999,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.002,
+      "loss": 5.7001,
+      "step": 693
+    },
+    {
+      "epoch": 0.07878781428383906,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.002,
+      "loss": 5.6996,
+      "step": 694
+    },
+    {
+      "epoch": 0.07890134139375814,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.002,
+      "loss": 5.7106,
+      "step": 695
+    },
+    {
+      "epoch": 0.07901486850367721,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.002,
+      "loss": 5.7326,
+      "step": 696
+    },
+    {
+      "epoch": 0.07912839561359629,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.002,
+      "loss": 5.7346,
+      "step": 697
+    },
+    {
+      "epoch": 0.07924192272351537,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.002,
+      "loss": 5.7037,
+      "step": 698
+    },
+    {
+      "epoch": 0.07935544983343444,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.002,
+      "loss": 5.7097,
+      "step": 699
+    },
+    {
+      "epoch": 0.07946897694335352,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.002,
+      "loss": 5.7179,
+      "step": 700
+    },
+    {
+      "epoch": 0.07958250405327259,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.7063,
+      "step": 701
+    },
+    {
+      "epoch": 0.07969603116319167,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.7066,
+      "step": 702
+    },
+    {
+      "epoch": 0.07980955827311075,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.6934,
+      "step": 703
+    },
+    {
+      "epoch": 0.07992308538302982,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.002,
+      "loss": 5.7136,
+      "step": 704
+    },
+    {
+      "epoch": 0.0800366124929489,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.002,
+      "loss": 5.7259,
+      "step": 705
+    },
+    {
+      "epoch": 0.08015013960286797,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.002,
+      "loss": 5.7111,
+      "step": 706
+    },
+    {
+      "epoch": 0.08026366671278705,
+      "grad_norm": 0.5,
+      "learning_rate": 0.002,
+      "loss": 5.7146,
+      "step": 707
+    },
+    {
+      "epoch": 0.08037719382270613,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.002,
+      "loss": 5.7274,
+      "step": 708
+    },
+    {
+      "epoch": 0.08049072093262521,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.002,
+      "loss": 5.7083,
+      "step": 709
+    },
+    {
+      "epoch": 0.08060424804254429,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.002,
+      "loss": 5.7025,
+      "step": 710
+    },
+    {
+      "epoch": 0.08071777515246337,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 5.7096,
+      "step": 711
+    },
+    {
+      "epoch": 0.08083130226238244,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.002,
+      "loss": 5.7122,
+      "step": 712
+    },
+    {
+      "epoch": 0.08094482937230152,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.002,
+      "loss": 5.7053,
+      "step": 713
+    },
+    {
+      "epoch": 0.0810583564822206,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.002,
+      "loss": 5.7054,
+      "step": 714
+    },
+    {
+      "epoch": 0.08117188359213967,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.7145,
+      "step": 715
+    },
+    {
+      "epoch": 0.08128541070205875,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.6956,
+      "step": 716
+    },
+    {
+      "epoch": 0.08139893781197782,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.714,
+      "step": 717
+    },
+    {
+      "epoch": 0.0815124649218969,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.6946,
+      "step": 718
+    },
+    {
+      "epoch": 0.08162599203181597,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.7053,
+      "step": 719
+    },
+    {
+      "epoch": 0.08173951914173505,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.7101,
+      "step": 720
+    },
+    {
+      "epoch": 0.08185304625165413,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 5.7074,
+      "step": 721
+    },
+    {
+      "epoch": 0.0819665733615732,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.002,
+      "loss": 5.7082,
+      "step": 722
+    },
+    {
+      "epoch": 0.08208010047149228,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.002,
+      "loss": 5.6899,
+      "step": 723
+    },
+    {
+      "epoch": 0.08219362758141135,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.002,
+      "loss": 5.7093,
+      "step": 724
+    },
+    {
+      "epoch": 0.08230715469133043,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 5.7094,
+      "step": 725
+    },
+    {
+      "epoch": 0.0824206818012495,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.7134,
+      "step": 726
+    },
+    {
+      "epoch": 0.08253420891116858,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.7131,
+      "step": 727
+    },
+    {
+      "epoch": 0.08264773602108766,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.7036,
+      "step": 728
+    },
+    {
+      "epoch": 0.08276126313100673,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.712,
+      "step": 729
+    },
+    {
+      "epoch": 0.08287479024092581,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.6783,
+      "step": 730
+    },
+    {
+      "epoch": 0.08298831735084489,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.6737,
+      "step": 731
+    },
+    {
+      "epoch": 0.08310184446076396,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.6974,
+      "step": 732
+    },
+    {
+      "epoch": 0.08321537157068304,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.7014,
+      "step": 733
+    },
+    {
+      "epoch": 0.08332889868060211,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.7028,
+      "step": 734
+    },
+    {
+      "epoch": 0.08344242579052119,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.6799,
+      "step": 735
+    },
+    {
+      "epoch": 0.08355595290044027,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.002,
+      "loss": 5.6847,
+      "step": 736
+    },
+    {
+      "epoch": 0.08366948001035934,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.6855,
+      "step": 737
+    },
+    {
+      "epoch": 0.08378300712027842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 5.6816,
+      "step": 738
+    },
+    {
+      "epoch": 0.0838965342301975,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.6893,
+      "step": 739
+    },
+    {
+      "epoch": 0.08401006134011658,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.6585,
+      "step": 740
+    },
+    {
+      "epoch": 0.08412358845003566,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.6878,
+      "step": 741
+    },
+    {
+      "epoch": 0.08423711555995474,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.002,
+      "loss": 5.7034,
+      "step": 742
+    },
+    {
+      "epoch": 0.08435064266987381,
+      "grad_norm": 0.5,
+      "learning_rate": 0.002,
+      "loss": 5.6927,
+      "step": 743
+    },
+    {
+      "epoch": 0.08446416977979289,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.002,
+      "loss": 5.6771,
+      "step": 744
+    },
+    {
+      "epoch": 0.08457769688971196,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.002,
+      "loss": 5.7051,
+      "step": 745
+    },
+    {
+      "epoch": 0.08469122399963104,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.002,
+      "loss": 5.6919,
+      "step": 746
+    },
+    {
+      "epoch": 0.08480475110955012,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.002,
+      "loss": 5.6853,
+      "step": 747
+    },
+    {
+      "epoch": 0.08491827821946919,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.002,
+      "loss": 5.6983,
+      "step": 748
+    },
+    {
+      "epoch": 0.08503180532938827,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.002,
+      "loss": 5.6894,
+      "step": 749
+    },
+    {
+      "epoch": 0.08514533243930734,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.002,
+      "loss": 5.6809,
+      "step": 750
+    },
+    {
+      "epoch": 0.08525885954922642,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.002,
+      "loss": 5.6604,
+      "step": 751
+    },
+    {
+      "epoch": 0.0853723866591455,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.002,
+      "loss": 5.6816,
+      "step": 752
+    },
+    {
+      "epoch": 0.08548591376906457,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 5.672,
+      "step": 753
+    },
+    {
+      "epoch": 0.08559944087898365,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.6886,
+      "step": 754
+    },
+    {
+      "epoch": 0.08571296798890272,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.6917,
+      "step": 755
+    },
+    {
+      "epoch": 0.0858264950988218,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.6703,
+      "step": 756
+    },
+    {
+      "epoch": 0.08594002220874088,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.6828,
+      "step": 757
+    },
+    {
+      "epoch": 0.08605354931865995,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.6771,
+      "step": 758
+    },
+    {
+      "epoch": 0.08616707642857903,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.6718,
+      "step": 759
+    },
+    {
+      "epoch": 0.0862806035384981,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.6722,
+      "step": 760
+    },
+    {
+      "epoch": 0.08639413064841718,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.6887,
+      "step": 761
+    },
+    {
+      "epoch": 0.08650765775833626,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.6836,
+      "step": 762
+    },
+    {
+      "epoch": 0.08662118486825533,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.002,
+      "loss": 5.678,
+      "step": 763
+    },
+    {
+      "epoch": 0.08673471197817441,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.6689,
+      "step": 764
+    },
+    {
+      "epoch": 0.08684823908809348,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.002,
+      "loss": 5.6632,
+      "step": 765
+    },
+    {
+      "epoch": 0.08696176619801256,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.6632,
+      "step": 766
+    },
+    {
+      "epoch": 0.08707529330793164,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.678,
+      "step": 767
+    },
+    {
+      "epoch": 0.08718882041785071,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.6778,
+      "step": 768
+    },
+    {
+      "epoch": 0.08730234752776979,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.6674,
+      "step": 769
+    },
+    {
+      "epoch": 0.08741587463768888,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.6487,
+      "step": 770
+    },
+    {
+      "epoch": 0.08752940174760795,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.6965,
+      "step": 771
+    },
+    {
+      "epoch": 0.08764292885752703,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.002,
+      "loss": 5.6662,
+      "step": 772
+    },
+    {
+      "epoch": 0.0877564559674461,
+      "grad_norm": 0.625,
+      "learning_rate": 0.002,
+      "loss": 5.6936,
+      "step": 773
+    },
+    {
+      "epoch": 0.08786998307736518,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 5.6424,
+      "step": 774
+    },
+    {
+      "epoch": 0.08798351018728426,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 5.6758,
+      "step": 775
+    },
+    {
+      "epoch": 0.08809703729720333,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 5.6707,
+      "step": 776
+    },
+    {
+      "epoch": 0.08821056440712241,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.6686,
+      "step": 777
+    },
+    {
+      "epoch": 0.08832409151704149,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.6656,
+      "step": 778
+    },
+    {
+      "epoch": 0.08843761862696056,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.6793,
+      "step": 779
+    },
+    {
+      "epoch": 0.08855114573687964,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.002,
+      "loss": 5.6657,
+      "step": 780
+    },
+    {
+      "epoch": 0.08866467284679871,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.002,
+      "loss": 5.6561,
+      "step": 781
+    },
+    {
+      "epoch": 0.08877819995671779,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.002,
+      "loss": 5.6558,
+      "step": 782
+    },
+    {
+      "epoch": 0.08889172706663687,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.6518,
+      "step": 783
+    },
+    {
+      "epoch": 0.08900525417655594,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.6836,
+      "step": 784
+    },
+    {
+      "epoch": 0.08911878128647502,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.6575,
+      "step": 785
+    },
+    {
+      "epoch": 0.0892323083963941,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.6695,
+      "step": 786
+    },
+    {
+      "epoch": 0.08934583550631317,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.6617,
+      "step": 787
+    },
+    {
+      "epoch": 0.08945936261623225,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.6731,
+      "step": 788
+    },
+    {
+      "epoch": 0.08957288972615132,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.6636,
+      "step": 789
+    },
+    {
+      "epoch": 0.0896864168360704,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.002,
+      "loss": 5.6592,
+      "step": 790
+    },
+    {
+      "epoch": 0.08979994394598947,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.002,
+      "loss": 5.6409,
+      "step": 791
+    },
+    {
+      "epoch": 0.08991347105590855,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.002,
+      "loss": 5.6797,
+      "step": 792
+    },
+    {
+      "epoch": 0.09002699816582763,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.002,
+      "loss": 5.6531,
+      "step": 793
+    },
+    {
+      "epoch": 0.0901405252757467,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.6628,
+      "step": 794
+    },
+    {
+      "epoch": 0.09025405238566578,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 5.6652,
+      "step": 795
+    },
+    {
+      "epoch": 0.09036757949558485,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.6619,
+      "step": 796
+    },
+    {
+      "epoch": 0.09048110660550393,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.6807,
+      "step": 797
+    },
+    {
+      "epoch": 0.090594633715423,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.6865,
+      "step": 798
+    },
+    {
+      "epoch": 0.09070816082534208,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.6616,
+      "step": 799
+    },
+    {
+      "epoch": 0.09082168793526117,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.6624,
+      "step": 800
+    },
+    {
+      "epoch": 0.09093521504518025,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.002,
+      "loss": 5.6658,
+      "step": 801
+    },
+    {
+      "epoch": 0.09104874215509932,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.002,
+      "loss": 5.6552,
+      "step": 802
+    },
+    {
+      "epoch": 0.0911622692650184,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.002,
+      "loss": 5.6669,
+      "step": 803
+    },
+    {
+      "epoch": 0.09127579637493748,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.675,
+      "step": 804
+    },
+    {
+      "epoch": 0.09138932348485655,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.6735,
+      "step": 805
+    },
+    {
+      "epoch": 0.09150285059477563,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.6651,
+      "step": 806
+    },
+    {
+      "epoch": 0.0916163777046947,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.6541,
+      "step": 807
+    },
+    {
+      "epoch": 0.09172990481461378,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.6562,
+      "step": 808
+    },
+    {
+      "epoch": 0.09184343192453286,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.002,
+      "loss": 5.6588,
+      "step": 809
+    },
+    {
+      "epoch": 0.09195695903445193,
+      "grad_norm": 0.5,
+      "learning_rate": 0.002,
+      "loss": 5.6664,
+      "step": 810
+    },
+    {
+      "epoch": 0.09207048614437101,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.002,
+      "loss": 5.65,
+      "step": 811
+    },
+    {
+      "epoch": 0.09218401325429008,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 5.6573,
+      "step": 812
+    },
+    {
+      "epoch": 0.09229754036420916,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.002,
+      "loss": 5.6691,
+      "step": 813
+    },
+    {
+      "epoch": 0.09241106747412824,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.6454,
+      "step": 814
+    },
+    {
+      "epoch": 0.09252459458404731,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.002,
+      "loss": 5.6413,
+      "step": 815
+    },
+    {
+      "epoch": 0.09263812169396639,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.002,
+      "loss": 5.6545,
+      "step": 816
+    },
+    {
+      "epoch": 0.09275164880388546,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.002,
+      "loss": 5.6425,
+      "step": 817
+    },
+    {
+      "epoch": 0.09286517591380454,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.6416,
+      "step": 818
+    },
+    {
+      "epoch": 0.09297870302372362,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.002,
+      "loss": 5.6595,
+      "step": 819
+    },
+    {
+      "epoch": 0.09309223013364269,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.6368,
+      "step": 820
+    },
+    {
+      "epoch": 0.09320575724356177,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.6714,
+      "step": 821
+    },
+    {
+      "epoch": 0.09331928435348084,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.6696,
+      "step": 822
+    },
+    {
+      "epoch": 0.09343281146339992,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.6734,
+      "step": 823
+    },
+    {
+      "epoch": 0.093546338573319,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.673,
+      "step": 824
+    },
+    {
+      "epoch": 0.09365986568323807,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.645,
+      "step": 825
+    },
+    {
+      "epoch": 0.09377339279315715,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.6506,
+      "step": 826
+    },
+    {
+      "epoch": 0.09388691990307622,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.6317,
+      "step": 827
+    },
+    {
+      "epoch": 0.0940004470129953,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.6394,
+      "step": 828
+    },
+    {
+      "epoch": 0.09411397412291438,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.6609,
+      "step": 829
+    },
+    {
+      "epoch": 0.09422750123283347,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.6418,
+      "step": 830
+    },
+    {
+      "epoch": 0.09434102834275254,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.6387,
+      "step": 831
+    },
+    {
+      "epoch": 0.09445455545267162,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.6421,
+      "step": 832
+    },
+    {
+      "epoch": 0.0945680825625907,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.656,
+      "step": 833
+    },
+    {
+      "epoch": 0.09468160967250977,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.6557,
+      "step": 834
+    },
+    {
+      "epoch": 0.09479513678242885,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.6472,
+      "step": 835
+    },
+    {
+      "epoch": 0.09490866389234792,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.002,
+      "loss": 5.6651,
+      "step": 836
+    },
+    {
+      "epoch": 0.095022191002267,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.002,
+      "loss": 5.6386,
+      "step": 837
+    },
+    {
+      "epoch": 0.09513571811218607,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.6418,
+      "step": 838
+    },
+    {
+      "epoch": 0.09524924522210515,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.6447,
+      "step": 839
+    },
+    {
+      "epoch": 0.09536277233202423,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.6292,
+      "step": 840
+    },
+    {
+      "epoch": 0.0954762994419433,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.6334,
+      "step": 841
+    },
+    {
+      "epoch": 0.09558982655186238,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.002,
+      "loss": 5.6577,
+      "step": 842
+    },
+    {
+      "epoch": 0.09570335366178145,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.002,
+      "loss": 5.654,
+      "step": 843
+    },
+    {
+      "epoch": 0.09581688077170053,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.002,
+      "loss": 5.624,
+      "step": 844
+    },
+    {
+      "epoch": 0.0959304078816196,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 5.6521,
+      "step": 845
+    },
+    {
+      "epoch": 0.09604393499153868,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.6229,
+      "step": 846
+    },
+    {
+      "epoch": 0.09615746210145776,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.6424,
+      "step": 847
+    },
+    {
+      "epoch": 0.09627098921137683,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.6545,
+      "step": 848
+    },
+    {
+      "epoch": 0.09638451632129591,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.6503,
+      "step": 849
+    },
+    {
+      "epoch": 0.09649804343121499,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.629,
+      "step": 850
+    },
+    {
+      "epoch": 0.09661157054113406,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.6165,
+      "step": 851
+    },
+    {
+      "epoch": 0.09672509765105314,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.6342,
+      "step": 852
+    },
+    {
+      "epoch": 0.09683862476097221,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.6343,
+      "step": 853
+    },
+    {
+      "epoch": 0.09695215187089129,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.6143,
+      "step": 854
+    },
+    {
+      "epoch": 0.09706567898081037,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.6072,
+      "step": 855
+    },
+    {
+      "epoch": 0.09717920609072944,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.6483,
+      "step": 856
+    },
+    {
+      "epoch": 0.09729273320064852,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.6203,
+      "step": 857
+    },
+    {
+      "epoch": 0.0974062603105676,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.6298,
+      "step": 858
+    },
+    {
+      "epoch": 0.09751978742048667,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.6195,
+      "step": 859
+    },
+    {
+      "epoch": 0.09763331453040576,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 5.6411,
+      "step": 860
+    },
+    {
+      "epoch": 0.09774684164032484,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 5.6273,
+      "step": 861
+    },
+    {
+      "epoch": 0.09786036875024391,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.002,
+      "loss": 5.6302,
+      "step": 862
+    },
+    {
+      "epoch": 0.09797389586016299,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.002,
+      "loss": 5.6143,
+      "step": 863
+    },
+    {
+      "epoch": 0.09808742297008206,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.002,
+      "loss": 5.6288,
+      "step": 864
+    },
+    {
+      "epoch": 0.09820095008000114,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.002,
+      "loss": 5.6417,
+      "step": 865
+    },
+    {
+      "epoch": 0.09831447718992022,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.6172,
+      "step": 866
+    },
+    {
+      "epoch": 0.09842800429983929,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.6289,
+      "step": 867
+    },
+    {
+      "epoch": 0.09854153140975837,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.6493,
+      "step": 868
+    },
+    {
+      "epoch": 0.09865505851967744,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.6282,
+      "step": 869
+    },
+    {
+      "epoch": 0.09876858562959652,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.6209,
+      "step": 870
+    },
+    {
+      "epoch": 0.0988821127395156,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.6253,
+      "step": 871
+    },
+    {
+      "epoch": 0.09899563984943467,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.6341,
+      "step": 872
+    },
+    {
+      "epoch": 0.09910916695935375,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.6358,
+      "step": 873
+    },
+    {
+      "epoch": 0.09922269406927282,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.6132,
+      "step": 874
+    },
+    {
+      "epoch": 0.0993362211791919,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.6295,
+      "step": 875
+    },
+    {
+      "epoch": 0.09944974828911098,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.002,
+      "loss": 5.6159,
+      "step": 876
+    },
+    {
+      "epoch": 0.09956327539903005,
+      "grad_norm": 0.625,
+      "learning_rate": 0.002,
+      "loss": 5.6369,
+      "step": 877
+    },
+    {
+      "epoch": 0.09967680250894913,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.002,
+      "loss": 5.6308,
+      "step": 878
+    },
+    {
+      "epoch": 0.0997903296188682,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.002,
+      "loss": 5.6488,
+      "step": 879
+    },
+    {
+      "epoch": 0.09990385672878728,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.002,
+      "loss": 5.6272,
+      "step": 880
+    },
+    {
+      "epoch": 0.10001738383870636,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.6217,
+      "step": 881
+    },
+    {
+      "epoch": 0.10013091094862543,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.6349,
+      "step": 882
+    },
+    {
+      "epoch": 0.10024443805854451,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.6253,
+      "step": 883
+    },
+    {
+      "epoch": 0.10035796516846358,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.6104,
+      "step": 884
+    },
+    {
+      "epoch": 0.10047149227838266,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.6363,
+      "step": 885
+    },
+    {
+      "epoch": 0.10058501938830174,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.6375,
+      "step": 886
+    },
+    {
+      "epoch": 0.10069854649822081,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.6283,
+      "step": 887
+    },
+    {
+      "epoch": 0.10081207360813989,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.6314,
+      "step": 888
+    },
+    {
+      "epoch": 0.10092560071805896,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.6346,
+      "step": 889
+    },
+    {
+      "epoch": 0.10103912782797805,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 5.6193,
+      "step": 890
+    },
+    {
+      "epoch": 0.10115265493789713,
+      "grad_norm": 0.625,
+      "learning_rate": 0.002,
+      "loss": 5.6268,
+      "step": 891
+    },
+    {
+      "epoch": 0.1012661820478162,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.002,
+      "loss": 5.6322,
+      "step": 892
+    },
+    {
+      "epoch": 0.10137970915773528,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.6032,
+      "step": 893
+    },
+    {
+      "epoch": 0.10149323626765436,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.6263,
+      "step": 894
+    },
+    {
+      "epoch": 0.10160676337757343,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.6228,
+      "step": 895
+    },
+    {
+      "epoch": 0.10172029048749251,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.6125,
+      "step": 896
+    },
+    {
+      "epoch": 0.10183381759741159,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.6185,
+      "step": 897
+    },
+    {
+      "epoch": 0.10194734470733066,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.6267,
+      "step": 898
+    },
+    {
+      "epoch": 0.10206087181724974,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.621,
+      "step": 899
+    },
+    {
+      "epoch": 0.10217439892716881,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.614,
+      "step": 900
+    },
+    {
+      "epoch": 0.10228792603708789,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.002,
+      "loss": 5.6343,
+      "step": 901
+    },
+    {
+      "epoch": 0.10240145314700697,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.002,
+      "loss": 5.6302,
+      "step": 902
+    },
+    {
+      "epoch": 0.10251498025692604,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.002,
+      "loss": 5.6186,
+      "step": 903
+    },
+    {
+      "epoch": 0.10262850736684512,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.002,
+      "loss": 5.6047,
+      "step": 904
+    },
+    {
+      "epoch": 0.1027420344767642,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.6217,
+      "step": 905
+    },
+    {
+      "epoch": 0.10285556158668327,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.6354,
+      "step": 906
+    },
+    {
+      "epoch": 0.10296908869660235,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.6317,
+      "step": 907
+    },
+    {
+      "epoch": 0.10308261580652142,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.5977,
+      "step": 908
+    },
+    {
+      "epoch": 0.1031961429164405,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.6094,
+      "step": 909
+    },
+    {
+      "epoch": 0.10330967002635957,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.604,
+      "step": 910
+    },
+    {
+      "epoch": 0.10342319713627865,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.6258,
+      "step": 911
+    },
+    {
+      "epoch": 0.10353672424619773,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.6154,
+      "step": 912
+    },
+    {
+      "epoch": 0.1036502513561168,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.6283,
+      "step": 913
+    },
+    {
+      "epoch": 0.10376377846603588,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.632,
+      "step": 914
+    },
+    {
+      "epoch": 0.10387730557595495,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.6283,
+      "step": 915
+    },
+    {
+      "epoch": 0.10399083268587403,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.6206,
+      "step": 916
+    },
+    {
+      "epoch": 0.1041043597957931,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.6153,
+      "step": 917
+    },
+    {
+      "epoch": 0.10421788690571218,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.5826,
+      "step": 918
+    },
+    {
+      "epoch": 0.10433141401563126,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.6171,
+      "step": 919
+    },
+    {
+      "epoch": 0.10444494112555033,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.6129,
+      "step": 920
+    },
+    {
+      "epoch": 0.10455846823546942,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.6148,
+      "step": 921
+    },
+    {
+      "epoch": 0.1046719953453885,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.6074,
+      "step": 922
+    },
+    {
+      "epoch": 0.10478552245530758,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.5983,
+      "step": 923
+    },
+    {
+      "epoch": 0.10489904956522665,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.6257,
+      "step": 924
+    },
+    {
+      "epoch": 0.10501257667514573,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.5954,
+      "step": 925
+    },
+    {
+      "epoch": 0.1051261037850648,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.6087,
+      "step": 926
+    },
+    {
+      "epoch": 0.10523963089498388,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.618,
+      "step": 927
+    },
+    {
+      "epoch": 0.10535315800490296,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.6049,
+      "step": 928
+    },
+    {
+      "epoch": 0.10546668511482203,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.5961,
+      "step": 929
+    },
+    {
+      "epoch": 0.10558021222474111,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.5927,
+      "step": 930
+    },
+    {
+      "epoch": 0.10569373933466018,
+      "grad_norm": 0.5,
+      "learning_rate": 0.002,
+      "loss": 5.6119,
+      "step": 931
+    },
+    {
+      "epoch": 0.10580726644457926,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.002,
+      "loss": 5.5959,
+      "step": 932
+    },
+    {
+      "epoch": 0.10592079355449834,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.002,
+      "loss": 5.6095,
+      "step": 933
+    },
+    {
+      "epoch": 0.10603432066441741,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002,
+      "loss": 5.6101,
+      "step": 934
+    },
+    {
+      "epoch": 0.10614784777433649,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.5955,
+      "step": 935
+    },
+    {
+      "epoch": 0.10626137488425556,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.6151,
+      "step": 936
+    },
+    {
+      "epoch": 0.10637490199417464,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.6209,
+      "step": 937
+    },
+    {
+      "epoch": 0.10648842910409372,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.591,
+      "step": 938
+    },
+    {
+      "epoch": 0.10660195621401279,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 5.6023,
+      "step": 939
+    },
+    {
+      "epoch": 0.10671548332393187,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.002,
+      "loss": 5.6064,
+      "step": 940
+    },
+    {
+      "epoch": 0.10682901043385094,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 5.6035,
+      "step": 941
+    },
+    {
+      "epoch": 0.10694253754377002,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.002,
+      "loss": 5.5801,
+      "step": 942
+    },
+    {
+      "epoch": 0.1070560646536891,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.002,
+      "loss": 5.6222,
+      "step": 943
+    },
+    {
+      "epoch": 0.10716959176360817,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.6049,
+      "step": 944
+    },
+    {
+      "epoch": 0.10728311887352725,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.6068,
+      "step": 945
+    },
+    {
+      "epoch": 0.10739664598344632,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.6129,
+      "step": 946
+    },
+    {
+      "epoch": 0.1075101730933654,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.6057,
+      "step": 947
+    },
+    {
+      "epoch": 0.10762370020328448,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.5867,
+      "step": 948
+    },
+    {
+      "epoch": 0.10773722731320355,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.6051,
+      "step": 949
+    },
+    {
+      "epoch": 0.10785075442312263,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.5915,
+      "step": 950
+    },
+    {
+      "epoch": 0.10796428153304172,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.5859,
+      "step": 951
+    },
+    {
+      "epoch": 0.1080778086429608,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.5956,
+      "step": 952
+    },
+    {
+      "epoch": 0.10819133575287987,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.6042,
+      "step": 953
+    },
+    {
+      "epoch": 0.10830486286279895,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.6142,
+      "step": 954
+    },
+    {
+      "epoch": 0.10841838997271802,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.5827,
+      "step": 955
+    },
+    {
+      "epoch": 0.1085319170826371,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.5824,
+      "step": 956
+    },
+    {
+      "epoch": 0.10864544419255617,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.6083,
+      "step": 957
+    },
+    {
+      "epoch": 0.10875897130247525,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.5804,
+      "step": 958
+    },
+    {
+      "epoch": 0.10887249841239433,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.6095,
+      "step": 959
+    },
+    {
+      "epoch": 0.1089860255223134,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.5909,
+      "step": 960
+    },
+    {
+      "epoch": 0.10909955263223248,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.002,
+      "loss": 5.5835,
+      "step": 961
+    },
+    {
+      "epoch": 0.10921307974215155,
+      "grad_norm": 0.5,
+      "learning_rate": 0.002,
+      "loss": 5.6202,
+      "step": 962
+    },
+    {
+      "epoch": 0.10932660685207063,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.002,
+      "loss": 5.5972,
+      "step": 963
+    },
+    {
+      "epoch": 0.1094401339619897,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.002,
+      "loss": 5.5943,
+      "step": 964
+    },
+    {
+      "epoch": 0.10955366107190878,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.5897,
+      "step": 965
+    },
+    {
+      "epoch": 0.10966718818182786,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.5808,
+      "step": 966
+    },
+    {
+      "epoch": 0.10978071529174693,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.6067,
+      "step": 967
+    },
+    {
+      "epoch": 0.10989424240166601,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.6222,
+      "step": 968
+    },
+    {
+      "epoch": 0.11000776951158509,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.5942,
+      "step": 969
+    },
+    {
+      "epoch": 0.11012129662150416,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.6148,
+      "step": 970
+    },
+    {
+      "epoch": 0.11023482373142324,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.5933,
+      "step": 971
+    },
+    {
+      "epoch": 0.11034835084134231,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.6097,
+      "step": 972
+    },
+    {
+      "epoch": 0.11046187795126139,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.6176,
+      "step": 973
+    },
+    {
+      "epoch": 0.11057540506118047,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.575,
+      "step": 974
+    },
+    {
+      "epoch": 0.11068893217109954,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.002,
+      "loss": 5.6118,
+      "step": 975
+    },
+    {
+      "epoch": 0.11080245928101862,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.002,
+      "loss": 5.5853,
+      "step": 976
+    },
+    {
+      "epoch": 0.1109159863909377,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.002,
+      "loss": 5.5862,
+      "step": 977
+    },
+    {
+      "epoch": 0.11102951350085677,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.002,
+      "loss": 5.5831,
+      "step": 978
+    },
+    {
+      "epoch": 0.11114304061077585,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.002,
+      "loss": 5.5997,
+      "step": 979
+    },
+    {
+      "epoch": 0.11125656772069492,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.002,
+      "loss": 5.5824,
+      "step": 980
+    },
+    {
+      "epoch": 0.11137009483061401,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.002,
+      "loss": 5.6081,
+      "step": 981
+    },
+    {
+      "epoch": 0.11148362194053309,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.002,
+      "loss": 5.5927,
+      "step": 982
+    },
+    {
+      "epoch": 0.11159714905045216,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.5977,
+      "step": 983
+    },
+    {
+      "epoch": 0.11171067616037124,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.5878,
+      "step": 984
+    },
+    {
+      "epoch": 0.11182420327029031,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.5938,
+      "step": 985
+    },
+    {
+      "epoch": 0.11193773038020939,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.5986,
+      "step": 986
+    },
+    {
+      "epoch": 0.11205125749012847,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.5747,
+      "step": 987
+    },
+    {
+      "epoch": 0.11216478460004754,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.5795,
+      "step": 988
+    },
+    {
+      "epoch": 0.11227831170996662,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.5969,
+      "step": 989
+    },
+    {
+      "epoch": 0.1123918388198857,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.602,
+      "step": 990
+    },
+    {
+      "epoch": 0.11250536592980477,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.5899,
+      "step": 991
+    },
+    {
+      "epoch": 0.11261889303972385,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.5868,
+      "step": 992
+    },
+    {
+      "epoch": 0.11273242014964292,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.5909,
+      "step": 993
+    },
+    {
+      "epoch": 0.112845947259562,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 5.6037,
+      "step": 994
+    },
+    {
+      "epoch": 0.11295947436948107,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.002,
+      "loss": 5.5669,
+      "step": 995
+    },
+    {
+      "epoch": 0.11307300147940015,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.5906,
+      "step": 996
+    },
+    {
+      "epoch": 0.11318652858931923,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.5805,
+      "step": 997
+    },
+    {
+      "epoch": 0.1133000556992383,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.5668,
+      "step": 998
+    },
+    {
+      "epoch": 0.11341358280915738,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.5836,
+      "step": 999
+    },
+    {
+      "epoch": 0.11352710991907645,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.5779,
+      "step": 1000
+    },
+    {
+      "epoch": 0.11364063702899553,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.5953,
+      "step": 1001
+    },
+    {
+      "epoch": 0.1137541641389146,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.5934,
+      "step": 1002
+    },
+    {
+      "epoch": 0.11386769124883368,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.5934,
+      "step": 1003
+    },
+    {
+      "epoch": 0.11398121835875276,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.5778,
+      "step": 1004
+    },
+    {
+      "epoch": 0.11409474546867183,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.5706,
+      "step": 1005
+    },
+    {
+      "epoch": 0.11420827257859091,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.5717,
+      "step": 1006
+    },
+    {
+      "epoch": 0.11432179968850999,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.5936,
+      "step": 1007
+    },
+    {
+      "epoch": 0.11443532679842906,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.002,
+      "loss": 5.578,
+      "step": 1008
+    },
+    {
+      "epoch": 0.11454885390834814,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.002,
+      "loss": 5.5532,
+      "step": 1009
+    },
+    {
+      "epoch": 0.11466238101826721,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.5636,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1147759081281863,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.5852,
+      "step": 1011
+    },
+    {
+      "epoch": 0.11488943523810538,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.5727,
+      "step": 1012
+    },
+    {
+      "epoch": 0.11500296234802446,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.578,
+      "step": 1013
+    },
+    {
+      "epoch": 0.11511648945794353,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.5804,
+      "step": 1014
+    },
+    {
+      "epoch": 0.11523001656786261,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.5649,
+      "step": 1015
+    },
+    {
+      "epoch": 0.11534354367778168,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.5849,
+      "step": 1016
+    },
+    {
+      "epoch": 0.11545707078770076,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.5747,
+      "step": 1017
+    },
+    {
+      "epoch": 0.11557059789761984,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.5729,
+      "step": 1018
+    },
+    {
+      "epoch": 0.11568412500753891,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.5544,
+      "step": 1019
+    },
+    {
+      "epoch": 0.11579765211745799,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.5804,
+      "step": 1020
+    },
+    {
+      "epoch": 0.11591117922737706,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.002,
+      "loss": 5.5695,
+      "step": 1021
+    },
+    {
+      "epoch": 0.11602470633729614,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.5706,
+      "step": 1022
+    },
+    {
+      "epoch": 0.11613823344721522,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.5751,
+      "step": 1023
+    },
+    {
+      "epoch": 0.11625176055713429,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.5699,
+      "step": 1024
+    },
+    {
+      "epoch": 0.11636528766705337,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.5722,
+      "step": 1025
+    },
+    {
+      "epoch": 0.11647881477697244,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.5785,
+      "step": 1026
+    },
+    {
+      "epoch": 0.11659234188689152,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.5789,
+      "step": 1027
+    },
+    {
+      "epoch": 0.1167058689968106,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.5705,
+      "step": 1028
+    },
+    {
+      "epoch": 0.11681939610672967,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.586,
+      "step": 1029
+    },
+    {
+      "epoch": 0.11693292321664875,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.5683,
+      "step": 1030
+    },
+    {
+      "epoch": 0.11704645032656782,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.5833,
+      "step": 1031
+    },
+    {
+      "epoch": 0.1171599774364869,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.5729,
+      "step": 1032
+    },
+    {
+      "epoch": 0.11727350454640598,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.5695,
+      "step": 1033
+    },
+    {
+      "epoch": 0.11738703165632505,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.5532,
+      "step": 1034
+    },
+    {
+      "epoch": 0.11750055876624413,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.5794,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1176140858761632,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.5763,
+      "step": 1036
+    },
+    {
+      "epoch": 0.11772761298608228,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.5686,
+      "step": 1037
+    },
+    {
+      "epoch": 0.11784114009600136,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.5819,
+      "step": 1038
+    },
+    {
+      "epoch": 0.11795466720592043,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 5.5806,
+      "step": 1039
+    },
+    {
+      "epoch": 0.11806819431583951,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.5665,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1181817214257586,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.5706,
+      "step": 1041
+    },
+    {
+      "epoch": 0.11829524853567767,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.562,
+      "step": 1042
+    },
+    {
+      "epoch": 0.11840877564559675,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.5604,
+      "step": 1043
+    },
+    {
+      "epoch": 0.11852230275551583,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.5737,
+      "step": 1044
+    },
+    {
+      "epoch": 0.1186358298654349,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.5703,
+      "step": 1045
+    },
+    {
+      "epoch": 0.11874935697535398,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.5501,
+      "step": 1046
+    },
+    {
+      "epoch": 0.11886288408527305,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.5648,
+      "step": 1047
+    },
+    {
+      "epoch": 0.11897641119519213,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.5639,
+      "step": 1048
+    },
+    {
+      "epoch": 0.1190899383051112,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.002,
+      "loss": 5.5697,
+      "step": 1049
+    },
+    {
+      "epoch": 0.11920346541503028,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.002,
+      "loss": 5.5672,
+      "step": 1050
+    },
+    {
+      "epoch": 0.11931699252494936,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.002,
+      "loss": 5.5607,
+      "step": 1051
+    },
+    {
+      "epoch": 0.11943051963486843,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.002,
+      "loss": 5.573,
+      "step": 1052
+    },
+    {
+      "epoch": 0.11954404674478751,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.5559,
+      "step": 1053
+    },
+    {
+      "epoch": 0.11965757385470659,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.5561,
+      "step": 1054
+    },
+    {
+      "epoch": 0.11977110096462566,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.5514,
+      "step": 1055
+    },
+    {
+      "epoch": 0.11988462807454474,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.5471,
+      "step": 1056
+    },
+    {
+      "epoch": 0.11999815518446381,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.5953,
+      "step": 1057
+    },
+    {
+      "epoch": 0.12011168229438289,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.5661,
+      "step": 1058
+    },
+    {
+      "epoch": 0.12022520940430197,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.5598,
+      "step": 1059
+    },
+    {
+      "epoch": 0.12033873651422104,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.5619,
+      "step": 1060
+    },
+    {
+      "epoch": 0.12045226362414012,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.5565,
+      "step": 1061
+    },
+    {
+      "epoch": 0.1205657907340592,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.5556,
+      "step": 1062
+    },
+    {
+      "epoch": 0.12067931784397827,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.5713,
+      "step": 1063
+    },
+    {
+      "epoch": 0.12079284495389735,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.5582,
+      "step": 1064
+    },
+    {
+      "epoch": 0.12090637206381642,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.5706,
+      "step": 1065
+    },
+    {
+      "epoch": 0.1210198991737355,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.5816,
+      "step": 1066
+    },
+    {
+      "epoch": 0.12113342628365457,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.5739,
+      "step": 1067
+    },
+    {
+      "epoch": 0.12124695339357365,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.574,
+      "step": 1068
+    },
+    {
+      "epoch": 0.12136048050349273,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.002,
+      "loss": 5.5521,
+      "step": 1069
+    },
+    {
+      "epoch": 0.1214740076134118,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.002,
+      "loss": 5.5583,
+      "step": 1070
+    },
+    {
+      "epoch": 0.12158753472333089,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.002,
+      "loss": 5.5464,
+      "step": 1071
+    },
+    {
+      "epoch": 0.12170106183324997,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 5.5743,
+      "step": 1072
+    },
+    {
+      "epoch": 0.12181458894316904,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.5292,
+      "step": 1073
+    },
+    {
+      "epoch": 0.12192811605308812,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.57,
+      "step": 1074
+    },
+    {
+      "epoch": 0.1220416431630072,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.5705,
+      "step": 1075
+    },
+    {
+      "epoch": 0.12215517027292627,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.5686,
+      "step": 1076
+    },
+    {
+      "epoch": 0.12226869738284535,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.5457,
+      "step": 1077
+    },
+    {
+      "epoch": 0.12238222449276442,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.5514,
+      "step": 1078
+    },
+    {
+      "epoch": 0.1224957516026835,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.5715,
+      "step": 1079
+    },
+    {
+      "epoch": 0.12260927871260258,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.5441,
+      "step": 1080
+    },
+    {
+      "epoch": 0.12272280582252165,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.002,
+      "loss": 5.5699,
+      "step": 1081
+    },
+    {
+      "epoch": 0.12283633293244073,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.5504,
+      "step": 1082
+    },
+    {
+      "epoch": 0.1229498600423598,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.5494,
+      "step": 1083
+    },
+    {
+      "epoch": 0.12306338715227888,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.5575,
+      "step": 1084
+    },
+    {
+      "epoch": 0.12317691426219796,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.563,
+      "step": 1085
+    },
+    {
+      "epoch": 0.12329044137211703,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.5622,
+      "step": 1086
+    },
+    {
+      "epoch": 0.12340396848203611,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.5618,
+      "step": 1087
+    },
+    {
+      "epoch": 0.12351749559195518,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.5402,
+      "step": 1088
+    },
+    {
+      "epoch": 0.12363102270187426,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.5557,
+      "step": 1089
+    },
+    {
+      "epoch": 0.12374454981179334,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.563,
+      "step": 1090
+    },
+    {
+      "epoch": 0.12385807692171241,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.5684,
+      "step": 1091
+    },
+    {
+      "epoch": 0.12397160403163149,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.5484,
+      "step": 1092
+    },
+    {
+      "epoch": 0.12408513114155056,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.5572,
+      "step": 1093
+    },
+    {
+      "epoch": 0.12419865825146964,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.5657,
+      "step": 1094
+    },
+    {
+      "epoch": 0.12431218536138872,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.002,
+      "loss": 5.5525,
+      "step": 1095
+    },
+    {
+      "epoch": 0.12442571247130779,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.002,
+      "loss": 5.5524,
+      "step": 1096
+    },
+    {
+      "epoch": 0.12453923958122687,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.002,
+      "loss": 5.5588,
+      "step": 1097
+    },
+    {
+      "epoch": 0.12465276669114594,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.002,
+      "loss": 5.5512,
+      "step": 1098
+    },
+    {
+      "epoch": 0.12476629380106502,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.5488,
+      "step": 1099
+    },
+    {
+      "epoch": 0.1248798209109841,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 5.5724,
+      "step": 1100
+    },
+    {
+      "epoch": 0.12499334802090317,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.5408,
+      "step": 1101
+    },
+    {
+      "epoch": 0.12510687513082225,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.5319,
+      "step": 1102
+    },
+    {
+      "epoch": 0.12522040224074132,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.5478,
+      "step": 1103
+    },
+    {
+      "epoch": 0.1253339293506604,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.002,
+      "loss": 5.5648,
+      "step": 1104
+    },
+    {
+      "epoch": 0.12544745646057948,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.5579,
+      "step": 1105
+    },
+    {
+      "epoch": 0.12556098357049855,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.5536,
+      "step": 1106
+    },
+    {
+      "epoch": 0.12567451068041763,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.5541,
+      "step": 1107
+    },
+    {
+      "epoch": 0.1257880377903367,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.5458,
+      "step": 1108
+    },
+    {
+      "epoch": 0.12590156490025578,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.5617,
+      "step": 1109
+    },
+    {
+      "epoch": 0.12601509201017486,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.5634,
+      "step": 1110
+    },
+    {
+      "epoch": 0.12612861912009393,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.5347,
+      "step": 1111
+    },
+    {
+      "epoch": 0.126242146230013,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.5753,
+      "step": 1112
+    },
+    {
+      "epoch": 0.12635567333993208,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.5607,
+      "step": 1113
+    },
+    {
+      "epoch": 0.12646920044985116,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.5263,
+      "step": 1114
+    },
+    {
+      "epoch": 0.12658272755977024,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.5605,
+      "step": 1115
+    },
+    {
+      "epoch": 0.1266962546696893,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.002,
+      "loss": 5.5603,
+      "step": 1116
+    },
+    {
+      "epoch": 0.12680978177960842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.54,
+      "step": 1117
+    },
+    {
+      "epoch": 0.1269233088895275,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.5457,
+      "step": 1118
+    },
+    {
+      "epoch": 0.12703683599944657,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.5271,
+      "step": 1119
+    },
+    {
+      "epoch": 0.12715036310936564,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.5395,
+      "step": 1120
+    },
+    {
+      "epoch": 0.12726389021928472,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.5356,
+      "step": 1121
+    },
+    {
+      "epoch": 0.1273774173292038,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.5575,
+      "step": 1122
+    },
+    {
+      "epoch": 0.12749094443912287,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.5514,
+      "step": 1123
+    },
+    {
+      "epoch": 0.12760447154904195,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.5386,
+      "step": 1124
+    },
+    {
+      "epoch": 0.12771799865896102,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.5548,
+      "step": 1125
+    },
+    {
+      "epoch": 0.1278315257688801,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.5535,
+      "step": 1126
+    },
+    {
+      "epoch": 0.12794505287879918,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.5349,
+      "step": 1127
+    },
+    {
+      "epoch": 0.12805857998871825,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.5458,
+      "step": 1128
+    },
+    {
+      "epoch": 0.12817210709863733,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.5261,
+      "step": 1129
+    },
+    {
+      "epoch": 0.1282856342085564,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.5365,
+      "step": 1130
+    },
+    {
+      "epoch": 0.12839916131847548,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.002,
+      "loss": 5.5481,
+      "step": 1131
+    },
+    {
+      "epoch": 0.12851268842839456,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.002,
+      "loss": 5.5552,
+      "step": 1132
+    },
+    {
+      "epoch": 0.12862621553831363,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002,
+      "loss": 5.5549,
+      "step": 1133
+    },
+    {
+      "epoch": 0.1287397426482327,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.002,
+      "loss": 5.5279,
+      "step": 1134
+    },
+    {
+      "epoch": 0.12885326975815178,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.002,
+      "loss": 5.5441,
+      "step": 1135
+    },
+    {
+      "epoch": 0.12896679686807086,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.5371,
+      "step": 1136
+    },
+    {
+      "epoch": 0.12908032397798994,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.002,
+      "loss": 5.5476,
+      "step": 1137
+    },
+    {
+      "epoch": 0.129193851087909,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.5427,
+      "step": 1138
+    },
+    {
+      "epoch": 0.1293073781978281,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.5379,
+      "step": 1139
+    },
+    {
+      "epoch": 0.12942090530774716,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.558,
+      "step": 1140
+    },
+    {
+      "epoch": 0.12953443241766624,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.5446,
+      "step": 1141
+    },
+    {
+      "epoch": 0.12964795952758532,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.5415,
+      "step": 1142
+    },
+    {
+      "epoch": 0.1297614866375044,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.5544,
+      "step": 1143
+    },
+    {
+      "epoch": 0.12987501374742347,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.5142,
+      "step": 1144
+    },
+    {
+      "epoch": 0.12998854085734254,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.5177,
+      "step": 1145
+    },
+    {
+      "epoch": 0.13010206796726162,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.5389,
+      "step": 1146
+    },
+    {
+      "epoch": 0.1302155950771807,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.5318,
+      "step": 1147
+    },
+    {
+      "epoch": 0.13032912218709977,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.5477,
+      "step": 1148
+    },
+    {
+      "epoch": 0.13044264929701885,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.5443,
+      "step": 1149
+    },
+    {
+      "epoch": 0.13055617640693792,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.5231,
+      "step": 1150
+    },
+    {
+      "epoch": 0.130669703516857,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.5401,
+      "step": 1151
+    },
+    {
+      "epoch": 0.13078323062677608,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.5374,
+      "step": 1152
+    },
+    {
+      "epoch": 0.13089675773669515,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.5306,
+      "step": 1153
+    },
+    {
+      "epoch": 0.13101028484661423,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.002,
+      "loss": 5.5381,
+      "step": 1154
+    },
+    {
+      "epoch": 0.1311238119565333,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 5.5337,
+      "step": 1155
+    },
+    {
+      "epoch": 0.13123733906645238,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.5434,
+      "step": 1156
+    },
+    {
+      "epoch": 0.13135086617637146,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.002,
+      "loss": 5.5441,
+      "step": 1157
+    },
+    {
+      "epoch": 0.13146439328629053,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.5277,
+      "step": 1158
+    },
+    {
+      "epoch": 0.1315779203962096,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.539,
+      "step": 1159
+    },
+    {
+      "epoch": 0.13169144750612868,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.5387,
+      "step": 1160
+    },
+    {
+      "epoch": 0.13180497461604776,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.5396,
+      "step": 1161
+    },
+    {
+      "epoch": 0.13191850172596684,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.5249,
+      "step": 1162
+    },
+    {
+      "epoch": 0.1320320288358859,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.5466,
+      "step": 1163
+    },
+    {
+      "epoch": 0.132145555945805,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.526,
+      "step": 1164
+    },
+    {
+      "epoch": 0.13225908305572406,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.5321,
+      "step": 1165
+    },
+    {
+      "epoch": 0.13237261016564314,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.5269,
+      "step": 1166
+    },
+    {
+      "epoch": 0.13248613727556222,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.002,
+      "loss": 5.5436,
+      "step": 1167
+    },
+    {
+      "epoch": 0.1325996643854813,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.002,
+      "loss": 5.5382,
+      "step": 1168
+    },
+    {
+      "epoch": 0.13271319149540037,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.5295,
+      "step": 1169
+    },
+    {
+      "epoch": 0.13282671860531944,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.5262,
+      "step": 1170
+    },
+    {
+      "epoch": 0.13294024571523852,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.5363,
+      "step": 1171
+    },
+    {
+      "epoch": 0.1330537728251576,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.5427,
+      "step": 1172
+    },
+    {
+      "epoch": 0.13316729993507667,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.5425,
+      "step": 1173
+    },
+    {
+      "epoch": 0.13328082704499575,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.5189,
+      "step": 1174
+    },
+    {
+      "epoch": 0.13339435415491482,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.554,
+      "step": 1175
+    },
+    {
+      "epoch": 0.1335078812648339,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.4912,
+      "step": 1176
+    },
+    {
+      "epoch": 0.133621408374753,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.5278,
+      "step": 1177
+    },
+    {
+      "epoch": 0.13373493548467208,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.5212,
+      "step": 1178
+    },
+    {
+      "epoch": 0.13384846259459116,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 5.5424,
+      "step": 1179
+    },
+    {
+      "epoch": 0.13396198970451023,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.5192,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1340755168144293,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.002,
+      "loss": 5.5283,
+      "step": 1181
+    },
+    {
+      "epoch": 0.13418904392434838,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.537,
+      "step": 1182
+    },
+    {
+      "epoch": 0.13430257103426746,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.5259,
+      "step": 1183
+    },
+    {
+      "epoch": 0.13441609814418654,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.5359,
+      "step": 1184
+    },
+    {
+      "epoch": 0.1345296252541056,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.5311,
+      "step": 1185
+    },
+    {
+      "epoch": 0.1346431523640247,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.5232,
+      "step": 1186
+    },
+    {
+      "epoch": 0.13475667947394376,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.5296,
+      "step": 1187
+    },
+    {
+      "epoch": 0.13487020658386284,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.5201,
+      "step": 1188
+    },
+    {
+      "epoch": 0.13498373369378192,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.5279,
+      "step": 1189
+    },
+    {
+      "epoch": 0.135097260803701,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.5238,
+      "step": 1190
+    },
+    {
+      "epoch": 0.13521078791362007,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.5171,
+      "step": 1191
+    },
+    {
+      "epoch": 0.13532431502353914,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.5474,
+      "step": 1192
+    },
+    {
+      "epoch": 0.13543784213345822,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.5475,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1355513692433773,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.5243,
+      "step": 1194
+    },
+    {
+      "epoch": 0.13566489635329637,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.5344,
+      "step": 1195
+    },
+    {
+      "epoch": 0.13577842346321545,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 5.53,
+      "step": 1196
+    },
+    {
+      "epoch": 0.13589195057313452,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.002,
+      "loss": 5.5222,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1360054776830536,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.002,
+      "loss": 5.5147,
+      "step": 1198
+    },
+    {
+      "epoch": 0.13611900479297268,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.002,
+      "loss": 5.5287,
+      "step": 1199
+    },
+    {
+      "epoch": 0.13623253190289175,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.5145,
+      "step": 1200
+    },
+    {
+      "epoch": 0.13634605901281083,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.529,
+      "step": 1201
+    },
+    {
+      "epoch": 0.1364595861227299,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.5366,
+      "step": 1202
+    },
+    {
+      "epoch": 0.13657311323264898,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.5282,
+      "step": 1203
+    },
+    {
+      "epoch": 0.13668664034256806,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.5166,
+      "step": 1204
+    },
+    {
+      "epoch": 0.13680016745248713,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.5217,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1369136945624062,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.002,
+      "loss": 5.5257,
+      "step": 1206
+    },
+    {
+      "epoch": 0.13702722167232528,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.5025,
+      "step": 1207
+    },
+    {
+      "epoch": 0.13714074878224436,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.5269,
+      "step": 1208
+    },
+    {
+      "epoch": 0.13725427589216344,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.5391,
+      "step": 1209
+    },
+    {
+      "epoch": 0.1373678030020825,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.5028,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1374813301120016,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.527,
+      "step": 1211
+    },
+    {
+      "epoch": 0.13759485722192066,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.5209,
+      "step": 1212
+    },
+    {
+      "epoch": 0.13770838433183974,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 5.5134,
+      "step": 1213
+    },
+    {
+      "epoch": 0.13782191144175882,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 5.5211,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1379354385516779,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.002,
+      "loss": 5.5359,
+      "step": 1215
+    },
+    {
+      "epoch": 0.13804896566159697,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.002,
+      "loss": 5.5179,
+      "step": 1216
+    },
+    {
+      "epoch": 0.13816249277151604,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.5178,
+      "step": 1217
+    },
+    {
+      "epoch": 0.13827601988143512,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.4937,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1383895469913542,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.5265,
+      "step": 1219
+    },
+    {
+      "epoch": 0.13850307410127327,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.5118,
+      "step": 1220
+    },
+    {
+      "epoch": 0.13861660121119235,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.5181,
+      "step": 1221
+    },
+    {
+      "epoch": 0.13873012832111142,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.5385,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1388436554310305,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.539,
+      "step": 1223
+    },
+    {
+      "epoch": 0.13895718254094958,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.5237,
+      "step": 1224
+    },
+    {
+      "epoch": 0.13907070965086865,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.5251,
+      "step": 1225
+    },
+    {
+      "epoch": 0.13918423676078773,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.5393,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1392977638707068,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.4971,
+      "step": 1227
+    },
+    {
+      "epoch": 0.13941129098062588,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.5192,
+      "step": 1228
+    },
+    {
+      "epoch": 0.13952481809054496,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.4989,
+      "step": 1229
+    },
+    {
+      "epoch": 0.13963834520046403,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.5368,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1397518723103831,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.5391,
+      "step": 1231
+    },
+    {
+      "epoch": 0.13986539942030218,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.519,
+      "step": 1232
+    },
+    {
+      "epoch": 0.13997892653022126,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.5221,
+      "step": 1233
+    },
+    {
+      "epoch": 0.14009245364014034,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.5069,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1402059807500594,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 5.53,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1403195078599785,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.5109,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1404330349698976,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.5188,
+      "step": 1237
+    },
+    {
+      "epoch": 0.14054656207981667,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.5151,
+      "step": 1238
+    },
+    {
+      "epoch": 0.14066008918973574,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.5115,
+      "step": 1239
+    },
+    {
+      "epoch": 0.14077361629965482,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.5396,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1408871434095739,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.5101,
+      "step": 1241
+    },
+    {
+      "epoch": 0.14100067051949297,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.5293,
+      "step": 1242
+    },
+    {
+      "epoch": 0.14111419762941205,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.507,
+      "step": 1243
+    },
+    {
+      "epoch": 0.14122772473933112,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.5201,
+      "step": 1244
+    },
+    {
+      "epoch": 0.1413412518492502,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.507,
+      "step": 1245
+    },
+    {
+      "epoch": 0.14145477895916928,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.002,
+      "loss": 5.5111,
+      "step": 1246
+    },
+    {
+      "epoch": 0.14156830606908835,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.002,
+      "loss": 5.5155,
+      "step": 1247
+    },
+    {
+      "epoch": 0.14168183317900743,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.002,
+      "loss": 5.5221,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1417953602889265,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 5.5118,
+      "step": 1249
+    },
+    {
+      "epoch": 0.14190888739884558,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 5.5088,
+      "step": 1250
+    },
+    {
+      "epoch": 0.14202241450876466,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.5193,
+      "step": 1251
+    },
+    {
+      "epoch": 0.14213594161868373,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.4872,
+      "step": 1252
+    },
+    {
+      "epoch": 0.1422494687286028,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.5207,
+      "step": 1253
+    },
+    {
+      "epoch": 0.14236299583852188,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.5089,
+      "step": 1254
+    },
+    {
+      "epoch": 0.14247652294844096,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.4982,
+      "step": 1255
+    },
+    {
+      "epoch": 0.14259005005836003,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.5159,
+      "step": 1256
+    },
+    {
+      "epoch": 0.1427035771682791,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 5.507,
+      "step": 1257
+    },
+    {
+      "epoch": 0.1428171042781982,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.002,
+      "loss": 5.5016,
+      "step": 1258
+    },
+    {
+      "epoch": 0.14293063138811726,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.002,
+      "loss": 5.5108,
+      "step": 1259
+    },
+    {
+      "epoch": 0.14304415849803634,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.002,
+      "loss": 5.5205,
+      "step": 1260
+    },
+    {
+      "epoch": 0.14315768560795541,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.5127,
+      "step": 1261
+    },
+    {
+      "epoch": 0.1432712127178745,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.4916,
+      "step": 1262
+    },
+    {
+      "epoch": 0.14338473982779357,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.5037,
+      "step": 1263
+    },
+    {
+      "epoch": 0.14349826693771264,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.518,
+      "step": 1264
+    },
+    {
+      "epoch": 0.14361179404763172,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.4998,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1437253211575508,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.5086,
+      "step": 1266
+    },
+    {
+      "epoch": 0.14383884826746987,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.5099,
+      "step": 1267
+    },
+    {
+      "epoch": 0.14395237537738895,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.5156,
+      "step": 1268
+    },
+    {
+      "epoch": 0.14406590248730802,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.4937,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1441794295972271,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.5108,
+      "step": 1270
+    },
+    {
+      "epoch": 0.14429295670714617,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.5208,
+      "step": 1271
+    },
+    {
+      "epoch": 0.14440648381706525,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.517,
+      "step": 1272
+    },
+    {
+      "epoch": 0.14452001092698433,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.514,
+      "step": 1273
+    },
+    {
+      "epoch": 0.1446335380369034,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.5225,
+      "step": 1274
+    },
+    {
+      "epoch": 0.14474706514682248,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.5213,
+      "step": 1275
+    },
+    {
+      "epoch": 0.14486059225674155,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.5253,
+      "step": 1276
+    },
+    {
+      "epoch": 0.14497411936666063,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.4806,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1450876464765797,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.4986,
+      "step": 1278
+    },
+    {
+      "epoch": 0.14520117358649878,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.4802,
+      "step": 1279
+    },
+    {
+      "epoch": 0.14531470069641786,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.509,
+      "step": 1280
+    },
+    {
+      "epoch": 0.14542822780633693,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.5083,
+      "step": 1281
+    },
+    {
+      "epoch": 0.145541754916256,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.4918,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1456552820261751,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.5365,
+      "step": 1283
+    },
+    {
+      "epoch": 0.14576880913609416,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.5178,
+      "step": 1284
+    },
+    {
+      "epoch": 0.14588233624601324,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.5097,
+      "step": 1285
+    },
+    {
+      "epoch": 0.14599586335593231,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.5035,
+      "step": 1286
+    },
+    {
+      "epoch": 0.1461093904658514,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.513,
+      "step": 1287
+    },
+    {
+      "epoch": 0.14622291757577047,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.5109,
+      "step": 1288
+    },
+    {
+      "epoch": 0.14633644468568954,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.5049,
+      "step": 1289
+    },
+    {
+      "epoch": 0.14644997179560862,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.5105,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1465634989055277,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.002,
+      "loss": 5.5136,
+      "step": 1291
+    },
+    {
+      "epoch": 0.14667702601544677,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.002,
+      "loss": 5.5088,
+      "step": 1292
+    },
+    {
+      "epoch": 0.14679055312536585,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.002,
+      "loss": 5.512,
+      "step": 1293
+    },
+    {
+      "epoch": 0.14690408023528492,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.002,
+      "loss": 5.4919,
+      "step": 1294
+    },
+    {
+      "epoch": 0.147017607345204,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.4931,
+      "step": 1295
+    },
+    {
+      "epoch": 0.14713113445512307,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.493,
+      "step": 1296
+    },
+    {
+      "epoch": 0.14724466156504215,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.4839,
+      "step": 1297
+    },
+    {
+      "epoch": 0.14735818867496125,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.4777,
+      "step": 1298
+    },
+    {
+      "epoch": 0.14747171578488033,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.5059,
+      "step": 1299
+    },
+    {
+      "epoch": 0.1475852428947994,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.5141,
+      "step": 1300
+    },
+    {
+      "epoch": 0.14769877000471848,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.5259,
+      "step": 1301
+    },
+    {
+      "epoch": 0.14781229711463756,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.5155,
+      "step": 1302
+    },
+    {
+      "epoch": 0.14792582422455663,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.503,
+      "step": 1303
+    },
+    {
+      "epoch": 0.1480393513344757,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.5032,
+      "step": 1304
+    },
+    {
+      "epoch": 0.1481528784443948,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4906,
+      "step": 1305
+    },
+    {
+      "epoch": 0.14826640555431386,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.4984,
+      "step": 1306
+    },
+    {
+      "epoch": 0.14837993266423294,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.5264,
+      "step": 1307
+    },
+    {
+      "epoch": 0.14849345977415201,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.002,
+      "loss": 5.5082,
+      "step": 1308
+    },
+    {
+      "epoch": 0.1486069868840711,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.002,
+      "loss": 5.4823,
+      "step": 1309
+    },
+    {
+      "epoch": 0.14872051399399017,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.002,
+      "loss": 5.5023,
+      "step": 1310
+    },
+    {
+      "epoch": 0.14883404110390924,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002,
+      "loss": 5.4978,
+      "step": 1311
+    },
+    {
+      "epoch": 0.14894756821382832,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.5204,
+      "step": 1312
+    },
+    {
+      "epoch": 0.1490610953237474,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.4882,
+      "step": 1313
+    },
+    {
+      "epoch": 0.14917462243366647,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.4701,
+      "step": 1314
+    },
+    {
+      "epoch": 0.14928814954358555,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.51,
+      "step": 1315
+    },
+    {
+      "epoch": 0.14940167665350462,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.4809,
+      "step": 1316
+    },
+    {
+      "epoch": 0.1495152037634237,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.4878,
+      "step": 1317
+    },
+    {
+      "epoch": 0.14962873087334277,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.4939,
+      "step": 1318
+    },
+    {
+      "epoch": 0.14974225798326185,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.4838,
+      "step": 1319
+    },
+    {
+      "epoch": 0.14985578509318093,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.4949,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1499693122031,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.4901,
+      "step": 1321
+    },
+    {
+      "epoch": 0.15008283931301908,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.4935,
+      "step": 1322
+    },
+    {
+      "epoch": 0.15019636642293815,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.002,
+      "loss": 5.5011,
+      "step": 1323
+    },
+    {
+      "epoch": 0.15030989353285723,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.002,
+      "loss": 5.4881,
+      "step": 1324
+    },
+    {
+      "epoch": 0.1504234206427763,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.002,
+      "loss": 5.4987,
+      "step": 1325
+    },
+    {
+      "epoch": 0.15053694775269538,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.4833,
+      "step": 1326
+    },
+    {
+      "epoch": 0.15065047486261446,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.5303,
+      "step": 1327
+    },
+    {
+      "epoch": 0.15076400197253353,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.5166,
+      "step": 1328
+    },
+    {
+      "epoch": 0.1508775290824526,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.4696,
+      "step": 1329
+    },
+    {
+      "epoch": 0.1509910561923717,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.4704,
+      "step": 1330
+    },
+    {
+      "epoch": 0.15110458330229076,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.4845,
+      "step": 1331
+    },
+    {
+      "epoch": 0.15121811041220984,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.4976,
+      "step": 1332
+    },
+    {
+      "epoch": 0.15133163752212891,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.4732,
+      "step": 1333
+    },
+    {
+      "epoch": 0.151445164632048,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.4948,
+      "step": 1334
+    },
+    {
+      "epoch": 0.15155869174196707,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.4892,
+      "step": 1335
+    },
+    {
+      "epoch": 0.15167221885188614,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.4983,
+      "step": 1336
+    },
+    {
+      "epoch": 0.15178574596180522,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.4859,
+      "step": 1337
+    },
+    {
+      "epoch": 0.1518992730717243,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.4646,
+      "step": 1338
+    },
+    {
+      "epoch": 0.15201280018164337,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.5193,
+      "step": 1339
+    },
+    {
+      "epoch": 0.15212632729156245,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.5059,
+      "step": 1340
+    },
+    {
+      "epoch": 0.15223985440148152,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.5149,
+      "step": 1341
+    },
+    {
+      "epoch": 0.1523533815114006,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.4976,
+      "step": 1342
+    },
+    {
+      "epoch": 0.15246690862131967,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.491,
+      "step": 1343
+    },
+    {
+      "epoch": 0.15258043573123875,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.4841,
+      "step": 1344
+    },
+    {
+      "epoch": 0.15269396284115783,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.4861,
+      "step": 1345
+    },
+    {
+      "epoch": 0.1528074899510769,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.4922,
+      "step": 1346
+    },
+    {
+      "epoch": 0.15292101706099598,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.5032,
+      "step": 1347
+    },
+    {
+      "epoch": 0.15303454417091505,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.4852,
+      "step": 1348
+    },
+    {
+      "epoch": 0.15314807128083413,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.5097,
+      "step": 1349
+    },
+    {
+      "epoch": 0.1532615983907532,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.4892,
+      "step": 1350
+    },
+    {
+      "epoch": 0.15337512550067228,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.4972,
+      "step": 1351
+    },
+    {
+      "epoch": 0.15348865261059136,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.5134,
+      "step": 1352
+    },
+    {
+      "epoch": 0.15360217972051043,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.4984,
+      "step": 1353
+    },
+    {
+      "epoch": 0.1537157068304295,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.4886,
+      "step": 1354
+    },
+    {
+      "epoch": 0.1538292339403486,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002,
+      "loss": 5.466,
+      "step": 1355
+    },
+    {
+      "epoch": 0.15394276105026766,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.002,
+      "loss": 5.4912,
+      "step": 1356
+    },
+    {
+      "epoch": 0.15405628816018674,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.002,
+      "loss": 5.5099,
+      "step": 1357
+    },
+    {
+      "epoch": 0.15416981527010584,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.002,
+      "loss": 5.5082,
+      "step": 1358
+    },
+    {
+      "epoch": 0.15428334238002492,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.4784,
+      "step": 1359
+    },
+    {
+      "epoch": 0.154396869489944,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.5001,
+      "step": 1360
+    },
+    {
+      "epoch": 0.15451039659986307,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.4859,
+      "step": 1361
+    },
+    {
+      "epoch": 0.15462392370978215,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.4926,
+      "step": 1362
+    },
+    {
+      "epoch": 0.15473745081970122,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.4977,
+      "step": 1363
+    },
+    {
+      "epoch": 0.1548509779296203,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.4775,
+      "step": 1364
+    },
+    {
+      "epoch": 0.15496450503953937,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.5012,
+      "step": 1365
+    },
+    {
+      "epoch": 0.15507803214945845,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.5044,
+      "step": 1366
+    },
+    {
+      "epoch": 0.15519155925937753,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.4905,
+      "step": 1367
+    },
+    {
+      "epoch": 0.1553050863692966,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.002,
+      "loss": 5.5009,
+      "step": 1368
+    },
+    {
+      "epoch": 0.15541861347921568,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.4886,
+      "step": 1369
+    },
+    {
+      "epoch": 0.15553214058913475,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.5077,
+      "step": 1370
+    },
+    {
+      "epoch": 0.15564566769905383,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.4751,
+      "step": 1371
+    },
+    {
+      "epoch": 0.1557591948089729,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.5031,
+      "step": 1372
+    },
+    {
+      "epoch": 0.15587272191889198,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.5057,
+      "step": 1373
+    },
+    {
+      "epoch": 0.15598624902881106,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.4909,
+      "step": 1374
+    },
+    {
+      "epoch": 0.15609977613873013,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.4869,
+      "step": 1375
+    },
+    {
+      "epoch": 0.1562133032486492,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.5038,
+      "step": 1376
+    },
+    {
+      "epoch": 0.1563268303585683,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.4862,
+      "step": 1377
+    },
+    {
+      "epoch": 0.15644035746848736,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.002,
+      "loss": 5.4859,
+      "step": 1378
+    },
+    {
+      "epoch": 0.15655388457840644,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.002,
+      "loss": 5.4768,
+      "step": 1379
+    },
+    {
+      "epoch": 0.15666741168832551,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.4793,
+      "step": 1380
+    },
+    {
+      "epoch": 0.1567809387982446,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.4765,
+      "step": 1381
+    },
+    {
+      "epoch": 0.15689446590816367,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.4869,
+      "step": 1382
+    },
+    {
+      "epoch": 0.15700799301808274,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.5028,
+      "step": 1383
+    },
+    {
+      "epoch": 0.15712152012800182,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.5071,
+      "step": 1384
+    },
+    {
+      "epoch": 0.1572350472379209,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.485,
+      "step": 1385
+    },
+    {
+      "epoch": 0.15734857434783997,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.4737,
+      "step": 1386
+    },
+    {
+      "epoch": 0.15746210145775905,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.4989,
+      "step": 1387
+    },
+    {
+      "epoch": 0.15757562856767812,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.4855,
+      "step": 1388
+    },
+    {
+      "epoch": 0.1576891556775972,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.4772,
+      "step": 1389
+    },
+    {
+      "epoch": 0.15780268278751627,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.4933,
+      "step": 1390
+    },
+    {
+      "epoch": 0.15791620989743535,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.4855,
+      "step": 1391
+    },
+    {
+      "epoch": 0.15802973700735443,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.4806,
+      "step": 1392
+    },
+    {
+      "epoch": 0.1581432641172735,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.4889,
+      "step": 1393
+    },
+    {
+      "epoch": 0.15825679122719258,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.4864,
+      "step": 1394
+    },
+    {
+      "epoch": 0.15837031833711165,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.4974,
+      "step": 1395
+    },
+    {
+      "epoch": 0.15848384544703073,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.4708,
+      "step": 1396
+    },
+    {
+      "epoch": 0.1585973725569498,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.4889,
+      "step": 1397
+    },
+    {
+      "epoch": 0.15871089966686888,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.4789,
+      "step": 1398
+    },
+    {
+      "epoch": 0.15882442677678796,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.5028,
+      "step": 1399
+    },
+    {
+      "epoch": 0.15893795388670703,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.4884,
+      "step": 1400
+    },
+    {
+      "epoch": 0.1590514809966261,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.4941,
+      "step": 1401
+    },
+    {
+      "epoch": 0.15916500810654519,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.002,
+      "loss": 5.476,
+      "step": 1402
+    },
+    {
+      "epoch": 0.15927853521646426,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 5.5018,
+      "step": 1403
+    },
+    {
+      "epoch": 0.15939206232638334,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.002,
+      "loss": 5.4796,
+      "step": 1404
+    },
+    {
+      "epoch": 0.15950558943630241,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.4904,
+      "step": 1405
+    },
+    {
+      "epoch": 0.1596191165462215,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.002,
+      "loss": 5.4859,
+      "step": 1406
+    },
+    {
+      "epoch": 0.15973264365614057,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.002,
+      "loss": 5.473,
+      "step": 1407
+    },
+    {
+      "epoch": 0.15984617076605964,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.002,
+      "loss": 5.4877,
+      "step": 1408
+    },
+    {
+      "epoch": 0.15995969787597872,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.002,
+      "loss": 5.495,
+      "step": 1409
+    },
+    {
+      "epoch": 0.1600732249858978,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.4757,
+      "step": 1410
+    },
+    {
+      "epoch": 0.16018675209581687,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.4813,
+      "step": 1411
+    },
+    {
+      "epoch": 0.16030027920573595,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.4831,
+      "step": 1412
+    },
+    {
+      "epoch": 0.16041380631565502,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.4812,
+      "step": 1413
+    },
+    {
+      "epoch": 0.1605273334255741,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.4656,
+      "step": 1414
+    },
+    {
+      "epoch": 0.16064086053549317,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.4795,
+      "step": 1415
+    },
+    {
+      "epoch": 0.16075438764541225,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.4759,
+      "step": 1416
+    },
+    {
+      "epoch": 0.16086791475533133,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.4771,
+      "step": 1417
+    },
+    {
+      "epoch": 0.16098144186525043,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.5008,
+      "step": 1418
+    },
+    {
+      "epoch": 0.1610949689751695,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.4532,
+      "step": 1419
+    },
+    {
+      "epoch": 0.16120849608508858,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002,
+      "loss": 5.4976,
+      "step": 1420
+    },
+    {
+      "epoch": 0.16132202319500766,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.002,
+      "loss": 5.4839,
+      "step": 1421
+    },
+    {
+      "epoch": 0.16143555030492673,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.002,
+      "loss": 5.4972,
+      "step": 1422
+    },
+    {
+      "epoch": 0.1615490774148458,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.002,
+      "loss": 5.4811,
+      "step": 1423
+    },
+    {
+      "epoch": 0.16166260452476489,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.4947,
+      "step": 1424
+    },
+    {
+      "epoch": 0.16177613163468396,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.4983,
+      "step": 1425
+    },
+    {
+      "epoch": 0.16188965874460304,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.4758,
+      "step": 1426
+    },
+    {
+      "epoch": 0.1620031858545221,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4805,
+      "step": 1427
+    },
+    {
+      "epoch": 0.1621167129644412,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.4665,
+      "step": 1428
+    },
+    {
+      "epoch": 0.16223024007436027,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.4805,
+      "step": 1429
+    },
+    {
+      "epoch": 0.16234376718427934,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.4692,
+      "step": 1430
+    },
+    {
+      "epoch": 0.16245729429419842,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.5048,
+      "step": 1431
+    },
+    {
+      "epoch": 0.1625708214041175,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.4789,
+      "step": 1432
+    },
+    {
+      "epoch": 0.16268434851403657,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.4673,
+      "step": 1433
+    },
+    {
+      "epoch": 0.16279787562395565,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.4844,
+      "step": 1434
+    },
+    {
+      "epoch": 0.16291140273387472,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.4937,
+      "step": 1435
+    },
+    {
+      "epoch": 0.1630249298437938,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.002,
+      "loss": 5.4909,
+      "step": 1436
+    },
+    {
+      "epoch": 0.16313845695371287,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.002,
+      "loss": 5.4605,
+      "step": 1437
+    },
+    {
+      "epoch": 0.16325198406363195,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.002,
+      "loss": 5.4732,
+      "step": 1438
+    },
+    {
+      "epoch": 0.16336551117355103,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.002,
+      "loss": 5.4838,
+      "step": 1439
+    },
+    {
+      "epoch": 0.1634790382834701,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.002,
+      "loss": 5.476,
+      "step": 1440
+    },
+    {
+      "epoch": 0.16359256539338918,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.002,
+      "loss": 5.4707,
+      "step": 1441
+    },
+    {
+      "epoch": 0.16370609250330825,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.491,
+      "step": 1442
+    },
+    {
+      "epoch": 0.16381961961322733,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.4845,
+      "step": 1443
+    },
+    {
+      "epoch": 0.1639331467231464,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.4903,
+      "step": 1444
+    },
+    {
+      "epoch": 0.16404667383306548,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4855,
+      "step": 1445
+    },
+    {
+      "epoch": 0.16416020094298456,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.4706,
+      "step": 1446
+    },
+    {
+      "epoch": 0.16427372805290363,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.4858,
+      "step": 1447
+    },
+    {
+      "epoch": 0.1643872551628227,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.4512,
+      "step": 1448
+    },
+    {
+      "epoch": 0.16450078227274179,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.4715,
+      "step": 1449
+    },
+    {
+      "epoch": 0.16461430938266086,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.4567,
+      "step": 1450
+    },
+    {
+      "epoch": 0.16472783649257994,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.4879,
+      "step": 1451
+    },
+    {
+      "epoch": 0.164841363602499,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.4386,
+      "step": 1452
+    },
+    {
+      "epoch": 0.1649548907124181,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.4727,
+      "step": 1453
+    },
+    {
+      "epoch": 0.16506841782233717,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.4587,
+      "step": 1454
+    },
+    {
+      "epoch": 0.16518194493225624,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.4848,
+      "step": 1455
+    },
+    {
+      "epoch": 0.16529547204217532,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.4719,
+      "step": 1456
+    },
+    {
+      "epoch": 0.1654089991520944,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.4634,
+      "step": 1457
+    },
+    {
+      "epoch": 0.16552252626201347,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.495,
+      "step": 1458
+    },
+    {
+      "epoch": 0.16563605337193255,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.4712,
+      "step": 1459
+    },
+    {
+      "epoch": 0.16574958048185162,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.4553,
+      "step": 1460
+    },
+    {
+      "epoch": 0.1658631075917707,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.4866,
+      "step": 1461
+    },
+    {
+      "epoch": 0.16597663470168977,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.4752,
+      "step": 1462
+    },
+    {
+      "epoch": 0.16609016181160885,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.4927,
+      "step": 1463
+    },
+    {
+      "epoch": 0.16620368892152793,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.4516,
+      "step": 1464
+    },
+    {
+      "epoch": 0.166317216031447,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.4769,
+      "step": 1465
+    },
+    {
+      "epoch": 0.16643074314136608,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.4526,
+      "step": 1466
+    },
+    {
+      "epoch": 0.16654427025128515,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.4606,
+      "step": 1467
+    },
+    {
+      "epoch": 0.16665779736120423,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4622,
+      "step": 1468
+    },
+    {
+      "epoch": 0.1667713244711233,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.4751,
+      "step": 1469
+    },
+    {
+      "epoch": 0.16688485158104238,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.4731,
+      "step": 1470
+    },
+    {
+      "epoch": 0.16699837869096146,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.4632,
+      "step": 1471
+    },
+    {
+      "epoch": 0.16711190580088053,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.472,
+      "step": 1472
+    },
+    {
+      "epoch": 0.1672254329107996,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.4779,
+      "step": 1473
+    },
+    {
+      "epoch": 0.16733896002071869,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.4662,
+      "step": 1474
+    },
+    {
+      "epoch": 0.16745248713063776,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.4617,
+      "step": 1475
+    },
+    {
+      "epoch": 0.16756601424055684,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.002,
+      "loss": 5.4877,
+      "step": 1476
+    },
+    {
+      "epoch": 0.1676795413504759,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.002,
+      "loss": 5.4593,
+      "step": 1477
+    },
+    {
+      "epoch": 0.167793068460395,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 5.4768,
+      "step": 1478
+    },
+    {
+      "epoch": 0.1679065955703141,
+      "grad_norm": 0.5,
+      "learning_rate": 0.002,
+      "loss": 5.4696,
+      "step": 1479
+    },
+    {
+      "epoch": 0.16802012268023317,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.002,
+      "loss": 5.4673,
+      "step": 1480
+    },
+    {
+      "epoch": 0.16813364979015225,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.4726,
+      "step": 1481
+    },
+    {
+      "epoch": 0.16824717690007132,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.4702,
+      "step": 1482
+    },
+    {
+      "epoch": 0.1683607040099904,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.485,
+      "step": 1483
+    },
+    {
+      "epoch": 0.16847423111990947,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.4598,
+      "step": 1484
+    },
+    {
+      "epoch": 0.16858775822982855,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.4861,
+      "step": 1485
+    },
+    {
+      "epoch": 0.16870128533974763,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.4659,
+      "step": 1486
+    },
+    {
+      "epoch": 0.1688148124496667,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.4681,
+      "step": 1487
+    },
+    {
+      "epoch": 0.16892833955958578,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.4692,
+      "step": 1488
+    },
+    {
+      "epoch": 0.16904186666950485,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 5.4862,
+      "step": 1489
+    },
+    {
+      "epoch": 0.16915539377942393,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.002,
+      "loss": 5.473,
+      "step": 1490
+    },
+    {
+      "epoch": 0.169268920889343,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.002,
+      "loss": 5.476,
+      "step": 1491
+    },
+    {
+      "epoch": 0.16938244799926208,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.4545,
+      "step": 1492
+    },
+    {
+      "epoch": 0.16949597510918116,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.471,
+      "step": 1493
+    },
+    {
+      "epoch": 0.16960950221910023,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.4775,
+      "step": 1494
+    },
+    {
+      "epoch": 0.1697230293290193,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.4582,
+      "step": 1495
+    },
+    {
+      "epoch": 0.16983655643893839,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.4853,
+      "step": 1496
+    },
+    {
+      "epoch": 0.16995008354885746,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.4723,
+      "step": 1497
+    },
+    {
+      "epoch": 0.17006361065877654,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.4633,
+      "step": 1498
+    },
+    {
+      "epoch": 0.1701771377686956,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.4763,
+      "step": 1499
+    },
+    {
+      "epoch": 0.1702906648786147,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.453,
+      "step": 1500
+    },
+    {
+      "epoch": 0.17040419198853377,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.4781,
+      "step": 1501
+    },
+    {
+      "epoch": 0.17051771909845284,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.4801,
+      "step": 1502
+    },
+    {
+      "epoch": 0.17063124620837192,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.465,
+      "step": 1503
+    },
+    {
+      "epoch": 0.170744773318291,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4826,
+      "step": 1504
+    },
+    {
+      "epoch": 0.17085830042821007,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.452,
+      "step": 1505
+    },
+    {
+      "epoch": 0.17097182753812915,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.4603,
+      "step": 1506
+    },
+    {
+      "epoch": 0.17108535464804822,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.4688,
+      "step": 1507
+    },
+    {
+      "epoch": 0.1711988817579673,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.4663,
+      "step": 1508
+    },
+    {
+      "epoch": 0.17131240886788637,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.4648,
+      "step": 1509
+    },
+    {
+      "epoch": 0.17142593597780545,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.4656,
+      "step": 1510
+    },
+    {
+      "epoch": 0.17153946308772453,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 5.4348,
+      "step": 1511
+    },
+    {
+      "epoch": 0.1716529901976436,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.002,
+      "loss": 5.4604,
+      "step": 1512
+    },
+    {
+      "epoch": 0.17176651730756268,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.002,
+      "loss": 5.4631,
+      "step": 1513
+    },
+    {
+      "epoch": 0.17188004441748175,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.4614,
+      "step": 1514
+    },
+    {
+      "epoch": 0.17199357152740083,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.4629,
+      "step": 1515
+    },
+    {
+      "epoch": 0.1721070986373199,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.4654,
+      "step": 1516
+    },
+    {
+      "epoch": 0.17222062574723898,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.477,
+      "step": 1517
+    },
+    {
+      "epoch": 0.17233415285715806,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.4586,
+      "step": 1518
+    },
+    {
+      "epoch": 0.17244767996707713,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.4732,
+      "step": 1519
+    },
+    {
+      "epoch": 0.1725612070769962,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.4437,
+      "step": 1520
+    },
+    {
+      "epoch": 0.17267473418691529,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.4716,
+      "step": 1521
+    },
+    {
+      "epoch": 0.17278826129683436,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.4551,
+      "step": 1522
+    },
+    {
+      "epoch": 0.17290178840675344,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.4778,
+      "step": 1523
+    },
+    {
+      "epoch": 0.1730153155166725,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.4687,
+      "step": 1524
+    },
+    {
+      "epoch": 0.1731288426265916,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.4522,
+      "step": 1525
+    },
+    {
+      "epoch": 0.17324236973651067,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.4479,
+      "step": 1526
+    },
+    {
+      "epoch": 0.17335589684642974,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4743,
+      "step": 1527
+    },
+    {
+      "epoch": 0.17346942395634882,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.472,
+      "step": 1528
+    },
+    {
+      "epoch": 0.1735829510662679,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.4773,
+      "step": 1529
+    },
+    {
+      "epoch": 0.17369647817618697,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.4717,
+      "step": 1530
+    },
+    {
+      "epoch": 0.17381000528610605,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.4617,
+      "step": 1531
+    },
+    {
+      "epoch": 0.17392353239602512,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.4705,
+      "step": 1532
+    },
+    {
+      "epoch": 0.1740370595059442,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.4663,
+      "step": 1533
+    },
+    {
+      "epoch": 0.17415058661586327,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.4957,
+      "step": 1534
+    },
+    {
+      "epoch": 0.17426411372578235,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.4572,
+      "step": 1535
+    },
+    {
+      "epoch": 0.17437764083570143,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.4683,
+      "step": 1536
+    },
+    {
+      "epoch": 0.1744911679456205,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.4949,
+      "step": 1537
+    },
+    {
+      "epoch": 0.17460469505553958,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.4521,
+      "step": 1538
+    },
+    {
+      "epoch": 0.17471822216545868,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.4502,
+      "step": 1539
+    },
+    {
+      "epoch": 0.17483174927537776,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.4611,
+      "step": 1540
+    },
+    {
+      "epoch": 0.17494527638529683,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.002,
+      "loss": 5.4778,
+      "step": 1541
+    },
+    {
+      "epoch": 0.1750588034952159,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.002,
+      "loss": 5.4469,
+      "step": 1542
+    },
+    {
+      "epoch": 0.17517233060513498,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.002,
+      "loss": 5.4497,
+      "step": 1543
+    },
+    {
+      "epoch": 0.17528585771505406,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 5.4443,
+      "step": 1544
+    },
+    {
+      "epoch": 0.17539938482497314,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.4545,
+      "step": 1545
+    },
+    {
+      "epoch": 0.1755129119348922,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.4699,
+      "step": 1546
+    },
+    {
+      "epoch": 0.1756264390448113,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.4678,
+      "step": 1547
+    },
+    {
+      "epoch": 0.17573996615473036,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.4652,
+      "step": 1548
+    },
+    {
+      "epoch": 0.17585349326464944,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.464,
+      "step": 1549
+    },
+    {
+      "epoch": 0.17596702037456852,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.002,
+      "loss": 5.483,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1760805474844876,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.002,
+      "loss": 5.4593,
+      "step": 1551
+    },
+    {
+      "epoch": 0.17619407459440667,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.4614,
+      "step": 1552
+    },
+    {
+      "epoch": 0.17630760170432574,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.4653,
+      "step": 1553
+    },
+    {
+      "epoch": 0.17642112881424482,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.4821,
+      "step": 1554
+    },
+    {
+      "epoch": 0.1765346559241639,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.4356,
+      "step": 1555
+    },
+    {
+      "epoch": 0.17664818303408297,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.4494,
+      "step": 1556
+    },
+    {
+      "epoch": 0.17676171014400205,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.4647,
+      "step": 1557
+    },
+    {
+      "epoch": 0.17687523725392112,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.4265,
+      "step": 1558
+    },
+    {
+      "epoch": 0.1769887643638402,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.4432,
+      "step": 1559
+    },
+    {
+      "epoch": 0.17710229147375928,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.4406,
+      "step": 1560
+    },
+    {
+      "epoch": 0.17721581858367835,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.4446,
+      "step": 1561
+    },
+    {
+      "epoch": 0.17732934569359743,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.467,
+      "step": 1562
+    },
+    {
+      "epoch": 0.1774428728035165,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.4731,
+      "step": 1563
+    },
+    {
+      "epoch": 0.17755639991343558,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.4614,
+      "step": 1564
+    },
+    {
+      "epoch": 0.17766992702335466,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.456,
+      "step": 1565
+    },
+    {
+      "epoch": 0.17778345413327373,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.002,
+      "loss": 5.4389,
+      "step": 1566
+    },
+    {
+      "epoch": 0.1778969812431928,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 5.4437,
+      "step": 1567
+    },
+    {
+      "epoch": 0.17801050835311188,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.4411,
+      "step": 1568
+    },
+    {
+      "epoch": 0.17812403546303096,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.4746,
+      "step": 1569
+    },
+    {
+      "epoch": 0.17823756257295004,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.4455,
+      "step": 1570
+    },
+    {
+      "epoch": 0.1783510896828691,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.4468,
+      "step": 1571
+    },
+    {
+      "epoch": 0.1784646167927882,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.4619,
+      "step": 1572
+    },
+    {
+      "epoch": 0.17857814390270726,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.4728,
+      "step": 1573
+    },
+    {
+      "epoch": 0.17869167101262634,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.4654,
+      "step": 1574
+    },
+    {
+      "epoch": 0.17880519812254542,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.4495,
+      "step": 1575
+    },
+    {
+      "epoch": 0.1789187252324645,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.4541,
+      "step": 1576
+    },
+    {
+      "epoch": 0.17903225234238357,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.4461,
+      "step": 1577
+    },
+    {
+      "epoch": 0.17914577945230264,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.4525,
+      "step": 1578
+    },
+    {
+      "epoch": 0.17925930656222172,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.4405,
+      "step": 1579
+    },
+    {
+      "epoch": 0.1793728336721408,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.45,
+      "step": 1580
+    },
+    {
+      "epoch": 0.17948636078205987,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.4674,
+      "step": 1581
+    },
+    {
+      "epoch": 0.17959988789197895,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.4532,
+      "step": 1582
+    },
+    {
+      "epoch": 0.17971341500189802,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.4626,
+      "step": 1583
+    },
+    {
+      "epoch": 0.1798269421118171,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.4286,
+      "step": 1584
+    },
+    {
+      "epoch": 0.17994046922173618,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.4414,
+      "step": 1585
+    },
+    {
+      "epoch": 0.18005399633165525,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.4642,
+      "step": 1586
+    },
+    {
+      "epoch": 0.18016752344157433,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.4289,
+      "step": 1587
+    },
+    {
+      "epoch": 0.1802810505514934,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.4624,
+      "step": 1588
+    },
+    {
+      "epoch": 0.18039457766141248,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.4389,
+      "step": 1589
+    },
+    {
+      "epoch": 0.18050810477133156,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.4468,
+      "step": 1590
+    },
+    {
+      "epoch": 0.18062163188125063,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.4455,
+      "step": 1591
+    },
+    {
+      "epoch": 0.1807351589911697,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.4618,
+      "step": 1592
+    },
+    {
+      "epoch": 0.18084868610108878,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.4449,
+      "step": 1593
+    },
+    {
+      "epoch": 0.18096221321100786,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.4244,
+      "step": 1594
+    },
+    {
+      "epoch": 0.18107574032092694,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.4657,
+      "step": 1595
+    },
+    {
+      "epoch": 0.181189267430846,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.4504,
+      "step": 1596
+    },
+    {
+      "epoch": 0.1813027945407651,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.4436,
+      "step": 1597
+    },
+    {
+      "epoch": 0.18141632165068416,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.4427,
+      "step": 1598
+    },
+    {
+      "epoch": 0.18152984876060327,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.4517,
+      "step": 1599
+    },
+    {
+      "epoch": 0.18164337587052234,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.4587,
+      "step": 1600
+    },
+    {
+      "epoch": 0.18175690298044142,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.002,
+      "loss": 5.4272,
+      "step": 1601
+    },
+    {
+      "epoch": 0.1818704300903605,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.4805,
+      "step": 1602
+    },
+    {
+      "epoch": 0.18198395720027957,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4365,
+      "step": 1603
+    },
+    {
+      "epoch": 0.18209748431019865,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.4694,
+      "step": 1604
+    },
+    {
+      "epoch": 0.18221101142011772,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.4463,
+      "step": 1605
+    },
+    {
+      "epoch": 0.1823245385300368,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.4463,
+      "step": 1606
+    },
+    {
+      "epoch": 0.18243806563995588,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.4485,
+      "step": 1607
+    },
+    {
+      "epoch": 0.18255159274987495,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.485,
+      "step": 1608
+    },
+    {
+      "epoch": 0.18266511985979403,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.453,
+      "step": 1609
+    },
+    {
+      "epoch": 0.1827786469697131,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.4696,
+      "step": 1610
+    },
+    {
+      "epoch": 0.18289217407963218,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.4293,
+      "step": 1611
+    },
+    {
+      "epoch": 0.18300570118955126,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.4609,
+      "step": 1612
+    },
+    {
+      "epoch": 0.18311922829947033,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.4205,
+      "step": 1613
+    },
+    {
+      "epoch": 0.1832327554093894,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.002,
+      "loss": 5.4594,
+      "step": 1614
+    },
+    {
+      "epoch": 0.18334628251930848,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.002,
+      "loss": 5.4291,
+      "step": 1615
+    },
+    {
+      "epoch": 0.18345980962922756,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 5.4399,
+      "step": 1616
+    },
+    {
+      "epoch": 0.18357333673914664,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.002,
+      "loss": 5.4586,
+      "step": 1617
+    },
+    {
+      "epoch": 0.1836868638490657,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.4633,
+      "step": 1618
+    },
+    {
+      "epoch": 0.1838003909589848,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.002,
+      "loss": 5.4631,
+      "step": 1619
+    },
+    {
+      "epoch": 0.18391391806890386,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.445,
+      "step": 1620
+    },
+    {
+      "epoch": 0.18402744517882294,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.4526,
+      "step": 1621
+    },
+    {
+      "epoch": 0.18414097228874202,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.4417,
+      "step": 1622
+    },
+    {
+      "epoch": 0.1842544993986611,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.454,
+      "step": 1623
+    },
+    {
+      "epoch": 0.18436802650858017,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.453,
+      "step": 1624
+    },
+    {
+      "epoch": 0.18448155361849924,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.4207,
+      "step": 1625
+    },
+    {
+      "epoch": 0.18459508072841832,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.4373,
+      "step": 1626
+    },
+    {
+      "epoch": 0.1847086078383374,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.4356,
+      "step": 1627
+    },
+    {
+      "epoch": 0.18482213494825647,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.4404,
+      "step": 1628
+    },
+    {
+      "epoch": 0.18493566205817555,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.4603,
+      "step": 1629
+    },
+    {
+      "epoch": 0.18504918916809462,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.002,
+      "loss": 5.4375,
+      "step": 1630
+    },
+    {
+      "epoch": 0.1851627162780137,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.4427,
+      "step": 1631
+    },
+    {
+      "epoch": 0.18527624338793278,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.4439,
+      "step": 1632
+    },
+    {
+      "epoch": 0.18538977049785185,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.4491,
+      "step": 1633
+    },
+    {
+      "epoch": 0.18550329760777093,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.4486,
+      "step": 1634
+    },
+    {
+      "epoch": 0.18561682471769,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.4597,
+      "step": 1635
+    },
+    {
+      "epoch": 0.18573035182760908,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.4445,
+      "step": 1636
+    },
+    {
+      "epoch": 0.18584387893752816,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4322,
+      "step": 1637
+    },
+    {
+      "epoch": 0.18595740604744723,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.4408,
+      "step": 1638
+    },
+    {
+      "epoch": 0.1860709331573663,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.4643,
+      "step": 1639
+    },
+    {
+      "epoch": 0.18618446026728538,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.4536,
+      "step": 1640
+    },
+    {
+      "epoch": 0.18629798737720446,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.4587,
+      "step": 1641
+    },
+    {
+      "epoch": 0.18641151448712354,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.4356,
+      "step": 1642
+    },
+    {
+      "epoch": 0.1865250415970426,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.4437,
+      "step": 1643
+    },
+    {
+      "epoch": 0.1866385687069617,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.4518,
+      "step": 1644
+    },
+    {
+      "epoch": 0.18675209581688076,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.4505,
+      "step": 1645
+    },
+    {
+      "epoch": 0.18686562292679984,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.4595,
+      "step": 1646
+    },
+    {
+      "epoch": 0.18697915003671892,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.432,
+      "step": 1647
+    },
+    {
+      "epoch": 0.187092677146638,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.4474,
+      "step": 1648
+    },
+    {
+      "epoch": 0.18720620425655707,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.4421,
+      "step": 1649
+    },
+    {
+      "epoch": 0.18731973136647614,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4705,
+      "step": 1650
+    },
+    {
+      "epoch": 0.18743325847639522,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 5.4448,
+      "step": 1651
+    },
+    {
+      "epoch": 0.1875467855863143,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.002,
+      "loss": 5.4486,
+      "step": 1652
+    },
+    {
+      "epoch": 0.18766031269623337,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.002,
+      "loss": 5.4351,
+      "step": 1653
+    },
+    {
+      "epoch": 0.18777383980615245,
+      "grad_norm": 0.5,
+      "learning_rate": 0.002,
+      "loss": 5.4437,
+      "step": 1654
+    },
+    {
+      "epoch": 0.18788736691607152,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.4236,
+      "step": 1655
+    },
+    {
+      "epoch": 0.1880008940259906,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.4349,
+      "step": 1656
+    },
+    {
+      "epoch": 0.18811442113590968,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.4658,
+      "step": 1657
+    },
+    {
+      "epoch": 0.18822794824582875,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.4347,
+      "step": 1658
+    },
+    {
+      "epoch": 0.18834147535574783,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.4195,
+      "step": 1659
+    },
+    {
+      "epoch": 0.18845500246566693,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.4737,
+      "step": 1660
+    },
+    {
+      "epoch": 0.188568529575586,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.4287,
+      "step": 1661
+    },
+    {
+      "epoch": 0.18868205668550508,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002,
+      "loss": 5.457,
+      "step": 1662
+    },
+    {
+      "epoch": 0.18879558379542416,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.4479,
+      "step": 1663
+    },
+    {
+      "epoch": 0.18890911090534324,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.4372,
+      "step": 1664
+    },
+    {
+      "epoch": 0.1890226380152623,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.4425,
+      "step": 1665
+    },
+    {
+      "epoch": 0.1891361651251814,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.428,
+      "step": 1666
+    },
+    {
+      "epoch": 0.18924969223510046,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.4491,
+      "step": 1667
+    },
+    {
+      "epoch": 0.18936321934501954,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.4045,
+      "step": 1668
+    },
+    {
+      "epoch": 0.18947674645493862,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.4429,
+      "step": 1669
+    },
+    {
+      "epoch": 0.1895902735648577,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.002,
+      "loss": 5.4346,
+      "step": 1670
+    },
+    {
+      "epoch": 0.18970380067477677,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.002,
+      "loss": 5.4292,
+      "step": 1671
+    },
+    {
+      "epoch": 0.18981732778469584,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.4513,
+      "step": 1672
+    },
+    {
+      "epoch": 0.18993085489461492,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.4545,
+      "step": 1673
+    },
+    {
+      "epoch": 0.190044382004534,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.4336,
+      "step": 1674
+    },
+    {
+      "epoch": 0.19015790911445307,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.4561,
+      "step": 1675
+    },
+    {
+      "epoch": 0.19027143622437215,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.4442,
+      "step": 1676
+    },
+    {
+      "epoch": 0.19038496333429122,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.4376,
+      "step": 1677
+    },
+    {
+      "epoch": 0.1904984904442103,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.4509,
+      "step": 1678
+    },
+    {
+      "epoch": 0.19061201755412938,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.4609,
+      "step": 1679
+    },
+    {
+      "epoch": 0.19072554466404845,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.434,
+      "step": 1680
+    },
+    {
+      "epoch": 0.19083907177396753,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.4381,
+      "step": 1681
+    },
+    {
+      "epoch": 0.1909525988838866,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4327,
+      "step": 1682
+    },
+    {
+      "epoch": 0.19106612599380568,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.4649,
+      "step": 1683
+    },
+    {
+      "epoch": 0.19117965310372476,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.4252,
+      "step": 1684
+    },
+    {
+      "epoch": 0.19129318021364383,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.425,
+      "step": 1685
+    },
+    {
+      "epoch": 0.1914067073235629,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.4499,
+      "step": 1686
+    },
+    {
+      "epoch": 0.19152023443348198,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.4485,
+      "step": 1687
+    },
+    {
+      "epoch": 0.19163376154340106,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.4428,
+      "step": 1688
+    },
+    {
+      "epoch": 0.19174728865332014,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.4555,
+      "step": 1689
+    },
+    {
+      "epoch": 0.1918608157632392,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.4447,
+      "step": 1690
+    },
+    {
+      "epoch": 0.1919743428731583,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.002,
+      "loss": 5.4394,
+      "step": 1691
+    },
+    {
+      "epoch": 0.19208786998307736,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.4614,
+      "step": 1692
+    },
+    {
+      "epoch": 0.19220139709299644,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.002,
+      "loss": 5.4278,
+      "step": 1693
+    },
+    {
+      "epoch": 0.19231492420291552,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.4706,
+      "step": 1694
+    },
+    {
+      "epoch": 0.1924284513128346,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.4458,
+      "step": 1695
+    },
+    {
+      "epoch": 0.19254197842275367,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.4402,
+      "step": 1696
+    },
+    {
+      "epoch": 0.19265550553267274,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.002,
+      "loss": 5.4508,
+      "step": 1697
+    },
+    {
+      "epoch": 0.19276903264259182,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.4455,
+      "step": 1698
+    },
+    {
+      "epoch": 0.1928825597525109,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.4254,
+      "step": 1699
+    },
+    {
+      "epoch": 0.19299608686242997,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.44,
+      "step": 1700
+    },
+    {
+      "epoch": 0.19310961397234905,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.451,
+      "step": 1701
+    },
+    {
+      "epoch": 0.19322314108226812,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.436,
+      "step": 1702
+    },
+    {
+      "epoch": 0.1933366681921872,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.4104,
+      "step": 1703
+    },
+    {
+      "epoch": 0.19345019530210628,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.4251,
+      "step": 1704
+    },
+    {
+      "epoch": 0.19356372241202535,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.4337,
+      "step": 1705
+    },
+    {
+      "epoch": 0.19367724952194443,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.4406,
+      "step": 1706
+    },
+    {
+      "epoch": 0.1937907766318635,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.439,
+      "step": 1707
+    },
+    {
+      "epoch": 0.19390430374178258,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.4343,
+      "step": 1708
+    },
+    {
+      "epoch": 0.19401783085170166,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.4254,
+      "step": 1709
+    },
+    {
+      "epoch": 0.19413135796162073,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.4301,
+      "step": 1710
+    },
+    {
+      "epoch": 0.1942448850715398,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.4377,
+      "step": 1711
+    },
+    {
+      "epoch": 0.19435841218145888,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.002,
+      "loss": 5.4302,
+      "step": 1712
+    },
+    {
+      "epoch": 0.19447193929137796,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.4458,
+      "step": 1713
+    },
+    {
+      "epoch": 0.19458546640129704,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.4197,
+      "step": 1714
+    },
+    {
+      "epoch": 0.1946989935112161,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.4456,
+      "step": 1715
+    },
+    {
+      "epoch": 0.1948125206211352,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.4335,
+      "step": 1716
+    },
+    {
+      "epoch": 0.19492604773105426,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.4399,
+      "step": 1717
+    },
+    {
+      "epoch": 0.19503957484097334,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.4432,
+      "step": 1718
+    },
+    {
+      "epoch": 0.19515310195089242,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.4308,
+      "step": 1719
+    },
+    {
+      "epoch": 0.19526662906081152,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.4433,
+      "step": 1720
+    },
+    {
+      "epoch": 0.1953801561707306,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.4424,
+      "step": 1721
+    },
+    {
+      "epoch": 0.19549368328064967,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.4281,
+      "step": 1722
+    },
+    {
+      "epoch": 0.19560721039056875,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.4521,
+      "step": 1723
+    },
+    {
+      "epoch": 0.19572073750048782,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.4382,
+      "step": 1724
+    },
+    {
+      "epoch": 0.1958342646104069,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.4436,
+      "step": 1725
+    },
+    {
+      "epoch": 0.19594779172032598,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.4182,
+      "step": 1726
+    },
+    {
+      "epoch": 0.19606131883024505,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.4215,
+      "step": 1727
+    },
+    {
+      "epoch": 0.19617484594016413,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.425,
+      "step": 1728
+    },
+    {
+      "epoch": 0.1962883730500832,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4334,
+      "step": 1729
+    },
+    {
+      "epoch": 0.19640190016000228,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.4219,
+      "step": 1730
+    },
+    {
+      "epoch": 0.19651542726992136,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.4228,
+      "step": 1731
+    },
+    {
+      "epoch": 0.19662895437984043,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.4405,
+      "step": 1732
+    },
+    {
+      "epoch": 0.1967424814897595,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.4157,
+      "step": 1733
+    },
+    {
+      "epoch": 0.19685600859967858,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.4272,
+      "step": 1734
+    },
+    {
+      "epoch": 0.19696953570959766,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.426,
+      "step": 1735
+    },
+    {
+      "epoch": 0.19708306281951674,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.4255,
+      "step": 1736
+    },
+    {
+      "epoch": 0.1971965899294358,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.44,
+      "step": 1737
+    },
+    {
+      "epoch": 0.1973101170393549,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.002,
+      "loss": 5.4355,
+      "step": 1738
+    },
+    {
+      "epoch": 0.19742364414927396,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 5.4303,
+      "step": 1739
+    },
+    {
+      "epoch": 0.19753717125919304,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.4257,
+      "step": 1740
+    },
+    {
+      "epoch": 0.19765069836911212,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.4174,
+      "step": 1741
+    },
+    {
+      "epoch": 0.1977642254790312,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.45,
+      "step": 1742
+    },
+    {
+      "epoch": 0.19787775258895027,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4156,
+      "step": 1743
+    },
+    {
+      "epoch": 0.19799127969886934,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.4177,
+      "step": 1744
+    },
+    {
+      "epoch": 0.19810480680878842,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.4242,
+      "step": 1745
+    },
+    {
+      "epoch": 0.1982183339187075,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.416,
+      "step": 1746
+    },
+    {
+      "epoch": 0.19833186102862657,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.4304,
+      "step": 1747
+    },
+    {
+      "epoch": 0.19844538813854565,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.4186,
+      "step": 1748
+    },
+    {
+      "epoch": 0.19855891524846472,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.002,
+      "loss": 5.4159,
+      "step": 1749
+    },
+    {
+      "epoch": 0.1986724423583838,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.4203,
+      "step": 1750
+    },
+    {
+      "epoch": 0.19878596946830288,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.4175,
+      "step": 1751
+    },
+    {
+      "epoch": 0.19889949657822195,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.4461,
+      "step": 1752
+    },
+    {
+      "epoch": 0.19901302368814103,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.4222,
+      "step": 1753
+    },
+    {
+      "epoch": 0.1991265507980601,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.4224,
+      "step": 1754
+    },
+    {
+      "epoch": 0.19924007790797918,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.4232,
+      "step": 1755
+    },
+    {
+      "epoch": 0.19935360501789826,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.4121,
+      "step": 1756
+    },
+    {
+      "epoch": 0.19946713212781733,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.4381,
+      "step": 1757
+    },
+    {
+      "epoch": 0.1995806592377364,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.4171,
+      "step": 1758
+    },
+    {
+      "epoch": 0.19969418634765548,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.4242,
+      "step": 1759
+    },
+    {
+      "epoch": 0.19980771345757456,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.4195,
+      "step": 1760
+    },
+    {
+      "epoch": 0.19992124056749364,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.4407,
+      "step": 1761
+    },
+    {
+      "epoch": 0.2000347676774127,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.4333,
+      "step": 1762
+    },
+    {
+      "epoch": 0.2001482947873318,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.3916,
+      "step": 1763
+    },
+    {
+      "epoch": 0.20026182189725086,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.4317,
+      "step": 1764
+    },
+    {
+      "epoch": 0.20037534900716994,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.4223,
+      "step": 1765
+    },
+    {
+      "epoch": 0.20048887611708902,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.4211,
+      "step": 1766
+    },
+    {
+      "epoch": 0.2006024032270081,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.443,
+      "step": 1767
+    },
+    {
+      "epoch": 0.20071593033692717,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.4414,
+      "step": 1768
+    },
+    {
+      "epoch": 0.20082945744684624,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.4441,
+      "step": 1769
+    },
+    {
+      "epoch": 0.20094298455676532,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.4197,
+      "step": 1770
+    },
+    {
+      "epoch": 0.2010565116666844,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.4335,
+      "step": 1771
+    },
+    {
+      "epoch": 0.20117003877660347,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.4135,
+      "step": 1772
+    },
+    {
+      "epoch": 0.20128356588652255,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.415,
+      "step": 1773
+    },
+    {
+      "epoch": 0.20139709299644162,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.4429,
+      "step": 1774
+    },
+    {
+      "epoch": 0.2015106201063607,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.4239,
+      "step": 1775
+    },
+    {
+      "epoch": 0.20162414721627978,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.4177,
+      "step": 1776
+    },
+    {
+      "epoch": 0.20173767432619885,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.4501,
+      "step": 1777
+    },
+    {
+      "epoch": 0.20185120143611793,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.002,
+      "loss": 5.4213,
+      "step": 1778
+    },
+    {
+      "epoch": 0.201964728546037,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.002,
+      "loss": 5.4293,
+      "step": 1779
+    },
+    {
+      "epoch": 0.2020782556559561,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.002,
+      "loss": 5.4367,
+      "step": 1780
+    },
+    {
+      "epoch": 0.20219178276587518,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.4392,
+      "step": 1781
+    },
+    {
+      "epoch": 0.20230530987579426,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.4157,
+      "step": 1782
+    },
+    {
+      "epoch": 0.20241883698571334,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.4281,
+      "step": 1783
+    },
+    {
+      "epoch": 0.2025323640956324,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.3975,
+      "step": 1784
+    },
+    {
+      "epoch": 0.2026458912055515,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.002,
+      "loss": 5.4414,
+      "step": 1785
+    },
+    {
+      "epoch": 0.20275941831547056,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.4281,
+      "step": 1786
+    },
+    {
+      "epoch": 0.20287294542538964,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.4183,
+      "step": 1787
+    },
+    {
+      "epoch": 0.20298647253530872,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.4122,
+      "step": 1788
+    },
+    {
+      "epoch": 0.2030999996452278,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.4049,
+      "step": 1789
+    },
+    {
+      "epoch": 0.20321352675514687,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.4121,
+      "step": 1790
+    },
+    {
+      "epoch": 0.20332705386506594,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.4204,
+      "step": 1791
+    },
+    {
+      "epoch": 0.20344058097498502,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.4223,
+      "step": 1792
+    },
+    {
+      "epoch": 0.2035541080849041,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.4206,
+      "step": 1793
+    },
+    {
+      "epoch": 0.20366763519482317,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.4167,
+      "step": 1794
+    },
+    {
+      "epoch": 0.20378116230474225,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.4226,
+      "step": 1795
+    },
+    {
+      "epoch": 0.20389468941466132,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.4224,
+      "step": 1796
+    },
+    {
+      "epoch": 0.2040082165245804,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.4074,
+      "step": 1797
+    },
+    {
+      "epoch": 0.20412174363449948,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.4349,
+      "step": 1798
+    },
+    {
+      "epoch": 0.20423527074441855,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3955,
+      "step": 1799
+    },
+    {
+      "epoch": 0.20434879785433763,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.4145,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2044623249642567,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 5.4225,
+      "step": 1801
+    },
+    {
+      "epoch": 0.20457585207417578,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.3982,
+      "step": 1802
+    },
+    {
+      "epoch": 0.20468937918409486,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.4133,
+      "step": 1803
+    },
+    {
+      "epoch": 0.20480290629401393,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.4301,
+      "step": 1804
+    },
+    {
+      "epoch": 0.204916433403933,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.4216,
+      "step": 1805
+    },
+    {
+      "epoch": 0.20502996051385208,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.4176,
+      "step": 1806
+    },
+    {
+      "epoch": 0.20514348762377116,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.4278,
+      "step": 1807
+    },
+    {
+      "epoch": 0.20525701473369024,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.4071,
+      "step": 1808
+    },
+    {
+      "epoch": 0.2053705418436093,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.4412,
+      "step": 1809
+    },
+    {
+      "epoch": 0.2054840689535284,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.4154,
+      "step": 1810
+    },
+    {
+      "epoch": 0.20559759606344746,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.4196,
+      "step": 1811
+    },
+    {
+      "epoch": 0.20571112317336654,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.4096,
+      "step": 1812
+    },
+    {
+      "epoch": 0.20582465028328562,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.411,
+      "step": 1813
+    },
+    {
+      "epoch": 0.2059381773932047,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.4079,
+      "step": 1814
+    },
+    {
+      "epoch": 0.20605170450312377,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 5.4057,
+      "step": 1815
+    },
+    {
+      "epoch": 0.20616523161304284,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.4242,
+      "step": 1816
+    },
+    {
+      "epoch": 0.20627875872296192,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.429,
+      "step": 1817
+    },
+    {
+      "epoch": 0.206392285832881,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.4308,
+      "step": 1818
+    },
+    {
+      "epoch": 0.20650581294280007,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.4171,
+      "step": 1819
+    },
+    {
+      "epoch": 0.20661934005271915,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.4254,
+      "step": 1820
+    },
+    {
+      "epoch": 0.20673286716263822,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.4287,
+      "step": 1821
+    },
+    {
+      "epoch": 0.2068463942725573,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.4344,
+      "step": 1822
+    },
+    {
+      "epoch": 0.20695992138247638,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.4188,
+      "step": 1823
+    },
+    {
+      "epoch": 0.20707344849239545,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 5.4039,
+      "step": 1824
+    },
+    {
+      "epoch": 0.20718697560231453,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.4286,
+      "step": 1825
+    },
+    {
+      "epoch": 0.2073005027122336,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.4135,
+      "step": 1826
+    },
+    {
+      "epoch": 0.20741402982215268,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.4281,
+      "step": 1827
+    },
+    {
+      "epoch": 0.20752755693207176,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.4046,
+      "step": 1828
+    },
+    {
+      "epoch": 0.20764108404199083,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.4162,
+      "step": 1829
+    },
+    {
+      "epoch": 0.2077546111519099,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.421,
+      "step": 1830
+    },
+    {
+      "epoch": 0.20786813826182898,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.002,
+      "loss": 5.4429,
+      "step": 1831
+    },
+    {
+      "epoch": 0.20798166537174806,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.002,
+      "loss": 5.418,
+      "step": 1832
+    },
+    {
+      "epoch": 0.20809519248166714,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.412,
+      "step": 1833
+    },
+    {
+      "epoch": 0.2082087195915862,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.4358,
+      "step": 1834
+    },
+    {
+      "epoch": 0.2083222467015053,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.4257,
+      "step": 1835
+    },
+    {
+      "epoch": 0.20843577381142436,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.4159,
+      "step": 1836
+    },
+    {
+      "epoch": 0.20854930092134344,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.4374,
+      "step": 1837
+    },
+    {
+      "epoch": 0.20866282803126251,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.4029,
+      "step": 1838
+    },
+    {
+      "epoch": 0.2087763551411816,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3989,
+      "step": 1839
+    },
+    {
+      "epoch": 0.20888988225110067,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.4171,
+      "step": 1840
+    },
+    {
+      "epoch": 0.20900340936101977,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.422,
+      "step": 1841
+    },
+    {
+      "epoch": 0.20911693647093885,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.4012,
+      "step": 1842
+    },
+    {
+      "epoch": 0.20923046358085792,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.002,
+      "loss": 5.4251,
+      "step": 1843
+    },
+    {
+      "epoch": 0.209343990690777,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.002,
+      "loss": 5.416,
+      "step": 1844
+    },
+    {
+      "epoch": 0.20945751780069607,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.002,
+      "loss": 5.4226,
+      "step": 1845
+    },
+    {
+      "epoch": 0.20957104491061515,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.427,
+      "step": 1846
+    },
+    {
+      "epoch": 0.20968457202053423,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.4247,
+      "step": 1847
+    },
+    {
+      "epoch": 0.2097980991304533,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.4358,
+      "step": 1848
+    },
+    {
+      "epoch": 0.20991162624037238,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.391,
+      "step": 1849
+    },
+    {
+      "epoch": 0.21002515335029145,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.419,
+      "step": 1850
+    },
+    {
+      "epoch": 0.21013868046021053,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.4086,
+      "step": 1851
+    },
+    {
+      "epoch": 0.2102522075701296,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.4202,
+      "step": 1852
+    },
+    {
+      "epoch": 0.21036573468004868,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.4064,
+      "step": 1853
+    },
+    {
+      "epoch": 0.21047926178996776,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.4271,
+      "step": 1854
+    },
+    {
+      "epoch": 0.21059278889988683,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.4143,
+      "step": 1855
+    },
+    {
+      "epoch": 0.2107063160098059,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3868,
+      "step": 1856
+    },
+    {
+      "epoch": 0.210819843119725,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.4158,
+      "step": 1857
+    },
+    {
+      "epoch": 0.21093337022964406,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.4217,
+      "step": 1858
+    },
+    {
+      "epoch": 0.21104689733956314,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.4094,
+      "step": 1859
+    },
+    {
+      "epoch": 0.21116042444948221,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.4318,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2112739515594013,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.4094,
+      "step": 1861
+    },
+    {
+      "epoch": 0.21138747866932037,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.4279,
+      "step": 1862
+    },
+    {
+      "epoch": 0.21150100577923944,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.408,
+      "step": 1863
+    },
+    {
+      "epoch": 0.21161453288915852,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.4187,
+      "step": 1864
+    },
+    {
+      "epoch": 0.2117280599990776,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.4209,
+      "step": 1865
+    },
+    {
+      "epoch": 0.21184158710899667,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.412,
+      "step": 1866
+    },
+    {
+      "epoch": 0.21195511421891575,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.4038,
+      "step": 1867
+    },
+    {
+      "epoch": 0.21206864132883482,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.415,
+      "step": 1868
+    },
+    {
+      "epoch": 0.2121821684387539,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.4166,
+      "step": 1869
+    },
+    {
+      "epoch": 0.21229569554867297,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.4355,
+      "step": 1870
+    },
+    {
+      "epoch": 0.21240922265859205,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.423,
+      "step": 1871
+    },
+    {
+      "epoch": 0.21252274976851113,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4043,
+      "step": 1872
+    },
+    {
+      "epoch": 0.2126362768784302,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.4141,
+      "step": 1873
+    },
+    {
+      "epoch": 0.21274980398834928,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.4215,
+      "step": 1874
+    },
+    {
+      "epoch": 0.21286333109826835,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.4197,
+      "step": 1875
+    },
+    {
+      "epoch": 0.21297685820818743,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.4246,
+      "step": 1876
+    },
+    {
+      "epoch": 0.2130903853181065,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.4259,
+      "step": 1877
+    },
+    {
+      "epoch": 0.21320391242802558,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.428,
+      "step": 1878
+    },
+    {
+      "epoch": 0.21331743953794466,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.407,
+      "step": 1879
+    },
+    {
+      "epoch": 0.21343096664786373,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.002,
+      "loss": 5.4029,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2135444937577828,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 5.4101,
+      "step": 1881
+    },
+    {
+      "epoch": 0.2136580208677019,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.4078,
+      "step": 1882
+    },
+    {
+      "epoch": 0.21377154797762096,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.441,
+      "step": 1883
+    },
+    {
+      "epoch": 0.21388507508754004,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.4151,
+      "step": 1884
+    },
+    {
+      "epoch": 0.21399860219745911,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.4206,
+      "step": 1885
+    },
+    {
+      "epoch": 0.2141121293073782,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.4111,
+      "step": 1886
+    },
+    {
+      "epoch": 0.21422565641729727,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.4172,
+      "step": 1887
+    },
+    {
+      "epoch": 0.21433918352721634,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.401,
+      "step": 1888
+    },
+    {
+      "epoch": 0.21445271063713542,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.4182,
+      "step": 1889
+    },
+    {
+      "epoch": 0.2145662377470545,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.3957,
+      "step": 1890
+    },
+    {
+      "epoch": 0.21467976485697357,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3993,
+      "step": 1891
+    },
+    {
+      "epoch": 0.21479329196689265,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.4031,
+      "step": 1892
+    },
+    {
+      "epoch": 0.21490681907681172,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.4166,
+      "step": 1893
+    },
+    {
+      "epoch": 0.2150203461867308,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.4179,
+      "step": 1894
+    },
+    {
+      "epoch": 0.21513387329664987,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.424,
+      "step": 1895
+    },
+    {
+      "epoch": 0.21524740040656895,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.4048,
+      "step": 1896
+    },
+    {
+      "epoch": 0.21536092751648803,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.4076,
+      "step": 1897
+    },
+    {
+      "epoch": 0.2154744546264071,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.4134,
+      "step": 1898
+    },
+    {
+      "epoch": 0.21558798173632618,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.4067,
+      "step": 1899
+    },
+    {
+      "epoch": 0.21570150884624525,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.3943,
+      "step": 1900
+    },
+    {
+      "epoch": 0.21581503595616436,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.406,
+      "step": 1901
+    },
+    {
+      "epoch": 0.21592856306608343,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.3992,
+      "step": 1902
+    },
+    {
+      "epoch": 0.2160420901760025,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.4058,
+      "step": 1903
+    },
+    {
+      "epoch": 0.2161556172859216,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.4055,
+      "step": 1904
+    },
+    {
+      "epoch": 0.21626914439584066,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.3939,
+      "step": 1905
+    },
+    {
+      "epoch": 0.21638267150575974,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.4179,
+      "step": 1906
+    },
+    {
+      "epoch": 0.21649619861567881,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.4217,
+      "step": 1907
+    },
+    {
+      "epoch": 0.2166097257255979,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.4194,
+      "step": 1908
+    },
+    {
+      "epoch": 0.21672325283551697,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3921,
+      "step": 1909
+    },
+    {
+      "epoch": 0.21683677994543604,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.4091,
+      "step": 1910
+    },
+    {
+      "epoch": 0.21695030705535512,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.41,
+      "step": 1911
+    },
+    {
+      "epoch": 0.2170638341652742,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.394,
+      "step": 1912
+    },
+    {
+      "epoch": 0.21717736127519327,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.4143,
+      "step": 1913
+    },
+    {
+      "epoch": 0.21729088838511235,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.002,
+      "loss": 5.4048,
+      "step": 1914
+    },
+    {
+      "epoch": 0.21740441549503142,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.4296,
+      "step": 1915
+    },
+    {
+      "epoch": 0.2175179426049505,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.3821,
+      "step": 1916
+    },
+    {
+      "epoch": 0.21763146971486957,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.4009,
+      "step": 1917
+    },
+    {
+      "epoch": 0.21774499682478865,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.4232,
+      "step": 1918
+    },
+    {
+      "epoch": 0.21785852393470773,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4186,
+      "step": 1919
+    },
+    {
+      "epoch": 0.2179720510446268,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.4007,
+      "step": 1920
+    },
+    {
+      "epoch": 0.21808557815454588,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.4072,
+      "step": 1921
+    },
+    {
+      "epoch": 0.21819910526446495,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.375,
+      "step": 1922
+    },
+    {
+      "epoch": 0.21831263237438403,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.4234,
+      "step": 1923
+    },
+    {
+      "epoch": 0.2184261594843031,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3999,
+      "step": 1924
+    },
+    {
+      "epoch": 0.21853968659422218,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.3978,
+      "step": 1925
+    },
+    {
+      "epoch": 0.21865321370414126,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.4143,
+      "step": 1926
+    },
+    {
+      "epoch": 0.21876674081406033,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.4105,
+      "step": 1927
+    },
+    {
+      "epoch": 0.2188802679239794,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002,
+      "loss": 5.4058,
+      "step": 1928
+    },
+    {
+      "epoch": 0.2189937950338985,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.4075,
+      "step": 1929
+    },
+    {
+      "epoch": 0.21910732214381756,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.4129,
+      "step": 1930
+    },
+    {
+      "epoch": 0.21922084925373664,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.4163,
+      "step": 1931
+    },
+    {
+      "epoch": 0.21933437636365571,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.4088,
+      "step": 1932
+    },
+    {
+      "epoch": 0.2194479034735748,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.002,
+      "loss": 5.3952,
+      "step": 1933
+    },
+    {
+      "epoch": 0.21956143058349387,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.382,
+      "step": 1934
+    },
+    {
+      "epoch": 0.21967495769341294,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.4053,
+      "step": 1935
+    },
+    {
+      "epoch": 0.21978848480333202,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.3779,
+      "step": 1936
+    },
+    {
+      "epoch": 0.2199020119132511,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.3973,
+      "step": 1937
+    },
+    {
+      "epoch": 0.22001553902317017,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.4245,
+      "step": 1938
+    },
+    {
+      "epoch": 0.22012906613308925,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.4045,
+      "step": 1939
+    },
+    {
+      "epoch": 0.22024259324300832,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3981,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2203561203529274,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.412,
+      "step": 1941
+    },
+    {
+      "epoch": 0.22046964746284647,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.4019,
+      "step": 1942
+    },
+    {
+      "epoch": 0.22058317457276555,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.4214,
+      "step": 1943
+    },
+    {
+      "epoch": 0.22069670168268463,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3993,
+      "step": 1944
+    },
+    {
+      "epoch": 0.2208102287926037,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.4013,
+      "step": 1945
+    },
+    {
+      "epoch": 0.22092375590252278,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.4114,
+      "step": 1946
+    },
+    {
+      "epoch": 0.22103728301244185,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.4081,
+      "step": 1947
+    },
+    {
+      "epoch": 0.22115081012236093,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.4156,
+      "step": 1948
+    },
+    {
+      "epoch": 0.22126433723228,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.002,
+      "loss": 5.4123,
+      "step": 1949
+    },
+    {
+      "epoch": 0.22137786434219908,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 5.4018,
+      "step": 1950
+    },
+    {
+      "epoch": 0.22149139145211816,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.387,
+      "step": 1951
+    },
+    {
+      "epoch": 0.22160491856203723,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.3916,
+      "step": 1952
+    },
+    {
+      "epoch": 0.2217184456719563,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.4162,
+      "step": 1953
+    },
+    {
+      "epoch": 0.2218319727818754,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.377,
+      "step": 1954
+    },
+    {
+      "epoch": 0.22194549989179446,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.3958,
+      "step": 1955
+    },
+    {
+      "epoch": 0.22205902700171354,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.4014,
+      "step": 1956
+    },
+    {
+      "epoch": 0.22217255411163261,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.4109,
+      "step": 1957
+    },
+    {
+      "epoch": 0.2222860812215517,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.4088,
+      "step": 1958
+    },
+    {
+      "epoch": 0.22239960833147077,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.4198,
+      "step": 1959
+    },
+    {
+      "epoch": 0.22251313544138984,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.4139,
+      "step": 1960
+    },
+    {
+      "epoch": 0.22262666255130895,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.3946,
+      "step": 1961
+    },
+    {
+      "epoch": 0.22274018966122802,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.4308,
+      "step": 1962
+    },
+    {
+      "epoch": 0.2228537167711471,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.4132,
+      "step": 1963
+    },
+    {
+      "epoch": 0.22296724388106617,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.4049,
+      "step": 1964
+    },
+    {
+      "epoch": 0.22308077099098525,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.4155,
+      "step": 1965
+    },
+    {
+      "epoch": 0.22319429810090433,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3751,
+      "step": 1966
+    },
+    {
+      "epoch": 0.2233078252108234,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3814,
+      "step": 1967
+    },
+    {
+      "epoch": 0.22342135232074248,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.4234,
+      "step": 1968
+    },
+    {
+      "epoch": 0.22353487943066155,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.3833,
+      "step": 1969
+    },
+    {
+      "epoch": 0.22364840654058063,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.4179,
+      "step": 1970
+    },
+    {
+      "epoch": 0.2237619336504997,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.388,
+      "step": 1971
+    },
+    {
+      "epoch": 0.22387546076041878,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.002,
+      "loss": 5.4069,
+      "step": 1972
+    },
+    {
+      "epoch": 0.22398898787033786,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.3993,
+      "step": 1973
+    },
+    {
+      "epoch": 0.22410251498025693,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.4097,
+      "step": 1974
+    },
+    {
+      "epoch": 0.224216042090176,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.4118,
+      "step": 1975
+    },
+    {
+      "epoch": 0.22432956920009509,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.4064,
+      "step": 1976
+    },
+    {
+      "epoch": 0.22444309631001416,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.4124,
+      "step": 1977
+    },
+    {
+      "epoch": 0.22455662341993324,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.4048,
+      "step": 1978
+    },
+    {
+      "epoch": 0.2246701505298523,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.4064,
+      "step": 1979
+    },
+    {
+      "epoch": 0.2247836776397714,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3858,
+      "step": 1980
+    },
+    {
+      "epoch": 0.22489720474969047,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.4007,
+      "step": 1981
+    },
+    {
+      "epoch": 0.22501073185960954,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.4041,
+      "step": 1982
+    },
+    {
+      "epoch": 0.22512425896952862,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.3858,
+      "step": 1983
+    },
+    {
+      "epoch": 0.2252377860794477,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.4045,
+      "step": 1984
+    },
+    {
+      "epoch": 0.22535131318936677,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.403,
+      "step": 1985
+    },
+    {
+      "epoch": 0.22546484029928585,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3755,
+      "step": 1986
+    },
+    {
+      "epoch": 0.22557836740920492,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3899,
+      "step": 1987
+    },
+    {
+      "epoch": 0.225691894519124,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.4022,
+      "step": 1988
+    },
+    {
+      "epoch": 0.22580542162904307,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.389,
+      "step": 1989
+    },
+    {
+      "epoch": 0.22591894873896215,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.4022,
+      "step": 1990
+    },
+    {
+      "epoch": 0.22603247584888123,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3991,
+      "step": 1991
+    },
+    {
+      "epoch": 0.2261460029588003,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3909,
+      "step": 1992
+    },
+    {
+      "epoch": 0.22625953006871938,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.3888,
+      "step": 1993
+    },
+    {
+      "epoch": 0.22637305717863845,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.4074,
+      "step": 1994
+    },
+    {
+      "epoch": 0.22648658428855753,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.002,
+      "loss": 5.4051,
+      "step": 1995
+    },
+    {
+      "epoch": 0.2266001113984766,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.3952,
+      "step": 1996
+    },
+    {
+      "epoch": 0.22671363850839568,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.4027,
+      "step": 1997
+    },
+    {
+      "epoch": 0.22682716561831476,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.4056,
+      "step": 1998
+    },
+    {
+      "epoch": 0.22694069272823383,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.4077,
+      "step": 1999
+    },
+    {
+      "epoch": 0.2270542198381529,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.4049,
+      "step": 2000
+    },
+    {
+      "epoch": 0.22716774694807199,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.4031,
+      "step": 2001
+    },
+    {
+      "epoch": 0.22728127405799106,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.4046,
+      "step": 2002
+    },
+    {
+      "epoch": 0.22739480116791014,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.3963,
+      "step": 2003
+    },
+    {
+      "epoch": 0.2275083282778292,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.3966,
+      "step": 2004
+    },
+    {
+      "epoch": 0.2276218553877483,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.3972,
+      "step": 2005
+    },
+    {
+      "epoch": 0.22773538249766737,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 5.4023,
+      "step": 2006
+    },
+    {
+      "epoch": 0.22784890960758644,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3935,
+      "step": 2007
+    },
+    {
+      "epoch": 0.22796243671750552,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.4097,
+      "step": 2008
+    },
+    {
+      "epoch": 0.2280759638274246,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3915,
+      "step": 2009
+    },
+    {
+      "epoch": 0.22818949093734367,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.002,
+      "loss": 5.4088,
+      "step": 2010
+    },
+    {
+      "epoch": 0.22830301804726275,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.3922,
+      "step": 2011
+    },
+    {
+      "epoch": 0.22841654515718182,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.002,
+      "loss": 5.3775,
+      "step": 2012
+    },
+    {
+      "epoch": 0.2285300722671009,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.002,
+      "loss": 5.401,
+      "step": 2013
+    },
+    {
+      "epoch": 0.22864359937701997,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.41,
+      "step": 2014
+    },
+    {
+      "epoch": 0.22875712648693905,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.4001,
+      "step": 2015
+    },
+    {
+      "epoch": 0.22887065359685813,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3902,
+      "step": 2016
+    },
+    {
+      "epoch": 0.2289841807067772,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.4054,
+      "step": 2017
+    },
+    {
+      "epoch": 0.22909770781669628,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.3748,
+      "step": 2018
+    },
+    {
+      "epoch": 0.22921123492661535,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.3984,
+      "step": 2019
+    },
+    {
+      "epoch": 0.22932476203653443,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.3988,
+      "step": 2020
+    },
+    {
+      "epoch": 0.2294382891464535,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3763,
+      "step": 2021
+    },
+    {
+      "epoch": 0.2295518162563726,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.385,
+      "step": 2022
+    },
+    {
+      "epoch": 0.22966534336629169,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.4149,
+      "step": 2023
+    },
+    {
+      "epoch": 0.22977887047621076,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4083,
+      "step": 2024
+    },
+    {
+      "epoch": 0.22989239758612984,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3961,
+      "step": 2025
+    },
+    {
+      "epoch": 0.2300059246960489,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.3852,
+      "step": 2026
+    },
+    {
+      "epoch": 0.230119451805968,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.385,
+      "step": 2027
+    },
+    {
+      "epoch": 0.23023297891588707,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.4009,
+      "step": 2028
+    },
+    {
+      "epoch": 0.23034650602580614,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4018,
+      "step": 2029
+    },
+    {
+      "epoch": 0.23046003313572522,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.4244,
+      "step": 2030
+    },
+    {
+      "epoch": 0.2305735602456443,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.4061,
+      "step": 2031
+    },
+    {
+      "epoch": 0.23068708735556337,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3881,
+      "step": 2032
+    },
+    {
+      "epoch": 0.23080061446548245,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.4068,
+      "step": 2033
+    },
+    {
+      "epoch": 0.23091414157540152,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3866,
+      "step": 2034
+    },
+    {
+      "epoch": 0.2310276686853206,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.385,
+      "step": 2035
+    },
+    {
+      "epoch": 0.23114119579523967,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.4174,
+      "step": 2036
+    },
+    {
+      "epoch": 0.23125472290515875,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.411,
+      "step": 2037
+    },
+    {
+      "epoch": 0.23136825001507783,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 5.3917,
+      "step": 2038
+    },
+    {
+      "epoch": 0.2314817771249969,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.4043,
+      "step": 2039
+    },
+    {
+      "epoch": 0.23159530423491598,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.002,
+      "loss": 5.4053,
+      "step": 2040
+    },
+    {
+      "epoch": 0.23170883134483505,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.4052,
+      "step": 2041
+    },
+    {
+      "epoch": 0.23182235845475413,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.4109,
+      "step": 2042
+    },
+    {
+      "epoch": 0.2319358855646732,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.3953,
+      "step": 2043
+    },
+    {
+      "epoch": 0.23204941267459228,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3913,
+      "step": 2044
+    },
+    {
+      "epoch": 0.23216293978451136,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3846,
+      "step": 2045
+    },
+    {
+      "epoch": 0.23227646689443043,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.4065,
+      "step": 2046
+    },
+    {
+      "epoch": 0.2323899940043495,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.3722,
+      "step": 2047
+    },
+    {
+      "epoch": 0.23250352111426859,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.402,
+      "step": 2048
+    },
+    {
+      "epoch": 0.23261704822418766,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3825,
+      "step": 2049
+    },
+    {
+      "epoch": 0.23273057533410674,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3987,
+      "step": 2050
+    },
+    {
+      "epoch": 0.2328441024440258,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.4239,
+      "step": 2051
+    },
+    {
+      "epoch": 0.2329576295539449,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.3948,
+      "step": 2052
+    },
+    {
+      "epoch": 0.23307115666386397,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3854,
+      "step": 2053
+    },
+    {
+      "epoch": 0.23318468377378304,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3738,
+      "step": 2054
+    },
+    {
+      "epoch": 0.23329821088370212,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3851,
+      "step": 2055
+    },
+    {
+      "epoch": 0.2334117379936212,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.4038,
+      "step": 2056
+    },
+    {
+      "epoch": 0.23352526510354027,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.401,
+      "step": 2057
+    },
+    {
+      "epoch": 0.23363879221345935,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.3975,
+      "step": 2058
+    },
+    {
+      "epoch": 0.23375231932337842,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.3579,
+      "step": 2059
+    },
+    {
+      "epoch": 0.2338658464332975,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.3699,
+      "step": 2060
+    },
+    {
+      "epoch": 0.23397937354321657,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.3939,
+      "step": 2061
+    },
+    {
+      "epoch": 0.23409290065313565,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.4082,
+      "step": 2062
+    },
+    {
+      "epoch": 0.23420642776305473,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.3757,
+      "step": 2063
+    },
+    {
+      "epoch": 0.2343199548729738,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.3791,
+      "step": 2064
+    },
+    {
+      "epoch": 0.23443348198289288,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 5.4254,
+      "step": 2065
+    },
+    {
+      "epoch": 0.23454700909281195,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.4129,
+      "step": 2066
+    },
+    {
+      "epoch": 0.23466053620273103,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.4029,
+      "step": 2067
+    },
+    {
+      "epoch": 0.2347740633126501,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3978,
+      "step": 2068
+    },
+    {
+      "epoch": 0.23488759042256918,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.3829,
+      "step": 2069
+    },
+    {
+      "epoch": 0.23500111753248826,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.419,
+      "step": 2070
+    },
+    {
+      "epoch": 0.23511464464240733,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.3789,
+      "step": 2071
+    },
+    {
+      "epoch": 0.2352281717523264,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.4004,
+      "step": 2072
+    },
+    {
+      "epoch": 0.23534169886224549,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.4018,
+      "step": 2073
+    },
+    {
+      "epoch": 0.23545522597216456,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.4067,
+      "step": 2074
+    },
+    {
+      "epoch": 0.23556875308208364,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.4032,
+      "step": 2075
+    },
+    {
+      "epoch": 0.2356822801920027,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.4079,
+      "step": 2076
+    },
+    {
+      "epoch": 0.2357958073019218,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.3909,
+      "step": 2077
+    },
+    {
+      "epoch": 0.23590933441184087,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.3813,
+      "step": 2078
+    },
+    {
+      "epoch": 0.23602286152175994,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3918,
+      "step": 2079
+    },
+    {
+      "epoch": 0.23613638863167902,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.3936,
+      "step": 2080
+    },
+    {
+      "epoch": 0.2362499157415981,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3989,
+      "step": 2081
+    },
+    {
+      "epoch": 0.2363634428515172,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.4042,
+      "step": 2082
+    },
+    {
+      "epoch": 0.23647696996143627,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.3853,
+      "step": 2083
+    },
+    {
+      "epoch": 0.23659049707135535,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3921,
+      "step": 2084
+    },
+    {
+      "epoch": 0.23670402418127443,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.4004,
+      "step": 2085
+    },
+    {
+      "epoch": 0.2368175512911935,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.3765,
+      "step": 2086
+    },
+    {
+      "epoch": 0.23693107840111258,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.4073,
+      "step": 2087
+    },
+    {
+      "epoch": 0.23704460551103165,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 5.393,
+      "step": 2088
+    },
+    {
+      "epoch": 0.23715813262095073,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.4015,
+      "step": 2089
+    },
+    {
+      "epoch": 0.2372716597308698,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3903,
+      "step": 2090
+    },
+    {
+      "epoch": 0.23738518684078888,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.4097,
+      "step": 2091
+    },
+    {
+      "epoch": 0.23749871395070796,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3742,
+      "step": 2092
+    },
+    {
+      "epoch": 0.23761224106062703,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3701,
+      "step": 2093
+    },
+    {
+      "epoch": 0.2377257681705461,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3913,
+      "step": 2094
+    },
+    {
+      "epoch": 0.23783929528046518,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3918,
+      "step": 2095
+    },
+    {
+      "epoch": 0.23795282239038426,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.3823,
+      "step": 2096
+    },
+    {
+      "epoch": 0.23806634950030334,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3811,
+      "step": 2097
+    },
+    {
+      "epoch": 0.2381798766102224,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.3858,
+      "step": 2098
+    },
+    {
+      "epoch": 0.2382934037201415,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3962,
+      "step": 2099
+    },
+    {
+      "epoch": 0.23840693083006056,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3937,
+      "step": 2100
+    },
+    {
+      "epoch": 0.23852045793997964,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.4131,
+      "step": 2101
+    },
+    {
+      "epoch": 0.23863398504989872,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3794,
+      "step": 2102
+    },
+    {
+      "epoch": 0.2387475121598178,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.3989,
+      "step": 2103
+    },
+    {
+      "epoch": 0.23886103926973687,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.3994,
+      "step": 2104
+    },
+    {
+      "epoch": 0.23897456637965594,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.3824,
+      "step": 2105
+    },
+    {
+      "epoch": 0.23908809348957502,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.397,
+      "step": 2106
+    },
+    {
+      "epoch": 0.2392016205994941,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.4024,
+      "step": 2107
+    },
+    {
+      "epoch": 0.23931514770941317,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.4112,
+      "step": 2108
+    },
+    {
+      "epoch": 0.23942867481933225,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.3826,
+      "step": 2109
+    },
+    {
+      "epoch": 0.23954220192925132,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.376,
+      "step": 2110
+    },
+    {
+      "epoch": 0.2396557290391704,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.3851,
+      "step": 2111
+    },
+    {
+      "epoch": 0.23976925614908948,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.3876,
+      "step": 2112
+    },
+    {
+      "epoch": 0.23988278325900855,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3778,
+      "step": 2113
+    },
+    {
+      "epoch": 0.23999631036892763,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3819,
+      "step": 2114
+    },
+    {
+      "epoch": 0.2401098374788467,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.392,
+      "step": 2115
+    },
+    {
+      "epoch": 0.24022336458876578,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.3932,
+      "step": 2116
+    },
+    {
+      "epoch": 0.24033689169868486,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3922,
+      "step": 2117
+    },
+    {
+      "epoch": 0.24045041880860393,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.408,
+      "step": 2118
+    },
+    {
+      "epoch": 0.240563945918523,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.3798,
+      "step": 2119
+    },
+    {
+      "epoch": 0.24067747302844208,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3892,
+      "step": 2120
+    },
+    {
+      "epoch": 0.24079100013836116,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.3991,
+      "step": 2121
+    },
+    {
+      "epoch": 0.24090452724828024,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.3892,
+      "step": 2122
+    },
+    {
+      "epoch": 0.2410180543581993,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.3834,
+      "step": 2123
+    },
+    {
+      "epoch": 0.2411315814681184,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.398,
+      "step": 2124
+    },
+    {
+      "epoch": 0.24124510857803746,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3822,
+      "step": 2125
+    },
+    {
+      "epoch": 0.24135863568795654,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.3883,
+      "step": 2126
+    },
+    {
+      "epoch": 0.24147216279787562,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3846,
+      "step": 2127
+    },
+    {
+      "epoch": 0.2415856899077947,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.3829,
+      "step": 2128
+    },
+    {
+      "epoch": 0.24169921701771377,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.002,
+      "loss": 5.3777,
+      "step": 2129
+    },
+    {
+      "epoch": 0.24181274412763284,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.4042,
+      "step": 2130
+    },
+    {
+      "epoch": 0.24192627123755192,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.3933,
+      "step": 2131
+    },
+    {
+      "epoch": 0.242039798347471,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.3824,
+      "step": 2132
+    },
+    {
+      "epoch": 0.24215332545739007,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3863,
+      "step": 2133
+    },
+    {
+      "epoch": 0.24226685256730915,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.3743,
+      "step": 2134
+    },
+    {
+      "epoch": 0.24238037967722822,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.382,
+      "step": 2135
+    },
+    {
+      "epoch": 0.2424939067871473,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.3797,
+      "step": 2136
+    },
+    {
+      "epoch": 0.24260743389706638,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3915,
+      "step": 2137
+    },
+    {
+      "epoch": 0.24272096100698545,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3877,
+      "step": 2138
+    },
+    {
+      "epoch": 0.24283448811690453,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3841,
+      "step": 2139
+    },
+    {
+      "epoch": 0.2429480152268236,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3565,
+      "step": 2140
+    },
+    {
+      "epoch": 0.24306154233674268,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.3835,
+      "step": 2141
+    },
+    {
+      "epoch": 0.24317506944666178,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.3726,
+      "step": 2142
+    },
+    {
+      "epoch": 0.24328859655658086,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3656,
+      "step": 2143
+    },
+    {
+      "epoch": 0.24340212366649994,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.3759,
+      "step": 2144
+    },
+    {
+      "epoch": 0.243515650776419,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3909,
+      "step": 2145
+    },
+    {
+      "epoch": 0.2436291778863381,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3735,
+      "step": 2146
+    },
+    {
+      "epoch": 0.24374270499625716,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3757,
+      "step": 2147
+    },
+    {
+      "epoch": 0.24385623210617624,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3903,
+      "step": 2148
+    },
+    {
+      "epoch": 0.24396975921609532,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.3924,
+      "step": 2149
+    },
+    {
+      "epoch": 0.2440832863260144,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.3989,
+      "step": 2150
+    },
+    {
+      "epoch": 0.24419681343593347,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3725,
+      "step": 2151
+    },
+    {
+      "epoch": 0.24431034054585254,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3858,
+      "step": 2152
+    },
+    {
+      "epoch": 0.24442386765577162,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3765,
+      "step": 2153
+    },
+    {
+      "epoch": 0.2445373947656907,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3717,
+      "step": 2154
+    },
+    {
+      "epoch": 0.24465092187560977,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.3982,
+      "step": 2155
+    },
+    {
+      "epoch": 0.24476444898552885,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.3693,
+      "step": 2156
+    },
+    {
+      "epoch": 0.24487797609544792,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.3722,
+      "step": 2157
+    },
+    {
+      "epoch": 0.244991503205367,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.3723,
+      "step": 2158
+    },
+    {
+      "epoch": 0.24510503031528608,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.4009,
+      "step": 2159
+    },
+    {
+      "epoch": 0.24521855742520515,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.3717,
+      "step": 2160
+    },
+    {
+      "epoch": 0.24533208453512423,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.378,
+      "step": 2161
+    },
+    {
+      "epoch": 0.2454456116450433,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3653,
+      "step": 2162
+    },
+    {
+      "epoch": 0.24555913875496238,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3694,
+      "step": 2163
+    },
+    {
+      "epoch": 0.24567266586488146,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3713,
+      "step": 2164
+    },
+    {
+      "epoch": 0.24578619297480053,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3894,
+      "step": 2165
+    },
+    {
+      "epoch": 0.2458997200847196,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3828,
+      "step": 2166
+    },
+    {
+      "epoch": 0.24601324719463868,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3866,
+      "step": 2167
+    },
+    {
+      "epoch": 0.24612677430455776,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3773,
+      "step": 2168
+    },
+    {
+      "epoch": 0.24624030141447684,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.3887,
+      "step": 2169
+    },
+    {
+      "epoch": 0.2463538285243959,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.3628,
+      "step": 2170
+    },
+    {
+      "epoch": 0.246467355634315,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.002,
+      "loss": 5.3883,
+      "step": 2171
+    },
+    {
+      "epoch": 0.24658088274423406,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.002,
+      "loss": 5.3939,
+      "step": 2172
+    },
+    {
+      "epoch": 0.24669440985415314,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.3899,
+      "step": 2173
+    },
+    {
+      "epoch": 0.24680793696407222,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.3814,
+      "step": 2174
+    },
+    {
+      "epoch": 0.2469214640739913,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.3869,
+      "step": 2175
+    },
+    {
+      "epoch": 0.24703499118391037,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.3761,
+      "step": 2176
+    },
+    {
+      "epoch": 0.24714851829382944,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.4045,
+      "step": 2177
+    },
+    {
+      "epoch": 0.24726204540374852,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.359,
+      "step": 2178
+    },
+    {
+      "epoch": 0.2473755725136676,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3757,
+      "step": 2179
+    },
+    {
+      "epoch": 0.24748909962358667,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.3918,
+      "step": 2180
+    },
+    {
+      "epoch": 0.24760262673350575,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.3606,
+      "step": 2181
+    },
+    {
+      "epoch": 0.24771615384342482,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.3714,
+      "step": 2182
+    },
+    {
+      "epoch": 0.2478296809533439,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3626,
+      "step": 2183
+    },
+    {
+      "epoch": 0.24794320806326298,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3666,
+      "step": 2184
+    },
+    {
+      "epoch": 0.24805673517318205,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.3706,
+      "step": 2185
+    },
+    {
+      "epoch": 0.24817026228310113,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3722,
+      "step": 2186
+    },
+    {
+      "epoch": 0.2482837893930202,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3818,
+      "step": 2187
+    },
+    {
+      "epoch": 0.24839731650293928,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.3888,
+      "step": 2188
+    },
+    {
+      "epoch": 0.24851084361285836,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.3918,
+      "step": 2189
+    },
+    {
+      "epoch": 0.24862437072277743,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.3825,
+      "step": 2190
+    },
+    {
+      "epoch": 0.2487378978326965,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.3804,
+      "step": 2191
+    },
+    {
+      "epoch": 0.24885142494261558,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.3776,
+      "step": 2192
+    },
+    {
+      "epoch": 0.24896495205253466,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.3857,
+      "step": 2193
+    },
+    {
+      "epoch": 0.24907847916245374,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.3733,
+      "step": 2194
+    },
+    {
+      "epoch": 0.2491920062723728,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.3628,
+      "step": 2195
+    },
+    {
+      "epoch": 0.2493055333822919,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3954,
+      "step": 2196
+    },
+    {
+      "epoch": 0.24941906049221096,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.38,
+      "step": 2197
+    },
+    {
+      "epoch": 0.24953258760213004,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.3786,
+      "step": 2198
+    },
+    {
+      "epoch": 0.24964611471204912,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.3641,
+      "step": 2199
+    },
+    {
+      "epoch": 0.2497596418219682,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3888,
+      "step": 2200
+    },
+    {
+      "epoch": 0.24987316893188727,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.3921,
+      "step": 2201
+    },
+    {
+      "epoch": 0.24998669604180634,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.3643,
+      "step": 2202
+    },
+    {
+      "epoch": 0.2501002231517254,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3762,
+      "step": 2203
+    },
+    {
+      "epoch": 0.2502137502616445,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3679,
+      "step": 2204
+    },
+    {
+      "epoch": 0.2503272773715636,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.358,
+      "step": 2205
+    },
+    {
+      "epoch": 0.25044080448148265,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.3759,
+      "step": 2206
+    },
+    {
+      "epoch": 0.2505543315914017,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.3565,
+      "step": 2207
+    },
+    {
+      "epoch": 0.2506678587013208,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3774,
+      "step": 2208
+    },
+    {
+      "epoch": 0.2507813858112399,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3796,
+      "step": 2209
+    },
+    {
+      "epoch": 0.25089491292115895,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3725,
+      "step": 2210
+    },
+    {
+      "epoch": 0.25100844003107803,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3791,
+      "step": 2211
+    },
+    {
+      "epoch": 0.2511219671409971,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3773,
+      "step": 2212
+    },
+    {
+      "epoch": 0.2512354942509162,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.3767,
+      "step": 2213
+    },
+    {
+      "epoch": 0.25134902136083526,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3853,
+      "step": 2214
+    },
+    {
+      "epoch": 0.25146254847075433,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.388,
+      "step": 2215
+    },
+    {
+      "epoch": 0.2515760755806734,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.3691,
+      "step": 2216
+    },
+    {
+      "epoch": 0.2516896026905925,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.3731,
+      "step": 2217
+    },
+    {
+      "epoch": 0.25180312980051156,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.4074,
+      "step": 2218
+    },
+    {
+      "epoch": 0.25191665691043064,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.3913,
+      "step": 2219
+    },
+    {
+      "epoch": 0.2520301840203497,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.3942,
+      "step": 2220
+    },
+    {
+      "epoch": 0.2521437111302688,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3758,
+      "step": 2221
+    },
+    {
+      "epoch": 0.25225723824018786,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.372,
+      "step": 2222
+    },
+    {
+      "epoch": 0.25237076535010694,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.3535,
+      "step": 2223
+    },
+    {
+      "epoch": 0.252484292460026,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.385,
+      "step": 2224
+    },
+    {
+      "epoch": 0.2525978195699451,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.3681,
+      "step": 2225
+    },
+    {
+      "epoch": 0.25271134667986417,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3894,
+      "step": 2226
+    },
+    {
+      "epoch": 0.25282487378978324,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.3829,
+      "step": 2227
+    },
+    {
+      "epoch": 0.2529384008997023,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3662,
+      "step": 2228
+    },
+    {
+      "epoch": 0.2530519280096214,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3852,
+      "step": 2229
+    },
+    {
+      "epoch": 0.2531654551195405,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3736,
+      "step": 2230
+    },
+    {
+      "epoch": 0.25327898222945955,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.3798,
+      "step": 2231
+    },
+    {
+      "epoch": 0.2533925093393786,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.3794,
+      "step": 2232
+    },
+    {
+      "epoch": 0.25350603644929776,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.3789,
+      "step": 2233
+    },
+    {
+      "epoch": 0.25361956355921683,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.3801,
+      "step": 2234
+    },
+    {
+      "epoch": 0.2537330906691359,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.3562,
+      "step": 2235
+    },
+    {
+      "epoch": 0.253846617779055,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.3646,
+      "step": 2236
+    },
+    {
+      "epoch": 0.25396014488897406,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.3798,
+      "step": 2237
+    },
+    {
+      "epoch": 0.25407367199889314,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.3857,
+      "step": 2238
+    },
+    {
+      "epoch": 0.2541871991088122,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.3707,
+      "step": 2239
+    },
+    {
+      "epoch": 0.2543007262187313,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 5.3886,
+      "step": 2240
+    },
+    {
+      "epoch": 0.25441425332865036,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.3673,
+      "step": 2241
+    },
+    {
+      "epoch": 0.25452778043856944,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.3728,
+      "step": 2242
+    },
+    {
+      "epoch": 0.2546413075484885,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.4026,
+      "step": 2243
+    },
+    {
+      "epoch": 0.2547548346584076,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3839,
+      "step": 2244
+    },
+    {
+      "epoch": 0.25486836176832667,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3835,
+      "step": 2245
+    },
+    {
+      "epoch": 0.25498188887824574,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.3616,
+      "step": 2246
+    },
+    {
+      "epoch": 0.2550954159881648,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3724,
+      "step": 2247
+    },
+    {
+      "epoch": 0.2552089430980839,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3684,
+      "step": 2248
+    },
+    {
+      "epoch": 0.25532247020800297,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3784,
+      "step": 2249
+    },
+    {
+      "epoch": 0.25543599731792205,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3811,
+      "step": 2250
+    },
+    {
+      "epoch": 0.2555495244278411,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3619,
+      "step": 2251
+    },
+    {
+      "epoch": 0.2556630515377602,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3966,
+      "step": 2252
+    },
+    {
+      "epoch": 0.2557765786476793,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3597,
+      "step": 2253
+    },
+    {
+      "epoch": 0.25589010575759835,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3571,
+      "step": 2254
+    },
+    {
+      "epoch": 0.2560036328675174,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.3814,
+      "step": 2255
+    },
+    {
+      "epoch": 0.2561171599774365,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.3683,
+      "step": 2256
+    },
+    {
+      "epoch": 0.2562306870873556,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.367,
+      "step": 2257
+    },
+    {
+      "epoch": 0.25634421419727466,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3781,
+      "step": 2258
+    },
+    {
+      "epoch": 0.25645774130719373,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.3797,
+      "step": 2259
+    },
+    {
+      "epoch": 0.2565712684171128,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.3751,
+      "step": 2260
+    },
+    {
+      "epoch": 0.2566847955270319,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3896,
+      "step": 2261
+    },
+    {
+      "epoch": 0.25679832263695096,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.3775,
+      "step": 2262
+    },
+    {
+      "epoch": 0.25691184974687004,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.002,
+      "loss": 5.3752,
+      "step": 2263
+    },
+    {
+      "epoch": 0.2570253768567891,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.3723,
+      "step": 2264
+    },
+    {
+      "epoch": 0.2571389039667082,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.3591,
+      "step": 2265
+    },
+    {
+      "epoch": 0.25725243107662726,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.3745,
+      "step": 2266
+    },
+    {
+      "epoch": 0.25736595818654634,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.3634,
+      "step": 2267
+    },
+    {
+      "epoch": 0.2574794852964654,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.3927,
+      "step": 2268
+    },
+    {
+      "epoch": 0.2575930124063845,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3695,
+      "step": 2269
+    },
+    {
+      "epoch": 0.25770653951630357,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.3797,
+      "step": 2270
+    },
+    {
+      "epoch": 0.25782006662622264,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.3603,
+      "step": 2271
+    },
+    {
+      "epoch": 0.2579335937361417,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.3724,
+      "step": 2272
+    },
+    {
+      "epoch": 0.2580471208460608,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.3726,
+      "step": 2273
+    },
+    {
+      "epoch": 0.25816064795597987,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.3601,
+      "step": 2274
+    },
+    {
+      "epoch": 0.25827417506589895,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3731,
+      "step": 2275
+    },
+    {
+      "epoch": 0.258387702175818,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3855,
+      "step": 2276
+    },
+    {
+      "epoch": 0.2585012292857371,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.3519,
+      "step": 2277
+    },
+    {
+      "epoch": 0.2586147563956562,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.3801,
+      "step": 2278
+    },
+    {
+      "epoch": 0.25872828350557525,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.002,
+      "loss": 5.3835,
+      "step": 2279
+    },
+    {
+      "epoch": 0.2588418106154943,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.002,
+      "loss": 5.3921,
+      "step": 2280
+    },
+    {
+      "epoch": 0.2589553377254134,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.3824,
+      "step": 2281
+    },
+    {
+      "epoch": 0.2590688648353325,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.3546,
+      "step": 2282
+    },
+    {
+      "epoch": 0.25918239194525156,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.3792,
+      "step": 2283
+    },
+    {
+      "epoch": 0.25929591905517063,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.002,
+      "loss": 5.3717,
+      "step": 2284
+    },
+    {
+      "epoch": 0.2594094461650897,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.3755,
+      "step": 2285
+    },
+    {
+      "epoch": 0.2595229732750088,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.3605,
+      "step": 2286
+    },
+    {
+      "epoch": 0.25963650038492786,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.3559,
+      "step": 2287
+    },
+    {
+      "epoch": 0.25975002749484694,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3695,
+      "step": 2288
+    },
+    {
+      "epoch": 0.259863554604766,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3625,
+      "step": 2289
+    },
+    {
+      "epoch": 0.2599770817146851,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3741,
+      "step": 2290
+    },
+    {
+      "epoch": 0.26009060882460416,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3482,
+      "step": 2291
+    },
+    {
+      "epoch": 0.26020413593452324,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3818,
+      "step": 2292
+    },
+    {
+      "epoch": 0.2603176630444423,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.3687,
+      "step": 2293
+    },
+    {
+      "epoch": 0.2604311901543614,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.3785,
+      "step": 2294
+    },
+    {
+      "epoch": 0.26054471726428047,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.3809,
+      "step": 2295
+    },
+    {
+      "epoch": 0.26065824437419954,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.3827,
+      "step": 2296
+    },
+    {
+      "epoch": 0.2607717714841186,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.3752,
+      "step": 2297
+    },
+    {
+      "epoch": 0.2608852985940377,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3729,
+      "step": 2298
+    },
+    {
+      "epoch": 0.26099882570395677,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3809,
+      "step": 2299
+    },
+    {
+      "epoch": 0.26111235281387585,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.3625,
+      "step": 2300
+    },
+    {
+      "epoch": 0.2612258799237949,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3655,
+      "step": 2301
+    },
+    {
+      "epoch": 0.261339407033714,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3651,
+      "step": 2302
+    },
+    {
+      "epoch": 0.2614529341436331,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3579,
+      "step": 2303
+    },
+    {
+      "epoch": 0.26156646125355215,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3849,
+      "step": 2304
+    },
+    {
+      "epoch": 0.2616799883634712,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3786,
+      "step": 2305
+    },
+    {
+      "epoch": 0.2617935154733903,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3629,
+      "step": 2306
+    },
+    {
+      "epoch": 0.2619070425833094,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.002,
+      "loss": 5.3586,
+      "step": 2307
+    },
+    {
+      "epoch": 0.26202056969322846,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.381,
+      "step": 2308
+    },
+    {
+      "epoch": 0.26213409680314753,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3641,
+      "step": 2309
+    },
+    {
+      "epoch": 0.2622476239130666,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.37,
+      "step": 2310
+    },
+    {
+      "epoch": 0.2623611510229857,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.393,
+      "step": 2311
+    },
+    {
+      "epoch": 0.26247467813290476,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3671,
+      "step": 2312
+    },
+    {
+      "epoch": 0.26258820524282384,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.3687,
+      "step": 2313
+    },
+    {
+      "epoch": 0.2627017323527429,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.3809,
+      "step": 2314
+    },
+    {
+      "epoch": 0.262815259462662,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.3712,
+      "step": 2315
+    },
+    {
+      "epoch": 0.26292878657258106,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.3449,
+      "step": 2316
+    },
+    {
+      "epoch": 0.26304231368250014,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.3721,
+      "step": 2317
+    },
+    {
+      "epoch": 0.2631558407924192,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.3897,
+      "step": 2318
+    },
+    {
+      "epoch": 0.2632693679023383,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.3531,
+      "step": 2319
+    },
+    {
+      "epoch": 0.26338289501225737,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.3663,
+      "step": 2320
+    },
+    {
+      "epoch": 0.26349642212217644,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3661,
+      "step": 2321
+    },
+    {
+      "epoch": 0.2636099492320955,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3477,
+      "step": 2322
+    },
+    {
+      "epoch": 0.2637234763420146,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3773,
+      "step": 2323
+    },
+    {
+      "epoch": 0.26383700345193367,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.3576,
+      "step": 2324
+    },
+    {
+      "epoch": 0.26395053056185275,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3664,
+      "step": 2325
+    },
+    {
+      "epoch": 0.2640640576717718,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3741,
+      "step": 2326
+    },
+    {
+      "epoch": 0.2641775847816909,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3485,
+      "step": 2327
+    },
+    {
+      "epoch": 0.26429111189161,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3624,
+      "step": 2328
+    },
+    {
+      "epoch": 0.26440463900152905,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3582,
+      "step": 2329
+    },
+    {
+      "epoch": 0.2645181661114481,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3683,
+      "step": 2330
+    },
+    {
+      "epoch": 0.2646316932213672,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3427,
+      "step": 2331
+    },
+    {
+      "epoch": 0.2647452203312863,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3618,
+      "step": 2332
+    },
+    {
+      "epoch": 0.26485874744120536,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3724,
+      "step": 2333
+    },
+    {
+      "epoch": 0.26497227455112443,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.3807,
+      "step": 2334
+    },
+    {
+      "epoch": 0.2650858016610435,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3608,
+      "step": 2335
+    },
+    {
+      "epoch": 0.2651993287709626,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.3692,
+      "step": 2336
+    },
+    {
+      "epoch": 0.26531285588088166,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.3816,
+      "step": 2337
+    },
+    {
+      "epoch": 0.26542638299080074,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.3728,
+      "step": 2338
+    },
+    {
+      "epoch": 0.2655399101007198,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.3493,
+      "step": 2339
+    },
+    {
+      "epoch": 0.2656534372106389,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.365,
+      "step": 2340
+    },
+    {
+      "epoch": 0.26576696432055796,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.3621,
+      "step": 2341
+    },
+    {
+      "epoch": 0.26588049143047704,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.3872,
+      "step": 2342
+    },
+    {
+      "epoch": 0.2659940185403961,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.356,
+      "step": 2343
+    },
+    {
+      "epoch": 0.2661075456503152,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.3688,
+      "step": 2344
+    },
+    {
+      "epoch": 0.26622107276023427,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.3744,
+      "step": 2345
+    },
+    {
+      "epoch": 0.26633459987015334,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.3645,
+      "step": 2346
+    },
+    {
+      "epoch": 0.2664481269800724,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3512,
+      "step": 2347
+    },
+    {
+      "epoch": 0.2665616540899915,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3718,
+      "step": 2348
+    },
+    {
+      "epoch": 0.26667518119991057,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3796,
+      "step": 2349
+    },
+    {
+      "epoch": 0.26678870830982965,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3652,
+      "step": 2350
+    },
+    {
+      "epoch": 0.2669022354197487,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 5.3977,
+      "step": 2351
+    },
+    {
+      "epoch": 0.2670157625296678,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.3615,
+      "step": 2352
+    },
+    {
+      "epoch": 0.26712928963958693,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3442,
+      "step": 2353
+    },
+    {
+      "epoch": 0.267242816749506,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3815,
+      "step": 2354
+    },
+    {
+      "epoch": 0.2673563438594251,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3595,
+      "step": 2355
+    },
+    {
+      "epoch": 0.26746987096934416,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.3547,
+      "step": 2356
+    },
+    {
+      "epoch": 0.26758339807926323,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3679,
+      "step": 2357
+    },
+    {
+      "epoch": 0.2676969251891823,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.355,
+      "step": 2358
+    },
+    {
+      "epoch": 0.2678104522991014,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.002,
+      "loss": 5.3619,
+      "step": 2359
+    },
+    {
+      "epoch": 0.26792397940902046,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.002,
+      "loss": 5.3802,
+      "step": 2360
+    },
+    {
+      "epoch": 0.26803750651893954,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.002,
+      "loss": 5.3428,
+      "step": 2361
+    },
+    {
+      "epoch": 0.2681510336288586,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.3605,
+      "step": 2362
+    },
+    {
+      "epoch": 0.2682645607387777,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3847,
+      "step": 2363
+    },
+    {
+      "epoch": 0.26837808784869677,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.3636,
+      "step": 2364
+    },
+    {
+      "epoch": 0.26849161495861584,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.3649,
+      "step": 2365
+    },
+    {
+      "epoch": 0.2686051420685349,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.3644,
+      "step": 2366
+    },
+    {
+      "epoch": 0.268718669178454,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.3542,
+      "step": 2367
+    },
+    {
+      "epoch": 0.26883219628837307,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.3625,
+      "step": 2368
+    },
+    {
+      "epoch": 0.26894572339829215,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.3676,
+      "step": 2369
+    },
+    {
+      "epoch": 0.2690592505082112,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3512,
+      "step": 2370
+    },
+    {
+      "epoch": 0.2691727776181303,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.3529,
+      "step": 2371
+    },
+    {
+      "epoch": 0.2692863047280494,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3475,
+      "step": 2372
+    },
+    {
+      "epoch": 0.26939983183796845,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3737,
+      "step": 2373
+    },
+    {
+      "epoch": 0.2695133589478875,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3523,
+      "step": 2374
+    },
+    {
+      "epoch": 0.2696268860578066,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.3521,
+      "step": 2375
+    },
+    {
+      "epoch": 0.2697404131677257,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.3541,
+      "step": 2376
+    },
+    {
+      "epoch": 0.26985394027764475,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3778,
+      "step": 2377
+    },
+    {
+      "epoch": 0.26996746738756383,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3716,
+      "step": 2378
+    },
+    {
+      "epoch": 0.2700809944974829,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.3657,
+      "step": 2379
+    },
+    {
+      "epoch": 0.270194521607402,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3762,
+      "step": 2380
+    },
+    {
+      "epoch": 0.27030804871732106,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.3495,
+      "step": 2381
+    },
+    {
+      "epoch": 0.27042157582724013,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3728,
+      "step": 2382
+    },
+    {
+      "epoch": 0.2705351029371592,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.3674,
+      "step": 2383
+    },
+    {
+      "epoch": 0.2706486300470783,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.3564,
+      "step": 2384
+    },
+    {
+      "epoch": 0.27076215715699736,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3791,
+      "step": 2385
+    },
+    {
+      "epoch": 0.27087568426691644,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.3709,
+      "step": 2386
+    },
+    {
+      "epoch": 0.2709892113768355,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3542,
+      "step": 2387
+    },
+    {
+      "epoch": 0.2711027384867546,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.3665,
+      "step": 2388
+    },
+    {
+      "epoch": 0.27121626559667367,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.3555,
+      "step": 2389
+    },
+    {
+      "epoch": 0.27132979270659274,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3479,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2714433198165118,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3534,
+      "step": 2391
+    },
+    {
+      "epoch": 0.2715568469264309,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3619,
+      "step": 2392
+    },
+    {
+      "epoch": 0.27167037403634997,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.368,
+      "step": 2393
+    },
+    {
+      "epoch": 0.27178390114626905,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.3835,
+      "step": 2394
+    },
+    {
+      "epoch": 0.2718974282561881,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3651,
+      "step": 2395
+    },
+    {
+      "epoch": 0.2720109553661072,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3643,
+      "step": 2396
+    },
+    {
+      "epoch": 0.2721244824760263,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3777,
+      "step": 2397
+    },
+    {
+      "epoch": 0.27223800958594535,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.3618,
+      "step": 2398
+    },
+    {
+      "epoch": 0.2723515366958644,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.3679,
+      "step": 2399
+    },
+    {
+      "epoch": 0.2724650638057835,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.3504,
+      "step": 2400
+    },
+    {
+      "epoch": 0.2725785909157026,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3545,
+      "step": 2401
+    },
+    {
+      "epoch": 0.27269211802562165,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.326,
+      "step": 2402
+    },
+    {
+      "epoch": 0.27280564513554073,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.3659,
+      "step": 2403
+    },
+    {
+      "epoch": 0.2729191722454598,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.002,
+      "loss": 5.3483,
+      "step": 2404
+    },
+    {
+      "epoch": 0.2730326993553789,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.002,
+      "loss": 5.3616,
+      "step": 2405
+    },
+    {
+      "epoch": 0.27314622646529796,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.002,
+      "loss": 5.3567,
+      "step": 2406
+    },
+    {
+      "epoch": 0.27325975357521703,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.002,
+      "loss": 5.3584,
+      "step": 2407
+    },
+    {
+      "epoch": 0.2733732806851361,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.3457,
+      "step": 2408
+    },
+    {
+      "epoch": 0.2734868077950552,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.3489,
+      "step": 2409
+    },
+    {
+      "epoch": 0.27360033490497426,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3529,
+      "step": 2410
+    },
+    {
+      "epoch": 0.27371386201489334,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3451,
+      "step": 2411
+    },
+    {
+      "epoch": 0.2738273891248124,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.3569,
+      "step": 2412
+    },
+    {
+      "epoch": 0.2739409162347315,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.3644,
+      "step": 2413
+    },
+    {
+      "epoch": 0.27405444334465057,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.3699,
+      "step": 2414
+    },
+    {
+      "epoch": 0.27416797045456964,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.335,
+      "step": 2415
+    },
+    {
+      "epoch": 0.2742814975644887,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3639,
+      "step": 2416
+    },
+    {
+      "epoch": 0.2743950246744078,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3572,
+      "step": 2417
+    },
+    {
+      "epoch": 0.27450855178432687,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3565,
+      "step": 2418
+    },
+    {
+      "epoch": 0.27462207889424595,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.3561,
+      "step": 2419
+    },
+    {
+      "epoch": 0.274735606004165,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3848,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2748491331140841,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.3925,
+      "step": 2421
+    },
+    {
+      "epoch": 0.2749626602240032,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.3607,
+      "step": 2422
+    },
+    {
+      "epoch": 0.27507618733392225,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.378,
+      "step": 2423
+    },
+    {
+      "epoch": 0.2751897144438413,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3582,
+      "step": 2424
+    },
+    {
+      "epoch": 0.2753032415537604,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3538,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2754167686636795,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3657,
+      "step": 2426
+    },
+    {
+      "epoch": 0.27553029577359855,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3457,
+      "step": 2427
+    },
+    {
+      "epoch": 0.27564382288351763,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3748,
+      "step": 2428
+    },
+    {
+      "epoch": 0.2757573499934367,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3652,
+      "step": 2429
+    },
+    {
+      "epoch": 0.2758708771033558,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.3376,
+      "step": 2430
+    },
+    {
+      "epoch": 0.27598440421327486,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3514,
+      "step": 2431
+    },
+    {
+      "epoch": 0.27609793132319393,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3831,
+      "step": 2432
+    },
+    {
+      "epoch": 0.276211458433113,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3605,
+      "step": 2433
+    },
+    {
+      "epoch": 0.2763249855430321,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.3646,
+      "step": 2434
+    },
+    {
+      "epoch": 0.27643851265295116,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.3773,
+      "step": 2435
+    },
+    {
+      "epoch": 0.27655203976287024,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.3634,
+      "step": 2436
+    },
+    {
+      "epoch": 0.2766655668727893,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3605,
+      "step": 2437
+    },
+    {
+      "epoch": 0.2767790939827084,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3524,
+      "step": 2438
+    },
+    {
+      "epoch": 0.27689262109262747,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.3584,
+      "step": 2439
+    },
+    {
+      "epoch": 0.27700614820254654,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.3535,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2771196753124656,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3707,
+      "step": 2441
+    },
+    {
+      "epoch": 0.2772332024223847,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.3441,
+      "step": 2442
+    },
+    {
+      "epoch": 0.27734672953230377,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.3677,
+      "step": 2443
+    },
+    {
+      "epoch": 0.27746025664222285,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.002,
+      "loss": 5.3463,
+      "step": 2444
+    },
+    {
+      "epoch": 0.2775737837521419,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.3485,
+      "step": 2445
+    },
+    {
+      "epoch": 0.277687310862061,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.3489,
+      "step": 2446
+    },
+    {
+      "epoch": 0.2778008379719801,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.3562,
+      "step": 2447
+    },
+    {
+      "epoch": 0.27791436508189915,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3534,
+      "step": 2448
+    },
+    {
+      "epoch": 0.2780278921918182,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.339,
+      "step": 2449
+    },
+    {
+      "epoch": 0.2781414193017373,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.3937,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2782549464116564,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3663,
+      "step": 2451
+    },
+    {
+      "epoch": 0.27836847352157545,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.355,
+      "step": 2452
+    },
+    {
+      "epoch": 0.27848200063149453,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3379,
+      "step": 2453
+    },
+    {
+      "epoch": 0.2785955277414136,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3444,
+      "step": 2454
+    },
+    {
+      "epoch": 0.2787090548513327,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.3576,
+      "step": 2455
+    },
+    {
+      "epoch": 0.27882258196125176,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3529,
+      "step": 2456
+    },
+    {
+      "epoch": 0.27893610907117083,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3495,
+      "step": 2457
+    },
+    {
+      "epoch": 0.2790496361810899,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.3736,
+      "step": 2458
+    },
+    {
+      "epoch": 0.279163163291009,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.3537,
+      "step": 2459
+    },
+    {
+      "epoch": 0.27927669040092806,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.3623,
+      "step": 2460
+    },
+    {
+      "epoch": 0.27939021751084714,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3488,
+      "step": 2461
+    },
+    {
+      "epoch": 0.2795037446207662,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.3732,
+      "step": 2462
+    },
+    {
+      "epoch": 0.2796172717306853,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.3481,
+      "step": 2463
+    },
+    {
+      "epoch": 0.27973079884060437,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3714,
+      "step": 2464
+    },
+    {
+      "epoch": 0.27984432595052344,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.3587,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2799578530604425,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3615,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2800713801703616,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3612,
+      "step": 2467
+    },
+    {
+      "epoch": 0.28018490728028067,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.3399,
+      "step": 2468
+    },
+    {
+      "epoch": 0.28029843439019975,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.3491,
+      "step": 2469
+    },
+    {
+      "epoch": 0.2804119615001188,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3483,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2805254886100379,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3832,
+      "step": 2471
+    },
+    {
+      "epoch": 0.280639015719957,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.3651,
+      "step": 2472
+    },
+    {
+      "epoch": 0.28075254282987605,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.3485,
+      "step": 2473
+    },
+    {
+      "epoch": 0.2808660699397952,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.3514,
+      "step": 2474
+    },
+    {
+      "epoch": 0.28097959704971426,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3549,
+      "step": 2475
+    },
+    {
+      "epoch": 0.28109312415963333,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.3659,
+      "step": 2476
+    },
+    {
+      "epoch": 0.2812066512695524,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3705,
+      "step": 2477
+    },
+    {
+      "epoch": 0.2813201783794715,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.3244,
+      "step": 2478
+    },
+    {
+      "epoch": 0.28143370548939056,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3508,
+      "step": 2479
+    },
+    {
+      "epoch": 0.28154723259930964,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.3657,
+      "step": 2480
+    },
+    {
+      "epoch": 0.2816607597092287,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.3507,
+      "step": 2481
+    },
+    {
+      "epoch": 0.2817742868191478,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.3579,
+      "step": 2482
+    },
+    {
+      "epoch": 0.28188781392906687,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3825,
+      "step": 2483
+    },
+    {
+      "epoch": 0.28200134103898594,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.3445,
+      "step": 2484
+    },
+    {
+      "epoch": 0.282114868148905,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.3444,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2822283952588241,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.3659,
+      "step": 2486
+    },
+    {
+      "epoch": 0.28234192236874317,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3711,
+      "step": 2487
+    },
+    {
+      "epoch": 0.28245544947866225,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3682,
+      "step": 2488
+    },
+    {
+      "epoch": 0.2825689765885813,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.36,
+      "step": 2489
+    },
+    {
+      "epoch": 0.2826825036985004,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.3365,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2827960308084195,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.3482,
+      "step": 2491
+    },
+    {
+      "epoch": 0.28290955791833855,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3279,
+      "step": 2492
+    },
+    {
+      "epoch": 0.2830230850282576,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3502,
+      "step": 2493
+    },
+    {
+      "epoch": 0.2831366121381767,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.3568,
+      "step": 2494
+    },
+    {
+      "epoch": 0.2832501392480958,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3691,
+      "step": 2495
+    },
+    {
+      "epoch": 0.28336366635801485,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.3665,
+      "step": 2496
+    },
+    {
+      "epoch": 0.28347719346793393,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3511,
+      "step": 2497
+    },
+    {
+      "epoch": 0.283590720577853,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3583,
+      "step": 2498
+    },
+    {
+      "epoch": 0.2837042476877721,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.356,
+      "step": 2499
+    },
+    {
+      "epoch": 0.28381777479769116,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3577,
+      "step": 2500
+    },
+    {
+      "epoch": 0.28393130190761023,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3445,
+      "step": 2501
+    },
+    {
+      "epoch": 0.2840448290175293,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3587,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2841583561274484,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3454,
+      "step": 2503
+    },
+    {
+      "epoch": 0.28427188323736746,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.3489,
+      "step": 2504
+    },
+    {
+      "epoch": 0.28438541034728654,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3634,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2844989374572056,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3582,
+      "step": 2506
+    },
+    {
+      "epoch": 0.2846124645671247,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.3541,
+      "step": 2507
+    },
+    {
+      "epoch": 0.28472599167704377,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 5.3565,
+      "step": 2508
+    },
+    {
+      "epoch": 0.28483951878696284,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3445,
+      "step": 2509
+    },
+    {
+      "epoch": 0.2849530458968819,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3612,
+      "step": 2510
+    },
+    {
+      "epoch": 0.285066573006801,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3661,
+      "step": 2511
+    },
+    {
+      "epoch": 0.28518010011672007,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3323,
+      "step": 2512
+    },
+    {
+      "epoch": 0.28529362722663915,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3676,
+      "step": 2513
+    },
+    {
+      "epoch": 0.2854071543365582,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.3548,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2855206814464773,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.3656,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2856342085563964,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.3388,
+      "step": 2516
+    },
+    {
+      "epoch": 0.28574773566631545,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.3548,
+      "step": 2517
+    },
+    {
+      "epoch": 0.2858612627762345,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.37,
+      "step": 2518
+    },
+    {
+      "epoch": 0.2859747898861536,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3515,
+      "step": 2519
+    },
+    {
+      "epoch": 0.2860883169960727,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3598,
+      "step": 2520
+    },
+    {
+      "epoch": 0.28620184410599175,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3482,
+      "step": 2521
+    },
+    {
+      "epoch": 0.28631537121591083,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3534,
+      "step": 2522
+    },
+    {
+      "epoch": 0.2864288983258299,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.3576,
+      "step": 2523
+    },
+    {
+      "epoch": 0.286542425435749,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.353,
+      "step": 2524
+    },
+    {
+      "epoch": 0.28665595254566806,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.359,
+      "step": 2525
+    },
+    {
+      "epoch": 0.28676947965558713,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3438,
+      "step": 2526
+    },
+    {
+      "epoch": 0.2868830067655062,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.3538,
+      "step": 2527
+    },
+    {
+      "epoch": 0.2869965338754253,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3563,
+      "step": 2528
+    },
+    {
+      "epoch": 0.28711006098534436,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3587,
+      "step": 2529
+    },
+    {
+      "epoch": 0.28722358809526344,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3341,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2873371152051825,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.3527,
+      "step": 2531
+    },
+    {
+      "epoch": 0.2874506423151016,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.3506,
+      "step": 2532
+    },
+    {
+      "epoch": 0.28756416942502067,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.3655,
+      "step": 2533
+    },
+    {
+      "epoch": 0.28767769653493974,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.3326,
+      "step": 2534
+    },
+    {
+      "epoch": 0.2877912236448588,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.3591,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2879047507547779,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3472,
+      "step": 2536
+    },
+    {
+      "epoch": 0.28801827786469697,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3472,
+      "step": 2537
+    },
+    {
+      "epoch": 0.28813180497461605,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3494,
+      "step": 2538
+    },
+    {
+      "epoch": 0.2882453320845351,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.3477,
+      "step": 2539
+    },
+    {
+      "epoch": 0.2883588591944542,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.002,
+      "loss": 5.3513,
+      "step": 2540
+    },
+    {
+      "epoch": 0.2884723863043733,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.002,
+      "loss": 5.3572,
+      "step": 2541
+    },
+    {
+      "epoch": 0.28858591341429235,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.3518,
+      "step": 2542
+    },
+    {
+      "epoch": 0.2886994405242114,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.3464,
+      "step": 2543
+    },
+    {
+      "epoch": 0.2888129676341305,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3441,
+      "step": 2544
+    },
+    {
+      "epoch": 0.2889264947440496,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3555,
+      "step": 2545
+    },
+    {
+      "epoch": 0.28904002185396865,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.3647,
+      "step": 2546
+    },
+    {
+      "epoch": 0.28915354896388773,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.3614,
+      "step": 2547
+    },
+    {
+      "epoch": 0.2892670760738068,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.3384,
+      "step": 2548
+    },
+    {
+      "epoch": 0.2893806031837259,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.3661,
+      "step": 2549
+    },
+    {
+      "epoch": 0.28949413029364496,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3477,
+      "step": 2550
+    },
+    {
+      "epoch": 0.28960765740356403,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3668,
+      "step": 2551
+    },
+    {
+      "epoch": 0.2897211845134831,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.3419,
+      "step": 2552
+    },
+    {
+      "epoch": 0.2898347116234022,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3398,
+      "step": 2553
+    },
+    {
+      "epoch": 0.28994823873332126,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3422,
+      "step": 2554
+    },
+    {
+      "epoch": 0.29006176584324034,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.3485,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2901752929531594,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3351,
+      "step": 2556
+    },
+    {
+      "epoch": 0.2902888200630785,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.3459,
+      "step": 2557
+    },
+    {
+      "epoch": 0.29040234717299757,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3607,
+      "step": 2558
+    },
+    {
+      "epoch": 0.29051587428291664,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3386,
+      "step": 2559
+    },
+    {
+      "epoch": 0.2906294013928357,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.361,
+      "step": 2560
+    },
+    {
+      "epoch": 0.2907429285027548,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3316,
+      "step": 2561
+    },
+    {
+      "epoch": 0.29085645561267387,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.3541,
+      "step": 2562
+    },
+    {
+      "epoch": 0.29096998272259295,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3345,
+      "step": 2563
+    },
+    {
+      "epoch": 0.291083509832512,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.3556,
+      "step": 2564
+    },
+    {
+      "epoch": 0.2911970369424311,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 5.3455,
+      "step": 2565
+    },
+    {
+      "epoch": 0.2913105640523502,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.3541,
+      "step": 2566
+    },
+    {
+      "epoch": 0.29142409116226925,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.3269,
+      "step": 2567
+    },
+    {
+      "epoch": 0.2915376182721883,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3498,
+      "step": 2568
+    },
+    {
+      "epoch": 0.2916511453821074,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.361,
+      "step": 2569
+    },
+    {
+      "epoch": 0.2917646724920265,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.3471,
+      "step": 2570
+    },
+    {
+      "epoch": 0.29187819960194555,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.3524,
+      "step": 2571
+    },
+    {
+      "epoch": 0.29199172671186463,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3462,
+      "step": 2572
+    },
+    {
+      "epoch": 0.2921052538217837,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3409,
+      "step": 2573
+    },
+    {
+      "epoch": 0.2922187809317028,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.3553,
+      "step": 2574
+    },
+    {
+      "epoch": 0.29233230804162186,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.3301,
+      "step": 2575
+    },
+    {
+      "epoch": 0.29244583515154093,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.3471,
+      "step": 2576
+    },
+    {
+      "epoch": 0.29255936226146,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3513,
+      "step": 2577
+    },
+    {
+      "epoch": 0.2926728893713791,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.356,
+      "step": 2578
+    },
+    {
+      "epoch": 0.29278641648129816,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.373,
+      "step": 2579
+    },
+    {
+      "epoch": 0.29289994359121724,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.342,
+      "step": 2580
+    },
+    {
+      "epoch": 0.2930134707011363,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.3498,
+      "step": 2581
+    },
+    {
+      "epoch": 0.2931269978110554,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.3701,
+      "step": 2582
+    },
+    {
+      "epoch": 0.29324052492097447,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.3586,
+      "step": 2583
+    },
+    {
+      "epoch": 0.29335405203089354,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.3188,
+      "step": 2584
+    },
+    {
+      "epoch": 0.2934675791408126,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3679,
+      "step": 2585
+    },
+    {
+      "epoch": 0.2935811062507317,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3458,
+      "step": 2586
+    },
+    {
+      "epoch": 0.29369463336065077,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3558,
+      "step": 2587
+    },
+    {
+      "epoch": 0.29380816047056985,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.331,
+      "step": 2588
+    },
+    {
+      "epoch": 0.2939216875804889,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.3614,
+      "step": 2589
+    },
+    {
+      "epoch": 0.294035214690408,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.3502,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2941487418003271,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.3417,
+      "step": 2591
+    },
+    {
+      "epoch": 0.29426226891024615,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.3353,
+      "step": 2592
+    },
+    {
+      "epoch": 0.2943757960201652,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.3257,
+      "step": 2593
+    },
+    {
+      "epoch": 0.2944893231300843,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3525,
+      "step": 2594
+    },
+    {
+      "epoch": 0.29460285024000343,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3291,
+      "step": 2595
+    },
+    {
+      "epoch": 0.2947163773499225,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.349,
+      "step": 2596
+    },
+    {
+      "epoch": 0.2948299044598416,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.345,
+      "step": 2597
+    },
+    {
+      "epoch": 0.29494343156976066,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3508,
+      "step": 2598
+    },
+    {
+      "epoch": 0.29505695867967974,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.3662,
+      "step": 2599
+    },
+    {
+      "epoch": 0.2951704857895988,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.3447,
+      "step": 2600
+    },
+    {
+      "epoch": 0.2952840128995179,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.3273,
+      "step": 2601
+    },
+    {
+      "epoch": 0.29539754000943697,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.3454,
+      "step": 2602
+    },
+    {
+      "epoch": 0.29551106711935604,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3529,
+      "step": 2603
+    },
+    {
+      "epoch": 0.2956245942292751,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.3538,
+      "step": 2604
+    },
+    {
+      "epoch": 0.2957381213391942,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.359,
+      "step": 2605
+    },
+    {
+      "epoch": 0.29585164844911327,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3371,
+      "step": 2606
+    },
+    {
+      "epoch": 0.29596517555903235,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3317,
+      "step": 2607
+    },
+    {
+      "epoch": 0.2960787026689514,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.339,
+      "step": 2608
+    },
+    {
+      "epoch": 0.2961922297788705,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.3311,
+      "step": 2609
+    },
+    {
+      "epoch": 0.2963057568887896,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3362,
+      "step": 2610
+    },
+    {
+      "epoch": 0.29641928399870865,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3661,
+      "step": 2611
+    },
+    {
+      "epoch": 0.2965328111086277,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 5.3678,
+      "step": 2612
+    },
+    {
+      "epoch": 0.2966463382185468,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3438,
+      "step": 2613
+    },
+    {
+      "epoch": 0.2967598653284659,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.3374,
+      "step": 2614
+    },
+    {
+      "epoch": 0.29687339243838495,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3207,
+      "step": 2615
+    },
+    {
+      "epoch": 0.29698691954830403,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.327,
+      "step": 2616
+    },
+    {
+      "epoch": 0.2971004466582231,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.3606,
+      "step": 2617
+    },
+    {
+      "epoch": 0.2972139737681422,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.3338,
+      "step": 2618
+    },
+    {
+      "epoch": 0.29732750087806126,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3313,
+      "step": 2619
+    },
+    {
+      "epoch": 0.29744102798798033,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3456,
+      "step": 2620
+    },
+    {
+      "epoch": 0.2975545550978994,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.33,
+      "step": 2621
+    },
+    {
+      "epoch": 0.2976680822078185,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.3694,
+      "step": 2622
+    },
+    {
+      "epoch": 0.29778160931773756,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.3583,
+      "step": 2623
+    },
+    {
+      "epoch": 0.29789513642765664,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.3354,
+      "step": 2624
+    },
+    {
+      "epoch": 0.2980086635375757,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3299,
+      "step": 2625
+    },
+    {
+      "epoch": 0.2981221906474948,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.3533,
+      "step": 2626
+    },
+    {
+      "epoch": 0.29823571775741387,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.3374,
+      "step": 2627
+    },
+    {
+      "epoch": 0.29834924486733294,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3585,
+      "step": 2628
+    },
+    {
+      "epoch": 0.298462771977252,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.341,
+      "step": 2629
+    },
+    {
+      "epoch": 0.2985762990871711,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.3304,
+      "step": 2630
+    },
+    {
+      "epoch": 0.29868982619709017,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.3343,
+      "step": 2631
+    },
+    {
+      "epoch": 0.29880335330700925,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.3329,
+      "step": 2632
+    },
+    {
+      "epoch": 0.2989168804169283,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.3447,
+      "step": 2633
+    },
+    {
+      "epoch": 0.2990304075268474,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.3568,
+      "step": 2634
+    },
+    {
+      "epoch": 0.2991439346367665,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.3096,
+      "step": 2635
+    },
+    {
+      "epoch": 0.29925746174668555,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.3422,
+      "step": 2636
+    },
+    {
+      "epoch": 0.2993709888566046,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.3535,
+      "step": 2637
+    },
+    {
+      "epoch": 0.2994845159665237,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3548,
+      "step": 2638
+    },
+    {
+      "epoch": 0.2995980430764428,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3577,
+      "step": 2639
+    },
+    {
+      "epoch": 0.29971157018636185,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.3489,
+      "step": 2640
+    },
+    {
+      "epoch": 0.29982509729628093,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.3419,
+      "step": 2641
+    },
+    {
+      "epoch": 0.2999386244062,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.324,
+      "step": 2642
+    },
+    {
+      "epoch": 0.3000521515161191,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.3465,
+      "step": 2643
+    },
+    {
+      "epoch": 0.30016567862603816,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3479,
+      "step": 2644
+    },
+    {
+      "epoch": 0.30027920573595723,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 5.3225,
+      "step": 2645
+    },
+    {
+      "epoch": 0.3003927328458763,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 5.3508,
+      "step": 2646
+    },
+    {
+      "epoch": 0.3005062599557954,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3504,
+      "step": 2647
+    },
+    {
+      "epoch": 0.30061978706571446,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.336,
+      "step": 2648
+    },
+    {
+      "epoch": 0.30073331417563354,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3425,
+      "step": 2649
+    },
+    {
+      "epoch": 0.3008468412855526,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3459,
+      "step": 2650
+    },
+    {
+      "epoch": 0.3009603683954717,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3418,
+      "step": 2651
+    },
+    {
+      "epoch": 0.30107389550539077,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3524,
+      "step": 2652
+    },
+    {
+      "epoch": 0.30118742261530984,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3242,
+      "step": 2653
+    },
+    {
+      "epoch": 0.3013009497252289,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3563,
+      "step": 2654
+    },
+    {
+      "epoch": 0.301414476835148,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3462,
+      "step": 2655
+    },
+    {
+      "epoch": 0.30152800394506707,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3461,
+      "step": 2656
+    },
+    {
+      "epoch": 0.30164153105498615,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.345,
+      "step": 2657
+    },
+    {
+      "epoch": 0.3017550581649052,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.3436,
+      "step": 2658
+    },
+    {
+      "epoch": 0.3018685852748243,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.3532,
+      "step": 2659
+    },
+    {
+      "epoch": 0.3019821123847434,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3572,
+      "step": 2660
+    },
+    {
+      "epoch": 0.30209563949466245,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3438,
+      "step": 2661
+    },
+    {
+      "epoch": 0.3022091666045815,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.3724,
+      "step": 2662
+    },
+    {
+      "epoch": 0.3023226937145006,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3409,
+      "step": 2663
+    },
+    {
+      "epoch": 0.3024362208244197,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.339,
+      "step": 2664
+    },
+    {
+      "epoch": 0.30254974793433875,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3297,
+      "step": 2665
+    },
+    {
+      "epoch": 0.30266327504425783,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3289,
+      "step": 2666
+    },
+    {
+      "epoch": 0.3027768021541769,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.3585,
+      "step": 2667
+    },
+    {
+      "epoch": 0.302890329264096,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.3271,
+      "step": 2668
+    },
+    {
+      "epoch": 0.30300385637401506,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.3465,
+      "step": 2669
+    },
+    {
+      "epoch": 0.30311738348393413,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.3523,
+      "step": 2670
+    },
+    {
+      "epoch": 0.3032309105938532,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3516,
+      "step": 2671
+    },
+    {
+      "epoch": 0.3033444377037723,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3454,
+      "step": 2672
+    },
+    {
+      "epoch": 0.30345796481369136,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.339,
+      "step": 2673
+    },
+    {
+      "epoch": 0.30357149192361044,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.3395,
+      "step": 2674
+    },
+    {
+      "epoch": 0.3036850190335295,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3396,
+      "step": 2675
+    },
+    {
+      "epoch": 0.3037985461434486,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3409,
+      "step": 2676
+    },
+    {
+      "epoch": 0.30391207325336766,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3437,
+      "step": 2677
+    },
+    {
+      "epoch": 0.30402560036328674,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3397,
+      "step": 2678
+    },
+    {
+      "epoch": 0.3041391274732058,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.3367,
+      "step": 2679
+    },
+    {
+      "epoch": 0.3042526545831249,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.002,
+      "loss": 5.3408,
+      "step": 2680
+    },
+    {
+      "epoch": 0.30436618169304397,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.342,
+      "step": 2681
+    },
+    {
+      "epoch": 0.30447970880296304,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.351,
+      "step": 2682
+    },
+    {
+      "epoch": 0.3045932359128821,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.3537,
+      "step": 2683
+    },
+    {
+      "epoch": 0.3047067630228012,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.3009,
+      "step": 2684
+    },
+    {
+      "epoch": 0.3048202901327203,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.3364,
+      "step": 2685
+    },
+    {
+      "epoch": 0.30493381724263935,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3383,
+      "step": 2686
+    },
+    {
+      "epoch": 0.3050473443525584,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3537,
+      "step": 2687
+    },
+    {
+      "epoch": 0.3051608714624775,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.353,
+      "step": 2688
+    },
+    {
+      "epoch": 0.3052743985723966,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3661,
+      "step": 2689
+    },
+    {
+      "epoch": 0.30538792568231565,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.3359,
+      "step": 2690
+    },
+    {
+      "epoch": 0.30550145279223473,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.3512,
+      "step": 2691
+    },
+    {
+      "epoch": 0.3056149799021538,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.345,
+      "step": 2692
+    },
+    {
+      "epoch": 0.3057285070120729,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.343,
+      "step": 2693
+    },
+    {
+      "epoch": 0.30584203412199196,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.3313,
+      "step": 2694
+    },
+    {
+      "epoch": 0.30595556123191103,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.3352,
+      "step": 2695
+    },
+    {
+      "epoch": 0.3060690883418301,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.3463,
+      "step": 2696
+    },
+    {
+      "epoch": 0.3061826154517492,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.365,
+      "step": 2697
+    },
+    {
+      "epoch": 0.30629614256166826,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 5.3402,
+      "step": 2698
+    },
+    {
+      "epoch": 0.30640966967158734,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 5.3273,
+      "step": 2699
+    },
+    {
+      "epoch": 0.3065231967815064,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.002,
+      "loss": 5.3307,
+      "step": 2700
+    },
+    {
+      "epoch": 0.3066367238914255,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 5.3397,
+      "step": 2701
+    },
+    {
+      "epoch": 0.30675025100134456,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3251,
+      "step": 2702
+    },
+    {
+      "epoch": 0.30686377811126364,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.3497,
+      "step": 2703
+    },
+    {
+      "epoch": 0.3069773052211827,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.3403,
+      "step": 2704
+    },
+    {
+      "epoch": 0.3070908323311018,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.3256,
+      "step": 2705
+    },
+    {
+      "epoch": 0.30720435944102087,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.3561,
+      "step": 2706
+    },
+    {
+      "epoch": 0.30731788655093994,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3235,
+      "step": 2707
+    },
+    {
+      "epoch": 0.307431413660859,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3304,
+      "step": 2708
+    },
+    {
+      "epoch": 0.3075449407707781,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.342,
+      "step": 2709
+    },
+    {
+      "epoch": 0.3076584678806972,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3503,
+      "step": 2710
+    },
+    {
+      "epoch": 0.30777199499061625,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.3484,
+      "step": 2711
+    },
+    {
+      "epoch": 0.3078855221005353,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.3268,
+      "step": 2712
+    },
+    {
+      "epoch": 0.3079990492104544,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.3628,
+      "step": 2713
+    },
+    {
+      "epoch": 0.3081125763203735,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3616,
+      "step": 2714
+    },
+    {
+      "epoch": 0.3082261034302926,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3459,
+      "step": 2715
+    },
+    {
+      "epoch": 0.3083396305402117,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.3393,
+      "step": 2716
+    },
+    {
+      "epoch": 0.30845315765013076,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.3488,
+      "step": 2717
+    },
+    {
+      "epoch": 0.30856668476004984,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.3409,
+      "step": 2718
+    },
+    {
+      "epoch": 0.3086802118699689,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.3536,
+      "step": 2719
+    },
+    {
+      "epoch": 0.308793738979888,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.3303,
+      "step": 2720
+    },
+    {
+      "epoch": 0.30890726608980706,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3454,
+      "step": 2721
+    },
+    {
+      "epoch": 0.30902079319972614,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.3333,
+      "step": 2722
+    },
+    {
+      "epoch": 0.3091343203096452,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.3361,
+      "step": 2723
+    },
+    {
+      "epoch": 0.3092478474195643,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.3314,
+      "step": 2724
+    },
+    {
+      "epoch": 0.30936137452948337,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.3335,
+      "step": 2725
+    },
+    {
+      "epoch": 0.30947490163940244,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.3457,
+      "step": 2726
+    },
+    {
+      "epoch": 0.3095884287493215,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.3325,
+      "step": 2727
+    },
+    {
+      "epoch": 0.3097019558592406,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.323,
+      "step": 2728
+    },
+    {
+      "epoch": 0.3098154829691597,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3496,
+      "step": 2729
+    },
+    {
+      "epoch": 0.30992901007907875,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3345,
+      "step": 2730
+    },
+    {
+      "epoch": 0.3100425371889978,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3401,
+      "step": 2731
+    },
+    {
+      "epoch": 0.3101560642989169,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.348,
+      "step": 2732
+    },
+    {
+      "epoch": 0.310269591408836,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3617,
+      "step": 2733
+    },
+    {
+      "epoch": 0.31038311851875505,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3244,
+      "step": 2734
+    },
+    {
+      "epoch": 0.31049664562867413,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3384,
+      "step": 2735
+    },
+    {
+      "epoch": 0.3106101727385932,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3517,
+      "step": 2736
+    },
+    {
+      "epoch": 0.3107236998485123,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3469,
+      "step": 2737
+    },
+    {
+      "epoch": 0.31083722695843136,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.326,
+      "step": 2738
+    },
+    {
+      "epoch": 0.31095075406835043,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3233,
+      "step": 2739
+    },
+    {
+      "epoch": 0.3110642811782695,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.3316,
+      "step": 2740
+    },
+    {
+      "epoch": 0.3111778082881886,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.3433,
+      "step": 2741
+    },
+    {
+      "epoch": 0.31129133539810766,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3288,
+      "step": 2742
+    },
+    {
+      "epoch": 0.31140486250802674,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3583,
+      "step": 2743
+    },
+    {
+      "epoch": 0.3115183896179458,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3372,
+      "step": 2744
+    },
+    {
+      "epoch": 0.3116319167278649,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3423,
+      "step": 2745
+    },
+    {
+      "epoch": 0.31174544383778396,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3474,
+      "step": 2746
+    },
+    {
+      "epoch": 0.31185897094770304,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.345,
+      "step": 2747
+    },
+    {
+      "epoch": 0.3119724980576221,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.3454,
+      "step": 2748
+    },
+    {
+      "epoch": 0.3120860251675412,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.3448,
+      "step": 2749
+    },
+    {
+      "epoch": 0.31219955227746027,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.322,
+      "step": 2750
+    },
+    {
+      "epoch": 0.31231307938737934,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.325,
+      "step": 2751
+    },
+    {
+      "epoch": 0.3124266064972984,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3631,
+      "step": 2752
+    },
+    {
+      "epoch": 0.3125401336072175,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3105,
+      "step": 2753
+    },
+    {
+      "epoch": 0.3126536607171366,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.339,
+      "step": 2754
+    },
+    {
+      "epoch": 0.31276718782705565,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3593,
+      "step": 2755
+    },
+    {
+      "epoch": 0.3128807149369747,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.354,
+      "step": 2756
+    },
+    {
+      "epoch": 0.3129942420468938,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3314,
+      "step": 2757
+    },
+    {
+      "epoch": 0.3131077691568129,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3324,
+      "step": 2758
+    },
+    {
+      "epoch": 0.31322129626673195,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.335,
+      "step": 2759
+    },
+    {
+      "epoch": 0.31333482337665103,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3499,
+      "step": 2760
+    },
+    {
+      "epoch": 0.3134483504865701,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3374,
+      "step": 2761
+    },
+    {
+      "epoch": 0.3135618775964892,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3346,
+      "step": 2762
+    },
+    {
+      "epoch": 0.31367540470640826,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3417,
+      "step": 2763
+    },
+    {
+      "epoch": 0.31378893181632733,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.337,
+      "step": 2764
+    },
+    {
+      "epoch": 0.3139024589262464,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.3368,
+      "step": 2765
+    },
+    {
+      "epoch": 0.3140159860361655,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.3337,
+      "step": 2766
+    },
+    {
+      "epoch": 0.31412951314608456,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.3574,
+      "step": 2767
+    },
+    {
+      "epoch": 0.31424304025600364,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3211,
+      "step": 2768
+    },
+    {
+      "epoch": 0.3143565673659227,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.3289,
+      "step": 2769
+    },
+    {
+      "epoch": 0.3144700944758418,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.3371,
+      "step": 2770
+    },
+    {
+      "epoch": 0.31458362158576086,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3471,
+      "step": 2771
+    },
+    {
+      "epoch": 0.31469714869567994,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3428,
+      "step": 2772
+    },
+    {
+      "epoch": 0.314810675805599,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.3197,
+      "step": 2773
+    },
+    {
+      "epoch": 0.3149242029155181,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3241,
+      "step": 2774
+    },
+    {
+      "epoch": 0.31503773002543717,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3403,
+      "step": 2775
+    },
+    {
+      "epoch": 0.31515125713535624,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3214,
+      "step": 2776
+    },
+    {
+      "epoch": 0.3152647842452753,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.32,
+      "step": 2777
+    },
+    {
+      "epoch": 0.3153783113551944,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3383,
+      "step": 2778
+    },
+    {
+      "epoch": 0.31549183846511347,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3028,
+      "step": 2779
+    },
+    {
+      "epoch": 0.31560536557503255,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3259,
+      "step": 2780
+    },
+    {
+      "epoch": 0.3157188926849516,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.339,
+      "step": 2781
+    },
+    {
+      "epoch": 0.3158324197948707,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.3177,
+      "step": 2782
+    },
+    {
+      "epoch": 0.3159459469047898,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3406,
+      "step": 2783
+    },
+    {
+      "epoch": 0.31605947401470885,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3191,
+      "step": 2784
+    },
+    {
+      "epoch": 0.31617300112462793,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.3351,
+      "step": 2785
+    },
+    {
+      "epoch": 0.316286528234547,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3275,
+      "step": 2786
+    },
+    {
+      "epoch": 0.3164000553444661,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3255,
+      "step": 2787
+    },
+    {
+      "epoch": 0.31651358245438516,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3377,
+      "step": 2788
+    },
+    {
+      "epoch": 0.31662710956430423,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3617,
+      "step": 2789
+    },
+    {
+      "epoch": 0.3167406366742233,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2955,
+      "step": 2790
+    },
+    {
+      "epoch": 0.3168541637841424,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3278,
+      "step": 2791
+    },
+    {
+      "epoch": 0.31696769089406146,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.349,
+      "step": 2792
+    },
+    {
+      "epoch": 0.31708121800398054,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.342,
+      "step": 2793
+    },
+    {
+      "epoch": 0.3171947451138996,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3315,
+      "step": 2794
+    },
+    {
+      "epoch": 0.3173082722238187,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.3461,
+      "step": 2795
+    },
+    {
+      "epoch": 0.31742179933373776,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3226,
+      "step": 2796
+    },
+    {
+      "epoch": 0.31753532644365684,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3329,
+      "step": 2797
+    },
+    {
+      "epoch": 0.3176488535535759,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3572,
+      "step": 2798
+    },
+    {
+      "epoch": 0.317762380663495,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.3387,
+      "step": 2799
+    },
+    {
+      "epoch": 0.31787590777341407,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.3468,
+      "step": 2800
+    },
+    {
+      "epoch": 0.31798943488333314,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.3465,
+      "step": 2801
+    },
+    {
+      "epoch": 0.3181029619932522,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3266,
+      "step": 2802
+    },
+    {
+      "epoch": 0.3182164891031713,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3318,
+      "step": 2803
+    },
+    {
+      "epoch": 0.31833001621309037,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.3378,
+      "step": 2804
+    },
+    {
+      "epoch": 0.31844354332300945,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.3332,
+      "step": 2805
+    },
+    {
+      "epoch": 0.3185570704329285,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.3283,
+      "step": 2806
+    },
+    {
+      "epoch": 0.3186705975428476,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.329,
+      "step": 2807
+    },
+    {
+      "epoch": 0.3187841246527667,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.3282,
+      "step": 2808
+    },
+    {
+      "epoch": 0.31889765176268575,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3157,
+      "step": 2809
+    },
+    {
+      "epoch": 0.31901117887260483,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.3247,
+      "step": 2810
+    },
+    {
+      "epoch": 0.3191247059825239,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3271,
+      "step": 2811
+    },
+    {
+      "epoch": 0.319238233092443,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3559,
+      "step": 2812
+    },
+    {
+      "epoch": 0.31935176020236206,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.316,
+      "step": 2813
+    },
+    {
+      "epoch": 0.31946528731228113,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.324,
+      "step": 2814
+    },
+    {
+      "epoch": 0.3195788144222002,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.3461,
+      "step": 2815
+    },
+    {
+      "epoch": 0.3196923415321193,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3474,
+      "step": 2816
+    },
+    {
+      "epoch": 0.31980586864203836,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3344,
+      "step": 2817
+    },
+    {
+      "epoch": 0.31991939575195744,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.3186,
+      "step": 2818
+    },
+    {
+      "epoch": 0.3200329228618765,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.3265,
+      "step": 2819
+    },
+    {
+      "epoch": 0.3201464499717956,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3307,
+      "step": 2820
+    },
+    {
+      "epoch": 0.32025997708171466,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.3419,
+      "step": 2821
+    },
+    {
+      "epoch": 0.32037350419163374,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.341,
+      "step": 2822
+    },
+    {
+      "epoch": 0.3204870313015528,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3275,
+      "step": 2823
+    },
+    {
+      "epoch": 0.3206005584114719,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.3393,
+      "step": 2824
+    },
+    {
+      "epoch": 0.32071408552139097,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 5.3532,
+      "step": 2825
+    },
+    {
+      "epoch": 0.32082761263131004,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.3255,
+      "step": 2826
+    },
+    {
+      "epoch": 0.3209411397412291,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.3202,
+      "step": 2827
+    },
+    {
+      "epoch": 0.3210546668511482,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3522,
+      "step": 2828
+    },
+    {
+      "epoch": 0.32116819396106727,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3643,
+      "step": 2829
+    },
+    {
+      "epoch": 0.32128172107098635,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.342,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3213952481809054,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.3307,
+      "step": 2831
+    },
+    {
+      "epoch": 0.3215087752908245,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3492,
+      "step": 2832
+    },
+    {
+      "epoch": 0.3216223024007436,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3297,
+      "step": 2833
+    },
+    {
+      "epoch": 0.32173582951066265,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.3318,
+      "step": 2834
+    },
+    {
+      "epoch": 0.32184935662058173,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.3228,
+      "step": 2835
+    },
+    {
+      "epoch": 0.32196288373050086,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.321,
+      "step": 2836
+    },
+    {
+      "epoch": 0.32207641084041994,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3407,
+      "step": 2837
+    },
+    {
+      "epoch": 0.322189937950339,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2928,
+      "step": 2838
+    },
+    {
+      "epoch": 0.3223034650602581,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3478,
+      "step": 2839
+    },
+    {
+      "epoch": 0.32241699217017716,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.3466,
+      "step": 2840
+    },
+    {
+      "epoch": 0.32253051928009624,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3429,
+      "step": 2841
+    },
+    {
+      "epoch": 0.3226440463900153,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.342,
+      "step": 2842
+    },
+    {
+      "epoch": 0.3227575734999344,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.3264,
+      "step": 2843
+    },
+    {
+      "epoch": 0.32287110060985347,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3213,
+      "step": 2844
+    },
+    {
+      "epoch": 0.32298462771977254,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3204,
+      "step": 2845
+    },
+    {
+      "epoch": 0.3230981548296916,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.3334,
+      "step": 2846
+    },
+    {
+      "epoch": 0.3232116819396107,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 5.3355,
+      "step": 2847
+    },
+    {
+      "epoch": 0.32332520904952977,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.3395,
+      "step": 2848
+    },
+    {
+      "epoch": 0.32343873615944885,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 5.3408,
+      "step": 2849
+    },
+    {
+      "epoch": 0.3235522632693679,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 5.3461,
+      "step": 2850
+    },
+    {
+      "epoch": 0.323665790379287,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.3486,
+      "step": 2851
+    },
+    {
+      "epoch": 0.3237793174892061,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.3219,
+      "step": 2852
+    },
+    {
+      "epoch": 0.32389284459912515,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.3319,
+      "step": 2853
+    },
+    {
+      "epoch": 0.3240063717090442,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3161,
+      "step": 2854
+    },
+    {
+      "epoch": 0.3241198988189633,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3457,
+      "step": 2855
+    },
+    {
+      "epoch": 0.3242334259288824,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3243,
+      "step": 2856
+    },
+    {
+      "epoch": 0.32434695303880146,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.3149,
+      "step": 2857
+    },
+    {
+      "epoch": 0.32446048014872053,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.3407,
+      "step": 2858
+    },
+    {
+      "epoch": 0.3245740072586396,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.3279,
+      "step": 2859
+    },
+    {
+      "epoch": 0.3246875343685587,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.3447,
+      "step": 2860
+    },
+    {
+      "epoch": 0.32480106147847776,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3372,
+      "step": 2861
+    },
+    {
+      "epoch": 0.32491458858839684,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.3345,
+      "step": 2862
+    },
+    {
+      "epoch": 0.3250281156983159,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.3291,
+      "step": 2863
+    },
+    {
+      "epoch": 0.325141642808235,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.325,
+      "step": 2864
+    },
+    {
+      "epoch": 0.32525516991815406,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3521,
+      "step": 2865
+    },
+    {
+      "epoch": 0.32536869702807314,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3367,
+      "step": 2866
+    },
+    {
+      "epoch": 0.3254822241379922,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.33,
+      "step": 2867
+    },
+    {
+      "epoch": 0.3255957512479113,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3272,
+      "step": 2868
+    },
+    {
+      "epoch": 0.32570927835783037,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3302,
+      "step": 2869
+    },
+    {
+      "epoch": 0.32582280546774944,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3327,
+      "step": 2870
+    },
+    {
+      "epoch": 0.3259363325776685,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3452,
+      "step": 2871
+    },
+    {
+      "epoch": 0.3260498596875876,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3306,
+      "step": 2872
+    },
+    {
+      "epoch": 0.32616338679750667,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.3186,
+      "step": 2873
+    },
+    {
+      "epoch": 0.32627691390742575,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.3402,
+      "step": 2874
+    },
+    {
+      "epoch": 0.3263904410173448,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3481,
+      "step": 2875
+    },
+    {
+      "epoch": 0.3265039681272639,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3264,
+      "step": 2876
+    },
+    {
+      "epoch": 0.326617495237183,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3282,
+      "step": 2877
+    },
+    {
+      "epoch": 0.32673102234710205,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.3183,
+      "step": 2878
+    },
+    {
+      "epoch": 0.3268445494570211,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3151,
+      "step": 2879
+    },
+    {
+      "epoch": 0.3269580765669402,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 5.3318,
+      "step": 2880
+    },
+    {
+      "epoch": 0.3270716036768593,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.3209,
+      "step": 2881
+    },
+    {
+      "epoch": 0.32718513078677836,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.337,
+      "step": 2882
+    },
+    {
+      "epoch": 0.32729865789669743,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3284,
+      "step": 2883
+    },
+    {
+      "epoch": 0.3274121850066165,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3222,
+      "step": 2884
+    },
+    {
+      "epoch": 0.3275257121165356,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.3146,
+      "step": 2885
+    },
+    {
+      "epoch": 0.32763923922645466,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.3317,
+      "step": 2886
+    },
+    {
+      "epoch": 0.32775276633637374,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.3342,
+      "step": 2887
+    },
+    {
+      "epoch": 0.3278662934462928,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.3026,
+      "step": 2888
+    },
+    {
+      "epoch": 0.3279798205562119,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.3248,
+      "step": 2889
+    },
+    {
+      "epoch": 0.32809334766613096,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.3417,
+      "step": 2890
+    },
+    {
+      "epoch": 0.32820687477605004,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.3284,
+      "step": 2891
+    },
+    {
+      "epoch": 0.3283204018859691,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.3465,
+      "step": 2892
+    },
+    {
+      "epoch": 0.3284339289958882,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.3252,
+      "step": 2893
+    },
+    {
+      "epoch": 0.32854745610580727,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3226,
+      "step": 2894
+    },
+    {
+      "epoch": 0.32866098321572634,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.3254,
+      "step": 2895
+    },
+    {
+      "epoch": 0.3287745103256454,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3207,
+      "step": 2896
+    },
+    {
+      "epoch": 0.3288880374355645,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3287,
+      "step": 2897
+    },
+    {
+      "epoch": 0.32900156454548357,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.3232,
+      "step": 2898
+    },
+    {
+      "epoch": 0.32911509165540265,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.319,
+      "step": 2899
+    },
+    {
+      "epoch": 0.3292286187653217,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.3463,
+      "step": 2900
+    },
+    {
+      "epoch": 0.3293421458752408,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3265,
+      "step": 2901
+    },
+    {
+      "epoch": 0.3294556729851599,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.3294,
+      "step": 2902
+    },
+    {
+      "epoch": 0.32956920009507895,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3316,
+      "step": 2903
+    },
+    {
+      "epoch": 0.329682727204998,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3094,
+      "step": 2904
+    },
+    {
+      "epoch": 0.3297962543149171,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3288,
+      "step": 2905
+    },
+    {
+      "epoch": 0.3299097814248362,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.317,
+      "step": 2906
+    },
+    {
+      "epoch": 0.33002330853475526,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3368,
+      "step": 2907
+    },
+    {
+      "epoch": 0.33013683564467433,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.352,
+      "step": 2908
+    },
+    {
+      "epoch": 0.3302503627545934,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.336,
+      "step": 2909
+    },
+    {
+      "epoch": 0.3303638898645125,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.3242,
+      "step": 2910
+    },
+    {
+      "epoch": 0.33047741697443156,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.3347,
+      "step": 2911
+    },
+    {
+      "epoch": 0.33059094408435064,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3348,
+      "step": 2912
+    },
+    {
+      "epoch": 0.3307044711942697,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3109,
+      "step": 2913
+    },
+    {
+      "epoch": 0.3308179983041888,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.3434,
+      "step": 2914
+    },
+    {
+      "epoch": 0.33093152541410786,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.3241,
+      "step": 2915
+    },
+    {
+      "epoch": 0.33104505252402694,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.3201,
+      "step": 2916
+    },
+    {
+      "epoch": 0.331158579633946,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3388,
+      "step": 2917
+    },
+    {
+      "epoch": 0.3312721067438651,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3363,
+      "step": 2918
+    },
+    {
+      "epoch": 0.33138563385378417,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3154,
+      "step": 2919
+    },
+    {
+      "epoch": 0.33149916096370324,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.3413,
+      "step": 2920
+    },
+    {
+      "epoch": 0.3316126880736223,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.3314,
+      "step": 2921
+    },
+    {
+      "epoch": 0.3317262151835414,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.3209,
+      "step": 2922
+    },
+    {
+      "epoch": 0.33183974229346047,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.3384,
+      "step": 2923
+    },
+    {
+      "epoch": 0.33195326940337955,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.3117,
+      "step": 2924
+    },
+    {
+      "epoch": 0.3320667965132986,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.331,
+      "step": 2925
+    },
+    {
+      "epoch": 0.3321803236232177,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.314,
+      "step": 2926
+    },
+    {
+      "epoch": 0.3322938507331368,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3115,
+      "step": 2927
+    },
+    {
+      "epoch": 0.33240737784305585,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.3048,
+      "step": 2928
+    },
+    {
+      "epoch": 0.3325209049529749,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3258,
+      "step": 2929
+    },
+    {
+      "epoch": 0.332634432062894,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3133,
+      "step": 2930
+    },
+    {
+      "epoch": 0.3327479591728131,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.3123,
+      "step": 2931
+    },
+    {
+      "epoch": 0.33286148628273216,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.3258,
+      "step": 2932
+    },
+    {
+      "epoch": 0.33297501339265123,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3353,
+      "step": 2933
+    },
+    {
+      "epoch": 0.3330885405025703,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3366,
+      "step": 2934
+    },
+    {
+      "epoch": 0.3332020676124894,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.3252,
+      "step": 2935
+    },
+    {
+      "epoch": 0.33331559472240846,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.3358,
+      "step": 2936
+    },
+    {
+      "epoch": 0.33342912183232754,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.3205,
+      "step": 2937
+    },
+    {
+      "epoch": 0.3335426489422466,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.3479,
+      "step": 2938
+    },
+    {
+      "epoch": 0.3336561760521657,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3189,
+      "step": 2939
+    },
+    {
+      "epoch": 0.33376970316208476,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3274,
+      "step": 2940
+    },
+    {
+      "epoch": 0.33388323027200384,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3057,
+      "step": 2941
+    },
+    {
+      "epoch": 0.3339967573819229,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3098,
+      "step": 2942
+    },
+    {
+      "epoch": 0.334110284491842,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3237,
+      "step": 2943
+    },
+    {
+      "epoch": 0.33422381160176107,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.3071,
+      "step": 2944
+    },
+    {
+      "epoch": 0.33433733871168014,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.336,
+      "step": 2945
+    },
+    {
+      "epoch": 0.3344508658215992,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.3172,
+      "step": 2946
+    },
+    {
+      "epoch": 0.3345643929315183,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.3317,
+      "step": 2947
+    },
+    {
+      "epoch": 0.33467792004143737,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.3274,
+      "step": 2948
+    },
+    {
+      "epoch": 0.33479144715135645,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.342,
+      "step": 2949
+    },
+    {
+      "epoch": 0.3349049742612755,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.3324,
+      "step": 2950
+    },
+    {
+      "epoch": 0.3350185013711946,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3296,
+      "step": 2951
+    },
+    {
+      "epoch": 0.3351320284811137,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3554,
+      "step": 2952
+    },
+    {
+      "epoch": 0.33524555559103275,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.3294,
+      "step": 2953
+    },
+    {
+      "epoch": 0.3353590827009518,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.314,
+      "step": 2954
+    },
+    {
+      "epoch": 0.3354726098108709,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3497,
+      "step": 2955
+    },
+    {
+      "epoch": 0.33558613692079,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.3047,
+      "step": 2956
+    },
+    {
+      "epoch": 0.3356996640307091,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3393,
+      "step": 2957
+    },
+    {
+      "epoch": 0.3358131911406282,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3344,
+      "step": 2958
+    },
+    {
+      "epoch": 0.33592671825054726,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3252,
+      "step": 2959
+    },
+    {
+      "epoch": 0.33604024536046634,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3176,
+      "step": 2960
+    },
+    {
+      "epoch": 0.3361537724703854,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.335,
+      "step": 2961
+    },
+    {
+      "epoch": 0.3362672995803045,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 5.3354,
+      "step": 2962
+    },
+    {
+      "epoch": 0.33638082669022357,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.3451,
+      "step": 2963
+    },
+    {
+      "epoch": 0.33649435380014264,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.3318,
+      "step": 2964
+    },
+    {
+      "epoch": 0.3366078809100617,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.3326,
+      "step": 2965
+    },
+    {
+      "epoch": 0.3367214080199808,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3261,
+      "step": 2966
+    },
+    {
+      "epoch": 0.33683493512989987,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.342,
+      "step": 2967
+    },
+    {
+      "epoch": 0.33694846223981895,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.3323,
+      "step": 2968
+    },
+    {
+      "epoch": 0.337061989349738,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 5.3338,
+      "step": 2969
+    },
+    {
+      "epoch": 0.3371755164596571,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.3543,
+      "step": 2970
+    },
+    {
+      "epoch": 0.3372890435695762,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.3273,
+      "step": 2971
+    },
+    {
+      "epoch": 0.33740257067949525,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3193,
+      "step": 2972
+    },
+    {
+      "epoch": 0.3375160977894143,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3292,
+      "step": 2973
+    },
+    {
+      "epoch": 0.3376296248993334,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3019,
+      "step": 2974
+    },
+    {
+      "epoch": 0.3377431520092525,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.3022,
+      "step": 2975
+    },
+    {
+      "epoch": 0.33785667911917155,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3382,
+      "step": 2976
+    },
+    {
+      "epoch": 0.33797020622909063,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.3395,
+      "step": 2977
+    },
+    {
+      "epoch": 0.3380837333390097,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.3156,
+      "step": 2978
+    },
+    {
+      "epoch": 0.3381972604489288,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 5.3168,
+      "step": 2979
+    },
+    {
+      "epoch": 0.33831078755884786,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.3191,
+      "step": 2980
+    },
+    {
+      "epoch": 0.33842431466876693,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3107,
+      "step": 2981
+    },
+    {
+      "epoch": 0.338537841778686,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3145,
+      "step": 2982
+    },
+    {
+      "epoch": 0.3386513688886051,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.3414,
+      "step": 2983
+    },
+    {
+      "epoch": 0.33876489599852416,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3291,
+      "step": 2984
+    },
+    {
+      "epoch": 0.33887842310844324,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.3253,
+      "step": 2985
+    },
+    {
+      "epoch": 0.3389919502183623,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3406,
+      "step": 2986
+    },
+    {
+      "epoch": 0.3391054773282814,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3289,
+      "step": 2987
+    },
+    {
+      "epoch": 0.33921900443820047,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.3112,
+      "step": 2988
+    },
+    {
+      "epoch": 0.33933253154811954,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.3254,
+      "step": 2989
+    },
+    {
+      "epoch": 0.3394460586580386,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.3211,
+      "step": 2990
+    },
+    {
+      "epoch": 0.3395595857679577,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 5.3269,
+      "step": 2991
+    },
+    {
+      "epoch": 0.33967311287787677,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3512,
+      "step": 2992
+    },
+    {
+      "epoch": 0.33978663998779585,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 5.3346,
+      "step": 2993
+    },
+    {
+      "epoch": 0.3399001670977149,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 5.3354,
+      "step": 2994
+    },
+    {
+      "epoch": 0.340013694207634,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3292,
+      "step": 2995
+    },
+    {
+      "epoch": 0.3401272213175531,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3184,
+      "step": 2996
+    },
+    {
+      "epoch": 0.34024074842747215,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.3216,
+      "step": 2997
+    },
+    {
+      "epoch": 0.3403542755373912,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.3234,
+      "step": 2998
+    },
+    {
+      "epoch": 0.3404678026473103,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 5.3348,
+      "step": 2999
+    },
+    {
+      "epoch": 0.3405813297572294,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.002,
+      "loss": 5.3241,
+      "step": 3000
+    },
+    {
+      "epoch": 0.34069485686714845,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.3281,
+      "step": 3001
+    },
+    {
+      "epoch": 0.34080838397706753,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.3485,
+      "step": 3002
+    },
+    {
+      "epoch": 0.3409219110869866,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.338,
+      "step": 3003
+    },
+    {
+      "epoch": 0.3410354381969057,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3191,
+      "step": 3004
+    },
+    {
+      "epoch": 0.34114896530682476,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3209,
+      "step": 3005
+    },
+    {
+      "epoch": 0.34126249241674383,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.308,
+      "step": 3006
+    },
+    {
+      "epoch": 0.3413760195266629,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.3311,
+      "step": 3007
+    },
+    {
+      "epoch": 0.341489546636582,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.3338,
+      "step": 3008
+    },
+    {
+      "epoch": 0.34160307374650106,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.3379,
+      "step": 3009
+    },
+    {
+      "epoch": 0.34171660085642014,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.3112,
+      "step": 3010
+    },
+    {
+      "epoch": 0.3418301279663392,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3238,
+      "step": 3011
+    },
+    {
+      "epoch": 0.3419436550762583,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.3304,
+      "step": 3012
+    },
+    {
+      "epoch": 0.34205718218617737,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3081,
+      "step": 3013
+    },
+    {
+      "epoch": 0.34217070929609644,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3116,
+      "step": 3014
+    },
+    {
+      "epoch": 0.3422842364060155,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.3105,
+      "step": 3015
+    },
+    {
+      "epoch": 0.3423977635159346,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3233,
+      "step": 3016
+    },
+    {
+      "epoch": 0.34251129062585367,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2972,
+      "step": 3017
+    },
+    {
+      "epoch": 0.34262481773577275,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.3195,
+      "step": 3018
+    },
+    {
+      "epoch": 0.3427383448456918,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.3048,
+      "step": 3019
+    },
+    {
+      "epoch": 0.3428518719556109,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.3248,
+      "step": 3020
+    },
+    {
+      "epoch": 0.34296539906553,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3049,
+      "step": 3021
+    },
+    {
+      "epoch": 0.34307892617544905,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3356,
+      "step": 3022
+    },
+    {
+      "epoch": 0.3431924532853681,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3357,
+      "step": 3023
+    },
+    {
+      "epoch": 0.3433059803952872,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3218,
+      "step": 3024
+    },
+    {
+      "epoch": 0.3434195075052063,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3314,
+      "step": 3025
+    },
+    {
+      "epoch": 0.34353303461512535,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.3361,
+      "step": 3026
+    },
+    {
+      "epoch": 0.34364656172504443,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.3372,
+      "step": 3027
+    },
+    {
+      "epoch": 0.3437600888349635,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3272,
+      "step": 3028
+    },
+    {
+      "epoch": 0.3438736159448826,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3178,
+      "step": 3029
+    },
+    {
+      "epoch": 0.34398714305480166,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3228,
+      "step": 3030
+    },
+    {
+      "epoch": 0.34410067016472073,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3156,
+      "step": 3031
+    },
+    {
+      "epoch": 0.3442141972746398,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3386,
+      "step": 3032
+    },
+    {
+      "epoch": 0.3443277243845589,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3128,
+      "step": 3033
+    },
+    {
+      "epoch": 0.34444125149447796,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.309,
+      "step": 3034
+    },
+    {
+      "epoch": 0.34455477860439704,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3168,
+      "step": 3035
+    },
+    {
+      "epoch": 0.3446683057143161,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3249,
+      "step": 3036
+    },
+    {
+      "epoch": 0.3447818328242352,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.3292,
+      "step": 3037
+    },
+    {
+      "epoch": 0.34489535993415427,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.3285,
+      "step": 3038
+    },
+    {
+      "epoch": 0.34500888704407334,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.3247,
+      "step": 3039
+    },
+    {
+      "epoch": 0.3451224141539924,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.3333,
+      "step": 3040
+    },
+    {
+      "epoch": 0.3452359412639115,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3243,
+      "step": 3041
+    },
+    {
+      "epoch": 0.34534946837383057,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3349,
+      "step": 3042
+    },
+    {
+      "epoch": 0.34546299548374965,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3014,
+      "step": 3043
+    },
+    {
+      "epoch": 0.3455765225936687,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3158,
+      "step": 3044
+    },
+    {
+      "epoch": 0.3456900497035878,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.34,
+      "step": 3045
+    },
+    {
+      "epoch": 0.3458035768135069,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.3326,
+      "step": 3046
+    },
+    {
+      "epoch": 0.34591710392342595,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.3482,
+      "step": 3047
+    },
+    {
+      "epoch": 0.346030631033345,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3207,
+      "step": 3048
+    },
+    {
+      "epoch": 0.3461441581432641,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.3131,
+      "step": 3049
+    },
+    {
+      "epoch": 0.3462576852531832,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.3176,
+      "step": 3050
+    },
+    {
+      "epoch": 0.34637121236310225,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.3297,
+      "step": 3051
+    },
+    {
+      "epoch": 0.34648473947302133,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3321,
+      "step": 3052
+    },
+    {
+      "epoch": 0.3465982665829404,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 5.319,
+      "step": 3053
+    },
+    {
+      "epoch": 0.3467117936928595,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.3066,
+      "step": 3054
+    },
+    {
+      "epoch": 0.34682532080277856,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3165,
+      "step": 3055
+    },
+    {
+      "epoch": 0.34693884791269763,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.3123,
+      "step": 3056
+    },
+    {
+      "epoch": 0.3470523750226167,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.3185,
+      "step": 3057
+    },
+    {
+      "epoch": 0.3471659021325358,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3288,
+      "step": 3058
+    },
+    {
+      "epoch": 0.34727942924245486,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3089,
+      "step": 3059
+    },
+    {
+      "epoch": 0.34739295635237394,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.314,
+      "step": 3060
+    },
+    {
+      "epoch": 0.347506483462293,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.314,
+      "step": 3061
+    },
+    {
+      "epoch": 0.3476200105722121,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.3109,
+      "step": 3062
+    },
+    {
+      "epoch": 0.34773353768213117,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.3109,
+      "step": 3063
+    },
+    {
+      "epoch": 0.34784706479205024,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3166,
+      "step": 3064
+    },
+    {
+      "epoch": 0.3479605919019693,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3213,
+      "step": 3065
+    },
+    {
+      "epoch": 0.3480741190118884,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3529,
+      "step": 3066
+    },
+    {
+      "epoch": 0.34818764612180747,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3441,
+      "step": 3067
+    },
+    {
+      "epoch": 0.34830117323172655,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.3203,
+      "step": 3068
+    },
+    {
+      "epoch": 0.3484147003416456,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3232,
+      "step": 3069
+    },
+    {
+      "epoch": 0.3485282274515647,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.3089,
+      "step": 3070
+    },
+    {
+      "epoch": 0.3486417545614838,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.3129,
+      "step": 3071
+    },
+    {
+      "epoch": 0.34875528167140285,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3029,
+      "step": 3072
+    },
+    {
+      "epoch": 0.3488688087813219,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.3018,
+      "step": 3073
+    },
+    {
+      "epoch": 0.348982335891241,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.316,
+      "step": 3074
+    },
+    {
+      "epoch": 0.3490958630011601,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2986,
+      "step": 3075
+    },
+    {
+      "epoch": 0.34920939011107915,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.312,
+      "step": 3076
+    },
+    {
+      "epoch": 0.3493229172209983,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.3352,
+      "step": 3077
+    },
+    {
+      "epoch": 0.34943644433091736,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3344,
+      "step": 3078
+    },
+    {
+      "epoch": 0.34954997144083644,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.329,
+      "step": 3079
+    },
+    {
+      "epoch": 0.3496634985507555,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.311,
+      "step": 3080
+    },
+    {
+      "epoch": 0.3497770256606746,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3275,
+      "step": 3081
+    },
+    {
+      "epoch": 0.34989055277059367,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.3306,
+      "step": 3082
+    },
+    {
+      "epoch": 0.35000407988051274,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.3277,
+      "step": 3083
+    },
+    {
+      "epoch": 0.3501176069904318,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.342,
+      "step": 3084
+    },
+    {
+      "epoch": 0.3502311341003509,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.3292,
+      "step": 3085
+    },
+    {
+      "epoch": 0.35034466121026997,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3075,
+      "step": 3086
+    },
+    {
+      "epoch": 0.35045818832018905,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.315,
+      "step": 3087
+    },
+    {
+      "epoch": 0.3505717154301081,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3349,
+      "step": 3088
+    },
+    {
+      "epoch": 0.3506852425400272,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3212,
+      "step": 3089
+    },
+    {
+      "epoch": 0.3507987696499463,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3096,
+      "step": 3090
+    },
+    {
+      "epoch": 0.35091229675986535,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.322,
+      "step": 3091
+    },
+    {
+      "epoch": 0.3510258238697844,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3226,
+      "step": 3092
+    },
+    {
+      "epoch": 0.3511393509797035,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.3051,
+      "step": 3093
+    },
+    {
+      "epoch": 0.3512528780896226,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3034,
+      "step": 3094
+    },
+    {
+      "epoch": 0.35136640519954165,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.3082,
+      "step": 3095
+    },
+    {
+      "epoch": 0.35147993230946073,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.3186,
+      "step": 3096
+    },
+    {
+      "epoch": 0.3515934594193798,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.3335,
+      "step": 3097
+    },
+    {
+      "epoch": 0.3517069865292989,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 5.2968,
+      "step": 3098
+    },
+    {
+      "epoch": 0.35182051363921796,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.3212,
+      "step": 3099
+    },
+    {
+      "epoch": 0.35193404074913703,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.3047,
+      "step": 3100
+    },
+    {
+      "epoch": 0.3520475678590561,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3045,
+      "step": 3101
+    },
+    {
+      "epoch": 0.3521610949689752,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.3172,
+      "step": 3102
+    },
+    {
+      "epoch": 0.35227462207889426,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3127,
+      "step": 3103
+    },
+    {
+      "epoch": 0.35238814918881334,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3219,
+      "step": 3104
+    },
+    {
+      "epoch": 0.3525016762987324,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.3156,
+      "step": 3105
+    },
+    {
+      "epoch": 0.3526152034086515,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.331,
+      "step": 3106
+    },
+    {
+      "epoch": 0.35272873051857057,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3147,
+      "step": 3107
+    },
+    {
+      "epoch": 0.35284225762848964,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.306,
+      "step": 3108
+    },
+    {
+      "epoch": 0.3529557847384087,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.3189,
+      "step": 3109
+    },
+    {
+      "epoch": 0.3530693118483278,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.328,
+      "step": 3110
+    },
+    {
+      "epoch": 0.35318283895824687,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3258,
+      "step": 3111
+    },
+    {
+      "epoch": 0.35329636606816595,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.3152,
+      "step": 3112
+    },
+    {
+      "epoch": 0.353409893178085,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.3245,
+      "step": 3113
+    },
+    {
+      "epoch": 0.3535234202880041,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3179,
+      "step": 3114
+    },
+    {
+      "epoch": 0.3536369473979232,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.3127,
+      "step": 3115
+    },
+    {
+      "epoch": 0.35375047450784225,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3291,
+      "step": 3116
+    },
+    {
+      "epoch": 0.3538640016177613,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.3073,
+      "step": 3117
+    },
+    {
+      "epoch": 0.3539775287276804,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3408,
+      "step": 3118
+    },
+    {
+      "epoch": 0.3540910558375995,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3225,
+      "step": 3119
+    },
+    {
+      "epoch": 0.35420458294751855,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3249,
+      "step": 3120
+    },
+    {
+      "epoch": 0.35431811005743763,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3279,
+      "step": 3121
+    },
+    {
+      "epoch": 0.3544316371673567,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3102,
+      "step": 3122
+    },
+    {
+      "epoch": 0.3545451642772758,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3192,
+      "step": 3123
+    },
+    {
+      "epoch": 0.35465869138719486,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3057,
+      "step": 3124
+    },
+    {
+      "epoch": 0.35477221849711393,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.309,
+      "step": 3125
+    },
+    {
+      "epoch": 0.354885745607033,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3038,
+      "step": 3126
+    },
+    {
+      "epoch": 0.3549992727169521,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3342,
+      "step": 3127
+    },
+    {
+      "epoch": 0.35511279982687116,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.3238,
+      "step": 3128
+    },
+    {
+      "epoch": 0.35522632693679024,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3152,
+      "step": 3129
+    },
+    {
+      "epoch": 0.3553398540467093,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3117,
+      "step": 3130
+    },
+    {
+      "epoch": 0.3554533811566284,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3244,
+      "step": 3131
+    },
+    {
+      "epoch": 0.35556690826654747,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2969,
+      "step": 3132
+    },
+    {
+      "epoch": 0.35568043537646654,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.3235,
+      "step": 3133
+    },
+    {
+      "epoch": 0.3557939624863856,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.3051,
+      "step": 3134
+    },
+    {
+      "epoch": 0.3559074895963047,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.2841,
+      "step": 3135
+    },
+    {
+      "epoch": 0.35602101670622377,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.3125,
+      "step": 3136
+    },
+    {
+      "epoch": 0.35613454381614285,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.3048,
+      "step": 3137
+    },
+    {
+      "epoch": 0.3562480709260619,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3234,
+      "step": 3138
+    },
+    {
+      "epoch": 0.356361598035981,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3029,
+      "step": 3139
+    },
+    {
+      "epoch": 0.3564751251459001,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.3099,
+      "step": 3140
+    },
+    {
+      "epoch": 0.35658865225581915,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3199,
+      "step": 3141
+    },
+    {
+      "epoch": 0.3567021793657382,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3249,
+      "step": 3142
+    },
+    {
+      "epoch": 0.3568157064756573,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3361,
+      "step": 3143
+    },
+    {
+      "epoch": 0.3569292335855764,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.317,
+      "step": 3144
+    },
+    {
+      "epoch": 0.35704276069549545,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3049,
+      "step": 3145
+    },
+    {
+      "epoch": 0.35715628780541453,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3286,
+      "step": 3146
+    },
+    {
+      "epoch": 0.3572698149153336,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3296,
+      "step": 3147
+    },
+    {
+      "epoch": 0.3573833420252527,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.3186,
+      "step": 3148
+    },
+    {
+      "epoch": 0.35749686913517176,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2932,
+      "step": 3149
+    },
+    {
+      "epoch": 0.35761039624509083,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2966,
+      "step": 3150
+    },
+    {
+      "epoch": 0.3577239233550099,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2995,
+      "step": 3151
+    },
+    {
+      "epoch": 0.357837450464929,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3147,
+      "step": 3152
+    },
+    {
+      "epoch": 0.35795097757484806,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.3156,
+      "step": 3153
+    },
+    {
+      "epoch": 0.35806450468476714,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2979,
+      "step": 3154
+    },
+    {
+      "epoch": 0.3581780317946862,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 5.3416,
+      "step": 3155
+    },
+    {
+      "epoch": 0.3582915589046053,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 5.3202,
+      "step": 3156
+    },
+    {
+      "epoch": 0.35840508601452437,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3069,
+      "step": 3157
+    },
+    {
+      "epoch": 0.35851861312444344,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3093,
+      "step": 3158
+    },
+    {
+      "epoch": 0.3586321402343625,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3125,
+      "step": 3159
+    },
+    {
+      "epoch": 0.3587456673442816,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.3123,
+      "step": 3160
+    },
+    {
+      "epoch": 0.35885919445420067,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.3165,
+      "step": 3161
+    },
+    {
+      "epoch": 0.35897272156411975,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.3106,
+      "step": 3162
+    },
+    {
+      "epoch": 0.3590862486740388,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.3291,
+      "step": 3163
+    },
+    {
+      "epoch": 0.3591997757839579,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.329,
+      "step": 3164
+    },
+    {
+      "epoch": 0.359313302893877,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.3282,
+      "step": 3165
+    },
+    {
+      "epoch": 0.35942683000379605,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.303,
+      "step": 3166
+    },
+    {
+      "epoch": 0.3595403571137151,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3195,
+      "step": 3167
+    },
+    {
+      "epoch": 0.3596538842236342,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.3166,
+      "step": 3168
+    },
+    {
+      "epoch": 0.3597674113335533,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.3063,
+      "step": 3169
+    },
+    {
+      "epoch": 0.35988093844347235,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.3031,
+      "step": 3170
+    },
+    {
+      "epoch": 0.35999446555339143,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3288,
+      "step": 3171
+    },
+    {
+      "epoch": 0.3601079926633105,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.3227,
+      "step": 3172
+    },
+    {
+      "epoch": 0.3602215197732296,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.3279,
+      "step": 3173
+    },
+    {
+      "epoch": 0.36033504688314866,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.326,
+      "step": 3174
+    },
+    {
+      "epoch": 0.36044857399306773,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.3191,
+      "step": 3175
+    },
+    {
+      "epoch": 0.3605621011029868,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3155,
+      "step": 3176
+    },
+    {
+      "epoch": 0.3606756282129059,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3063,
+      "step": 3177
+    },
+    {
+      "epoch": 0.36078915532282496,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.3223,
+      "step": 3178
+    },
+    {
+      "epoch": 0.36090268243274404,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3163,
+      "step": 3179
+    },
+    {
+      "epoch": 0.3610162095426631,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3038,
+      "step": 3180
+    },
+    {
+      "epoch": 0.3611297366525822,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3241,
+      "step": 3181
+    },
+    {
+      "epoch": 0.36124326376250127,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.3181,
+      "step": 3182
+    },
+    {
+      "epoch": 0.36135679087242034,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3181,
+      "step": 3183
+    },
+    {
+      "epoch": 0.3614703179823394,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3372,
+      "step": 3184
+    },
+    {
+      "epoch": 0.3615838450922585,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.3051,
+      "step": 3185
+    },
+    {
+      "epoch": 0.36169737220217757,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.309,
+      "step": 3186
+    },
+    {
+      "epoch": 0.36181089931209665,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3042,
+      "step": 3187
+    },
+    {
+      "epoch": 0.3619244264220157,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2981,
+      "step": 3188
+    },
+    {
+      "epoch": 0.3620379535319348,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3048,
+      "step": 3189
+    },
+    {
+      "epoch": 0.3621514806418539,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3262,
+      "step": 3190
+    },
+    {
+      "epoch": 0.36226500775177295,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3297,
+      "step": 3191
+    },
+    {
+      "epoch": 0.362378534861692,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3192,
+      "step": 3192
+    },
+    {
+      "epoch": 0.3624920619716111,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3149,
+      "step": 3193
+    },
+    {
+      "epoch": 0.3626055890815302,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.3346,
+      "step": 3194
+    },
+    {
+      "epoch": 0.36271911619144925,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3045,
+      "step": 3195
+    },
+    {
+      "epoch": 0.36283264330136833,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2903,
+      "step": 3196
+    },
+    {
+      "epoch": 0.3629461704112874,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3244,
+      "step": 3197
+    },
+    {
+      "epoch": 0.36305969752120654,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2937,
+      "step": 3198
+    },
+    {
+      "epoch": 0.3631732246311256,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3175,
+      "step": 3199
+    },
+    {
+      "epoch": 0.3632867517410447,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3192,
+      "step": 3200
+    },
+    {
+      "epoch": 0.36340027885096376,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.305,
+      "step": 3201
+    },
+    {
+      "epoch": 0.36351380596088284,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.3004,
+      "step": 3202
+    },
+    {
+      "epoch": 0.3636273330708019,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 5.3369,
+      "step": 3203
+    },
+    {
+      "epoch": 0.363740860180721,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3221,
+      "step": 3204
+    },
+    {
+      "epoch": 0.36385438729064007,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3127,
+      "step": 3205
+    },
+    {
+      "epoch": 0.36396791440055914,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.2887,
+      "step": 3206
+    },
+    {
+      "epoch": 0.3640814415104782,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.002,
+      "loss": 5.3157,
+      "step": 3207
+    },
+    {
+      "epoch": 0.3641949686203973,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.3085,
+      "step": 3208
+    },
+    {
+      "epoch": 0.3643084957303164,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.3228,
+      "step": 3209
+    },
+    {
+      "epoch": 0.36442202284023545,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.3223,
+      "step": 3210
+    },
+    {
+      "epoch": 0.3645355499501545,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3314,
+      "step": 3211
+    },
+    {
+      "epoch": 0.3646490770600736,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3099,
+      "step": 3212
+    },
+    {
+      "epoch": 0.3647626041699927,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3108,
+      "step": 3213
+    },
+    {
+      "epoch": 0.36487613127991175,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.3112,
+      "step": 3214
+    },
+    {
+      "epoch": 0.36498965838983083,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.3203,
+      "step": 3215
+    },
+    {
+      "epoch": 0.3651031854997499,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2962,
+      "step": 3216
+    },
+    {
+      "epoch": 0.365216712609669,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.3033,
+      "step": 3217
+    },
+    {
+      "epoch": 0.36533023971958806,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.323,
+      "step": 3218
+    },
+    {
+      "epoch": 0.36544376682950713,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.3085,
+      "step": 3219
+    },
+    {
+      "epoch": 0.3655572939394262,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3226,
+      "step": 3220
+    },
+    {
+      "epoch": 0.3656708210493453,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.3234,
+      "step": 3221
+    },
+    {
+      "epoch": 0.36578434815926436,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.3133,
+      "step": 3222
+    },
+    {
+      "epoch": 0.36589787526918344,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.3283,
+      "step": 3223
+    },
+    {
+      "epoch": 0.3660114023791025,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.3181,
+      "step": 3224
+    },
+    {
+      "epoch": 0.3661249294890216,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2906,
+      "step": 3225
+    },
+    {
+      "epoch": 0.36623845659894066,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.311,
+      "step": 3226
+    },
+    {
+      "epoch": 0.36635198370885974,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3213,
+      "step": 3227
+    },
+    {
+      "epoch": 0.3664655108187788,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.3002,
+      "step": 3228
+    },
+    {
+      "epoch": 0.3665790379286979,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.3069,
+      "step": 3229
+    },
+    {
+      "epoch": 0.36669256503861697,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2891,
+      "step": 3230
+    },
+    {
+      "epoch": 0.36680609214853604,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.3018,
+      "step": 3231
+    },
+    {
+      "epoch": 0.3669196192584551,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2983,
+      "step": 3232
+    },
+    {
+      "epoch": 0.3670331463683742,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3005,
+      "step": 3233
+    },
+    {
+      "epoch": 0.3671466734782933,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3322,
+      "step": 3234
+    },
+    {
+      "epoch": 0.36726020058821235,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3175,
+      "step": 3235
+    },
+    {
+      "epoch": 0.3673737276981314,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3278,
+      "step": 3236
+    },
+    {
+      "epoch": 0.3674872548080505,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.3255,
+      "step": 3237
+    },
+    {
+      "epoch": 0.3676007819179696,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.313,
+      "step": 3238
+    },
+    {
+      "epoch": 0.36771430902788865,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.3094,
+      "step": 3239
+    },
+    {
+      "epoch": 0.36782783613780773,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3037,
+      "step": 3240
+    },
+    {
+      "epoch": 0.3679413632477268,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3215,
+      "step": 3241
+    },
+    {
+      "epoch": 0.3680548903576459,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2966,
+      "step": 3242
+    },
+    {
+      "epoch": 0.36816841746756496,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2997,
+      "step": 3243
+    },
+    {
+      "epoch": 0.36828194457748403,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3044,
+      "step": 3244
+    },
+    {
+      "epoch": 0.3683954716874031,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2939,
+      "step": 3245
+    },
+    {
+      "epoch": 0.3685089987973222,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.3082,
+      "step": 3246
+    },
+    {
+      "epoch": 0.36862252590724126,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.3151,
+      "step": 3247
+    },
+    {
+      "epoch": 0.36873605301716034,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.3188,
+      "step": 3248
+    },
+    {
+      "epoch": 0.3688495801270794,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.316,
+      "step": 3249
+    },
+    {
+      "epoch": 0.3689631072369985,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2889,
+      "step": 3250
+    },
+    {
+      "epoch": 0.36907663434691756,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2942,
+      "step": 3251
+    },
+    {
+      "epoch": 0.36919016145683664,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.3072,
+      "step": 3252
+    },
+    {
+      "epoch": 0.3693036885667557,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 5.32,
+      "step": 3253
+    },
+    {
+      "epoch": 0.3694172156766748,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3133,
+      "step": 3254
+    },
+    {
+      "epoch": 0.36953074278659387,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.3143,
+      "step": 3255
+    },
+    {
+      "epoch": 0.36964426989651294,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.3153,
+      "step": 3256
+    },
+    {
+      "epoch": 0.369757797006432,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.002,
+      "loss": 5.3181,
+      "step": 3257
+    },
+    {
+      "epoch": 0.3698713241163511,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.002,
+      "loss": 5.3032,
+      "step": 3258
+    },
+    {
+      "epoch": 0.3699848512262702,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.2921,
+      "step": 3259
+    },
+    {
+      "epoch": 0.37009837833618925,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.3105,
+      "step": 3260
+    },
+    {
+      "epoch": 0.3702119054461083,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3234,
+      "step": 3261
+    },
+    {
+      "epoch": 0.3703254325560274,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3226,
+      "step": 3262
+    },
+    {
+      "epoch": 0.3704389596659465,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3033,
+      "step": 3263
+    },
+    {
+      "epoch": 0.37055248677586555,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3234,
+      "step": 3264
+    },
+    {
+      "epoch": 0.37066601388578463,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3089,
+      "step": 3265
+    },
+    {
+      "epoch": 0.3707795409957037,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3144,
+      "step": 3266
+    },
+    {
+      "epoch": 0.3708930681056228,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3109,
+      "step": 3267
+    },
+    {
+      "epoch": 0.37100659521554186,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.3064,
+      "step": 3268
+    },
+    {
+      "epoch": 0.37112012232546093,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.312,
+      "step": 3269
+    },
+    {
+      "epoch": 0.37123364943538,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.317,
+      "step": 3270
+    },
+    {
+      "epoch": 0.3713471765452991,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3365,
+      "step": 3271
+    },
+    {
+      "epoch": 0.37146070365521816,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2891,
+      "step": 3272
+    },
+    {
+      "epoch": 0.37157423076513724,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.3191,
+      "step": 3273
+    },
+    {
+      "epoch": 0.3716877578750563,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3033,
+      "step": 3274
+    },
+    {
+      "epoch": 0.3718012849849754,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3263,
+      "step": 3275
+    },
+    {
+      "epoch": 0.37191481209489446,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.3138,
+      "step": 3276
+    },
+    {
+      "epoch": 0.37202833920481354,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3031,
+      "step": 3277
+    },
+    {
+      "epoch": 0.3721418663147326,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.3059,
+      "step": 3278
+    },
+    {
+      "epoch": 0.3722553934246517,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2929,
+      "step": 3279
+    },
+    {
+      "epoch": 0.37236892053457077,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.317,
+      "step": 3280
+    },
+    {
+      "epoch": 0.37248244764448984,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.3127,
+      "step": 3281
+    },
+    {
+      "epoch": 0.3725959747544089,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3216,
+      "step": 3282
+    },
+    {
+      "epoch": 0.372709501864328,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2894,
+      "step": 3283
+    },
+    {
+      "epoch": 0.3728230289742471,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2889,
+      "step": 3284
+    },
+    {
+      "epoch": 0.37293655608416615,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3017,
+      "step": 3285
+    },
+    {
+      "epoch": 0.3730500831940852,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3205,
+      "step": 3286
+    },
+    {
+      "epoch": 0.3731636103040043,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.3003,
+      "step": 3287
+    },
+    {
+      "epoch": 0.3732771374139234,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3238,
+      "step": 3288
+    },
+    {
+      "epoch": 0.37339066452384245,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2894,
+      "step": 3289
+    },
+    {
+      "epoch": 0.37350419163376153,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.3041,
+      "step": 3290
+    },
+    {
+      "epoch": 0.3736177187436806,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2978,
+      "step": 3291
+    },
+    {
+      "epoch": 0.3737312458535997,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3051,
+      "step": 3292
+    },
+    {
+      "epoch": 0.37384477296351876,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3247,
+      "step": 3293
+    },
+    {
+      "epoch": 0.37395830007343783,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2791,
+      "step": 3294
+    },
+    {
+      "epoch": 0.3740718271833569,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 5.3095,
+      "step": 3295
+    },
+    {
+      "epoch": 0.374185354293276,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.002,
+      "loss": 5.2932,
+      "step": 3296
+    },
+    {
+      "epoch": 0.37429888140319506,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 5.3145,
+      "step": 3297
+    },
+    {
+      "epoch": 0.37441240851311414,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 5.3071,
+      "step": 3298
+    },
+    {
+      "epoch": 0.3745259356230332,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2972,
+      "step": 3299
+    },
+    {
+      "epoch": 0.3746394627329523,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3253,
+      "step": 3300
+    },
+    {
+      "epoch": 0.37475298984287136,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.3118,
+      "step": 3301
+    },
+    {
+      "epoch": 0.37486651695279044,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2949,
+      "step": 3302
+    },
+    {
+      "epoch": 0.3749800440627095,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3095,
+      "step": 3303
+    },
+    {
+      "epoch": 0.3750935711726286,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.307,
+      "step": 3304
+    },
+    {
+      "epoch": 0.37520709828254767,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.3159,
+      "step": 3305
+    },
+    {
+      "epoch": 0.37532062539246674,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.2928,
+      "step": 3306
+    },
+    {
+      "epoch": 0.3754341525023858,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.3079,
+      "step": 3307
+    },
+    {
+      "epoch": 0.3755476796123049,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.3211,
+      "step": 3308
+    },
+    {
+      "epoch": 0.375661206722224,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.3178,
+      "step": 3309
+    },
+    {
+      "epoch": 0.37577473383214305,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2976,
+      "step": 3310
+    },
+    {
+      "epoch": 0.3758882609420621,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3059,
+      "step": 3311
+    },
+    {
+      "epoch": 0.3760017880519812,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3111,
+      "step": 3312
+    },
+    {
+      "epoch": 0.3761153151619003,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3226,
+      "step": 3313
+    },
+    {
+      "epoch": 0.37622884227181935,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.3025,
+      "step": 3314
+    },
+    {
+      "epoch": 0.37634236938173843,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2729,
+      "step": 3315
+    },
+    {
+      "epoch": 0.3764558964916575,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.296,
+      "step": 3316
+    },
+    {
+      "epoch": 0.3765694236015766,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.306,
+      "step": 3317
+    },
+    {
+      "epoch": 0.37668295071149566,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3082,
+      "step": 3318
+    },
+    {
+      "epoch": 0.3767964778214148,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3173,
+      "step": 3319
+    },
+    {
+      "epoch": 0.37691000493133386,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.3247,
+      "step": 3320
+    },
+    {
+      "epoch": 0.37702353204125294,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3139,
+      "step": 3321
+    },
+    {
+      "epoch": 0.377137059151172,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.3161,
+      "step": 3322
+    },
+    {
+      "epoch": 0.3772505862610911,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3052,
+      "step": 3323
+    },
+    {
+      "epoch": 0.37736411337101017,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2994,
+      "step": 3324
+    },
+    {
+      "epoch": 0.37747764048092924,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.3116,
+      "step": 3325
+    },
+    {
+      "epoch": 0.3775911675908483,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3029,
+      "step": 3326
+    },
+    {
+      "epoch": 0.3777046947007674,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.317,
+      "step": 3327
+    },
+    {
+      "epoch": 0.37781822181068647,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.3163,
+      "step": 3328
+    },
+    {
+      "epoch": 0.37793174892060555,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.3218,
+      "step": 3329
+    },
+    {
+      "epoch": 0.3780452760305246,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3068,
+      "step": 3330
+    },
+    {
+      "epoch": 0.3781588031404437,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2996,
+      "step": 3331
+    },
+    {
+      "epoch": 0.3782723302503628,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.3009,
+      "step": 3332
+    },
+    {
+      "epoch": 0.37838585736028185,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3158,
+      "step": 3333
+    },
+    {
+      "epoch": 0.37849938447020093,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2897,
+      "step": 3334
+    },
+    {
+      "epoch": 0.37861291158012,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2965,
+      "step": 3335
+    },
+    {
+      "epoch": 0.3787264386900391,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.3016,
+      "step": 3336
+    },
+    {
+      "epoch": 0.37883996579995816,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3235,
+      "step": 3337
+    },
+    {
+      "epoch": 0.37895349290987723,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 5.2885,
+      "step": 3338
+    },
+    {
+      "epoch": 0.3790670200197963,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.3143,
+      "step": 3339
+    },
+    {
+      "epoch": 0.3791805471297154,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2938,
+      "step": 3340
+    },
+    {
+      "epoch": 0.37929407423963446,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3177,
+      "step": 3341
+    },
+    {
+      "epoch": 0.37940760134955354,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.3067,
+      "step": 3342
+    },
+    {
+      "epoch": 0.3795211284594726,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3083,
+      "step": 3343
+    },
+    {
+      "epoch": 0.3796346555693917,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.3052,
+      "step": 3344
+    },
+    {
+      "epoch": 0.37974818267931076,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.2903,
+      "step": 3345
+    },
+    {
+      "epoch": 0.37986170978922984,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.3206,
+      "step": 3346
+    },
+    {
+      "epoch": 0.3799752368991489,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.3173,
+      "step": 3347
+    },
+    {
+      "epoch": 0.380088764009068,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3131,
+      "step": 3348
+    },
+    {
+      "epoch": 0.38020229111898707,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3136,
+      "step": 3349
+    },
+    {
+      "epoch": 0.38031581822890614,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.3236,
+      "step": 3350
+    },
+    {
+      "epoch": 0.3804293453388252,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3102,
+      "step": 3351
+    },
+    {
+      "epoch": 0.3805428724487443,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.3174,
+      "step": 3352
+    },
+    {
+      "epoch": 0.38065639955866337,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.3144,
+      "step": 3353
+    },
+    {
+      "epoch": 0.38076992666858245,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.2903,
+      "step": 3354
+    },
+    {
+      "epoch": 0.3808834537785015,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3128,
+      "step": 3355
+    },
+    {
+      "epoch": 0.3809969808884206,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2917,
+      "step": 3356
+    },
+    {
+      "epoch": 0.3811105079983397,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 5.3145,
+      "step": 3357
+    },
+    {
+      "epoch": 0.38122403510825875,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2955,
+      "step": 3358
+    },
+    {
+      "epoch": 0.38133756221817783,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 5.2991,
+      "step": 3359
+    },
+    {
+      "epoch": 0.3814510893280969,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.3034,
+      "step": 3360
+    },
+    {
+      "epoch": 0.381564616438016,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.3114,
+      "step": 3361
+    },
+    {
+      "epoch": 0.38167814354793506,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.3143,
+      "step": 3362
+    },
+    {
+      "epoch": 0.38179167065785413,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2907,
+      "step": 3363
+    },
+    {
+      "epoch": 0.3819051977677732,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3209,
+      "step": 3364
+    },
+    {
+      "epoch": 0.3820187248776923,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2979,
+      "step": 3365
+    },
+    {
+      "epoch": 0.38213225198761136,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2874,
+      "step": 3366
+    },
+    {
+      "epoch": 0.38224577909753044,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.3126,
+      "step": 3367
+    },
+    {
+      "epoch": 0.3823593062074495,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2988,
+      "step": 3368
+    },
+    {
+      "epoch": 0.3824728333173686,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.3126,
+      "step": 3369
+    },
+    {
+      "epoch": 0.38258636042728766,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3096,
+      "step": 3370
+    },
+    {
+      "epoch": 0.38269988753720674,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.3,
+      "step": 3371
+    },
+    {
+      "epoch": 0.3828134146471258,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2919,
+      "step": 3372
+    },
+    {
+      "epoch": 0.3829269417570449,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3093,
+      "step": 3373
+    },
+    {
+      "epoch": 0.38304046886696397,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3092,
+      "step": 3374
+    },
+    {
+      "epoch": 0.38315399597688304,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3262,
+      "step": 3375
+    },
+    {
+      "epoch": 0.3832675230868021,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.302,
+      "step": 3376
+    },
+    {
+      "epoch": 0.3833810501967212,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3078,
+      "step": 3377
+    },
+    {
+      "epoch": 0.38349457730664027,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2978,
+      "step": 3378
+    },
+    {
+      "epoch": 0.38360810441655935,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3158,
+      "step": 3379
+    },
+    {
+      "epoch": 0.3837216315264784,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3061,
+      "step": 3380
+    },
+    {
+      "epoch": 0.3838351586363975,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.2861,
+      "step": 3381
+    },
+    {
+      "epoch": 0.3839486857463166,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.3241,
+      "step": 3382
+    },
+    {
+      "epoch": 0.38406221285623565,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.2907,
+      "step": 3383
+    },
+    {
+      "epoch": 0.38417573996615473,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.309,
+      "step": 3384
+    },
+    {
+      "epoch": 0.3842892670760738,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.3131,
+      "step": 3385
+    },
+    {
+      "epoch": 0.3844027941859929,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.3057,
+      "step": 3386
+    },
+    {
+      "epoch": 0.38451632129591196,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3159,
+      "step": 3387
+    },
+    {
+      "epoch": 0.38462984840583103,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3073,
+      "step": 3388
+    },
+    {
+      "epoch": 0.3847433755157501,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.3024,
+      "step": 3389
+    },
+    {
+      "epoch": 0.3848569026256692,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.3007,
+      "step": 3390
+    },
+    {
+      "epoch": 0.38497042973558826,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.3183,
+      "step": 3391
+    },
+    {
+      "epoch": 0.38508395684550734,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3184,
+      "step": 3392
+    },
+    {
+      "epoch": 0.3851974839554264,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2942,
+      "step": 3393
+    },
+    {
+      "epoch": 0.3853110110653455,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.306,
+      "step": 3394
+    },
+    {
+      "epoch": 0.38542453817526456,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.306,
+      "step": 3395
+    },
+    {
+      "epoch": 0.38553806528518364,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2898,
+      "step": 3396
+    },
+    {
+      "epoch": 0.3856515923951027,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.3073,
+      "step": 3397
+    },
+    {
+      "epoch": 0.3857651195050218,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2993,
+      "step": 3398
+    },
+    {
+      "epoch": 0.38587864661494087,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.316,
+      "step": 3399
+    },
+    {
+      "epoch": 0.38599217372485994,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2966,
+      "step": 3400
+    },
+    {
+      "epoch": 0.386105700834779,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 5.2976,
+      "step": 3401
+    },
+    {
+      "epoch": 0.3862192279446981,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.2977,
+      "step": 3402
+    },
+    {
+      "epoch": 0.38633275505461717,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.3054,
+      "step": 3403
+    },
+    {
+      "epoch": 0.38644628216453625,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3014,
+      "step": 3404
+    },
+    {
+      "epoch": 0.3865598092744553,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2918,
+      "step": 3405
+    },
+    {
+      "epoch": 0.3866733363843744,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2981,
+      "step": 3406
+    },
+    {
+      "epoch": 0.3867868634942935,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.3097,
+      "step": 3407
+    },
+    {
+      "epoch": 0.38690039060421255,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.302,
+      "step": 3408
+    },
+    {
+      "epoch": 0.38701391771413163,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.3007,
+      "step": 3409
+    },
+    {
+      "epoch": 0.3871274448240507,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.3025,
+      "step": 3410
+    },
+    {
+      "epoch": 0.3872409719339698,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3125,
+      "step": 3411
+    },
+    {
+      "epoch": 0.38735449904388886,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.3009,
+      "step": 3412
+    },
+    {
+      "epoch": 0.38746802615380793,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2904,
+      "step": 3413
+    },
+    {
+      "epoch": 0.387581553263727,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3,
+      "step": 3414
+    },
+    {
+      "epoch": 0.3876950803736461,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3193,
+      "step": 3415
+    },
+    {
+      "epoch": 0.38780860748356516,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3055,
+      "step": 3416
+    },
+    {
+      "epoch": 0.38792213459348424,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.3106,
+      "step": 3417
+    },
+    {
+      "epoch": 0.3880356617034033,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2826,
+      "step": 3418
+    },
+    {
+      "epoch": 0.3881491888133224,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2808,
+      "step": 3419
+    },
+    {
+      "epoch": 0.38826271592324146,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3091,
+      "step": 3420
+    },
+    {
+      "epoch": 0.38837624303316054,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.308,
+      "step": 3421
+    },
+    {
+      "epoch": 0.3884897701430796,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3116,
+      "step": 3422
+    },
+    {
+      "epoch": 0.3886032972529987,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2926,
+      "step": 3423
+    },
+    {
+      "epoch": 0.38871682436291777,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3214,
+      "step": 3424
+    },
+    {
+      "epoch": 0.38883035147283684,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.32,
+      "step": 3425
+    },
+    {
+      "epoch": 0.3889438785827559,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3081,
+      "step": 3426
+    },
+    {
+      "epoch": 0.389057405692675,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 5.3099,
+      "step": 3427
+    },
+    {
+      "epoch": 0.38917093280259407,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 5.3099,
+      "step": 3428
+    },
+    {
+      "epoch": 0.38928445991251315,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.3123,
+      "step": 3429
+    },
+    {
+      "epoch": 0.3893979870224322,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3189,
+      "step": 3430
+    },
+    {
+      "epoch": 0.3895115141323513,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2955,
+      "step": 3431
+    },
+    {
+      "epoch": 0.3896250412422704,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2989,
+      "step": 3432
+    },
+    {
+      "epoch": 0.38973856835218945,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.3033,
+      "step": 3433
+    },
+    {
+      "epoch": 0.3898520954621085,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3075,
+      "step": 3434
+    },
+    {
+      "epoch": 0.3899656225720276,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3146,
+      "step": 3435
+    },
+    {
+      "epoch": 0.3900791496819467,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2914,
+      "step": 3436
+    },
+    {
+      "epoch": 0.39019267679186576,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3067,
+      "step": 3437
+    },
+    {
+      "epoch": 0.39030620390178483,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2906,
+      "step": 3438
+    },
+    {
+      "epoch": 0.39041973101170396,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.296,
+      "step": 3439
+    },
+    {
+      "epoch": 0.39053325812162304,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3257,
+      "step": 3440
+    },
+    {
+      "epoch": 0.3906467852315421,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 5.3018,
+      "step": 3441
+    },
+    {
+      "epoch": 0.3907603123414612,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 5.2783,
+      "step": 3442
+    },
+    {
+      "epoch": 0.39087383945138027,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3081,
+      "step": 3443
+    },
+    {
+      "epoch": 0.39098736656129934,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 5.268,
+      "step": 3444
+    },
+    {
+      "epoch": 0.3911008936712184,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.3024,
+      "step": 3445
+    },
+    {
+      "epoch": 0.3912144207811375,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.286,
+      "step": 3446
+    },
+    {
+      "epoch": 0.39132794789105657,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.3,
+      "step": 3447
+    },
+    {
+      "epoch": 0.39144147500097565,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.3078,
+      "step": 3448
+    },
+    {
+      "epoch": 0.3915550021108947,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 5.3012,
+      "step": 3449
+    },
+    {
+      "epoch": 0.3916685292208138,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.3216,
+      "step": 3450
+    },
+    {
+      "epoch": 0.3917820563307329,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.2945,
+      "step": 3451
+    },
+    {
+      "epoch": 0.39189558344065195,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.299,
+      "step": 3452
+    },
+    {
+      "epoch": 0.392009110550571,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2933,
+      "step": 3453
+    },
+    {
+      "epoch": 0.3921226376604901,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.3217,
+      "step": 3454
+    },
+    {
+      "epoch": 0.3922361647704092,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 5.3,
+      "step": 3455
+    },
+    {
+      "epoch": 0.39234969188032826,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.002,
+      "loss": 5.2957,
+      "step": 3456
+    },
+    {
+      "epoch": 0.39246321899024733,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.3054,
+      "step": 3457
+    },
+    {
+      "epoch": 0.3925767461001664,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002,
+      "loss": 5.2942,
+      "step": 3458
+    },
+    {
+      "epoch": 0.3926902732100855,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2916,
+      "step": 3459
+    },
+    {
+      "epoch": 0.39280380032000456,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.3069,
+      "step": 3460
+    },
+    {
+      "epoch": 0.39291732742992364,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.3339,
+      "step": 3461
+    },
+    {
+      "epoch": 0.3930308545398427,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2988,
+      "step": 3462
+    },
+    {
+      "epoch": 0.3931443816497618,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2844,
+      "step": 3463
+    },
+    {
+      "epoch": 0.39325790875968086,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.293,
+      "step": 3464
+    },
+    {
+      "epoch": 0.39337143586959994,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 5.3167,
+      "step": 3465
+    },
+    {
+      "epoch": 0.393484962979519,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 5.2966,
+      "step": 3466
+    },
+    {
+      "epoch": 0.3935984900894381,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.3009,
+      "step": 3467
+    },
+    {
+      "epoch": 0.39371201719935717,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.3213,
+      "step": 3468
+    },
+    {
+      "epoch": 0.39382554430927624,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3021,
+      "step": 3469
+    },
+    {
+      "epoch": 0.3939390714191953,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3151,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3940525985291144,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3003,
+      "step": 3471
+    },
+    {
+      "epoch": 0.39416612563903347,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2806,
+      "step": 3472
+    },
+    {
+      "epoch": 0.39427965274895255,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2983,
+      "step": 3473
+    },
+    {
+      "epoch": 0.3943931798588716,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.2978,
+      "step": 3474
+    },
+    {
+      "epoch": 0.3945067069687907,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.3034,
+      "step": 3475
+    },
+    {
+      "epoch": 0.3946202340787098,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.3052,
+      "step": 3476
+    },
+    {
+      "epoch": 0.39473376118862885,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.3185,
+      "step": 3477
+    },
+    {
+      "epoch": 0.3948472882985479,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.3108,
+      "step": 3478
+    },
+    {
+      "epoch": 0.394960815408467,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.3001,
+      "step": 3479
+    },
+    {
+      "epoch": 0.3950743425183861,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3039,
+      "step": 3480
+    },
+    {
+      "epoch": 0.39518786962830516,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3097,
+      "step": 3481
+    },
+    {
+      "epoch": 0.39530139673822423,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.305,
+      "step": 3482
+    },
+    {
+      "epoch": 0.3954149238481433,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2938,
+      "step": 3483
+    },
+    {
+      "epoch": 0.3955284509580624,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2985,
+      "step": 3484
+    },
+    {
+      "epoch": 0.39564197806798146,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.3001,
+      "step": 3485
+    },
+    {
+      "epoch": 0.39575550517790054,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2935,
+      "step": 3486
+    },
+    {
+      "epoch": 0.3958690322878196,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2927,
+      "step": 3487
+    },
+    {
+      "epoch": 0.3959825593977387,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3173,
+      "step": 3488
+    },
+    {
+      "epoch": 0.39609608650765776,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.298,
+      "step": 3489
+    },
+    {
+      "epoch": 0.39620961361757684,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.3123,
+      "step": 3490
+    },
+    {
+      "epoch": 0.3963231407274959,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2835,
+      "step": 3491
+    },
+    {
+      "epoch": 0.396436667837415,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2936,
+      "step": 3492
+    },
+    {
+      "epoch": 0.39655019494733407,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2981,
+      "step": 3493
+    },
+    {
+      "epoch": 0.39666372205725314,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.3031,
+      "step": 3494
+    },
+    {
+      "epoch": 0.3967772491671722,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2965,
+      "step": 3495
+    },
+    {
+      "epoch": 0.3968907762770913,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 5.3012,
+      "step": 3496
+    },
+    {
+      "epoch": 0.39700430338701037,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.29,
+      "step": 3497
+    },
+    {
+      "epoch": 0.39711783049692945,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.3111,
+      "step": 3498
+    },
+    {
+      "epoch": 0.3972313576068485,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.3024,
+      "step": 3499
+    },
+    {
+      "epoch": 0.3973448847167676,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 5.3032,
+      "step": 3500
+    },
+    {
+      "epoch": 0.3974584118266867,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3017,
+      "step": 3501
+    },
+    {
+      "epoch": 0.39757193893660575,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3034,
+      "step": 3502
+    },
+    {
+      "epoch": 0.3976854660465248,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3329,
+      "step": 3503
+    },
+    {
+      "epoch": 0.3977989931564439,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3098,
+      "step": 3504
+    },
+    {
+      "epoch": 0.397912520266363,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.295,
+      "step": 3505
+    },
+    {
+      "epoch": 0.39802604737628206,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3028,
+      "step": 3506
+    },
+    {
+      "epoch": 0.39813957448620113,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2981,
+      "step": 3507
+    },
+    {
+      "epoch": 0.3982531015961202,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2944,
+      "step": 3508
+    },
+    {
+      "epoch": 0.3983666287060393,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.3039,
+      "step": 3509
+    },
+    {
+      "epoch": 0.39848015581595836,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.285,
+      "step": 3510
+    },
+    {
+      "epoch": 0.39859368292587743,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.317,
+      "step": 3511
+    },
+    {
+      "epoch": 0.3987072100357965,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2916,
+      "step": 3512
+    },
+    {
+      "epoch": 0.3988207371457156,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.3098,
+      "step": 3513
+    },
+    {
+      "epoch": 0.39893426425563466,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.3116,
+      "step": 3514
+    },
+    {
+      "epoch": 0.39904779136555374,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3072,
+      "step": 3515
+    },
+    {
+      "epoch": 0.3991613184754728,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2904,
+      "step": 3516
+    },
+    {
+      "epoch": 0.3992748455853919,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.2925,
+      "step": 3517
+    },
+    {
+      "epoch": 0.39938837269531097,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2987,
+      "step": 3518
+    },
+    {
+      "epoch": 0.39950189980523004,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.301,
+      "step": 3519
+    },
+    {
+      "epoch": 0.3996154269151491,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2903,
+      "step": 3520
+    },
+    {
+      "epoch": 0.3997289540250682,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.3041,
+      "step": 3521
+    },
+    {
+      "epoch": 0.39984248113498727,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.3045,
+      "step": 3522
+    },
+    {
+      "epoch": 0.39995600824490635,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2957,
+      "step": 3523
+    },
+    {
+      "epoch": 0.4000695353548254,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.3071,
+      "step": 3524
+    },
+    {
+      "epoch": 0.4001830624647445,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.2996,
+      "step": 3525
+    },
+    {
+      "epoch": 0.4002965895746636,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3032,
+      "step": 3526
+    },
+    {
+      "epoch": 0.40041011668458265,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2764,
+      "step": 3527
+    },
+    {
+      "epoch": 0.4005236437945017,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2986,
+      "step": 3528
+    },
+    {
+      "epoch": 0.4006371709044208,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3104,
+      "step": 3529
+    },
+    {
+      "epoch": 0.4007506980143399,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2727,
+      "step": 3530
+    },
+    {
+      "epoch": 0.40086422512425895,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.293,
+      "step": 3531
+    },
+    {
+      "epoch": 0.40097775223417803,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.3032,
+      "step": 3532
+    },
+    {
+      "epoch": 0.4010912793440971,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2937,
+      "step": 3533
+    },
+    {
+      "epoch": 0.4012048064540162,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3117,
+      "step": 3534
+    },
+    {
+      "epoch": 0.40131833356393526,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.3185,
+      "step": 3535
+    },
+    {
+      "epoch": 0.40143186067385433,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2879,
+      "step": 3536
+    },
+    {
+      "epoch": 0.4015453877837734,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2881,
+      "step": 3537
+    },
+    {
+      "epoch": 0.4016589148936925,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.3153,
+      "step": 3538
+    },
+    {
+      "epoch": 0.40177244200361156,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3068,
+      "step": 3539
+    },
+    {
+      "epoch": 0.40188596911353064,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2959,
+      "step": 3540
+    },
+    {
+      "epoch": 0.4019994962234497,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 5.3149,
+      "step": 3541
+    },
+    {
+      "epoch": 0.4021130233333688,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 5.3011,
+      "step": 3542
+    },
+    {
+      "epoch": 0.40222655044328787,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.307,
+      "step": 3543
+    },
+    {
+      "epoch": 0.40234007755320694,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.278,
+      "step": 3544
+    },
+    {
+      "epoch": 0.402453604663126,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2931,
+      "step": 3545
+    },
+    {
+      "epoch": 0.4025671317730451,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2876,
+      "step": 3546
+    },
+    {
+      "epoch": 0.40268065888296417,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3264,
+      "step": 3547
+    },
+    {
+      "epoch": 0.40279418599288325,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3071,
+      "step": 3548
+    },
+    {
+      "epoch": 0.4029077131028023,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3006,
+      "step": 3549
+    },
+    {
+      "epoch": 0.4030212402127214,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2991,
+      "step": 3550
+    },
+    {
+      "epoch": 0.4031347673226405,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.3044,
+      "step": 3551
+    },
+    {
+      "epoch": 0.40324829443255955,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.284,
+      "step": 3552
+    },
+    {
+      "epoch": 0.4033618215424786,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2986,
+      "step": 3553
+    },
+    {
+      "epoch": 0.4034753486523977,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3029,
+      "step": 3554
+    },
+    {
+      "epoch": 0.4035888757623168,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2868,
+      "step": 3555
+    },
+    {
+      "epoch": 0.40370240287223585,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.3079,
+      "step": 3556
+    },
+    {
+      "epoch": 0.40381592998215493,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3003,
+      "step": 3557
+    },
+    {
+      "epoch": 0.403929457092074,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3022,
+      "step": 3558
+    },
+    {
+      "epoch": 0.4040429842019931,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 5.2988,
+      "step": 3559
+    },
+    {
+      "epoch": 0.4041565113119122,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 5.3062,
+      "step": 3560
+    },
+    {
+      "epoch": 0.4042700384218313,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2976,
+      "step": 3561
+    },
+    {
+      "epoch": 0.40438356553175037,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2917,
+      "step": 3562
+    },
+    {
+      "epoch": 0.40449709264166944,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2905,
+      "step": 3563
+    },
+    {
+      "epoch": 0.4046106197515885,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.297,
+      "step": 3564
+    },
+    {
+      "epoch": 0.4047241468615076,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3048,
+      "step": 3565
+    },
+    {
+      "epoch": 0.40483767397142667,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2934,
+      "step": 3566
+    },
+    {
+      "epoch": 0.40495120108134575,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.312,
+      "step": 3567
+    },
+    {
+      "epoch": 0.4050647281912648,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3029,
+      "step": 3568
+    },
+    {
+      "epoch": 0.4051782553011839,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2987,
+      "step": 3569
+    },
+    {
+      "epoch": 0.405291782411103,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.3009,
+      "step": 3570
+    },
+    {
+      "epoch": 0.40540530952102205,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2972,
+      "step": 3571
+    },
+    {
+      "epoch": 0.4055188366309411,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.294,
+      "step": 3572
+    },
+    {
+      "epoch": 0.4056323637408602,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.308,
+      "step": 3573
+    },
+    {
+      "epoch": 0.4057458908507793,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2801,
+      "step": 3574
+    },
+    {
+      "epoch": 0.40585941796069835,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2828,
+      "step": 3575
+    },
+    {
+      "epoch": 0.40597294507061743,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2875,
+      "step": 3576
+    },
+    {
+      "epoch": 0.4060864721805365,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.3125,
+      "step": 3577
+    },
+    {
+      "epoch": 0.4061999992904556,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2972,
+      "step": 3578
+    },
+    {
+      "epoch": 0.40631352640037466,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.3157,
+      "step": 3579
+    },
+    {
+      "epoch": 0.40642705351029373,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3299,
+      "step": 3580
+    },
+    {
+      "epoch": 0.4065405806202128,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2951,
+      "step": 3581
+    },
+    {
+      "epoch": 0.4066541077301319,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2847,
+      "step": 3582
+    },
+    {
+      "epoch": 0.40676763484005096,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2937,
+      "step": 3583
+    },
+    {
+      "epoch": 0.40688116194997004,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2945,
+      "step": 3584
+    },
+    {
+      "epoch": 0.4069946890598891,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3009,
+      "step": 3585
+    },
+    {
+      "epoch": 0.4071082161698082,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2816,
+      "step": 3586
+    },
+    {
+      "epoch": 0.40722174327972727,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.286,
+      "step": 3587
+    },
+    {
+      "epoch": 0.40733527038964634,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2799,
+      "step": 3588
+    },
+    {
+      "epoch": 0.4074487974995654,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3011,
+      "step": 3589
+    },
+    {
+      "epoch": 0.4075623246094845,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.3049,
+      "step": 3590
+    },
+    {
+      "epoch": 0.40767585171940357,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2948,
+      "step": 3591
+    },
+    {
+      "epoch": 0.40778937882932265,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2978,
+      "step": 3592
+    },
+    {
+      "epoch": 0.4079029059392417,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.3154,
+      "step": 3593
+    },
+    {
+      "epoch": 0.4080164330491608,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.3182,
+      "step": 3594
+    },
+    {
+      "epoch": 0.4081299601590799,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2984,
+      "step": 3595
+    },
+    {
+      "epoch": 0.40824348726899895,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3003,
+      "step": 3596
+    },
+    {
+      "epoch": 0.408357014378918,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2951,
+      "step": 3597
+    },
+    {
+      "epoch": 0.4084705414888371,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.3,
+      "step": 3598
+    },
+    {
+      "epoch": 0.4085840685987562,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2876,
+      "step": 3599
+    },
+    {
+      "epoch": 0.40869759570867525,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.3136,
+      "step": 3600
+    },
+    {
+      "epoch": 0.40881112281859433,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.2939,
+      "step": 3601
+    },
+    {
+      "epoch": 0.4089246499285134,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.3082,
+      "step": 3602
+    },
+    {
+      "epoch": 0.4090381770384325,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.3045,
+      "step": 3603
+    },
+    {
+      "epoch": 0.40915170414835156,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.3018,
+      "step": 3604
+    },
+    {
+      "epoch": 0.40926523125827063,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2978,
+      "step": 3605
+    },
+    {
+      "epoch": 0.4093787583681897,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.3034,
+      "step": 3606
+    },
+    {
+      "epoch": 0.4094922854781088,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3036,
+      "step": 3607
+    },
+    {
+      "epoch": 0.40960581258802786,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2857,
+      "step": 3608
+    },
+    {
+      "epoch": 0.40971933969794694,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2897,
+      "step": 3609
+    },
+    {
+      "epoch": 0.409832866807866,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.294,
+      "step": 3610
+    },
+    {
+      "epoch": 0.4099463939177851,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.3061,
+      "step": 3611
+    },
+    {
+      "epoch": 0.41005992102770417,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2969,
+      "step": 3612
+    },
+    {
+      "epoch": 0.41017344813762324,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2939,
+      "step": 3613
+    },
+    {
+      "epoch": 0.4102869752475423,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2876,
+      "step": 3614
+    },
+    {
+      "epoch": 0.4104005023574614,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.3014,
+      "step": 3615
+    },
+    {
+      "epoch": 0.41051402946738047,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.3089,
+      "step": 3616
+    },
+    {
+      "epoch": 0.41062755657729955,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2978,
+      "step": 3617
+    },
+    {
+      "epoch": 0.4107410836872186,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3136,
+      "step": 3618
+    },
+    {
+      "epoch": 0.4108546107971377,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2935,
+      "step": 3619
+    },
+    {
+      "epoch": 0.4109681379070568,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2758,
+      "step": 3620
+    },
+    {
+      "epoch": 0.41108166501697585,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2908,
+      "step": 3621
+    },
+    {
+      "epoch": 0.4111951921268949,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.3007,
+      "step": 3622
+    },
+    {
+      "epoch": 0.411308719236814,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2993,
+      "step": 3623
+    },
+    {
+      "epoch": 0.4114222463467331,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2886,
+      "step": 3624
+    },
+    {
+      "epoch": 0.41153577345665215,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2864,
+      "step": 3625
+    },
+    {
+      "epoch": 0.41164930056657123,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2879,
+      "step": 3626
+    },
+    {
+      "epoch": 0.4117628276764903,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2943,
+      "step": 3627
+    },
+    {
+      "epoch": 0.4118763547864094,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2781,
+      "step": 3628
+    },
+    {
+      "epoch": 0.41198988189632846,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.308,
+      "step": 3629
+    },
+    {
+      "epoch": 0.41210340900624753,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.3124,
+      "step": 3630
+    },
+    {
+      "epoch": 0.4122169361161666,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2937,
+      "step": 3631
+    },
+    {
+      "epoch": 0.4123304632260857,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2811,
+      "step": 3632
+    },
+    {
+      "epoch": 0.41244399033600476,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3014,
+      "step": 3633
+    },
+    {
+      "epoch": 0.41255751744592384,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2935,
+      "step": 3634
+    },
+    {
+      "epoch": 0.4126710445558429,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.3082,
+      "step": 3635
+    },
+    {
+      "epoch": 0.412784571665762,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2898,
+      "step": 3636
+    },
+    {
+      "epoch": 0.41289809877568107,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2917,
+      "step": 3637
+    },
+    {
+      "epoch": 0.41301162588560014,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.291,
+      "step": 3638
+    },
+    {
+      "epoch": 0.4131251529955192,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2981,
+      "step": 3639
+    },
+    {
+      "epoch": 0.4132386801054383,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.2854,
+      "step": 3640
+    },
+    {
+      "epoch": 0.41335220721535737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.002,
+      "loss": 5.3038,
+      "step": 3641
+    },
+    {
+      "epoch": 0.41346573432527645,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.297,
+      "step": 3642
+    },
+    {
+      "epoch": 0.4135792614351955,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.2583,
+      "step": 3643
+    },
+    {
+      "epoch": 0.4136927885451146,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.2793,
+      "step": 3644
+    },
+    {
+      "epoch": 0.4138063156550337,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2934,
+      "step": 3645
+    },
+    {
+      "epoch": 0.41391984276495275,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2896,
+      "step": 3646
+    },
+    {
+      "epoch": 0.4140333698748718,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2604,
+      "step": 3647
+    },
+    {
+      "epoch": 0.4141468969847909,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3003,
+      "step": 3648
+    },
+    {
+      "epoch": 0.41426042409471,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.3102,
+      "step": 3649
+    },
+    {
+      "epoch": 0.41437395120462905,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2925,
+      "step": 3650
+    },
+    {
+      "epoch": 0.41448747831454813,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.3125,
+      "step": 3651
+    },
+    {
+      "epoch": 0.4146010054244672,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.2844,
+      "step": 3652
+    },
+    {
+      "epoch": 0.4147145325343863,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 5.2941,
+      "step": 3653
+    },
+    {
+      "epoch": 0.41482805964430536,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 5.283,
+      "step": 3654
+    },
+    {
+      "epoch": 0.41494158675422443,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 5.307,
+      "step": 3655
+    },
+    {
+      "epoch": 0.4150551138641435,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 5.2789,
+      "step": 3656
+    },
+    {
+      "epoch": 0.4151686409740626,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 5.3001,
+      "step": 3657
+    },
+    {
+      "epoch": 0.41528216808398166,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.2774,
+      "step": 3658
+    },
+    {
+      "epoch": 0.41539569519390074,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3043,
+      "step": 3659
+    },
+    {
+      "epoch": 0.4155092223038198,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3003,
+      "step": 3660
+    },
+    {
+      "epoch": 0.4156227494137389,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2833,
+      "step": 3661
+    },
+    {
+      "epoch": 0.41573627652365797,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.2877,
+      "step": 3662
+    },
+    {
+      "epoch": 0.41584980363357704,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.2747,
+      "step": 3663
+    },
+    {
+      "epoch": 0.4159633307434961,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 5.2928,
+      "step": 3664
+    },
+    {
+      "epoch": 0.4160768578534152,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 5.2731,
+      "step": 3665
+    },
+    {
+      "epoch": 0.41619038496333427,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 5.2717,
+      "step": 3666
+    },
+    {
+      "epoch": 0.41630391207325335,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.3072,
+      "step": 3667
+    },
+    {
+      "epoch": 0.4164174391831724,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2901,
+      "step": 3668
+    },
+    {
+      "epoch": 0.4165309662930915,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.3037,
+      "step": 3669
+    },
+    {
+      "epoch": 0.4166444934030106,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 5.2888,
+      "step": 3670
+    },
+    {
+      "epoch": 0.41675802051292965,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.002,
+      "loss": 5.3021,
+      "step": 3671
+    },
+    {
+      "epoch": 0.4168715476228487,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.002,
+      "loss": 5.3102,
+      "step": 3672
+    },
+    {
+      "epoch": 0.4169850747327678,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.2874,
+      "step": 3673
+    },
+    {
+      "epoch": 0.4170986018426869,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2871,
+      "step": 3674
+    },
+    {
+      "epoch": 0.41721212895260595,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.2872,
+      "step": 3675
+    },
+    {
+      "epoch": 0.41732565606252503,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.291,
+      "step": 3676
+    },
+    {
+      "epoch": 0.4174391831724441,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2627,
+      "step": 3677
+    },
+    {
+      "epoch": 0.4175527102823632,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2903,
+      "step": 3678
+    },
+    {
+      "epoch": 0.41766623739228226,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2882,
+      "step": 3679
+    },
+    {
+      "epoch": 0.41777976450220133,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3105,
+      "step": 3680
+    },
+    {
+      "epoch": 0.41789329161212047,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2805,
+      "step": 3681
+    },
+    {
+      "epoch": 0.41800681872203954,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 5.2992,
+      "step": 3682
+    },
+    {
+      "epoch": 0.4181203458319586,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.3232,
+      "step": 3683
+    },
+    {
+      "epoch": 0.4182338729418777,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.301,
+      "step": 3684
+    },
+    {
+      "epoch": 0.41834740005179677,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.271,
+      "step": 3685
+    },
+    {
+      "epoch": 0.41846092716171585,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.3078,
+      "step": 3686
+    },
+    {
+      "epoch": 0.4185744542716349,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3062,
+      "step": 3687
+    },
+    {
+      "epoch": 0.418687981381554,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2966,
+      "step": 3688
+    },
+    {
+      "epoch": 0.4188015084914731,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.302,
+      "step": 3689
+    },
+    {
+      "epoch": 0.41891503560139215,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.305,
+      "step": 3690
+    },
+    {
+      "epoch": 0.4190285627113112,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.296,
+      "step": 3691
+    },
+    {
+      "epoch": 0.4191420898212303,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2692,
+      "step": 3692
+    },
+    {
+      "epoch": 0.4192556169311494,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3002,
+      "step": 3693
+    },
+    {
+      "epoch": 0.41936914404106845,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2997,
+      "step": 3694
+    },
+    {
+      "epoch": 0.41948267115098753,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2662,
+      "step": 3695
+    },
+    {
+      "epoch": 0.4195961982609066,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.2947,
+      "step": 3696
+    },
+    {
+      "epoch": 0.4197097253708257,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.3024,
+      "step": 3697
+    },
+    {
+      "epoch": 0.41982325248074476,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2864,
+      "step": 3698
+    },
+    {
+      "epoch": 0.41993677959066383,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.2949,
+      "step": 3699
+    },
+    {
+      "epoch": 0.4200503067005829,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2846,
+      "step": 3700
+    },
+    {
+      "epoch": 0.420163833810502,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.3051,
+      "step": 3701
+    },
+    {
+      "epoch": 0.42027736092042106,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.3035,
+      "step": 3702
+    },
+    {
+      "epoch": 0.42039088803034014,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.299,
+      "step": 3703
+    },
+    {
+      "epoch": 0.4205044151402592,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2945,
+      "step": 3704
+    },
+    {
+      "epoch": 0.4206179422501783,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2941,
+      "step": 3705
+    },
+    {
+      "epoch": 0.42073146936009737,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.3235,
+      "step": 3706
+    },
+    {
+      "epoch": 0.42084499647001644,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2808,
+      "step": 3707
+    },
+    {
+      "epoch": 0.4209585235799355,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3046,
+      "step": 3708
+    },
+    {
+      "epoch": 0.4210720506898546,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2938,
+      "step": 3709
+    },
+    {
+      "epoch": 0.42118557779977367,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2916,
+      "step": 3710
+    },
+    {
+      "epoch": 0.42129910490969275,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.3209,
+      "step": 3711
+    },
+    {
+      "epoch": 0.4214126320196118,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.292,
+      "step": 3712
+    },
+    {
+      "epoch": 0.4215261591295309,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2895,
+      "step": 3713
+    },
+    {
+      "epoch": 0.42163968623945,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2933,
+      "step": 3714
+    },
+    {
+      "epoch": 0.42175321334936905,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2959,
+      "step": 3715
+    },
+    {
+      "epoch": 0.4218667404592881,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2868,
+      "step": 3716
+    },
+    {
+      "epoch": 0.4219802675692072,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2855,
+      "step": 3717
+    },
+    {
+      "epoch": 0.4220937946791263,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2936,
+      "step": 3718
+    },
+    {
+      "epoch": 0.42220732178904535,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3081,
+      "step": 3719
+    },
+    {
+      "epoch": 0.42232084889896443,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3117,
+      "step": 3720
+    },
+    {
+      "epoch": 0.4224343760088835,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.3087,
+      "step": 3721
+    },
+    {
+      "epoch": 0.4225479031188026,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.307,
+      "step": 3722
+    },
+    {
+      "epoch": 0.42266143022872166,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.2946,
+      "step": 3723
+    },
+    {
+      "epoch": 0.42277495733864073,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.3042,
+      "step": 3724
+    },
+    {
+      "epoch": 0.4228884844485598,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.2966,
+      "step": 3725
+    },
+    {
+      "epoch": 0.4230020115584789,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.282,
+      "step": 3726
+    },
+    {
+      "epoch": 0.42311553866839796,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.2891,
+      "step": 3727
+    },
+    {
+      "epoch": 0.42322906577831704,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.2962,
+      "step": 3728
+    },
+    {
+      "epoch": 0.4233425928882361,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.301,
+      "step": 3729
+    },
+    {
+      "epoch": 0.4234561199981552,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2796,
+      "step": 3730
+    },
+    {
+      "epoch": 0.42356964710807427,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2876,
+      "step": 3731
+    },
+    {
+      "epoch": 0.42368317421799334,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2825,
+      "step": 3732
+    },
+    {
+      "epoch": 0.4237967013279124,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 5.2911,
+      "step": 3733
+    },
+    {
+      "epoch": 0.4239102284378315,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 5.3028,
+      "step": 3734
+    },
+    {
+      "epoch": 0.42402375554775057,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 5.283,
+      "step": 3735
+    },
+    {
+      "epoch": 0.42413728265766965,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 5.3015,
+      "step": 3736
+    },
+    {
+      "epoch": 0.4242508097675887,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3062,
+      "step": 3737
+    },
+    {
+      "epoch": 0.4243643368775078,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.3115,
+      "step": 3738
+    },
+    {
+      "epoch": 0.4244778639874269,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3011,
+      "step": 3739
+    },
+    {
+      "epoch": 0.42459139109734595,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2951,
+      "step": 3740
+    },
+    {
+      "epoch": 0.424704918207265,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.3041,
+      "step": 3741
+    },
+    {
+      "epoch": 0.4248184453171841,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2975,
+      "step": 3742
+    },
+    {
+      "epoch": 0.4249319724271032,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2928,
+      "step": 3743
+    },
+    {
+      "epoch": 0.42504549953702225,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2868,
+      "step": 3744
+    },
+    {
+      "epoch": 0.42515902664694133,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3056,
+      "step": 3745
+    },
+    {
+      "epoch": 0.4252725537568604,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.3118,
+      "step": 3746
+    },
+    {
+      "epoch": 0.4253860808667795,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2948,
+      "step": 3747
+    },
+    {
+      "epoch": 0.42549960797669856,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.2825,
+      "step": 3748
+    },
+    {
+      "epoch": 0.42561313508661763,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2953,
+      "step": 3749
+    },
+    {
+      "epoch": 0.4257266621965367,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2879,
+      "step": 3750
+    },
+    {
+      "epoch": 0.4258401893064558,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2991,
+      "step": 3751
+    },
+    {
+      "epoch": 0.42595371641637486,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.306,
+      "step": 3752
+    },
+    {
+      "epoch": 0.42606724352629394,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.2958,
+      "step": 3753
+    },
+    {
+      "epoch": 0.426180770636213,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.291,
+      "step": 3754
+    },
+    {
+      "epoch": 0.4262942977461321,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2881,
+      "step": 3755
+    },
+    {
+      "epoch": 0.42640782485605117,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2915,
+      "step": 3756
+    },
+    {
+      "epoch": 0.42652135196597024,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.2984,
+      "step": 3757
+    },
+    {
+      "epoch": 0.4266348790758893,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2982,
+      "step": 3758
+    },
+    {
+      "epoch": 0.4267484061858084,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 5.3014,
+      "step": 3759
+    },
+    {
+      "epoch": 0.42686193329572747,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 5.2723,
+      "step": 3760
+    },
+    {
+      "epoch": 0.42697546040564655,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 5.2868,
+      "step": 3761
+    },
+    {
+      "epoch": 0.4270889875155656,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2959,
+      "step": 3762
+    },
+    {
+      "epoch": 0.4272025146254847,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2997,
+      "step": 3763
+    },
+    {
+      "epoch": 0.4273160417354038,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2964,
+      "step": 3764
+    },
+    {
+      "epoch": 0.42742956884532285,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2984,
+      "step": 3765
+    },
+    {
+      "epoch": 0.4275430959552419,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2935,
+      "step": 3766
+    },
+    {
+      "epoch": 0.427656623065161,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2907,
+      "step": 3767
+    },
+    {
+      "epoch": 0.4277701501750801,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2951,
+      "step": 3768
+    },
+    {
+      "epoch": 0.42788367728499915,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2967,
+      "step": 3769
+    },
+    {
+      "epoch": 0.42799720439491823,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2795,
+      "step": 3770
+    },
+    {
+      "epoch": 0.4281107315048373,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.287,
+      "step": 3771
+    },
+    {
+      "epoch": 0.4282242586147564,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2671,
+      "step": 3772
+    },
+    {
+      "epoch": 0.42833778572467546,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.3001,
+      "step": 3773
+    },
+    {
+      "epoch": 0.42845131283459453,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.2821,
+      "step": 3774
+    },
+    {
+      "epoch": 0.4285648399445136,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.3044,
+      "step": 3775
+    },
+    {
+      "epoch": 0.4286783670544327,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.2747,
+      "step": 3776
+    },
+    {
+      "epoch": 0.42879189416435176,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2895,
+      "step": 3777
+    },
+    {
+      "epoch": 0.42890542127427084,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.294,
+      "step": 3778
+    },
+    {
+      "epoch": 0.4290189483841899,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.3025,
+      "step": 3779
+    },
+    {
+      "epoch": 0.429132475494109,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2991,
+      "step": 3780
+    },
+    {
+      "epoch": 0.42924600260402807,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2889,
+      "step": 3781
+    },
+    {
+      "epoch": 0.42935952971394714,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2656,
+      "step": 3782
+    },
+    {
+      "epoch": 0.4294730568238662,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3098,
+      "step": 3783
+    },
+    {
+      "epoch": 0.4295865839337853,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2785,
+      "step": 3784
+    },
+    {
+      "epoch": 0.42970011104370437,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2952,
+      "step": 3785
+    },
+    {
+      "epoch": 0.42981363815362345,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2848,
+      "step": 3786
+    },
+    {
+      "epoch": 0.4299271652635425,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.2969,
+      "step": 3787
+    },
+    {
+      "epoch": 0.4300406923734616,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.2926,
+      "step": 3788
+    },
+    {
+      "epoch": 0.4301542194833807,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.2859,
+      "step": 3789
+    },
+    {
+      "epoch": 0.43026774659329975,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.2677,
+      "step": 3790
+    },
+    {
+      "epoch": 0.4303812737032188,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2999,
+      "step": 3791
+    },
+    {
+      "epoch": 0.4304948008131379,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3005,
+      "step": 3792
+    },
+    {
+      "epoch": 0.430608327923057,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2825,
+      "step": 3793
+    },
+    {
+      "epoch": 0.43072185503297605,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.299,
+      "step": 3794
+    },
+    {
+      "epoch": 0.43083538214289513,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2773,
+      "step": 3795
+    },
+    {
+      "epoch": 0.4309489092528142,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2918,
+      "step": 3796
+    },
+    {
+      "epoch": 0.4310624363627333,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.3051,
+      "step": 3797
+    },
+    {
+      "epoch": 0.43117596347265236,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2897,
+      "step": 3798
+    },
+    {
+      "epoch": 0.43128949058257143,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2962,
+      "step": 3799
+    },
+    {
+      "epoch": 0.4314030176924905,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2889,
+      "step": 3800
+    },
+    {
+      "epoch": 0.43151654480240964,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2915,
+      "step": 3801
+    },
+    {
+      "epoch": 0.4316300719123287,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.284,
+      "step": 3802
+    },
+    {
+      "epoch": 0.4317435990222478,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2944,
+      "step": 3803
+    },
+    {
+      "epoch": 0.43185712613216687,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2712,
+      "step": 3804
+    },
+    {
+      "epoch": 0.43197065324208594,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 5.31,
+      "step": 3805
+    },
+    {
+      "epoch": 0.432084180352005,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2562,
+      "step": 3806
+    },
+    {
+      "epoch": 0.4321977074619241,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 5.295,
+      "step": 3807
+    },
+    {
+      "epoch": 0.4323112345718432,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2754,
+      "step": 3808
+    },
+    {
+      "epoch": 0.43242476168176225,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.297,
+      "step": 3809
+    },
+    {
+      "epoch": 0.4325382887916813,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.283,
+      "step": 3810
+    },
+    {
+      "epoch": 0.4326518159016004,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.3044,
+      "step": 3811
+    },
+    {
+      "epoch": 0.4327653430115195,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3017,
+      "step": 3812
+    },
+    {
+      "epoch": 0.43287887012143855,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2844,
+      "step": 3813
+    },
+    {
+      "epoch": 0.43299239723135763,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.281,
+      "step": 3814
+    },
+    {
+      "epoch": 0.4331059243412767,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2785,
+      "step": 3815
+    },
+    {
+      "epoch": 0.4332194514511958,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2897,
+      "step": 3816
+    },
+    {
+      "epoch": 0.43333297856111486,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2632,
+      "step": 3817
+    },
+    {
+      "epoch": 0.43344650567103393,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2693,
+      "step": 3818
+    },
+    {
+      "epoch": 0.433560032780953,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.2938,
+      "step": 3819
+    },
+    {
+      "epoch": 0.4336735598908721,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2851,
+      "step": 3820
+    },
+    {
+      "epoch": 0.43378708700079116,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2821,
+      "step": 3821
+    },
+    {
+      "epoch": 0.43390061411071024,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2816,
+      "step": 3822
+    },
+    {
+      "epoch": 0.4340141412206293,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2922,
+      "step": 3823
+    },
+    {
+      "epoch": 0.4341276683305484,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2939,
+      "step": 3824
+    },
+    {
+      "epoch": 0.43424119544046746,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.274,
+      "step": 3825
+    },
+    {
+      "epoch": 0.43435472255038654,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2963,
+      "step": 3826
+    },
+    {
+      "epoch": 0.4344682496603056,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.305,
+      "step": 3827
+    },
+    {
+      "epoch": 0.4345817767702247,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.3021,
+      "step": 3828
+    },
+    {
+      "epoch": 0.43469530388014377,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2887,
+      "step": 3829
+    },
+    {
+      "epoch": 0.43480883099006284,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2871,
+      "step": 3830
+    },
+    {
+      "epoch": 0.4349223580999819,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.2864,
+      "step": 3831
+    },
+    {
+      "epoch": 0.435035885209901,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.308,
+      "step": 3832
+    },
+    {
+      "epoch": 0.4351494123198201,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2937,
+      "step": 3833
+    },
+    {
+      "epoch": 0.43526293942973915,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2912,
+      "step": 3834
+    },
+    {
+      "epoch": 0.4353764665396582,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2915,
+      "step": 3835
+    },
+    {
+      "epoch": 0.4354899936495773,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.3117,
+      "step": 3836
+    },
+    {
+      "epoch": 0.4356035207594964,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2952,
+      "step": 3837
+    },
+    {
+      "epoch": 0.43571704786941545,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2897,
+      "step": 3838
+    },
+    {
+      "epoch": 0.43583057497933453,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2984,
+      "step": 3839
+    },
+    {
+      "epoch": 0.4359441020892536,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2844,
+      "step": 3840
+    },
+    {
+      "epoch": 0.4360576291991727,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2814,
+      "step": 3841
+    },
+    {
+      "epoch": 0.43617115630909176,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2826,
+      "step": 3842
+    },
+    {
+      "epoch": 0.43628468341901083,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2849,
+      "step": 3843
+    },
+    {
+      "epoch": 0.4363982105289299,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.3129,
+      "step": 3844
+    },
+    {
+      "epoch": 0.436511737638849,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2702,
+      "step": 3845
+    },
+    {
+      "epoch": 0.43662526474876806,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.307,
+      "step": 3846
+    },
+    {
+      "epoch": 0.43673879185868714,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2912,
+      "step": 3847
+    },
+    {
+      "epoch": 0.4368523189686062,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2948,
+      "step": 3848
+    },
+    {
+      "epoch": 0.4369658460785253,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2832,
+      "step": 3849
+    },
+    {
+      "epoch": 0.43707937318844436,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.285,
+      "step": 3850
+    },
+    {
+      "epoch": 0.43719290029836344,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2932,
+      "step": 3851
+    },
+    {
+      "epoch": 0.4373064274082825,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.3,
+      "step": 3852
+    },
+    {
+      "epoch": 0.4374199545182016,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2956,
+      "step": 3853
+    },
+    {
+      "epoch": 0.43753348162812067,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3149,
+      "step": 3854
+    },
+    {
+      "epoch": 0.43764700873803974,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.3006,
+      "step": 3855
+    },
+    {
+      "epoch": 0.4377605358479588,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2646,
+      "step": 3856
+    },
+    {
+      "epoch": 0.4378740629578779,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.281,
+      "step": 3857
+    },
+    {
+      "epoch": 0.437987590067797,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2853,
+      "step": 3858
+    },
+    {
+      "epoch": 0.43810111717771605,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.3017,
+      "step": 3859
+    },
+    {
+      "epoch": 0.4382146442876351,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2803,
+      "step": 3860
+    },
+    {
+      "epoch": 0.4383281713975542,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2938,
+      "step": 3861
+    },
+    {
+      "epoch": 0.4384416985074733,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.3087,
+      "step": 3862
+    },
+    {
+      "epoch": 0.43855522561739235,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2709,
+      "step": 3863
+    },
+    {
+      "epoch": 0.43866875272731143,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.279,
+      "step": 3864
+    },
+    {
+      "epoch": 0.4387822798372305,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2843,
+      "step": 3865
+    },
+    {
+      "epoch": 0.4388958069471496,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2765,
+      "step": 3866
+    },
+    {
+      "epoch": 0.43900933405706866,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2856,
+      "step": 3867
+    },
+    {
+      "epoch": 0.43912286116698773,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.2983,
+      "step": 3868
+    },
+    {
+      "epoch": 0.4392363882769068,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.2912,
+      "step": 3869
+    },
+    {
+      "epoch": 0.4393499153868259,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.2999,
+      "step": 3870
+    },
+    {
+      "epoch": 0.43946344249674496,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2997,
+      "step": 3871
+    },
+    {
+      "epoch": 0.43957696960666404,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2761,
+      "step": 3872
+    },
+    {
+      "epoch": 0.4396904967165831,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2813,
+      "step": 3873
+    },
+    {
+      "epoch": 0.4398040238265022,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2742,
+      "step": 3874
+    },
+    {
+      "epoch": 0.43991755093642126,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2915,
+      "step": 3875
+    },
+    {
+      "epoch": 0.44003107804634034,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2783,
+      "step": 3876
+    },
+    {
+      "epoch": 0.4401446051562594,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2887,
+      "step": 3877
+    },
+    {
+      "epoch": 0.4402581322661785,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2969,
+      "step": 3878
+    },
+    {
+      "epoch": 0.44037165937609757,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2546,
+      "step": 3879
+    },
+    {
+      "epoch": 0.44048518648601664,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2898,
+      "step": 3880
+    },
+    {
+      "epoch": 0.4405987135959357,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2849,
+      "step": 3881
+    },
+    {
+      "epoch": 0.4407122407058548,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2905,
+      "step": 3882
+    },
+    {
+      "epoch": 0.4408257678157739,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2861,
+      "step": 3883
+    },
+    {
+      "epoch": 0.44093929492569295,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2859,
+      "step": 3884
+    },
+    {
+      "epoch": 0.441052822035612,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2764,
+      "step": 3885
+    },
+    {
+      "epoch": 0.4411663491455311,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2883,
+      "step": 3886
+    },
+    {
+      "epoch": 0.4412798762554502,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2796,
+      "step": 3887
+    },
+    {
+      "epoch": 0.44139340336536925,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2888,
+      "step": 3888
+    },
+    {
+      "epoch": 0.44150693047528833,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2913,
+      "step": 3889
+    },
+    {
+      "epoch": 0.4416204575852074,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2887,
+      "step": 3890
+    },
+    {
+      "epoch": 0.4417339846951265,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.297,
+      "step": 3891
+    },
+    {
+      "epoch": 0.44184751180504556,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2846,
+      "step": 3892
+    },
+    {
+      "epoch": 0.44196103891496463,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2845,
+      "step": 3893
+    },
+    {
+      "epoch": 0.4420745660248837,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.3043,
+      "step": 3894
+    },
+    {
+      "epoch": 0.4421880931348028,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2902,
+      "step": 3895
+    },
+    {
+      "epoch": 0.44230162024472186,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.2924,
+      "step": 3896
+    },
+    {
+      "epoch": 0.44241514735464094,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.3143,
+      "step": 3897
+    },
+    {
+      "epoch": 0.44252867446456,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3083,
+      "step": 3898
+    },
+    {
+      "epoch": 0.4426422015744791,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2736,
+      "step": 3899
+    },
+    {
+      "epoch": 0.44275572868439816,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2645,
+      "step": 3900
+    },
+    {
+      "epoch": 0.44286925579431724,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.2921,
+      "step": 3901
+    },
+    {
+      "epoch": 0.4429827829042363,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2833,
+      "step": 3902
+    },
+    {
+      "epoch": 0.4430963100141554,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.3081,
+      "step": 3903
+    },
+    {
+      "epoch": 0.44320983712407447,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2944,
+      "step": 3904
+    },
+    {
+      "epoch": 0.44332336423399354,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2932,
+      "step": 3905
+    },
+    {
+      "epoch": 0.4434368913439126,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2769,
+      "step": 3906
+    },
+    {
+      "epoch": 0.4435504184538317,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2969,
+      "step": 3907
+    },
+    {
+      "epoch": 0.4436639455637508,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.29,
+      "step": 3908
+    },
+    {
+      "epoch": 0.44377747267366985,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2713,
+      "step": 3909
+    },
+    {
+      "epoch": 0.4438909997835889,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2919,
+      "step": 3910
+    },
+    {
+      "epoch": 0.444004526893508,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.296,
+      "step": 3911
+    },
+    {
+      "epoch": 0.4441180540034271,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2732,
+      "step": 3912
+    },
+    {
+      "epoch": 0.44423158111334615,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2838,
+      "step": 3913
+    },
+    {
+      "epoch": 0.44434510822326523,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2893,
+      "step": 3914
+    },
+    {
+      "epoch": 0.4444586353331843,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2903,
+      "step": 3915
+    },
+    {
+      "epoch": 0.4445721624431034,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.2994,
+      "step": 3916
+    },
+    {
+      "epoch": 0.44468568955302246,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.2762,
+      "step": 3917
+    },
+    {
+      "epoch": 0.44479921666294153,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.3021,
+      "step": 3918
+    },
+    {
+      "epoch": 0.4449127437728606,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.2769,
+      "step": 3919
+    },
+    {
+      "epoch": 0.4450262708827797,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.2904,
+      "step": 3920
+    },
+    {
+      "epoch": 0.44513979799269876,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2882,
+      "step": 3921
+    },
+    {
+      "epoch": 0.4452533251026179,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2838,
+      "step": 3922
+    },
+    {
+      "epoch": 0.44536685221253697,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2931,
+      "step": 3923
+    },
+    {
+      "epoch": 0.44548037932245604,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2976,
+      "step": 3924
+    },
+    {
+      "epoch": 0.4455939064323751,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2809,
+      "step": 3925
+    },
+    {
+      "epoch": 0.4457074335422942,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2923,
+      "step": 3926
+    },
+    {
+      "epoch": 0.44582096065221327,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3006,
+      "step": 3927
+    },
+    {
+      "epoch": 0.44593448776213235,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2877,
+      "step": 3928
+    },
+    {
+      "epoch": 0.4460480148720514,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2896,
+      "step": 3929
+    },
+    {
+      "epoch": 0.4461615419819705,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2913,
+      "step": 3930
+    },
+    {
+      "epoch": 0.4462750690918896,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.251,
+      "step": 3931
+    },
+    {
+      "epoch": 0.44638859620180865,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2739,
+      "step": 3932
+    },
+    {
+      "epoch": 0.4465021233117277,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2822,
+      "step": 3933
+    },
+    {
+      "epoch": 0.4466156504216468,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2813,
+      "step": 3934
+    },
+    {
+      "epoch": 0.4467291775315659,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2803,
+      "step": 3935
+    },
+    {
+      "epoch": 0.44684270464148496,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2984,
+      "step": 3936
+    },
+    {
+      "epoch": 0.44695623175140403,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2748,
+      "step": 3937
+    },
+    {
+      "epoch": 0.4470697588613231,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.287,
+      "step": 3938
+    },
+    {
+      "epoch": 0.4471832859712422,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2815,
+      "step": 3939
+    },
+    {
+      "epoch": 0.44729681308116126,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2993,
+      "step": 3940
+    },
+    {
+      "epoch": 0.44741034019108034,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2724,
+      "step": 3941
+    },
+    {
+      "epoch": 0.4475238673009994,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2768,
+      "step": 3942
+    },
+    {
+      "epoch": 0.4476373944109185,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2772,
+      "step": 3943
+    },
+    {
+      "epoch": 0.44775092152083756,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2781,
+      "step": 3944
+    },
+    {
+      "epoch": 0.44786444863075664,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.284,
+      "step": 3945
+    },
+    {
+      "epoch": 0.4479779757406757,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2891,
+      "step": 3946
+    },
+    {
+      "epoch": 0.4480915028505948,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 5.3101,
+      "step": 3947
+    },
+    {
+      "epoch": 0.44820502996051387,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 5.266,
+      "step": 3948
+    },
+    {
+      "epoch": 0.44831855707043294,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2694,
+      "step": 3949
+    },
+    {
+      "epoch": 0.448432084180352,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.2755,
+      "step": 3950
+    },
+    {
+      "epoch": 0.4485456112902711,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2878,
+      "step": 3951
+    },
+    {
+      "epoch": 0.44865913840019017,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.3013,
+      "step": 3952
+    },
+    {
+      "epoch": 0.44877266551010925,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2806,
+      "step": 3953
+    },
+    {
+      "epoch": 0.4488861926200283,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.3035,
+      "step": 3954
+    },
+    {
+      "epoch": 0.4489997197299474,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2878,
+      "step": 3955
+    },
+    {
+      "epoch": 0.4491132468398665,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2928,
+      "step": 3956
+    },
+    {
+      "epoch": 0.44922677394978555,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2913,
+      "step": 3957
+    },
+    {
+      "epoch": 0.4493403010597046,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2884,
+      "step": 3958
+    },
+    {
+      "epoch": 0.4494538281696237,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.284,
+      "step": 3959
+    },
+    {
+      "epoch": 0.4495673552795428,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2891,
+      "step": 3960
+    },
+    {
+      "epoch": 0.44968088238946186,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.278,
+      "step": 3961
+    },
+    {
+      "epoch": 0.44979440949938093,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2729,
+      "step": 3962
+    },
+    {
+      "epoch": 0.4499079366093,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2766,
+      "step": 3963
+    },
+    {
+      "epoch": 0.4500214637192191,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.3046,
+      "step": 3964
+    },
+    {
+      "epoch": 0.45013499082913816,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2786,
+      "step": 3965
+    },
+    {
+      "epoch": 0.45024851793905724,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2798,
+      "step": 3966
+    },
+    {
+      "epoch": 0.4503620450489763,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2921,
+      "step": 3967
+    },
+    {
+      "epoch": 0.4504755721588954,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2837,
+      "step": 3968
+    },
+    {
+      "epoch": 0.45058909926881446,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.296,
+      "step": 3969
+    },
+    {
+      "epoch": 0.45070262637873354,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.3064,
+      "step": 3970
+    },
+    {
+      "epoch": 0.4508161534886526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.2776,
+      "step": 3971
+    },
+    {
+      "epoch": 0.4509296805985717,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.2819,
+      "step": 3972
+    },
+    {
+      "epoch": 0.45104320770849077,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.2912,
+      "step": 3973
+    },
+    {
+      "epoch": 0.45115673481840984,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2886,
+      "step": 3974
+    },
+    {
+      "epoch": 0.4512702619283289,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3105,
+      "step": 3975
+    },
+    {
+      "epoch": 0.451383789038248,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2944,
+      "step": 3976
+    },
+    {
+      "epoch": 0.45149731614816707,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2867,
+      "step": 3977
+    },
+    {
+      "epoch": 0.45161084325808615,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2758,
+      "step": 3978
+    },
+    {
+      "epoch": 0.4517243703680052,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2941,
+      "step": 3979
+    },
+    {
+      "epoch": 0.4518378974779243,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2723,
+      "step": 3980
+    },
+    {
+      "epoch": 0.4519514245878434,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2902,
+      "step": 3981
+    },
+    {
+      "epoch": 0.45206495169776245,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.28,
+      "step": 3982
+    },
+    {
+      "epoch": 0.4521784788076815,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2714,
+      "step": 3983
+    },
+    {
+      "epoch": 0.4522920059176006,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.3035,
+      "step": 3984
+    },
+    {
+      "epoch": 0.4524055330275197,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2948,
+      "step": 3985
+    },
+    {
+      "epoch": 0.45251906013743876,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 5.2786,
+      "step": 3986
+    },
+    {
+      "epoch": 0.45263258724735783,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 5.2893,
+      "step": 3987
+    },
+    {
+      "epoch": 0.4527461143572769,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2721,
+      "step": 3988
+    },
+    {
+      "epoch": 0.452859641467196,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2958,
+      "step": 3989
+    },
+    {
+      "epoch": 0.45297316857711506,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2866,
+      "step": 3990
+    },
+    {
+      "epoch": 0.45308669568703414,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.3043,
+      "step": 3991
+    },
+    {
+      "epoch": 0.4532002227969532,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2802,
+      "step": 3992
+    },
+    {
+      "epoch": 0.4533137499068723,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2818,
+      "step": 3993
+    },
+    {
+      "epoch": 0.45342727701679136,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.3004,
+      "step": 3994
+    },
+    {
+      "epoch": 0.45354080412671044,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2793,
+      "step": 3995
+    },
+    {
+      "epoch": 0.4536543312366295,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2803,
+      "step": 3996
+    },
+    {
+      "epoch": 0.4537678583465486,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2796,
+      "step": 3997
+    },
+    {
+      "epoch": 0.45388138545646767,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2712,
+      "step": 3998
+    },
+    {
+      "epoch": 0.45399491256638674,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2955,
+      "step": 3999
+    },
+    {
+      "epoch": 0.4541084396763058,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2819,
+      "step": 4000
+    },
+    {
+      "epoch": 0.4542219667862249,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2817,
+      "step": 4001
+    },
+    {
+      "epoch": 0.45433549389614397,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2827,
+      "step": 4002
+    },
+    {
+      "epoch": 0.45444902100606305,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2932,
+      "step": 4003
+    },
+    {
+      "epoch": 0.4545625481159821,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2668,
+      "step": 4004
+    },
+    {
+      "epoch": 0.4546760752259012,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.2457,
+      "step": 4005
+    },
+    {
+      "epoch": 0.4547896023358203,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.3042,
+      "step": 4006
+    },
+    {
+      "epoch": 0.45490312944573935,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.284,
+      "step": 4007
+    },
+    {
+      "epoch": 0.4550166565556584,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2686,
+      "step": 4008
+    },
+    {
+      "epoch": 0.4551301836655775,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2903,
+      "step": 4009
+    },
+    {
+      "epoch": 0.4552437107754966,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2873,
+      "step": 4010
+    },
+    {
+      "epoch": 0.45535723788541566,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2706,
+      "step": 4011
+    },
+    {
+      "epoch": 0.45547076499533473,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.301,
+      "step": 4012
+    },
+    {
+      "epoch": 0.4555842921052538,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2909,
+      "step": 4013
+    },
+    {
+      "epoch": 0.4556978192151729,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2811,
+      "step": 4014
+    },
+    {
+      "epoch": 0.45581134632509196,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.289,
+      "step": 4015
+    },
+    {
+      "epoch": 0.45592487343501104,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2723,
+      "step": 4016
+    },
+    {
+      "epoch": 0.4560384005449301,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 5.2955,
+      "step": 4017
+    },
+    {
+      "epoch": 0.4561519276548492,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.292,
+      "step": 4018
+    },
+    {
+      "epoch": 0.45626545476476826,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2962,
+      "step": 4019
+    },
+    {
+      "epoch": 0.45637898187468734,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2682,
+      "step": 4020
+    },
+    {
+      "epoch": 0.4564925089846064,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.002,
+      "loss": 5.2834,
+      "step": 4021
+    },
+    {
+      "epoch": 0.4566060360945255,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.2766,
+      "step": 4022
+    },
+    {
+      "epoch": 0.45671956320444457,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.2896,
+      "step": 4023
+    },
+    {
+      "epoch": 0.45683309031436364,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2806,
+      "step": 4024
+    },
+    {
+      "epoch": 0.4569466174242827,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2889,
+      "step": 4025
+    },
+    {
+      "epoch": 0.4570601445342018,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2772,
+      "step": 4026
+    },
+    {
+      "epoch": 0.45717367164412087,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2828,
+      "step": 4027
+    },
+    {
+      "epoch": 0.45728719875403995,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2946,
+      "step": 4028
+    },
+    {
+      "epoch": 0.457400725863959,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2919,
+      "step": 4029
+    },
+    {
+      "epoch": 0.4575142529738781,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2979,
+      "step": 4030
+    },
+    {
+      "epoch": 0.4576277800837972,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2864,
+      "step": 4031
+    },
+    {
+      "epoch": 0.45774130719371625,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2572,
+      "step": 4032
+    },
+    {
+      "epoch": 0.4578548343036353,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2861,
+      "step": 4033
+    },
+    {
+      "epoch": 0.4579683614135544,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2833,
+      "step": 4034
+    },
+    {
+      "epoch": 0.4580818885234735,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.2831,
+      "step": 4035
+    },
+    {
+      "epoch": 0.45819541563339256,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.2842,
+      "step": 4036
+    },
+    {
+      "epoch": 0.45830894274331163,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2789,
+      "step": 4037
+    },
+    {
+      "epoch": 0.4584224698532307,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3044,
+      "step": 4038
+    },
+    {
+      "epoch": 0.4585359969631498,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2897,
+      "step": 4039
+    },
+    {
+      "epoch": 0.45864952407306886,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2436,
+      "step": 4040
+    },
+    {
+      "epoch": 0.45876305118298794,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2704,
+      "step": 4041
+    },
+    {
+      "epoch": 0.458876578292907,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.26,
+      "step": 4042
+    },
+    {
+      "epoch": 0.45899010540282614,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2919,
+      "step": 4043
+    },
+    {
+      "epoch": 0.4591036325127452,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 5.2863,
+      "step": 4044
+    },
+    {
+      "epoch": 0.4592171596226643,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 5.3056,
+      "step": 4045
+    },
+    {
+      "epoch": 0.45933068673258337,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.002,
+      "loss": 5.2907,
+      "step": 4046
+    },
+    {
+      "epoch": 0.45944421384250245,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 5.2774,
+      "step": 4047
+    },
+    {
+      "epoch": 0.4595577409524215,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2928,
+      "step": 4048
+    },
+    {
+      "epoch": 0.4596712680623406,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.3004,
+      "step": 4049
+    },
+    {
+      "epoch": 0.4597847951722597,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2973,
+      "step": 4050
+    },
+    {
+      "epoch": 0.45989832228217875,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2725,
+      "step": 4051
+    },
+    {
+      "epoch": 0.4600118493920978,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2891,
+      "step": 4052
+    },
+    {
+      "epoch": 0.4601253765020169,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2991,
+      "step": 4053
+    },
+    {
+      "epoch": 0.460238903611936,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2829,
+      "step": 4054
+    },
+    {
+      "epoch": 0.46035243072185505,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2852,
+      "step": 4055
+    },
+    {
+      "epoch": 0.46046595783177413,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2875,
+      "step": 4056
+    },
+    {
+      "epoch": 0.4605794849416932,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.3042,
+      "step": 4057
+    },
+    {
+      "epoch": 0.4606930120516123,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2936,
+      "step": 4058
+    },
+    {
+      "epoch": 0.46080653916153136,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2652,
+      "step": 4059
+    },
+    {
+      "epoch": 0.46092006627145043,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2743,
+      "step": 4060
+    },
+    {
+      "epoch": 0.4610335933813695,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2879,
+      "step": 4061
+    },
+    {
+      "epoch": 0.4611471204912886,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2528,
+      "step": 4062
+    },
+    {
+      "epoch": 0.46126064760120766,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.3061,
+      "step": 4063
+    },
+    {
+      "epoch": 0.46137417471112674,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2697,
+      "step": 4064
+    },
+    {
+      "epoch": 0.4614877018210458,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.266,
+      "step": 4065
+    },
+    {
+      "epoch": 0.4616012289309649,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2803,
+      "step": 4066
+    },
+    {
+      "epoch": 0.46171475604088397,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2868,
+      "step": 4067
+    },
+    {
+      "epoch": 0.46182828315080304,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2661,
+      "step": 4068
+    },
+    {
+      "epoch": 0.4619418102607221,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2778,
+      "step": 4069
+    },
+    {
+      "epoch": 0.4620553373706412,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.3047,
+      "step": 4070
+    },
+    {
+      "epoch": 0.46216886448056027,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 5.257,
+      "step": 4071
+    },
+    {
+      "epoch": 0.46228239159047935,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.274,
+      "step": 4072
+    },
+    {
+      "epoch": 0.4623959187003984,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2878,
+      "step": 4073
+    },
+    {
+      "epoch": 0.4625094458103175,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2696,
+      "step": 4074
+    },
+    {
+      "epoch": 0.4626229729202366,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2632,
+      "step": 4075
+    },
+    {
+      "epoch": 0.46273650003015565,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.3005,
+      "step": 4076
+    },
+    {
+      "epoch": 0.4628500271400747,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2812,
+      "step": 4077
+    },
+    {
+      "epoch": 0.4629635542499938,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2902,
+      "step": 4078
+    },
+    {
+      "epoch": 0.4630770813599129,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2795,
+      "step": 4079
+    },
+    {
+      "epoch": 0.46319060846983195,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2712,
+      "step": 4080
+    },
+    {
+      "epoch": 0.46330413557975103,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2934,
+      "step": 4081
+    },
+    {
+      "epoch": 0.4634176626896701,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2806,
+      "step": 4082
+    },
+    {
+      "epoch": 0.4635311897995892,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2759,
+      "step": 4083
+    },
+    {
+      "epoch": 0.46364471690950826,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2846,
+      "step": 4084
+    },
+    {
+      "epoch": 0.46375824401942733,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2951,
+      "step": 4085
+    },
+    {
+      "epoch": 0.4638717711293464,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2864,
+      "step": 4086
+    },
+    {
+      "epoch": 0.4639852982392655,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2733,
+      "step": 4087
+    },
+    {
+      "epoch": 0.46409882534918456,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3129,
+      "step": 4088
+    },
+    {
+      "epoch": 0.46421235245910364,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.2757,
+      "step": 4089
+    },
+    {
+      "epoch": 0.4643258795690227,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2806,
+      "step": 4090
+    },
+    {
+      "epoch": 0.4644394066789418,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2868,
+      "step": 4091
+    },
+    {
+      "epoch": 0.46455293378886087,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2857,
+      "step": 4092
+    },
+    {
+      "epoch": 0.46466646089877994,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2733,
+      "step": 4093
+    },
+    {
+      "epoch": 0.464779988008699,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2803,
+      "step": 4094
+    },
+    {
+      "epoch": 0.4648935151186181,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2663,
+      "step": 4095
+    },
+    {
+      "epoch": 0.46500704222853717,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.002,
+      "loss": 5.2682,
+      "step": 4096
+    },
+    {
+      "epoch": 0.46512056933845625,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 5.2762,
+      "step": 4097
+    },
+    {
+      "epoch": 0.4652340964483753,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2764,
+      "step": 4098
+    },
+    {
+      "epoch": 0.4653476235582944,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.258,
+      "step": 4099
+    },
+    {
+      "epoch": 0.4654611506682135,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2764,
+      "step": 4100
+    },
+    {
+      "epoch": 0.46557467777813255,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2613,
+      "step": 4101
+    },
+    {
+      "epoch": 0.4656882048880516,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2942,
+      "step": 4102
+    },
+    {
+      "epoch": 0.4658017319979707,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2912,
+      "step": 4103
+    },
+    {
+      "epoch": 0.4659152591078898,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2793,
+      "step": 4104
+    },
+    {
+      "epoch": 0.46602878621780885,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2731,
+      "step": 4105
+    },
+    {
+      "epoch": 0.46614231332772793,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2958,
+      "step": 4106
+    },
+    {
+      "epoch": 0.466255840437647,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.249,
+      "step": 4107
+    },
+    {
+      "epoch": 0.4663693675475661,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.261,
+      "step": 4108
+    },
+    {
+      "epoch": 0.46648289465748516,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2911,
+      "step": 4109
+    },
+    {
+      "epoch": 0.46659642176740423,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2865,
+      "step": 4110
+    },
+    {
+      "epoch": 0.4667099488773233,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2802,
+      "step": 4111
+    },
+    {
+      "epoch": 0.4668234759872424,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2738,
+      "step": 4112
+    },
+    {
+      "epoch": 0.46693700309716146,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 5.2809,
+      "step": 4113
+    },
+    {
+      "epoch": 0.46705053020708054,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2591,
+      "step": 4114
+    },
+    {
+      "epoch": 0.4671640573169996,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2916,
+      "step": 4115
+    },
+    {
+      "epoch": 0.4672775844269187,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2841,
+      "step": 4116
+    },
+    {
+      "epoch": 0.46739111153683777,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3,
+      "step": 4117
+    },
+    {
+      "epoch": 0.46750463864675684,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2533,
+      "step": 4118
+    },
+    {
+      "epoch": 0.4676181657566759,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2761,
+      "step": 4119
+    },
+    {
+      "epoch": 0.467731692866595,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2845,
+      "step": 4120
+    },
+    {
+      "epoch": 0.46784521997651407,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2878,
+      "step": 4121
+    },
+    {
+      "epoch": 0.46795874708643315,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2692,
+      "step": 4122
+    },
+    {
+      "epoch": 0.4680722741963522,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2755,
+      "step": 4123
+    },
+    {
+      "epoch": 0.4681858013062713,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2861,
+      "step": 4124
+    },
+    {
+      "epoch": 0.4682993284161904,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2918,
+      "step": 4125
+    },
+    {
+      "epoch": 0.46841285552610945,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2607,
+      "step": 4126
+    },
+    {
+      "epoch": 0.4685263826360285,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2812,
+      "step": 4127
+    },
+    {
+      "epoch": 0.4686399097459476,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2799,
+      "step": 4128
+    },
+    {
+      "epoch": 0.4687534368558667,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2948,
+      "step": 4129
+    },
+    {
+      "epoch": 0.46886696396578575,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2733,
+      "step": 4130
+    },
+    {
+      "epoch": 0.46898049107570483,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2945,
+      "step": 4131
+    },
+    {
+      "epoch": 0.4690940181856239,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2952,
+      "step": 4132
+    },
+    {
+      "epoch": 0.469207545295543,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.3019,
+      "step": 4133
+    },
+    {
+      "epoch": 0.46932107240546206,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2648,
+      "step": 4134
+    },
+    {
+      "epoch": 0.46943459951538113,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2838,
+      "step": 4135
+    },
+    {
+      "epoch": 0.4695481266253002,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2817,
+      "step": 4136
+    },
+    {
+      "epoch": 0.4696616537352193,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2793,
+      "step": 4137
+    },
+    {
+      "epoch": 0.46977518084513836,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2858,
+      "step": 4138
+    },
+    {
+      "epoch": 0.46988870795505744,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2728,
+      "step": 4139
+    },
+    {
+      "epoch": 0.4700022350649765,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2713,
+      "step": 4140
+    },
+    {
+      "epoch": 0.4701157621748956,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2692,
+      "step": 4141
+    },
+    {
+      "epoch": 0.47022928928481467,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.276,
+      "step": 4142
+    },
+    {
+      "epoch": 0.47034281639473374,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2783,
+      "step": 4143
+    },
+    {
+      "epoch": 0.4704563435046528,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2683,
+      "step": 4144
+    },
+    {
+      "epoch": 0.4705698706145719,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2619,
+      "step": 4145
+    },
+    {
+      "epoch": 0.47068339772449097,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.271,
+      "step": 4146
+    },
+    {
+      "epoch": 0.47079692483441005,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2878,
+      "step": 4147
+    },
+    {
+      "epoch": 0.4709104519443291,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2918,
+      "step": 4148
+    },
+    {
+      "epoch": 0.4710239790542482,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2767,
+      "step": 4149
+    },
+    {
+      "epoch": 0.4711375061641673,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2808,
+      "step": 4150
+    },
+    {
+      "epoch": 0.47125103327408635,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.273,
+      "step": 4151
+    },
+    {
+      "epoch": 0.4713645603840054,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2629,
+      "step": 4152
+    },
+    {
+      "epoch": 0.4714780874939245,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2921,
+      "step": 4153
+    },
+    {
+      "epoch": 0.4715916146038436,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2972,
+      "step": 4154
+    },
+    {
+      "epoch": 0.47170514171376265,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.3023,
+      "step": 4155
+    },
+    {
+      "epoch": 0.47181866882368173,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.2756,
+      "step": 4156
+    },
+    {
+      "epoch": 0.4719321959336008,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2768,
+      "step": 4157
+    },
+    {
+      "epoch": 0.4720457230435199,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2821,
+      "step": 4158
+    },
+    {
+      "epoch": 0.47215925015343896,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2838,
+      "step": 4159
+    },
+    {
+      "epoch": 0.47227277726335803,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2825,
+      "step": 4160
+    },
+    {
+      "epoch": 0.4723863043732771,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.278,
+      "step": 4161
+    },
+    {
+      "epoch": 0.4724998314831962,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2833,
+      "step": 4162
+    },
+    {
+      "epoch": 0.4726133585931153,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2699,
+      "step": 4163
+    },
+    {
+      "epoch": 0.4727268857030344,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.3013,
+      "step": 4164
+    },
+    {
+      "epoch": 0.47284041281295347,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2818,
+      "step": 4165
+    },
+    {
+      "epoch": 0.47295393992287255,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2717,
+      "step": 4166
+    },
+    {
+      "epoch": 0.4730674670327916,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2927,
+      "step": 4167
+    },
+    {
+      "epoch": 0.4731809941427107,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2719,
+      "step": 4168
+    },
+    {
+      "epoch": 0.4732945212526298,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2836,
+      "step": 4169
+    },
+    {
+      "epoch": 0.47340804836254885,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2851,
+      "step": 4170
+    },
+    {
+      "epoch": 0.4735215754724679,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.29,
+      "step": 4171
+    },
+    {
+      "epoch": 0.473635102582387,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2873,
+      "step": 4172
+    },
+    {
+      "epoch": 0.4737486296923061,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.3,
+      "step": 4173
+    },
+    {
+      "epoch": 0.47386215680222515,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2803,
+      "step": 4174
+    },
+    {
+      "epoch": 0.47397568391214423,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.2982,
+      "step": 4175
+    },
+    {
+      "epoch": 0.4740892110220633,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.2887,
+      "step": 4176
+    },
+    {
+      "epoch": 0.4742027381319824,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2971,
+      "step": 4177
+    },
+    {
+      "epoch": 0.47431626524190146,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2501,
+      "step": 4178
+    },
+    {
+      "epoch": 0.47442979235182053,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2807,
+      "step": 4179
+    },
+    {
+      "epoch": 0.4745433194617396,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2766,
+      "step": 4180
+    },
+    {
+      "epoch": 0.4746568465716587,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2766,
+      "step": 4181
+    },
+    {
+      "epoch": 0.47477037368157776,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2923,
+      "step": 4182
+    },
+    {
+      "epoch": 0.47488390079149684,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2686,
+      "step": 4183
+    },
+    {
+      "epoch": 0.4749974279014159,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2881,
+      "step": 4184
+    },
+    {
+      "epoch": 0.475110955011335,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2908,
+      "step": 4185
+    },
+    {
+      "epoch": 0.47522448212125407,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2915,
+      "step": 4186
+    },
+    {
+      "epoch": 0.47533800923117314,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2809,
+      "step": 4187
+    },
+    {
+      "epoch": 0.4754515363410922,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.277,
+      "step": 4188
+    },
+    {
+      "epoch": 0.4755650634510113,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2705,
+      "step": 4189
+    },
+    {
+      "epoch": 0.47567859056093037,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2763,
+      "step": 4190
+    },
+    {
+      "epoch": 0.47579211767084945,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2898,
+      "step": 4191
+    },
+    {
+      "epoch": 0.4759056447807685,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2893,
+      "step": 4192
+    },
+    {
+      "epoch": 0.4760191718906876,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2652,
+      "step": 4193
+    },
+    {
+      "epoch": 0.4761326990006067,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2725,
+      "step": 4194
+    },
+    {
+      "epoch": 0.47624622611052575,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2617,
+      "step": 4195
+    },
+    {
+      "epoch": 0.4763597532204448,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2768,
+      "step": 4196
+    },
+    {
+      "epoch": 0.4764732803303639,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2846,
+      "step": 4197
+    },
+    {
+      "epoch": 0.476586807440283,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2661,
+      "step": 4198
+    },
+    {
+      "epoch": 0.47670033455020205,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2783,
+      "step": 4199
+    },
+    {
+      "epoch": 0.47681386166012113,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.278,
+      "step": 4200
+    },
+    {
+      "epoch": 0.4769273887700402,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2849,
+      "step": 4201
+    },
+    {
+      "epoch": 0.4770409158799593,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2701,
+      "step": 4202
+    },
+    {
+      "epoch": 0.47715444298987836,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2718,
+      "step": 4203
+    },
+    {
+      "epoch": 0.47726797009979743,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2737,
+      "step": 4204
+    },
+    {
+      "epoch": 0.4773814972097165,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2646,
+      "step": 4205
+    },
+    {
+      "epoch": 0.4774950243196356,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2951,
+      "step": 4206
+    },
+    {
+      "epoch": 0.47760855142955466,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2739,
+      "step": 4207
+    },
+    {
+      "epoch": 0.47772207853947374,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.271,
+      "step": 4208
+    },
+    {
+      "epoch": 0.4778356056493928,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2796,
+      "step": 4209
+    },
+    {
+      "epoch": 0.4779491327593119,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2817,
+      "step": 4210
+    },
+    {
+      "epoch": 0.47806265986923097,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2706,
+      "step": 4211
+    },
+    {
+      "epoch": 0.47817618697915004,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.3022,
+      "step": 4212
+    },
+    {
+      "epoch": 0.4782897140890691,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.276,
+      "step": 4213
+    },
+    {
+      "epoch": 0.4784032411989882,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2765,
+      "step": 4214
+    },
+    {
+      "epoch": 0.47851676830890727,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.272,
+      "step": 4215
+    },
+    {
+      "epoch": 0.47863029541882635,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2928,
+      "step": 4216
+    },
+    {
+      "epoch": 0.4787438225287454,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2649,
+      "step": 4217
+    },
+    {
+      "epoch": 0.4788573496386645,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2961,
+      "step": 4218
+    },
+    {
+      "epoch": 0.4789708767485836,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2805,
+      "step": 4219
+    },
+    {
+      "epoch": 0.47908440385850265,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2756,
+      "step": 4220
+    },
+    {
+      "epoch": 0.4791979309684217,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2657,
+      "step": 4221
+    },
+    {
+      "epoch": 0.4793114580783408,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.3049,
+      "step": 4222
+    },
+    {
+      "epoch": 0.4794249851882599,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2901,
+      "step": 4223
+    },
+    {
+      "epoch": 0.47953851229817895,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2723,
+      "step": 4224
+    },
+    {
+      "epoch": 0.47965203940809803,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2568,
+      "step": 4225
+    },
+    {
+      "epoch": 0.4797655665180171,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.2746,
+      "step": 4226
+    },
+    {
+      "epoch": 0.4798790936279362,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2832,
+      "step": 4227
+    },
+    {
+      "epoch": 0.47999262073785526,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2877,
+      "step": 4228
+    },
+    {
+      "epoch": 0.48010614784777433,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.3013,
+      "step": 4229
+    },
+    {
+      "epoch": 0.4802196749576934,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2827,
+      "step": 4230
+    },
+    {
+      "epoch": 0.4803332020676125,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2846,
+      "step": 4231
+    },
+    {
+      "epoch": 0.48044672917753156,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2667,
+      "step": 4232
+    },
+    {
+      "epoch": 0.48056025628745064,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.275,
+      "step": 4233
+    },
+    {
+      "epoch": 0.4806737833973697,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2753,
+      "step": 4234
+    },
+    {
+      "epoch": 0.4807873105072888,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2781,
+      "step": 4235
+    },
+    {
+      "epoch": 0.48090083761720787,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2808,
+      "step": 4236
+    },
+    {
+      "epoch": 0.48101436472712694,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2913,
+      "step": 4237
+    },
+    {
+      "epoch": 0.481127891837046,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2696,
+      "step": 4238
+    },
+    {
+      "epoch": 0.4812414189469651,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2676,
+      "step": 4239
+    },
+    {
+      "epoch": 0.48135494605688417,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.285,
+      "step": 4240
+    },
+    {
+      "epoch": 0.48146847316680325,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.3024,
+      "step": 4241
+    },
+    {
+      "epoch": 0.4815820002767223,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.3009,
+      "step": 4242
+    },
+    {
+      "epoch": 0.4816955273866414,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2913,
+      "step": 4243
+    },
+    {
+      "epoch": 0.4818090544965605,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2837,
+      "step": 4244
+    },
+    {
+      "epoch": 0.48192258160647955,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2719,
+      "step": 4245
+    },
+    {
+      "epoch": 0.4820361087163986,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2745,
+      "step": 4246
+    },
+    {
+      "epoch": 0.4821496358263177,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2769,
+      "step": 4247
+    },
+    {
+      "epoch": 0.4822631629362368,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2741,
+      "step": 4248
+    },
+    {
+      "epoch": 0.48237669004615585,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2689,
+      "step": 4249
+    },
+    {
+      "epoch": 0.48249021715607493,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2889,
+      "step": 4250
+    },
+    {
+      "epoch": 0.482603744265994,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2725,
+      "step": 4251
+    },
+    {
+      "epoch": 0.4827172713759131,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2645,
+      "step": 4252
+    },
+    {
+      "epoch": 0.48283079848583216,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2783,
+      "step": 4253
+    },
+    {
+      "epoch": 0.48294432559575123,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2719,
+      "step": 4254
+    },
+    {
+      "epoch": 0.4830578527056703,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.2734,
+      "step": 4255
+    },
+    {
+      "epoch": 0.4831713798155894,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.264,
+      "step": 4256
+    },
+    {
+      "epoch": 0.48328490692550846,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.2741,
+      "step": 4257
+    },
+    {
+      "epoch": 0.48339843403542754,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.271,
+      "step": 4258
+    },
+    {
+      "epoch": 0.4835119611453466,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.278,
+      "step": 4259
+    },
+    {
+      "epoch": 0.4836254882552657,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2848,
+      "step": 4260
+    },
+    {
+      "epoch": 0.48373901536518477,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2806,
+      "step": 4261
+    },
+    {
+      "epoch": 0.48385254247510384,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2913,
+      "step": 4262
+    },
+    {
+      "epoch": 0.4839660695850229,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2763,
+      "step": 4263
+    },
+    {
+      "epoch": 0.484079596694942,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2919,
+      "step": 4264
+    },
+    {
+      "epoch": 0.48419312380486107,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2589,
+      "step": 4265
+    },
+    {
+      "epoch": 0.48430665091478015,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2554,
+      "step": 4266
+    },
+    {
+      "epoch": 0.4844201780246992,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.281,
+      "step": 4267
+    },
+    {
+      "epoch": 0.4845337051346183,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.2589,
+      "step": 4268
+    },
+    {
+      "epoch": 0.4846472322445374,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2809,
+      "step": 4269
+    },
+    {
+      "epoch": 0.48476075935445645,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2884,
+      "step": 4270
+    },
+    {
+      "epoch": 0.4848742864643755,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.3009,
+      "step": 4271
+    },
+    {
+      "epoch": 0.4849878135742946,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2942,
+      "step": 4272
+    },
+    {
+      "epoch": 0.4851013406842137,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 5.2842,
+      "step": 4273
+    },
+    {
+      "epoch": 0.48521486779413275,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.002,
+      "loss": 5.2685,
+      "step": 4274
+    },
+    {
+      "epoch": 0.48532839490405183,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.002,
+      "loss": 5.2721,
+      "step": 4275
+    },
+    {
+      "epoch": 0.4854419220139709,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2798,
+      "step": 4276
+    },
+    {
+      "epoch": 0.48555544912389,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2674,
+      "step": 4277
+    },
+    {
+      "epoch": 0.48566897623380906,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2881,
+      "step": 4278
+    },
+    {
+      "epoch": 0.48578250334372813,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.289,
+      "step": 4279
+    },
+    {
+      "epoch": 0.4858960304536472,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2972,
+      "step": 4280
+    },
+    {
+      "epoch": 0.4860095575635663,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.2825,
+      "step": 4281
+    },
+    {
+      "epoch": 0.48612308467348536,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.2715,
+      "step": 4282
+    },
+    {
+      "epoch": 0.48623661178340444,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.002,
+      "loss": 5.2685,
+      "step": 4283
+    },
+    {
+      "epoch": 0.48635013889332357,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.275,
+      "step": 4284
+    },
+    {
+      "epoch": 0.48646366600324265,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.002,
+      "loss": 5.2644,
+      "step": 4285
+    },
+    {
+      "epoch": 0.4865771931131617,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.2604,
+      "step": 4286
+    },
+    {
+      "epoch": 0.4866907202230808,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.2898,
+      "step": 4287
+    },
+    {
+      "epoch": 0.4868042473329999,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2765,
+      "step": 4288
+    },
+    {
+      "epoch": 0.48691777444291895,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2545,
+      "step": 4289
+    },
+    {
+      "epoch": 0.487031301552838,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.274,
+      "step": 4290
+    },
+    {
+      "epoch": 0.4871448286627571,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.272,
+      "step": 4291
+    },
+    {
+      "epoch": 0.4872583557726762,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2914,
+      "step": 4292
+    },
+    {
+      "epoch": 0.48737188288259525,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2612,
+      "step": 4293
+    },
+    {
+      "epoch": 0.48748540999251433,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2782,
+      "step": 4294
+    },
+    {
+      "epoch": 0.4875989371024334,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2715,
+      "step": 4295
+    },
+    {
+      "epoch": 0.4877124642123525,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.2713,
+      "step": 4296
+    },
+    {
+      "epoch": 0.48782599132227156,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.2512,
+      "step": 4297
+    },
+    {
+      "epoch": 0.48793951843219063,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 5.2924,
+      "step": 4298
+    },
+    {
+      "epoch": 0.4880530455421097,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2789,
+      "step": 4299
+    },
+    {
+      "epoch": 0.4881665726520288,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 5.2764,
+      "step": 4300
+    },
+    {
+      "epoch": 0.48828009976194786,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.2655,
+      "step": 4301
+    },
+    {
+      "epoch": 0.48839362687186694,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2864,
+      "step": 4302
+    },
+    {
+      "epoch": 0.488507153981786,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2633,
+      "step": 4303
+    },
+    {
+      "epoch": 0.4886206810917051,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2908,
+      "step": 4304
+    },
+    {
+      "epoch": 0.48873420820162417,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.3012,
+      "step": 4305
+    },
+    {
+      "epoch": 0.48884773531154324,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2654,
+      "step": 4306
+    },
+    {
+      "epoch": 0.4889612624214623,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2833,
+      "step": 4307
+    },
+    {
+      "epoch": 0.4890747895313814,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.239,
+      "step": 4308
+    },
+    {
+      "epoch": 0.48918831664130047,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2725,
+      "step": 4309
+    },
+    {
+      "epoch": 0.48930184375121955,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2812,
+      "step": 4310
+    },
+    {
+      "epoch": 0.4894153708611386,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2751,
+      "step": 4311
+    },
+    {
+      "epoch": 0.4895288979710577,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2899,
+      "step": 4312
+    },
+    {
+      "epoch": 0.4896424250809768,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.27,
+      "step": 4313
+    },
+    {
+      "epoch": 0.48975595219089585,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2771,
+      "step": 4314
+    },
+    {
+      "epoch": 0.4898694793008149,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2809,
+      "step": 4315
+    },
+    {
+      "epoch": 0.489983006410734,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2721,
+      "step": 4316
+    },
+    {
+      "epoch": 0.4900965335206531,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2596,
+      "step": 4317
+    },
+    {
+      "epoch": 0.49021006063057215,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.2667,
+      "step": 4318
+    },
+    {
+      "epoch": 0.49032358774049123,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2727,
+      "step": 4319
+    },
+    {
+      "epoch": 0.4904371148504103,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2581,
+      "step": 4320
+    },
+    {
+      "epoch": 0.4905506419603294,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.2885,
+      "step": 4321
+    },
+    {
+      "epoch": 0.49066416907024846,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2668,
+      "step": 4322
+    },
+    {
+      "epoch": 0.49077769618016753,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2866,
+      "step": 4323
+    },
+    {
+      "epoch": 0.4908912232900866,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.285,
+      "step": 4324
+    },
+    {
+      "epoch": 0.4910047504000057,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.3013,
+      "step": 4325
+    },
+    {
+      "epoch": 0.49111827750992476,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2688,
+      "step": 4326
+    },
+    {
+      "epoch": 0.49123180461984384,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2867,
+      "step": 4327
+    },
+    {
+      "epoch": 0.4913453317297629,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2655,
+      "step": 4328
+    },
+    {
+      "epoch": 0.491458858839682,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2834,
+      "step": 4329
+    },
+    {
+      "epoch": 0.49157238594960107,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2866,
+      "step": 4330
+    },
+    {
+      "epoch": 0.49168591305952014,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.293,
+      "step": 4331
+    },
+    {
+      "epoch": 0.4917994401694392,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2685,
+      "step": 4332
+    },
+    {
+      "epoch": 0.4919129672793583,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2922,
+      "step": 4333
+    },
+    {
+      "epoch": 0.49202649438927737,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2881,
+      "step": 4334
+    },
+    {
+      "epoch": 0.49214002149919645,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2684,
+      "step": 4335
+    },
+    {
+      "epoch": 0.4922535486091155,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2828,
+      "step": 4336
+    },
+    {
+      "epoch": 0.4923670757190346,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.265,
+      "step": 4337
+    },
+    {
+      "epoch": 0.4924806028289537,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2705,
+      "step": 4338
+    },
+    {
+      "epoch": 0.49259412993887275,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.2789,
+      "step": 4339
+    },
+    {
+      "epoch": 0.4927076570487918,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.2735,
+      "step": 4340
+    },
+    {
+      "epoch": 0.4928211841587109,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2954,
+      "step": 4341
+    },
+    {
+      "epoch": 0.49293471126863,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2829,
+      "step": 4342
+    },
+    {
+      "epoch": 0.49304823837854905,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2729,
+      "step": 4343
+    },
+    {
+      "epoch": 0.49316176548846813,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2861,
+      "step": 4344
+    },
+    {
+      "epoch": 0.4932752925983872,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2911,
+      "step": 4345
+    },
+    {
+      "epoch": 0.4933888197083063,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2675,
+      "step": 4346
+    },
+    {
+      "epoch": 0.49350234681822536,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2774,
+      "step": 4347
+    },
+    {
+      "epoch": 0.49361587392814443,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2792,
+      "step": 4348
+    },
+    {
+      "epoch": 0.4937294010380635,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2845,
+      "step": 4349
+    },
+    {
+      "epoch": 0.4938429281479826,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.259,
+      "step": 4350
+    },
+    {
+      "epoch": 0.49395645525790166,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2573,
+      "step": 4351
+    },
+    {
+      "epoch": 0.49406998236782074,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.259,
+      "step": 4352
+    },
+    {
+      "epoch": 0.4941835094777398,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2878,
+      "step": 4353
+    },
+    {
+      "epoch": 0.4942970365876589,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2687,
+      "step": 4354
+    },
+    {
+      "epoch": 0.49441056369757796,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2737,
+      "step": 4355
+    },
+    {
+      "epoch": 0.49452409080749704,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2714,
+      "step": 4356
+    },
+    {
+      "epoch": 0.4946376179174161,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2749,
+      "step": 4357
+    },
+    {
+      "epoch": 0.4947511450273352,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.276,
+      "step": 4358
+    },
+    {
+      "epoch": 0.49486467213725427,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2784,
+      "step": 4359
+    },
+    {
+      "epoch": 0.49497819924717334,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2792,
+      "step": 4360
+    },
+    {
+      "epoch": 0.4950917263570924,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2767,
+      "step": 4361
+    },
+    {
+      "epoch": 0.4952052534670115,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 5.2824,
+      "step": 4362
+    },
+    {
+      "epoch": 0.4953187805769306,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2581,
+      "step": 4363
+    },
+    {
+      "epoch": 0.49543230768684965,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2796,
+      "step": 4364
+    },
+    {
+      "epoch": 0.4955458347967687,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2757,
+      "step": 4365
+    },
+    {
+      "epoch": 0.4956593619066878,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2657,
+      "step": 4366
+    },
+    {
+      "epoch": 0.4957728890166069,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2829,
+      "step": 4367
+    },
+    {
+      "epoch": 0.49588641612652595,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2794,
+      "step": 4368
+    },
+    {
+      "epoch": 0.49599994323644503,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2744,
+      "step": 4369
+    },
+    {
+      "epoch": 0.4961134703463641,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2651,
+      "step": 4370
+    },
+    {
+      "epoch": 0.4962269974562832,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2864,
+      "step": 4371
+    },
+    {
+      "epoch": 0.49634052456620226,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2515,
+      "step": 4372
+    },
+    {
+      "epoch": 0.49645405167612133,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2747,
+      "step": 4373
+    },
+    {
+      "epoch": 0.4965675787860404,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2938,
+      "step": 4374
+    },
+    {
+      "epoch": 0.4966811058959595,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2761,
+      "step": 4375
+    },
+    {
+      "epoch": 0.49679463300587856,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2859,
+      "step": 4376
+    },
+    {
+      "epoch": 0.49690816011579764,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2499,
+      "step": 4377
+    },
+    {
+      "epoch": 0.4970216872257167,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2793,
+      "step": 4378
+    },
+    {
+      "epoch": 0.4971352143356358,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.2832,
+      "step": 4379
+    },
+    {
+      "epoch": 0.49724874144555486,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2891,
+      "step": 4380
+    },
+    {
+      "epoch": 0.49736226855547394,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2661,
+      "step": 4381
+    },
+    {
+      "epoch": 0.497475795665393,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2564,
+      "step": 4382
+    },
+    {
+      "epoch": 0.4975893227753121,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2822,
+      "step": 4383
+    },
+    {
+      "epoch": 0.49770284988523117,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2446,
+      "step": 4384
+    },
+    {
+      "epoch": 0.49781637699515024,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2813,
+      "step": 4385
+    },
+    {
+      "epoch": 0.4979299041050693,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.277,
+      "step": 4386
+    },
+    {
+      "epoch": 0.4980434312149884,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2752,
+      "step": 4387
+    },
+    {
+      "epoch": 0.4981569583249075,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2691,
+      "step": 4388
+    },
+    {
+      "epoch": 0.49827048543482655,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.2692,
+      "step": 4389
+    },
+    {
+      "epoch": 0.4983840125447456,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 5.2716,
+      "step": 4390
+    },
+    {
+      "epoch": 0.4984975396546647,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 5.2966,
+      "step": 4391
+    },
+    {
+      "epoch": 0.4986110667645838,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2567,
+      "step": 4392
+    },
+    {
+      "epoch": 0.49872459387450285,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.2874,
+      "step": 4393
+    },
+    {
+      "epoch": 0.49883812098442193,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.297,
+      "step": 4394
+    },
+    {
+      "epoch": 0.498951648094341,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2886,
+      "step": 4395
+    },
+    {
+      "epoch": 0.4990651752042601,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2704,
+      "step": 4396
+    },
+    {
+      "epoch": 0.49917870231417916,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2579,
+      "step": 4397
+    },
+    {
+      "epoch": 0.49929222942409823,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2935,
+      "step": 4398
+    },
+    {
+      "epoch": 0.4994057565340173,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2758,
+      "step": 4399
+    },
+    {
+      "epoch": 0.4995192836439364,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2413,
+      "step": 4400
+    },
+    {
+      "epoch": 0.49963281075385546,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2778,
+      "step": 4401
+    },
+    {
+      "epoch": 0.49974633786377454,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2682,
+      "step": 4402
+    },
+    {
+      "epoch": 0.4998598649736936,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2754,
+      "step": 4403
+    },
+    {
+      "epoch": 0.4999733920836127,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2544,
+      "step": 4404
+    },
+    {
+      "epoch": 0.5000869191935318,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2859,
+      "step": 4405
+    },
+    {
+      "epoch": 0.5002004463034508,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2678,
+      "step": 4406
+    },
+    {
+      "epoch": 0.5003139734133699,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2839,
+      "step": 4407
+    },
+    {
+      "epoch": 0.500427500523289,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2951,
+      "step": 4408
+    },
+    {
+      "epoch": 0.5005410276332081,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 5.262,
+      "step": 4409
+    },
+    {
+      "epoch": 0.5006545547431271,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.002,
+      "loss": 5.2515,
+      "step": 4410
+    },
+    {
+      "epoch": 0.5007680818530462,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2542,
+      "step": 4411
+    },
+    {
+      "epoch": 0.5008816089629653,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2926,
+      "step": 4412
+    },
+    {
+      "epoch": 0.5009951360728844,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2731,
+      "step": 4413
+    },
+    {
+      "epoch": 0.5011086631828034,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2616,
+      "step": 4414
+    },
+    {
+      "epoch": 0.5012221902927225,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.2803,
+      "step": 4415
+    },
+    {
+      "epoch": 0.5013357174026416,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2849,
+      "step": 4416
+    },
+    {
+      "epoch": 0.5014492445125607,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2535,
+      "step": 4417
+    },
+    {
+      "epoch": 0.5015627716224798,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2637,
+      "step": 4418
+    },
+    {
+      "epoch": 0.5016762987323988,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.2848,
+      "step": 4419
+    },
+    {
+      "epoch": 0.5017898258423179,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2842,
+      "step": 4420
+    },
+    {
+      "epoch": 0.501903352952237,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2852,
+      "step": 4421
+    },
+    {
+      "epoch": 0.5020168800621561,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2745,
+      "step": 4422
+    },
+    {
+      "epoch": 0.5021304071720751,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2847,
+      "step": 4423
+    },
+    {
+      "epoch": 0.5022439342819942,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2735,
+      "step": 4424
+    },
+    {
+      "epoch": 0.5023574613919133,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2885,
+      "step": 4425
+    },
+    {
+      "epoch": 0.5024709885018324,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.2935,
+      "step": 4426
+    },
+    {
+      "epoch": 0.5025845156117514,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 5.2628,
+      "step": 4427
+    },
+    {
+      "epoch": 0.5026980427216705,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 5.2868,
+      "step": 4428
+    },
+    {
+      "epoch": 0.5028115698315896,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2769,
+      "step": 4429
+    },
+    {
+      "epoch": 0.5029250969415087,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.2724,
+      "step": 4430
+    },
+    {
+      "epoch": 0.5030386240514277,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2838,
+      "step": 4431
+    },
+    {
+      "epoch": 0.5031521511613468,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2899,
+      "step": 4432
+    },
+    {
+      "epoch": 0.5032656782712659,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2833,
+      "step": 4433
+    },
+    {
+      "epoch": 0.503379205381185,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.2755,
+      "step": 4434
+    },
+    {
+      "epoch": 0.503492732491104,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.002,
+      "loss": 5.2613,
+      "step": 4435
+    },
+    {
+      "epoch": 0.5036062596010231,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.002,
+      "loss": 5.2894,
+      "step": 4436
+    },
+    {
+      "epoch": 0.5037197867109422,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.2776,
+      "step": 4437
+    },
+    {
+      "epoch": 0.5038333138208613,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.2769,
+      "step": 4438
+    },
+    {
+      "epoch": 0.5039468409307803,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.257,
+      "step": 4439
+    },
+    {
+      "epoch": 0.5040603680406994,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2543,
+      "step": 4440
+    },
+    {
+      "epoch": 0.5041738951506185,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 5.2533,
+      "step": 4441
+    },
+    {
+      "epoch": 0.5042874222605376,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2752,
+      "step": 4442
+    },
+    {
+      "epoch": 0.5044009493704567,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.2787,
+      "step": 4443
+    },
+    {
+      "epoch": 0.5045144764803757,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2588,
+      "step": 4444
+    },
+    {
+      "epoch": 0.5046280035902948,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2626,
+      "step": 4445
+    },
+    {
+      "epoch": 0.5047415307002139,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2998,
+      "step": 4446
+    },
+    {
+      "epoch": 0.504855057810133,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2842,
+      "step": 4447
+    },
+    {
+      "epoch": 0.504968584920052,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.269,
+      "step": 4448
+    },
+    {
+      "epoch": 0.5050821120299711,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.2676,
+      "step": 4449
+    },
+    {
+      "epoch": 0.5051956391398902,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2915,
+      "step": 4450
+    },
+    {
+      "epoch": 0.5053091662498093,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2783,
+      "step": 4451
+    },
+    {
+      "epoch": 0.5054226933597283,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.251,
+      "step": 4452
+    },
+    {
+      "epoch": 0.5055362204696474,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2532,
+      "step": 4453
+    },
+    {
+      "epoch": 0.5056497475795665,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2629,
+      "step": 4454
+    },
+    {
+      "epoch": 0.5057632746894856,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.2785,
+      "step": 4455
+    },
+    {
+      "epoch": 0.5058768017994046,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.2722,
+      "step": 4456
+    },
+    {
+      "epoch": 0.5059903289093237,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.2931,
+      "step": 4457
+    },
+    {
+      "epoch": 0.5061038560192428,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2488,
+      "step": 4458
+    },
+    {
+      "epoch": 0.5062173831291619,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2554,
+      "step": 4459
+    },
+    {
+      "epoch": 0.506330910239081,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2653,
+      "step": 4460
+    },
+    {
+      "epoch": 0.506444437349,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.2808,
+      "step": 4461
+    },
+    {
+      "epoch": 0.5065579644589191,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2679,
+      "step": 4462
+    },
+    {
+      "epoch": 0.5066714915688382,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2781,
+      "step": 4463
+    },
+    {
+      "epoch": 0.5067850186787572,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2764,
+      "step": 4464
+    },
+    {
+      "epoch": 0.5068985457886764,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2822,
+      "step": 4465
+    },
+    {
+      "epoch": 0.5070120728985955,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2914,
+      "step": 4466
+    },
+    {
+      "epoch": 0.5071256000085146,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2817,
+      "step": 4467
+    },
+    {
+      "epoch": 0.5072391271184337,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2562,
+      "step": 4468
+    },
+    {
+      "epoch": 0.5073526542283527,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2588,
+      "step": 4469
+    },
+    {
+      "epoch": 0.5074661813382718,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2821,
+      "step": 4470
+    },
+    {
+      "epoch": 0.5075797084481909,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.2679,
+      "step": 4471
+    },
+    {
+      "epoch": 0.50769323555811,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2778,
+      "step": 4472
+    },
+    {
+      "epoch": 0.507806762668029,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2561,
+      "step": 4473
+    },
+    {
+      "epoch": 0.5079202897779481,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2599,
+      "step": 4474
+    },
+    {
+      "epoch": 0.5080338168878672,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.264,
+      "step": 4475
+    },
+    {
+      "epoch": 0.5081473439977863,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2594,
+      "step": 4476
+    },
+    {
+      "epoch": 0.5082608711077053,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.257,
+      "step": 4477
+    },
+    {
+      "epoch": 0.5083743982176244,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2827,
+      "step": 4478
+    },
+    {
+      "epoch": 0.5084879253275435,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2675,
+      "step": 4479
+    },
+    {
+      "epoch": 0.5086014524374626,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.275,
+      "step": 4480
+    },
+    {
+      "epoch": 0.5087149795473817,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2754,
+      "step": 4481
+    },
+    {
+      "epoch": 0.5088285066573007,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.285,
+      "step": 4482
+    },
+    {
+      "epoch": 0.5089420337672198,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.263,
+      "step": 4483
+    },
+    {
+      "epoch": 0.5090555608771389,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.292,
+      "step": 4484
+    },
+    {
+      "epoch": 0.509169087987058,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2537,
+      "step": 4485
+    },
+    {
+      "epoch": 0.509282615096977,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.2826,
+      "step": 4486
+    },
+    {
+      "epoch": 0.5093961422068961,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2706,
+      "step": 4487
+    },
+    {
+      "epoch": 0.5095096693168152,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2787,
+      "step": 4488
+    },
+    {
+      "epoch": 0.5096231964267343,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2816,
+      "step": 4489
+    },
+    {
+      "epoch": 0.5097367235366533,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2761,
+      "step": 4490
+    },
+    {
+      "epoch": 0.5098502506465724,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2401,
+      "step": 4491
+    },
+    {
+      "epoch": 0.5099637777564915,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2777,
+      "step": 4492
+    },
+    {
+      "epoch": 0.5100773048664106,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2774,
+      "step": 4493
+    },
+    {
+      "epoch": 0.5101908319763296,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2584,
+      "step": 4494
+    },
+    {
+      "epoch": 0.5103043590862487,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2725,
+      "step": 4495
+    },
+    {
+      "epoch": 0.5104178861961678,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2768,
+      "step": 4496
+    },
+    {
+      "epoch": 0.5105314133060869,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2623,
+      "step": 4497
+    },
+    {
+      "epoch": 0.5106449404160059,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.249,
+      "step": 4498
+    },
+    {
+      "epoch": 0.510758467525925,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2789,
+      "step": 4499
+    },
+    {
+      "epoch": 0.5108719946358441,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2577,
+      "step": 4500
+    },
+    {
+      "epoch": 0.5109855217457632,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.24,
+      "step": 4501
+    },
+    {
+      "epoch": 0.5110990488556822,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2747,
+      "step": 4502
+    },
+    {
+      "epoch": 0.5112125759656013,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2761,
+      "step": 4503
+    },
+    {
+      "epoch": 0.5113261030755204,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2614,
+      "step": 4504
+    },
+    {
+      "epoch": 0.5114396301854395,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2729,
+      "step": 4505
+    },
+    {
+      "epoch": 0.5115531572953586,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.276,
+      "step": 4506
+    },
+    {
+      "epoch": 0.5116666844052776,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2727,
+      "step": 4507
+    },
+    {
+      "epoch": 0.5117802115151967,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.277,
+      "step": 4508
+    },
+    {
+      "epoch": 0.5118937386251158,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.002,
+      "loss": 5.2764,
+      "step": 4509
+    },
+    {
+      "epoch": 0.5120072657350349,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.2687,
+      "step": 4510
+    },
+    {
+      "epoch": 0.5121207928449539,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2862,
+      "step": 4511
+    },
+    {
+      "epoch": 0.512234319954873,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2527,
+      "step": 4512
+    },
+    {
+      "epoch": 0.5123478470647921,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2478,
+      "step": 4513
+    },
+    {
+      "epoch": 0.5124613741747112,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 5.272,
+      "step": 4514
+    },
+    {
+      "epoch": 0.5125749012846302,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.2608,
+      "step": 4515
+    },
+    {
+      "epoch": 0.5126884283945493,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2636,
+      "step": 4516
+    },
+    {
+      "epoch": 0.5128019555044684,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.266,
+      "step": 4517
+    },
+    {
+      "epoch": 0.5129154826143875,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2682,
+      "step": 4518
+    },
+    {
+      "epoch": 0.5130290097243065,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2658,
+      "step": 4519
+    },
+    {
+      "epoch": 0.5131425368342256,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.274,
+      "step": 4520
+    },
+    {
+      "epoch": 0.5132560639441447,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2478,
+      "step": 4521
+    },
+    {
+      "epoch": 0.5133695910540638,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2583,
+      "step": 4522
+    },
+    {
+      "epoch": 0.5134831181639828,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.267,
+      "step": 4523
+    },
+    {
+      "epoch": 0.5135966452739019,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2777,
+      "step": 4524
+    },
+    {
+      "epoch": 0.513710172383821,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2638,
+      "step": 4525
+    },
+    {
+      "epoch": 0.5138236994937401,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2852,
+      "step": 4526
+    },
+    {
+      "epoch": 0.5139372266036591,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2706,
+      "step": 4527
+    },
+    {
+      "epoch": 0.5140507537135782,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2595,
+      "step": 4528
+    },
+    {
+      "epoch": 0.5141642808234973,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2785,
+      "step": 4529
+    },
+    {
+      "epoch": 0.5142778079334164,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.268,
+      "step": 4530
+    },
+    {
+      "epoch": 0.5143913350433355,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2768,
+      "step": 4531
+    },
+    {
+      "epoch": 0.5145048621532545,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2795,
+      "step": 4532
+    },
+    {
+      "epoch": 0.5146183892631736,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2858,
+      "step": 4533
+    },
+    {
+      "epoch": 0.5147319163730927,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2685,
+      "step": 4534
+    },
+    {
+      "epoch": 0.5148454434830118,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.257,
+      "step": 4535
+    },
+    {
+      "epoch": 0.5149589705929308,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2677,
+      "step": 4536
+    },
+    {
+      "epoch": 0.5150724977028499,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2891,
+      "step": 4537
+    },
+    {
+      "epoch": 0.515186024812769,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2767,
+      "step": 4538
+    },
+    {
+      "epoch": 0.5152995519226881,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2675,
+      "step": 4539
+    },
+    {
+      "epoch": 0.5154130790326071,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2758,
+      "step": 4540
+    },
+    {
+      "epoch": 0.5155266061425262,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2728,
+      "step": 4541
+    },
+    {
+      "epoch": 0.5156401332524453,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2665,
+      "step": 4542
+    },
+    {
+      "epoch": 0.5157536603623644,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2827,
+      "step": 4543
+    },
+    {
+      "epoch": 0.5158671874722834,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2654,
+      "step": 4544
+    },
+    {
+      "epoch": 0.5159807145822025,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.2621,
+      "step": 4545
+    },
+    {
+      "epoch": 0.5160942416921216,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2575,
+      "step": 4546
+    },
+    {
+      "epoch": 0.5162077688020407,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2613,
+      "step": 4547
+    },
+    {
+      "epoch": 0.5163212959119597,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2578,
+      "step": 4548
+    },
+    {
+      "epoch": 0.5164348230218788,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.2759,
+      "step": 4549
+    },
+    {
+      "epoch": 0.5165483501317979,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.2896,
+      "step": 4550
+    },
+    {
+      "epoch": 0.516661877241717,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 5.2496,
+      "step": 4551
+    },
+    {
+      "epoch": 0.516775404351636,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2649,
+      "step": 4552
+    },
+    {
+      "epoch": 0.5168889314615551,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2708,
+      "step": 4553
+    },
+    {
+      "epoch": 0.5170024585714742,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2592,
+      "step": 4554
+    },
+    {
+      "epoch": 0.5171159856813933,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.252,
+      "step": 4555
+    },
+    {
+      "epoch": 0.5172295127913124,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2828,
+      "step": 4556
+    },
+    {
+      "epoch": 0.5173430399012314,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2921,
+      "step": 4557
+    },
+    {
+      "epoch": 0.5174565670111505,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2711,
+      "step": 4558
+    },
+    {
+      "epoch": 0.5175700941210696,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2884,
+      "step": 4559
+    },
+    {
+      "epoch": 0.5176836212309887,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2598,
+      "step": 4560
+    },
+    {
+      "epoch": 0.5177971483409077,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2724,
+      "step": 4561
+    },
+    {
+      "epoch": 0.5179106754508268,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2781,
+      "step": 4562
+    },
+    {
+      "epoch": 0.5180242025607459,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2622,
+      "step": 4563
+    },
+    {
+      "epoch": 0.518137729670665,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2789,
+      "step": 4564
+    },
+    {
+      "epoch": 0.518251256780584,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2813,
+      "step": 4565
+    },
+    {
+      "epoch": 0.5183647838905031,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2625,
+      "step": 4566
+    },
+    {
+      "epoch": 0.5184783110004222,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2652,
+      "step": 4567
+    },
+    {
+      "epoch": 0.5185918381103413,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2826,
+      "step": 4568
+    },
+    {
+      "epoch": 0.5187053652202603,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2973,
+      "step": 4569
+    },
+    {
+      "epoch": 0.5188188923301794,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2557,
+      "step": 4570
+    },
+    {
+      "epoch": 0.5189324194400985,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2761,
+      "step": 4571
+    },
+    {
+      "epoch": 0.5190459465500176,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2556,
+      "step": 4572
+    },
+    {
+      "epoch": 0.5191594736599366,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.262,
+      "step": 4573
+    },
+    {
+      "epoch": 0.5192730007698557,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2994,
+      "step": 4574
+    },
+    {
+      "epoch": 0.5193865278797748,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.002,
+      "loss": 5.2627,
+      "step": 4575
+    },
+    {
+      "epoch": 0.5195000549896939,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 5.2736,
+      "step": 4576
+    },
+    {
+      "epoch": 0.519613582099613,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.2822,
+      "step": 4577
+    },
+    {
+      "epoch": 0.519727109209532,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2554,
+      "step": 4578
+    },
+    {
+      "epoch": 0.5198406363194511,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2642,
+      "step": 4579
+    },
+    {
+      "epoch": 0.5199541634293702,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2858,
+      "step": 4580
+    },
+    {
+      "epoch": 0.5200676905392893,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2613,
+      "step": 4581
+    },
+    {
+      "epoch": 0.5201812176492083,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2585,
+      "step": 4582
+    },
+    {
+      "epoch": 0.5202947447591274,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2865,
+      "step": 4583
+    },
+    {
+      "epoch": 0.5204082718690465,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2731,
+      "step": 4584
+    },
+    {
+      "epoch": 0.5205217989789656,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2687,
+      "step": 4585
+    },
+    {
+      "epoch": 0.5206353260888846,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2729,
+      "step": 4586
+    },
+    {
+      "epoch": 0.5207488531988037,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2746,
+      "step": 4587
+    },
+    {
+      "epoch": 0.5208623803087228,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2524,
+      "step": 4588
+    },
+    {
+      "epoch": 0.5209759074186419,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2844,
+      "step": 4589
+    },
+    {
+      "epoch": 0.5210894345285609,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2802,
+      "step": 4590
+    },
+    {
+      "epoch": 0.52120296163848,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2804,
+      "step": 4591
+    },
+    {
+      "epoch": 0.5213164887483991,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2526,
+      "step": 4592
+    },
+    {
+      "epoch": 0.5214300158583182,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2675,
+      "step": 4593
+    },
+    {
+      "epoch": 0.5215435429682372,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.267,
+      "step": 4594
+    },
+    {
+      "epoch": 0.5216570700781563,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2815,
+      "step": 4595
+    },
+    {
+      "epoch": 0.5217705971880754,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.267,
+      "step": 4596
+    },
+    {
+      "epoch": 0.5218841242979945,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2847,
+      "step": 4597
+    },
+    {
+      "epoch": 0.5219976514079135,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2788,
+      "step": 4598
+    },
+    {
+      "epoch": 0.5221111785178326,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2759,
+      "step": 4599
+    },
+    {
+      "epoch": 0.5222247056277517,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2738,
+      "step": 4600
+    },
+    {
+      "epoch": 0.5223382327376708,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2744,
+      "step": 4601
+    },
+    {
+      "epoch": 0.5224517598475898,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.287,
+      "step": 4602
+    },
+    {
+      "epoch": 0.5225652869575089,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2836,
+      "step": 4603
+    },
+    {
+      "epoch": 0.522678814067428,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2817,
+      "step": 4604
+    },
+    {
+      "epoch": 0.5227923411773471,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2503,
+      "step": 4605
+    },
+    {
+      "epoch": 0.5229058682872662,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2808,
+      "step": 4606
+    },
+    {
+      "epoch": 0.5230193953971852,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2864,
+      "step": 4607
+    },
+    {
+      "epoch": 0.5231329225071043,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2666,
+      "step": 4608
+    },
+    {
+      "epoch": 0.5232464496170234,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2798,
+      "step": 4609
+    },
+    {
+      "epoch": 0.5233599767269425,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2871,
+      "step": 4610
+    },
+    {
+      "epoch": 0.5234735038368615,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 5.2519,
+      "step": 4611
+    },
+    {
+      "epoch": 0.5235870309467806,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2784,
+      "step": 4612
+    },
+    {
+      "epoch": 0.5237005580566997,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2724,
+      "step": 4613
+    },
+    {
+      "epoch": 0.5238140851666188,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2588,
+      "step": 4614
+    },
+    {
+      "epoch": 0.5239276122765378,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2797,
+      "step": 4615
+    },
+    {
+      "epoch": 0.5240411393864569,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2612,
+      "step": 4616
+    },
+    {
+      "epoch": 0.524154666496376,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2873,
+      "step": 4617
+    },
+    {
+      "epoch": 0.5242681936062951,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2602,
+      "step": 4618
+    },
+    {
+      "epoch": 0.5243817207162141,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2693,
+      "step": 4619
+    },
+    {
+      "epoch": 0.5244952478261332,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2778,
+      "step": 4620
+    },
+    {
+      "epoch": 0.5246087749360523,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2516,
+      "step": 4621
+    },
+    {
+      "epoch": 0.5247223020459714,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.287,
+      "step": 4622
+    },
+    {
+      "epoch": 0.5248358291558904,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2522,
+      "step": 4623
+    },
+    {
+      "epoch": 0.5249493562658095,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2626,
+      "step": 4624
+    },
+    {
+      "epoch": 0.5250628833757286,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2707,
+      "step": 4625
+    },
+    {
+      "epoch": 0.5251764104856477,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.269,
+      "step": 4626
+    },
+    {
+      "epoch": 0.5252899375955667,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.2843,
+      "step": 4627
+    },
+    {
+      "epoch": 0.5254034647054858,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.2642,
+      "step": 4628
+    },
+    {
+      "epoch": 0.5255169918154049,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2685,
+      "step": 4629
+    },
+    {
+      "epoch": 0.525630518925324,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2693,
+      "step": 4630
+    },
+    {
+      "epoch": 0.525744046035243,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2869,
+      "step": 4631
+    },
+    {
+      "epoch": 0.5258575731451621,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2516,
+      "step": 4632
+    },
+    {
+      "epoch": 0.5259711002550812,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2452,
+      "step": 4633
+    },
+    {
+      "epoch": 0.5260846273650003,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2657,
+      "step": 4634
+    },
+    {
+      "epoch": 0.5261981544749194,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2556,
+      "step": 4635
+    },
+    {
+      "epoch": 0.5263116815848384,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.2759,
+      "step": 4636
+    },
+    {
+      "epoch": 0.5264252086947575,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 5.295,
+      "step": 4637
+    },
+    {
+      "epoch": 0.5265387358046766,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.002,
+      "loss": 5.2628,
+      "step": 4638
+    },
+    {
+      "epoch": 0.5266522629145957,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.002,
+      "loss": 5.2665,
+      "step": 4639
+    },
+    {
+      "epoch": 0.5267657900245147,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002,
+      "loss": 5.2609,
+      "step": 4640
+    },
+    {
+      "epoch": 0.5268793171344338,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.2637,
+      "step": 4641
+    },
+    {
+      "epoch": 0.5269928442443529,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.2693,
+      "step": 4642
+    },
+    {
+      "epoch": 0.527106371354272,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2838,
+      "step": 4643
+    },
+    {
+      "epoch": 0.527219898464191,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2675,
+      "step": 4644
+    },
+    {
+      "epoch": 0.5273334255741101,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2669,
+      "step": 4645
+    },
+    {
+      "epoch": 0.5274469526840292,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2685,
+      "step": 4646
+    },
+    {
+      "epoch": 0.5275604797939483,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2591,
+      "step": 4647
+    },
+    {
+      "epoch": 0.5276740069038673,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2725,
+      "step": 4648
+    },
+    {
+      "epoch": 0.5277875340137864,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2576,
+      "step": 4649
+    },
+    {
+      "epoch": 0.5279010611237055,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2674,
+      "step": 4650
+    },
+    {
+      "epoch": 0.5280145882336246,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2735,
+      "step": 4651
+    },
+    {
+      "epoch": 0.5281281153435436,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2575,
+      "step": 4652
+    },
+    {
+      "epoch": 0.5282416424534627,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2633,
+      "step": 4653
+    },
+    {
+      "epoch": 0.5283551695633818,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2644,
+      "step": 4654
+    },
+    {
+      "epoch": 0.5284686966733009,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.275,
+      "step": 4655
+    },
+    {
+      "epoch": 0.52858222378322,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.2638,
+      "step": 4656
+    },
+    {
+      "epoch": 0.528695750893139,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2458,
+      "step": 4657
+    },
+    {
+      "epoch": 0.5288092780030581,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2556,
+      "step": 4658
+    },
+    {
+      "epoch": 0.5289228051129772,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.2712,
+      "step": 4659
+    },
+    {
+      "epoch": 0.5290363322228963,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 5.2643,
+      "step": 4660
+    },
+    {
+      "epoch": 0.5291498593328153,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.291,
+      "step": 4661
+    },
+    {
+      "epoch": 0.5292633864427344,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2924,
+      "step": 4662
+    },
+    {
+      "epoch": 0.5293769135526535,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2956,
+      "step": 4663
+    },
+    {
+      "epoch": 0.5294904406625726,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2764,
+      "step": 4664
+    },
+    {
+      "epoch": 0.5296039677724916,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2938,
+      "step": 4665
+    },
+    {
+      "epoch": 0.5297174948824107,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.252,
+      "step": 4666
+    },
+    {
+      "epoch": 0.5298310219923298,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2632,
+      "step": 4667
+    },
+    {
+      "epoch": 0.5299445491022489,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2767,
+      "step": 4668
+    },
+    {
+      "epoch": 0.5300580762121679,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2681,
+      "step": 4669
+    },
+    {
+      "epoch": 0.530171603322087,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2644,
+      "step": 4670
+    },
+    {
+      "epoch": 0.5302851304320061,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2742,
+      "step": 4671
+    },
+    {
+      "epoch": 0.5303986575419252,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2872,
+      "step": 4672
+    },
+    {
+      "epoch": 0.5305121846518442,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2853,
+      "step": 4673
+    },
+    {
+      "epoch": 0.5306257117617633,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2595,
+      "step": 4674
+    },
+    {
+      "epoch": 0.5307392388716824,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.2788,
+      "step": 4675
+    },
+    {
+      "epoch": 0.5308527659816015,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.2727,
+      "step": 4676
+    },
+    {
+      "epoch": 0.5309662930915205,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2789,
+      "step": 4677
+    },
+    {
+      "epoch": 0.5310798202014396,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.2691,
+      "step": 4678
+    },
+    {
+      "epoch": 0.5311933473113587,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2821,
+      "step": 4679
+    },
+    {
+      "epoch": 0.5313068744212778,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2646,
+      "step": 4680
+    },
+    {
+      "epoch": 0.5314204015311969,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2663,
+      "step": 4681
+    },
+    {
+      "epoch": 0.5315339286411159,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2468,
+      "step": 4682
+    },
+    {
+      "epoch": 0.531647455751035,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2525,
+      "step": 4683
+    },
+    {
+      "epoch": 0.5317609828609541,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2644,
+      "step": 4684
+    },
+    {
+      "epoch": 0.5318745099708732,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2676,
+      "step": 4685
+    },
+    {
+      "epoch": 0.5319880370807922,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.2533,
+      "step": 4686
+    },
+    {
+      "epoch": 0.5321015641907113,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2618,
+      "step": 4687
+    },
+    {
+      "epoch": 0.5322150913006304,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.2868,
+      "step": 4688
+    },
+    {
+      "epoch": 0.5323286184105495,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.286,
+      "step": 4689
+    },
+    {
+      "epoch": 0.5324421455204685,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2853,
+      "step": 4690
+    },
+    {
+      "epoch": 0.5325556726303876,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2608,
+      "step": 4691
+    },
+    {
+      "epoch": 0.5326691997403067,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2534,
+      "step": 4692
+    },
+    {
+      "epoch": 0.5327827268502258,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2533,
+      "step": 4693
+    },
+    {
+      "epoch": 0.5328962539601448,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2674,
+      "step": 4694
+    },
+    {
+      "epoch": 0.5330097810700639,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2653,
+      "step": 4695
+    },
+    {
+      "epoch": 0.533123308179983,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2691,
+      "step": 4696
+    },
+    {
+      "epoch": 0.5332368352899021,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2582,
+      "step": 4697
+    },
+    {
+      "epoch": 0.5333503623998211,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2716,
+      "step": 4698
+    },
+    {
+      "epoch": 0.5334638895097402,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2584,
+      "step": 4699
+    },
+    {
+      "epoch": 0.5335774166196593,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2741,
+      "step": 4700
+    },
+    {
+      "epoch": 0.5336909437295784,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2861,
+      "step": 4701
+    },
+    {
+      "epoch": 0.5338044708394974,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2831,
+      "step": 4702
+    },
+    {
+      "epoch": 0.5339179979494165,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 5.268,
+      "step": 4703
+    },
+    {
+      "epoch": 0.5340315250593356,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2463,
+      "step": 4704
+    },
+    {
+      "epoch": 0.5341450521692547,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.002,
+      "loss": 5.2673,
+      "step": 4705
+    },
+    {
+      "epoch": 0.5342585792791739,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 5.289,
+      "step": 4706
+    },
+    {
+      "epoch": 0.5343721063890929,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2744,
+      "step": 4707
+    },
+    {
+      "epoch": 0.534485633499012,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2681,
+      "step": 4708
+    },
+    {
+      "epoch": 0.5345991606089311,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2951,
+      "step": 4709
+    },
+    {
+      "epoch": 0.5347126877188502,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2497,
+      "step": 4710
+    },
+    {
+      "epoch": 0.5348262148287692,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2657,
+      "step": 4711
+    },
+    {
+      "epoch": 0.5349397419386883,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2862,
+      "step": 4712
+    },
+    {
+      "epoch": 0.5350532690486074,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2674,
+      "step": 4713
+    },
+    {
+      "epoch": 0.5351667961585265,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.264,
+      "step": 4714
+    },
+    {
+      "epoch": 0.5352803232684455,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2677,
+      "step": 4715
+    },
+    {
+      "epoch": 0.5353938503783646,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2734,
+      "step": 4716
+    },
+    {
+      "epoch": 0.5355073774882837,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2763,
+      "step": 4717
+    },
+    {
+      "epoch": 0.5356209045982028,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2434,
+      "step": 4718
+    },
+    {
+      "epoch": 0.5357344317081218,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.266,
+      "step": 4719
+    },
+    {
+      "epoch": 0.5358479588180409,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2668,
+      "step": 4720
+    },
+    {
+      "epoch": 0.53596148592796,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2604,
+      "step": 4721
+    },
+    {
+      "epoch": 0.5360750130378791,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2683,
+      "step": 4722
+    },
+    {
+      "epoch": 0.5361885401477982,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.2601,
+      "step": 4723
+    },
+    {
+      "epoch": 0.5363020672577172,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2473,
+      "step": 4724
+    },
+    {
+      "epoch": 0.5364155943676363,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2721,
+      "step": 4725
+    },
+    {
+      "epoch": 0.5365291214775554,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.291,
+      "step": 4726
+    },
+    {
+      "epoch": 0.5366426485874745,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2722,
+      "step": 4727
+    },
+    {
+      "epoch": 0.5367561756973935,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2896,
+      "step": 4728
+    },
+    {
+      "epoch": 0.5368697028073126,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2558,
+      "step": 4729
+    },
+    {
+      "epoch": 0.5369832299172317,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2672,
+      "step": 4730
+    },
+    {
+      "epoch": 0.5370967570271508,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.279,
+      "step": 4731
+    },
+    {
+      "epoch": 0.5372102841370698,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2847,
+      "step": 4732
+    },
+    {
+      "epoch": 0.5373238112469889,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2575,
+      "step": 4733
+    },
+    {
+      "epoch": 0.537437338356908,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2789,
+      "step": 4734
+    },
+    {
+      "epoch": 0.5375508654668271,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2528,
+      "step": 4735
+    },
+    {
+      "epoch": 0.5376643925767461,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2638,
+      "step": 4736
+    },
+    {
+      "epoch": 0.5377779196866652,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.002,
+      "loss": 5.2796,
+      "step": 4737
+    },
+    {
+      "epoch": 0.5378914467965843,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.2742,
+      "step": 4738
+    },
+    {
+      "epoch": 0.5380049739065034,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.3038,
+      "step": 4739
+    },
+    {
+      "epoch": 0.5381185010164224,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2565,
+      "step": 4740
+    },
+    {
+      "epoch": 0.5382320281263415,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2664,
+      "step": 4741
+    },
+    {
+      "epoch": 0.5383455552362606,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2826,
+      "step": 4742
+    },
+    {
+      "epoch": 0.5384590823461797,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2574,
+      "step": 4743
+    },
+    {
+      "epoch": 0.5385726094560987,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2592,
+      "step": 4744
+    },
+    {
+      "epoch": 0.5386861365660178,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2493,
+      "step": 4745
+    },
+    {
+      "epoch": 0.5387996636759369,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2667,
+      "step": 4746
+    },
+    {
+      "epoch": 0.538913190785856,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2665,
+      "step": 4747
+    },
+    {
+      "epoch": 0.539026717895775,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2505,
+      "step": 4748
+    },
+    {
+      "epoch": 0.5391402450056941,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2598,
+      "step": 4749
+    },
+    {
+      "epoch": 0.5392537721156132,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2859,
+      "step": 4750
+    },
+    {
+      "epoch": 0.5393672992255323,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2975,
+      "step": 4751
+    },
+    {
+      "epoch": 0.5394808263354514,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2652,
+      "step": 4752
+    },
+    {
+      "epoch": 0.5395943534453704,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2702,
+      "step": 4753
+    },
+    {
+      "epoch": 0.5397078805552895,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2767,
+      "step": 4754
+    },
+    {
+      "epoch": 0.5398214076652086,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2694,
+      "step": 4755
+    },
+    {
+      "epoch": 0.5399349347751277,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.261,
+      "step": 4756
+    },
+    {
+      "epoch": 0.5400484618850467,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2656,
+      "step": 4757
+    },
+    {
+      "epoch": 0.5401619889949658,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2525,
+      "step": 4758
+    },
+    {
+      "epoch": 0.5402755161048849,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2514,
+      "step": 4759
+    },
+    {
+      "epoch": 0.540389043214804,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.27,
+      "step": 4760
+    },
+    {
+      "epoch": 0.540502570324723,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2491,
+      "step": 4761
+    },
+    {
+      "epoch": 0.5406160974346421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.2621,
+      "step": 4762
+    },
+    {
+      "epoch": 0.5407296245445612,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.2698,
+      "step": 4763
+    },
+    {
+      "epoch": 0.5408431516544803,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2601,
+      "step": 4764
+    },
+    {
+      "epoch": 0.5409566787643993,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2681,
+      "step": 4765
+    },
+    {
+      "epoch": 0.5410702058743184,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2899,
+      "step": 4766
+    },
+    {
+      "epoch": 0.5411837329842375,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2805,
+      "step": 4767
+    },
+    {
+      "epoch": 0.5412972600941566,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2656,
+      "step": 4768
+    },
+    {
+      "epoch": 0.5414107872040756,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2657,
+      "step": 4769
+    },
+    {
+      "epoch": 0.5415243143139947,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2539,
+      "step": 4770
+    },
+    {
+      "epoch": 0.5416378414239138,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.262,
+      "step": 4771
+    },
+    {
+      "epoch": 0.5417513685338329,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2779,
+      "step": 4772
+    },
+    {
+      "epoch": 0.541864895643752,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2721,
+      "step": 4773
+    },
+    {
+      "epoch": 0.541978422753671,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2482,
+      "step": 4774
+    },
+    {
+      "epoch": 0.5420919498635901,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2841,
+      "step": 4775
+    },
+    {
+      "epoch": 0.5422054769735092,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2692,
+      "step": 4776
+    },
+    {
+      "epoch": 0.5423190040834283,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.288,
+      "step": 4777
+    },
+    {
+      "epoch": 0.5424325311933473,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2471,
+      "step": 4778
+    },
+    {
+      "epoch": 0.5425460583032664,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2591,
+      "step": 4779
+    },
+    {
+      "epoch": 0.5426595854131855,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2679,
+      "step": 4780
+    },
+    {
+      "epoch": 0.5427731125231046,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2839,
+      "step": 4781
+    },
+    {
+      "epoch": 0.5428866396330236,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2534,
+      "step": 4782
+    },
+    {
+      "epoch": 0.5430001667429427,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2566,
+      "step": 4783
+    },
+    {
+      "epoch": 0.5431136938528618,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2677,
+      "step": 4784
+    },
+    {
+      "epoch": 0.5432272209627809,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2848,
+      "step": 4785
+    },
+    {
+      "epoch": 0.5433407480726999,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2614,
+      "step": 4786
+    },
+    {
+      "epoch": 0.543454275182619,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2702,
+      "step": 4787
+    },
+    {
+      "epoch": 0.5435678022925381,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2501,
+      "step": 4788
+    },
+    {
+      "epoch": 0.5436813294024572,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2469,
+      "step": 4789
+    },
+    {
+      "epoch": 0.5437948565123762,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2621,
+      "step": 4790
+    },
+    {
+      "epoch": 0.5439083836222953,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.2656,
+      "step": 4791
+    },
+    {
+      "epoch": 0.5440219107322144,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.274,
+      "step": 4792
+    },
+    {
+      "epoch": 0.5441354378421335,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.2694,
+      "step": 4793
+    },
+    {
+      "epoch": 0.5442489649520525,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.2536,
+      "step": 4794
+    },
+    {
+      "epoch": 0.5443624920619716,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2815,
+      "step": 4795
+    },
+    {
+      "epoch": 0.5444760191718907,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2686,
+      "step": 4796
+    },
+    {
+      "epoch": 0.5445895462818098,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2558,
+      "step": 4797
+    },
+    {
+      "epoch": 0.5447030733917289,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2573,
+      "step": 4798
+    },
+    {
+      "epoch": 0.5448166005016479,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2705,
+      "step": 4799
+    },
+    {
+      "epoch": 0.544930127611567,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2681,
+      "step": 4800
+    },
+    {
+      "epoch": 0.5450436547214861,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2624,
+      "step": 4801
+    },
+    {
+      "epoch": 0.5451571818314052,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2727,
+      "step": 4802
+    },
+    {
+      "epoch": 0.5452707089413242,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2668,
+      "step": 4803
+    },
+    {
+      "epoch": 0.5453842360512433,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2578,
+      "step": 4804
+    },
+    {
+      "epoch": 0.5454977631611624,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2728,
+      "step": 4805
+    },
+    {
+      "epoch": 0.5456112902710815,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2691,
+      "step": 4806
+    },
+    {
+      "epoch": 0.5457248173810005,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2667,
+      "step": 4807
+    },
+    {
+      "epoch": 0.5458383444909196,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2356,
+      "step": 4808
+    },
+    {
+      "epoch": 0.5459518716008387,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2789,
+      "step": 4809
+    },
+    {
+      "epoch": 0.5460653987107578,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2598,
+      "step": 4810
+    },
+    {
+      "epoch": 0.5461789258206768,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2812,
+      "step": 4811
+    },
+    {
+      "epoch": 0.5462924529305959,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2471,
+      "step": 4812
+    },
+    {
+      "epoch": 0.546405980040515,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2602,
+      "step": 4813
+    },
+    {
+      "epoch": 0.5465195071504341,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.259,
+      "step": 4814
+    },
+    {
+      "epoch": 0.5466330342603531,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2572,
+      "step": 4815
+    },
+    {
+      "epoch": 0.5467465613702722,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2786,
+      "step": 4816
+    },
+    {
+      "epoch": 0.5468600884801913,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2753,
+      "step": 4817
+    },
+    {
+      "epoch": 0.5469736155901104,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2718,
+      "step": 4818
+    },
+    {
+      "epoch": 0.5470871427000294,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2534,
+      "step": 4819
+    },
+    {
+      "epoch": 0.5472006698099485,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.2783,
+      "step": 4820
+    },
+    {
+      "epoch": 0.5473141969198676,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.2803,
+      "step": 4821
+    },
+    {
+      "epoch": 0.5474277240297867,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2757,
+      "step": 4822
+    },
+    {
+      "epoch": 0.5475412511397058,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.27,
+      "step": 4823
+    },
+    {
+      "epoch": 0.5476547782496248,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2766,
+      "step": 4824
+    },
+    {
+      "epoch": 0.5477683053595439,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 5.2794,
+      "step": 4825
+    },
+    {
+      "epoch": 0.547881832469463,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2864,
+      "step": 4826
+    },
+    {
+      "epoch": 0.5479953595793821,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 5.2738,
+      "step": 4827
+    },
+    {
+      "epoch": 0.5481088866893011,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 5.265,
+      "step": 4828
+    },
+    {
+      "epoch": 0.5482224137992202,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2708,
+      "step": 4829
+    },
+    {
+      "epoch": 0.5483359409091393,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2446,
+      "step": 4830
+    },
+    {
+      "epoch": 0.5484494680190584,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2725,
+      "step": 4831
+    },
+    {
+      "epoch": 0.5485629951289774,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2428,
+      "step": 4832
+    },
+    {
+      "epoch": 0.5486765222388965,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2532,
+      "step": 4833
+    },
+    {
+      "epoch": 0.5487900493488156,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2713,
+      "step": 4834
+    },
+    {
+      "epoch": 0.5489035764587347,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2614,
+      "step": 4835
+    },
+    {
+      "epoch": 0.5490171035686537,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2593,
+      "step": 4836
+    },
+    {
+      "epoch": 0.5491306306785728,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2472,
+      "step": 4837
+    },
+    {
+      "epoch": 0.5492441577884919,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2648,
+      "step": 4838
+    },
+    {
+      "epoch": 0.549357684898411,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2384,
+      "step": 4839
+    },
+    {
+      "epoch": 0.54947121200833,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2734,
+      "step": 4840
+    },
+    {
+      "epoch": 0.5495847391182491,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2606,
+      "step": 4841
+    },
+    {
+      "epoch": 0.5496982662281682,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2821,
+      "step": 4842
+    },
+    {
+      "epoch": 0.5498117933380873,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2549,
+      "step": 4843
+    },
+    {
+      "epoch": 0.5499253204480063,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2668,
+      "step": 4844
+    },
+    {
+      "epoch": 0.5500388475579254,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2737,
+      "step": 4845
+    },
+    {
+      "epoch": 0.5501523746678445,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2557,
+      "step": 4846
+    },
+    {
+      "epoch": 0.5502659017777636,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2451,
+      "step": 4847
+    },
+    {
+      "epoch": 0.5503794288876827,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2861,
+      "step": 4848
+    },
+    {
+      "epoch": 0.5504929559976017,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2678,
+      "step": 4849
+    },
+    {
+      "epoch": 0.5506064831075208,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2563,
+      "step": 4850
+    },
+    {
+      "epoch": 0.5507200102174399,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2791,
+      "step": 4851
+    },
+    {
+      "epoch": 0.550833537327359,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2773,
+      "step": 4852
+    },
+    {
+      "epoch": 0.550947064437278,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2514,
+      "step": 4853
+    },
+    {
+      "epoch": 0.5510605915471971,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2465,
+      "step": 4854
+    },
+    {
+      "epoch": 0.5511741186571162,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2707,
+      "step": 4855
+    },
+    {
+      "epoch": 0.5512876457670353,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.288,
+      "step": 4856
+    },
+    {
+      "epoch": 0.5514011728769543,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2801,
+      "step": 4857
+    },
+    {
+      "epoch": 0.5515146999868734,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2546,
+      "step": 4858
+    },
+    {
+      "epoch": 0.5516282270967925,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2608,
+      "step": 4859
+    },
+    {
+      "epoch": 0.5517417542067116,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2584,
+      "step": 4860
+    },
+    {
+      "epoch": 0.5518552813166306,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2602,
+      "step": 4861
+    },
+    {
+      "epoch": 0.5519688084265497,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2732,
+      "step": 4862
+    },
+    {
+      "epoch": 0.5520823355364688,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.289,
+      "step": 4863
+    },
+    {
+      "epoch": 0.5521958626463879,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2686,
+      "step": 4864
+    },
+    {
+      "epoch": 0.552309389756307,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 5.2674,
+      "step": 4865
+    },
+    {
+      "epoch": 0.552422916866226,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 5.2552,
+      "step": 4866
+    },
+    {
+      "epoch": 0.5525364439761451,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.2819,
+      "step": 4867
+    },
+    {
+      "epoch": 0.5526499710860642,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2762,
+      "step": 4868
+    },
+    {
+      "epoch": 0.5527634981959832,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2731,
+      "step": 4869
+    },
+    {
+      "epoch": 0.5528770253059023,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.268,
+      "step": 4870
+    },
+    {
+      "epoch": 0.5529905524158214,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2658,
+      "step": 4871
+    },
+    {
+      "epoch": 0.5531040795257405,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2427,
+      "step": 4872
+    },
+    {
+      "epoch": 0.5532176066356596,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2461,
+      "step": 4873
+    },
+    {
+      "epoch": 0.5533311337455786,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2468,
+      "step": 4874
+    },
+    {
+      "epoch": 0.5534446608554977,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2644,
+      "step": 4875
+    },
+    {
+      "epoch": 0.5535581879654168,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2604,
+      "step": 4876
+    },
+    {
+      "epoch": 0.5536717150753359,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.2765,
+      "step": 4877
+    },
+    {
+      "epoch": 0.5537852421852549,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2489,
+      "step": 4878
+    },
+    {
+      "epoch": 0.553898769295174,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.249,
+      "step": 4879
+    },
+    {
+      "epoch": 0.5540122964050931,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2556,
+      "step": 4880
+    },
+    {
+      "epoch": 0.5541258235150122,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.252,
+      "step": 4881
+    },
+    {
+      "epoch": 0.5542393506249312,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2476,
+      "step": 4882
+    },
+    {
+      "epoch": 0.5543528777348503,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2588,
+      "step": 4883
+    },
+    {
+      "epoch": 0.5544664048447694,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2708,
+      "step": 4884
+    },
+    {
+      "epoch": 0.5545799319546885,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2757,
+      "step": 4885
+    },
+    {
+      "epoch": 0.5546934590646075,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2743,
+      "step": 4886
+    },
+    {
+      "epoch": 0.5548069861745266,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.2978,
+      "step": 4887
+    },
+    {
+      "epoch": 0.5549205132844457,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.28,
+      "step": 4888
+    },
+    {
+      "epoch": 0.5550340403943648,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2753,
+      "step": 4889
+    },
+    {
+      "epoch": 0.5551475675042838,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2616,
+      "step": 4890
+    },
+    {
+      "epoch": 0.5552610946142029,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2672,
+      "step": 4891
+    },
+    {
+      "epoch": 0.555374621724122,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2489,
+      "step": 4892
+    },
+    {
+      "epoch": 0.5554881488340411,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2742,
+      "step": 4893
+    },
+    {
+      "epoch": 0.5556016759439601,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.259,
+      "step": 4894
+    },
+    {
+      "epoch": 0.5557152030538792,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2753,
+      "step": 4895
+    },
+    {
+      "epoch": 0.5558287301637983,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2791,
+      "step": 4896
+    },
+    {
+      "epoch": 0.5559422572737174,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2427,
+      "step": 4897
+    },
+    {
+      "epoch": 0.5560557843836365,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2631,
+      "step": 4898
+    },
+    {
+      "epoch": 0.5561693114935555,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2643,
+      "step": 4899
+    },
+    {
+      "epoch": 0.5562828386034746,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2618,
+      "step": 4900
+    },
+    {
+      "epoch": 0.5563963657133937,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2575,
+      "step": 4901
+    },
+    {
+      "epoch": 0.5565098928233128,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2599,
+      "step": 4902
+    },
+    {
+      "epoch": 0.5566234199332318,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2515,
+      "step": 4903
+    },
+    {
+      "epoch": 0.5567369470431509,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2714,
+      "step": 4904
+    },
+    {
+      "epoch": 0.55685047415307,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.284,
+      "step": 4905
+    },
+    {
+      "epoch": 0.5569640012629891,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.2701,
+      "step": 4906
+    },
+    {
+      "epoch": 0.5570775283729081,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2642,
+      "step": 4907
+    },
+    {
+      "epoch": 0.5571910554828272,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.266,
+      "step": 4908
+    },
+    {
+      "epoch": 0.5573045825927463,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.2678,
+      "step": 4909
+    },
+    {
+      "epoch": 0.5574181097026654,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.2475,
+      "step": 4910
+    },
+    {
+      "epoch": 0.5575316368125844,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.2458,
+      "step": 4911
+    },
+    {
+      "epoch": 0.5576451639225035,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.2837,
+      "step": 4912
+    },
+    {
+      "epoch": 0.5577586910324226,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2611,
+      "step": 4913
+    },
+    {
+      "epoch": 0.5578722181423417,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2806,
+      "step": 4914
+    },
+    {
+      "epoch": 0.5579857452522607,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.274,
+      "step": 4915
+    },
+    {
+      "epoch": 0.5580992723621798,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2491,
+      "step": 4916
+    },
+    {
+      "epoch": 0.5582127994720989,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2842,
+      "step": 4917
+    },
+    {
+      "epoch": 0.558326326582018,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2646,
+      "step": 4918
+    },
+    {
+      "epoch": 0.558439853691937,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.254,
+      "step": 4919
+    },
+    {
+      "epoch": 0.5585533808018561,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2335,
+      "step": 4920
+    },
+    {
+      "epoch": 0.5586669079117752,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2735,
+      "step": 4921
+    },
+    {
+      "epoch": 0.5587804350216943,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.002,
+      "loss": 5.2751,
+      "step": 4922
+    },
+    {
+      "epoch": 0.5588939621316134,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 5.2644,
+      "step": 4923
+    },
+    {
+      "epoch": 0.5590074892415324,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.2791,
+      "step": 4924
+    },
+    {
+      "epoch": 0.5591210163514515,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2762,
+      "step": 4925
+    },
+    {
+      "epoch": 0.5592345434613706,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2835,
+      "step": 4926
+    },
+    {
+      "epoch": 0.5593480705712897,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2573,
+      "step": 4927
+    },
+    {
+      "epoch": 0.5594615976812087,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2572,
+      "step": 4928
+    },
+    {
+      "epoch": 0.5595751247911278,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2578,
+      "step": 4929
+    },
+    {
+      "epoch": 0.5596886519010469,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2523,
+      "step": 4930
+    },
+    {
+      "epoch": 0.559802179010966,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2722,
+      "step": 4931
+    },
+    {
+      "epoch": 0.559915706120885,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2724,
+      "step": 4932
+    },
+    {
+      "epoch": 0.5600292332308041,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2436,
+      "step": 4933
+    },
+    {
+      "epoch": 0.5601427603407232,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2627,
+      "step": 4934
+    },
+    {
+      "epoch": 0.5602562874506423,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2679,
+      "step": 4935
+    },
+    {
+      "epoch": 0.5603698145605613,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2673,
+      "step": 4936
+    },
+    {
+      "epoch": 0.5604833416704804,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2613,
+      "step": 4937
+    },
+    {
+      "epoch": 0.5605968687803995,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.2709,
+      "step": 4938
+    },
+    {
+      "epoch": 0.5607103958903186,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.2534,
+      "step": 4939
+    },
+    {
+      "epoch": 0.5608239230002376,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.2902,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5609374501101567,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.2548,
+      "step": 4941
+    },
+    {
+      "epoch": 0.5610509772200758,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.2546,
+      "step": 4942
+    },
+    {
+      "epoch": 0.5611645043299949,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.279,
+      "step": 4943
+    },
+    {
+      "epoch": 0.561278031439914,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2489,
+      "step": 4944
+    },
+    {
+      "epoch": 0.561391558549833,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2567,
+      "step": 4945
+    },
+    {
+      "epoch": 0.5615050856597521,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.2749,
+      "step": 4946
+    },
+    {
+      "epoch": 0.5616186127696712,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2626,
+      "step": 4947
+    },
+    {
+      "epoch": 0.5617321398795904,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.272,
+      "step": 4948
+    },
+    {
+      "epoch": 0.5618456669895094,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2681,
+      "step": 4949
+    },
+    {
+      "epoch": 0.5619591940994285,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2593,
+      "step": 4950
+    },
+    {
+      "epoch": 0.5620727212093476,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2599,
+      "step": 4951
+    },
+    {
+      "epoch": 0.5621862483192667,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2606,
+      "step": 4952
+    },
+    {
+      "epoch": 0.5622997754291857,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2742,
+      "step": 4953
+    },
+    {
+      "epoch": 0.5624133025391048,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 5.2869,
+      "step": 4954
+    },
+    {
+      "epoch": 0.5625268296490239,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2519,
+      "step": 4955
+    },
+    {
+      "epoch": 0.562640356758943,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2702,
+      "step": 4956
+    },
+    {
+      "epoch": 0.562753883868862,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2523,
+      "step": 4957
+    },
+    {
+      "epoch": 0.5628674109787811,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2656,
+      "step": 4958
+    },
+    {
+      "epoch": 0.5629809380887002,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2715,
+      "step": 4959
+    },
+    {
+      "epoch": 0.5630944651986193,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2514,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5632079923085384,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2739,
+      "step": 4961
+    },
+    {
+      "epoch": 0.5633215194184574,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2516,
+      "step": 4962
+    },
+    {
+      "epoch": 0.5634350465283765,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2959,
+      "step": 4963
+    },
+    {
+      "epoch": 0.5635485736382956,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2617,
+      "step": 4964
+    },
+    {
+      "epoch": 0.5636621007482147,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2576,
+      "step": 4965
+    },
+    {
+      "epoch": 0.5637756278581337,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2649,
+      "step": 4966
+    },
+    {
+      "epoch": 0.5638891549680528,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2706,
+      "step": 4967
+    },
+    {
+      "epoch": 0.5640026820779719,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2523,
+      "step": 4968
+    },
+    {
+      "epoch": 0.564116209187891,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2626,
+      "step": 4969
+    },
+    {
+      "epoch": 0.56422973629781,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2693,
+      "step": 4970
+    },
+    {
+      "epoch": 0.5643432634077291,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2718,
+      "step": 4971
+    },
+    {
+      "epoch": 0.5644567905176482,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2836,
+      "step": 4972
+    },
+    {
+      "epoch": 0.5645703176275673,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2616,
+      "step": 4973
+    },
+    {
+      "epoch": 0.5646838447374863,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.27,
+      "step": 4974
+    },
+    {
+      "epoch": 0.5647973718474054,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2288,
+      "step": 4975
+    },
+    {
+      "epoch": 0.5649108989573245,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.272,
+      "step": 4976
+    },
+    {
+      "epoch": 0.5650244260672436,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2348,
+      "step": 4977
+    },
+    {
+      "epoch": 0.5651379531771626,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2764,
+      "step": 4978
+    },
+    {
+      "epoch": 0.5652514802870817,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2546,
+      "step": 4979
+    },
+    {
+      "epoch": 0.5653650073970008,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.249,
+      "step": 4980
+    },
+    {
+      "epoch": 0.5654785345069199,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2379,
+      "step": 4981
+    },
+    {
+      "epoch": 0.565592061616839,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2597,
+      "step": 4982
+    },
+    {
+      "epoch": 0.565705588726758,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2864,
+      "step": 4983
+    },
+    {
+      "epoch": 0.5658191158366771,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2508,
+      "step": 4984
+    },
+    {
+      "epoch": 0.5659326429465962,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2698,
+      "step": 4985
+    },
+    {
+      "epoch": 0.5660461700565153,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2628,
+      "step": 4986
+    },
+    {
+      "epoch": 0.5661596971664343,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2731,
+      "step": 4987
+    },
+    {
+      "epoch": 0.5662732242763534,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.276,
+      "step": 4988
+    },
+    {
+      "epoch": 0.5663867513862725,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.287,
+      "step": 4989
+    },
+    {
+      "epoch": 0.5665002784961916,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2646,
+      "step": 4990
+    },
+    {
+      "epoch": 0.5666138056061106,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2592,
+      "step": 4991
+    },
+    {
+      "epoch": 0.5667273327160297,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2439,
+      "step": 4992
+    },
+    {
+      "epoch": 0.5668408598259488,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.2753,
+      "step": 4993
+    },
+    {
+      "epoch": 0.5669543869358679,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.265,
+      "step": 4994
+    },
+    {
+      "epoch": 0.5670679140457869,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2626,
+      "step": 4995
+    },
+    {
+      "epoch": 0.567181441155706,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2856,
+      "step": 4996
+    },
+    {
+      "epoch": 0.5672949682656251,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2589,
+      "step": 4997
+    },
+    {
+      "epoch": 0.5674084953755442,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2614,
+      "step": 4998
+    },
+    {
+      "epoch": 0.5675220224854632,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2521,
+      "step": 4999
+    },
+    {
+      "epoch": 0.5676355495953823,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2686,
+      "step": 5000
+    },
+    {
+      "epoch": 0.5677490767053014,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2578,
+      "step": 5001
+    },
+    {
+      "epoch": 0.5678626038152205,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.2686,
+      "step": 5002
+    },
+    {
+      "epoch": 0.5679761309251395,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.2643,
+      "step": 5003
+    },
+    {
+      "epoch": 0.5680896580350586,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2545,
+      "step": 5004
+    },
+    {
+      "epoch": 0.5682031851449777,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2767,
+      "step": 5005
+    },
+    {
+      "epoch": 0.5683167122548968,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2592,
+      "step": 5006
+    },
+    {
+      "epoch": 0.5684302393648158,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.2521,
+      "step": 5007
+    },
+    {
+      "epoch": 0.5685437664747349,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.2628,
+      "step": 5008
+    },
+    {
+      "epoch": 0.568657293584654,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 5.2698,
+      "step": 5009
+    },
+    {
+      "epoch": 0.5687708206945731,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.2457,
+      "step": 5010
+    },
+    {
+      "epoch": 0.5688843478044922,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2523,
+      "step": 5011
+    },
+    {
+      "epoch": 0.5689978749144112,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2672,
+      "step": 5012
+    },
+    {
+      "epoch": 0.5691114020243303,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2653,
+      "step": 5013
+    },
+    {
+      "epoch": 0.5692249291342494,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2666,
+      "step": 5014
+    },
+    {
+      "epoch": 0.5693384562441685,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2546,
+      "step": 5015
+    },
+    {
+      "epoch": 0.5694519833540875,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2734,
+      "step": 5016
+    },
+    {
+      "epoch": 0.5695655104640066,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2583,
+      "step": 5017
+    },
+    {
+      "epoch": 0.5696790375739257,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2635,
+      "step": 5018
+    },
+    {
+      "epoch": 0.5697925646838448,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2713,
+      "step": 5019
+    },
+    {
+      "epoch": 0.5699060917937638,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.261,
+      "step": 5020
+    },
+    {
+      "epoch": 0.5700196189036829,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.252,
+      "step": 5021
+    },
+    {
+      "epoch": 0.570133146013602,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2588,
+      "step": 5022
+    },
+    {
+      "epoch": 0.5702466731235211,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2708,
+      "step": 5023
+    },
+    {
+      "epoch": 0.5703602002334401,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2664,
+      "step": 5024
+    },
+    {
+      "epoch": 0.5704737273433592,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2676,
+      "step": 5025
+    },
+    {
+      "epoch": 0.5705872544532783,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2706,
+      "step": 5026
+    },
+    {
+      "epoch": 0.5707007815631974,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2566,
+      "step": 5027
+    },
+    {
+      "epoch": 0.5708143086731164,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2639,
+      "step": 5028
+    },
+    {
+      "epoch": 0.5709278357830355,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2503,
+      "step": 5029
+    },
+    {
+      "epoch": 0.5710413628929546,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 5.2927,
+      "step": 5030
+    },
+    {
+      "epoch": 0.5711548900028737,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2496,
+      "step": 5031
+    },
+    {
+      "epoch": 0.5712684171127927,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.002,
+      "loss": 5.2634,
+      "step": 5032
+    },
+    {
+      "epoch": 0.5713819442227118,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.267,
+      "step": 5033
+    },
+    {
+      "epoch": 0.5714954713326309,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2813,
+      "step": 5034
+    },
+    {
+      "epoch": 0.57160899844255,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2546,
+      "step": 5035
+    },
+    {
+      "epoch": 0.571722525552469,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2491,
+      "step": 5036
+    },
+    {
+      "epoch": 0.5718360526623881,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2434,
+      "step": 5037
+    },
+    {
+      "epoch": 0.5719495797723072,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2554,
+      "step": 5038
+    },
+    {
+      "epoch": 0.5720631068822263,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2426,
+      "step": 5039
+    },
+    {
+      "epoch": 0.5721766339921454,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2456,
+      "step": 5040
+    },
+    {
+      "epoch": 0.5722901611020644,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2911,
+      "step": 5041
+    },
+    {
+      "epoch": 0.5724036882119835,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2565,
+      "step": 5042
+    },
+    {
+      "epoch": 0.5725172153219026,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2585,
+      "step": 5043
+    },
+    {
+      "epoch": 0.5726307424318217,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2553,
+      "step": 5044
+    },
+    {
+      "epoch": 0.5727442695417407,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2581,
+      "step": 5045
+    },
+    {
+      "epoch": 0.5728577966516598,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.2628,
+      "step": 5046
+    },
+    {
+      "epoch": 0.5729713237615789,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.002,
+      "loss": 5.2466,
+      "step": 5047
+    },
+    {
+      "epoch": 0.573084850871498,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.2531,
+      "step": 5048
+    },
+    {
+      "epoch": 0.573198377981417,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2689,
+      "step": 5049
+    },
+    {
+      "epoch": 0.5733119050913361,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2583,
+      "step": 5050
+    },
+    {
+      "epoch": 0.5734254322012552,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.254,
+      "step": 5051
+    },
+    {
+      "epoch": 0.5735389593111743,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2539,
+      "step": 5052
+    },
+    {
+      "epoch": 0.5736524864210933,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2861,
+      "step": 5053
+    },
+    {
+      "epoch": 0.5737660135310124,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2428,
+      "step": 5054
+    },
+    {
+      "epoch": 0.5738795406409315,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2471,
+      "step": 5055
+    },
+    {
+      "epoch": 0.5739930677508506,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2761,
+      "step": 5056
+    },
+    {
+      "epoch": 0.5741065948607696,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2671,
+      "step": 5057
+    },
+    {
+      "epoch": 0.5742201219706887,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2595,
+      "step": 5058
+    },
+    {
+      "epoch": 0.5743336490806078,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2505,
+      "step": 5059
+    },
+    {
+      "epoch": 0.5744471761905269,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2721,
+      "step": 5060
+    },
+    {
+      "epoch": 0.574560703300446,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2642,
+      "step": 5061
+    },
+    {
+      "epoch": 0.574674230410365,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2767,
+      "step": 5062
+    },
+    {
+      "epoch": 0.5747877575202841,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2783,
+      "step": 5063
+    },
+    {
+      "epoch": 0.5749012846302032,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2391,
+      "step": 5064
+    },
+    {
+      "epoch": 0.5750148117401223,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2576,
+      "step": 5065
+    },
+    {
+      "epoch": 0.5751283388500413,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2672,
+      "step": 5066
+    },
+    {
+      "epoch": 0.5752418659599604,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.002,
+      "loss": 5.277,
+      "step": 5067
+    },
+    {
+      "epoch": 0.5753553930698795,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2663,
+      "step": 5068
+    },
+    {
+      "epoch": 0.5754689201797986,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2694,
+      "step": 5069
+    },
+    {
+      "epoch": 0.5755824472897176,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2438,
+      "step": 5070
+    },
+    {
+      "epoch": 0.5756959743996367,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2638,
+      "step": 5071
+    },
+    {
+      "epoch": 0.5758095015095558,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2409,
+      "step": 5072
+    },
+    {
+      "epoch": 0.5759230286194749,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 5.245,
+      "step": 5073
+    },
+    {
+      "epoch": 0.5760365557293939,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.272,
+      "step": 5074
+    },
+    {
+      "epoch": 0.576150082839313,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2674,
+      "step": 5075
+    },
+    {
+      "epoch": 0.5762636099492321,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.261,
+      "step": 5076
+    },
+    {
+      "epoch": 0.5763771370591512,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.258,
+      "step": 5077
+    },
+    {
+      "epoch": 0.5764906641690702,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 5.2687,
+      "step": 5078
+    },
+    {
+      "epoch": 0.5766041912789893,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2592,
+      "step": 5079
+    },
+    {
+      "epoch": 0.5767177183889084,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2566,
+      "step": 5080
+    },
+    {
+      "epoch": 0.5768312454988275,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2544,
+      "step": 5081
+    },
+    {
+      "epoch": 0.5769447726087465,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2699,
+      "step": 5082
+    },
+    {
+      "epoch": 0.5770582997186656,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2321,
+      "step": 5083
+    },
+    {
+      "epoch": 0.5771718268285847,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2409,
+      "step": 5084
+    },
+    {
+      "epoch": 0.5772853539385038,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2617,
+      "step": 5085
+    },
+    {
+      "epoch": 0.5773988810484229,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2462,
+      "step": 5086
+    },
+    {
+      "epoch": 0.5775124081583419,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.2618,
+      "step": 5087
+    },
+    {
+      "epoch": 0.577625935268261,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2622,
+      "step": 5088
+    },
+    {
+      "epoch": 0.5777394623781801,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2508,
+      "step": 5089
+    },
+    {
+      "epoch": 0.5778529894880992,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2617,
+      "step": 5090
+    },
+    {
+      "epoch": 0.5779665165980182,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2326,
+      "step": 5091
+    },
+    {
+      "epoch": 0.5780800437079373,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2541,
+      "step": 5092
+    },
+    {
+      "epoch": 0.5781935708178564,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2562,
+      "step": 5093
+    },
+    {
+      "epoch": 0.5783070979277755,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2614,
+      "step": 5094
+    },
+    {
+      "epoch": 0.5784206250376945,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2566,
+      "step": 5095
+    },
+    {
+      "epoch": 0.5785341521476136,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2463,
+      "step": 5096
+    },
+    {
+      "epoch": 0.5786476792575327,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2718,
+      "step": 5097
+    },
+    {
+      "epoch": 0.5787612063674518,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2538,
+      "step": 5098
+    },
+    {
+      "epoch": 0.5788747334773708,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2363,
+      "step": 5099
+    },
+    {
+      "epoch": 0.5789882605872899,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2567,
+      "step": 5100
+    },
+    {
+      "epoch": 0.579101787697209,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2499,
+      "step": 5101
+    },
+    {
+      "epoch": 0.5792153148071281,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2622,
+      "step": 5102
+    },
+    {
+      "epoch": 0.5793288419170471,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2555,
+      "step": 5103
+    },
+    {
+      "epoch": 0.5794423690269662,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2627,
+      "step": 5104
+    },
+    {
+      "epoch": 0.5795558961368853,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2719,
+      "step": 5105
+    },
+    {
+      "epoch": 0.5796694232468044,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2705,
+      "step": 5106
+    },
+    {
+      "epoch": 0.5797829503567234,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.2734,
+      "step": 5107
+    },
+    {
+      "epoch": 0.5798964774666425,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.2648,
+      "step": 5108
+    },
+    {
+      "epoch": 0.5800100045765616,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.2785,
+      "step": 5109
+    },
+    {
+      "epoch": 0.5801235316864807,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.2624,
+      "step": 5110
+    },
+    {
+      "epoch": 0.5802370587963998,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2666,
+      "step": 5111
+    },
+    {
+      "epoch": 0.5803505859063188,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2467,
+      "step": 5112
+    },
+    {
+      "epoch": 0.5804641130162379,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2524,
+      "step": 5113
+    },
+    {
+      "epoch": 0.580577640126157,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2508,
+      "step": 5114
+    },
+    {
+      "epoch": 0.5806911672360761,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2689,
+      "step": 5115
+    },
+    {
+      "epoch": 0.5808046943459951,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2446,
+      "step": 5116
+    },
+    {
+      "epoch": 0.5809182214559142,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2664,
+      "step": 5117
+    },
+    {
+      "epoch": 0.5810317485658333,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.261,
+      "step": 5118
+    },
+    {
+      "epoch": 0.5811452756757524,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2615,
+      "step": 5119
+    },
+    {
+      "epoch": 0.5812588027856714,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2589,
+      "step": 5120
+    },
+    {
+      "epoch": 0.5813723298955905,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2528,
+      "step": 5121
+    },
+    {
+      "epoch": 0.5814858570055096,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2696,
+      "step": 5122
+    },
+    {
+      "epoch": 0.5815993841154287,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2502,
+      "step": 5123
+    },
+    {
+      "epoch": 0.5817129112253477,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2365,
+      "step": 5124
+    },
+    {
+      "epoch": 0.5818264383352668,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.2585,
+      "step": 5125
+    },
+    {
+      "epoch": 0.5819399654451859,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2575,
+      "step": 5126
+    },
+    {
+      "epoch": 0.582053492555105,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2662,
+      "step": 5127
+    },
+    {
+      "epoch": 0.582167019665024,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2464,
+      "step": 5128
+    },
+    {
+      "epoch": 0.5822805467749431,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2513,
+      "step": 5129
+    },
+    {
+      "epoch": 0.5823940738848622,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2852,
+      "step": 5130
+    },
+    {
+      "epoch": 0.5825076009947813,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2704,
+      "step": 5131
+    },
+    {
+      "epoch": 0.5826211281047003,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2662,
+      "step": 5132
+    },
+    {
+      "epoch": 0.5827346552146194,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2529,
+      "step": 5133
+    },
+    {
+      "epoch": 0.5828481823245385,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2883,
+      "step": 5134
+    },
+    {
+      "epoch": 0.5829617094344576,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2539,
+      "step": 5135
+    },
+    {
+      "epoch": 0.5830752365443767,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2347,
+      "step": 5136
+    },
+    {
+      "epoch": 0.5831887636542957,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2561,
+      "step": 5137
+    },
+    {
+      "epoch": 0.5833022907642148,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2385,
+      "step": 5138
+    },
+    {
+      "epoch": 0.5834158178741339,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.225,
+      "step": 5139
+    },
+    {
+      "epoch": 0.583529344984053,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2732,
+      "step": 5140
+    },
+    {
+      "epoch": 0.583642872093972,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.257,
+      "step": 5141
+    },
+    {
+      "epoch": 0.5837563992038911,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2609,
+      "step": 5142
+    },
+    {
+      "epoch": 0.5838699263138102,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2737,
+      "step": 5143
+    },
+    {
+      "epoch": 0.5839834534237293,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2533,
+      "step": 5144
+    },
+    {
+      "epoch": 0.5840969805336483,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.2559,
+      "step": 5145
+    },
+    {
+      "epoch": 0.5842105076435674,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.002,
+      "loss": 5.2473,
+      "step": 5146
+    },
+    {
+      "epoch": 0.5843240347534865,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.002,
+      "loss": 5.2597,
+      "step": 5147
+    },
+    {
+      "epoch": 0.5844375618634056,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 5.2586,
+      "step": 5148
+    },
+    {
+      "epoch": 0.5845510889733246,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2581,
+      "step": 5149
+    },
+    {
+      "epoch": 0.5846646160832437,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.002,
+      "loss": 5.245,
+      "step": 5150
+    },
+    {
+      "epoch": 0.5847781431931628,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 5.2721,
+      "step": 5151
+    },
+    {
+      "epoch": 0.5848916703030819,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2573,
+      "step": 5152
+    },
+    {
+      "epoch": 0.5850051974130009,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2539,
+      "step": 5153
+    },
+    {
+      "epoch": 0.58511872452292,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2722,
+      "step": 5154
+    },
+    {
+      "epoch": 0.5852322516328391,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2545,
+      "step": 5155
+    },
+    {
+      "epoch": 0.5853457787427582,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2622,
+      "step": 5156
+    },
+    {
+      "epoch": 0.5854593058526772,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2466,
+      "step": 5157
+    },
+    {
+      "epoch": 0.5855728329625963,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.257,
+      "step": 5158
+    },
+    {
+      "epoch": 0.5856863600725154,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2528,
+      "step": 5159
+    },
+    {
+      "epoch": 0.5857998871824345,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2651,
+      "step": 5160
+    },
+    {
+      "epoch": 0.5859134142923536,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2415,
+      "step": 5161
+    },
+    {
+      "epoch": 0.5860269414022726,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2645,
+      "step": 5162
+    },
+    {
+      "epoch": 0.5861404685121917,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2317,
+      "step": 5163
+    },
+    {
+      "epoch": 0.5862539956221108,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2435,
+      "step": 5164
+    },
+    {
+      "epoch": 0.5863675227320299,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2519,
+      "step": 5165
+    },
+    {
+      "epoch": 0.5864810498419489,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2518,
+      "step": 5166
+    },
+    {
+      "epoch": 0.586594576951868,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2673,
+      "step": 5167
+    },
+    {
+      "epoch": 0.5867081040617871,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2631,
+      "step": 5168
+    },
+    {
+      "epoch": 0.5868216311717062,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.271,
+      "step": 5169
+    },
+    {
+      "epoch": 0.5869351582816252,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2642,
+      "step": 5170
+    },
+    {
+      "epoch": 0.5870486853915443,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2781,
+      "step": 5171
+    },
+    {
+      "epoch": 0.5871622125014634,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2708,
+      "step": 5172
+    },
+    {
+      "epoch": 0.5872757396113825,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2594,
+      "step": 5173
+    },
+    {
+      "epoch": 0.5873892667213015,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2595,
+      "step": 5174
+    },
+    {
+      "epoch": 0.5875027938312206,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2582,
+      "step": 5175
+    },
+    {
+      "epoch": 0.5876163209411397,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.275,
+      "step": 5176
+    },
+    {
+      "epoch": 0.5877298480510588,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2654,
+      "step": 5177
+    },
+    {
+      "epoch": 0.5878433751609778,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.234,
+      "step": 5178
+    },
+    {
+      "epoch": 0.5879569022708969,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2704,
+      "step": 5179
+    },
+    {
+      "epoch": 0.588070429380816,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2387,
+      "step": 5180
+    },
+    {
+      "epoch": 0.5881839564907351,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2716,
+      "step": 5181
+    },
+    {
+      "epoch": 0.5882974836006541,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2482,
+      "step": 5182
+    },
+    {
+      "epoch": 0.5884110107105732,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2657,
+      "step": 5183
+    },
+    {
+      "epoch": 0.5885245378204923,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2374,
+      "step": 5184
+    },
+    {
+      "epoch": 0.5886380649304114,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2697,
+      "step": 5185
+    },
+    {
+      "epoch": 0.5887515920403305,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2309,
+      "step": 5186
+    },
+    {
+      "epoch": 0.5888651191502495,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2389,
+      "step": 5187
+    },
+    {
+      "epoch": 0.5889786462601686,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2659,
+      "step": 5188
+    },
+    {
+      "epoch": 0.5890921733700878,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2663,
+      "step": 5189
+    },
+    {
+      "epoch": 0.5892057004800069,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2498,
+      "step": 5190
+    },
+    {
+      "epoch": 0.5893192275899259,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2371,
+      "step": 5191
+    },
+    {
+      "epoch": 0.589432754699845,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.252,
+      "step": 5192
+    },
+    {
+      "epoch": 0.5895462818097641,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2727,
+      "step": 5193
+    },
+    {
+      "epoch": 0.5896598089196832,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.2689,
+      "step": 5194
+    },
+    {
+      "epoch": 0.5897733360296022,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.26,
+      "step": 5195
+    },
+    {
+      "epoch": 0.5898868631395213,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.2429,
+      "step": 5196
+    },
+    {
+      "epoch": 0.5900003902494404,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2411,
+      "step": 5197
+    },
+    {
+      "epoch": 0.5901139173593595,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2626,
+      "step": 5198
+    },
+    {
+      "epoch": 0.5902274444692786,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2747,
+      "step": 5199
+    },
+    {
+      "epoch": 0.5903409715791976,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2296,
+      "step": 5200
+    },
+    {
+      "epoch": 0.5904544986891167,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2764,
+      "step": 5201
+    },
+    {
+      "epoch": 0.5905680257990358,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2299,
+      "step": 5202
+    },
+    {
+      "epoch": 0.5906815529089549,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2282,
+      "step": 5203
+    },
+    {
+      "epoch": 0.5907950800188739,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2665,
+      "step": 5204
+    },
+    {
+      "epoch": 0.590908607128793,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2421,
+      "step": 5205
+    },
+    {
+      "epoch": 0.5910221342387121,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.254,
+      "step": 5206
+    },
+    {
+      "epoch": 0.5911356613486312,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.262,
+      "step": 5207
+    },
+    {
+      "epoch": 0.5912491884585502,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2487,
+      "step": 5208
+    },
+    {
+      "epoch": 0.5913627155684693,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2488,
+      "step": 5209
+    },
+    {
+      "epoch": 0.5914762426783884,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2583,
+      "step": 5210
+    },
+    {
+      "epoch": 0.5915897697883075,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2607,
+      "step": 5211
+    },
+    {
+      "epoch": 0.5917032968982265,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2359,
+      "step": 5212
+    },
+    {
+      "epoch": 0.5918168240081456,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2655,
+      "step": 5213
+    },
+    {
+      "epoch": 0.5919303511180647,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2623,
+      "step": 5214
+    },
+    {
+      "epoch": 0.5920438782279838,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2612,
+      "step": 5215
+    },
+    {
+      "epoch": 0.5921574053379028,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.242,
+      "step": 5216
+    },
+    {
+      "epoch": 0.5922709324478219,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2404,
+      "step": 5217
+    },
+    {
+      "epoch": 0.592384459557741,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2644,
+      "step": 5218
+    },
+    {
+      "epoch": 0.5924979866676601,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2483,
+      "step": 5219
+    },
+    {
+      "epoch": 0.5926115137775791,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2536,
+      "step": 5220
+    },
+    {
+      "epoch": 0.5927250408874982,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2664,
+      "step": 5221
+    },
+    {
+      "epoch": 0.5928385679974173,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2746,
+      "step": 5222
+    },
+    {
+      "epoch": 0.5929520951073364,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2493,
+      "step": 5223
+    },
+    {
+      "epoch": 0.5930656222172555,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.258,
+      "step": 5224
+    },
+    {
+      "epoch": 0.5931791493271745,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.2746,
+      "step": 5225
+    },
+    {
+      "epoch": 0.5932926764370936,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 5.2684,
+      "step": 5226
+    },
+    {
+      "epoch": 0.5934062035470127,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.2521,
+      "step": 5227
+    },
+    {
+      "epoch": 0.5935197306569318,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2577,
+      "step": 5228
+    },
+    {
+      "epoch": 0.5936332577668508,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.272,
+      "step": 5229
+    },
+    {
+      "epoch": 0.5937467848767699,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2604,
+      "step": 5230
+    },
+    {
+      "epoch": 0.593860311986689,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.2616,
+      "step": 5231
+    },
+    {
+      "epoch": 0.5939738390966081,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.2644,
+      "step": 5232
+    },
+    {
+      "epoch": 0.5940873662065271,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 5.2245,
+      "step": 5233
+    },
+    {
+      "epoch": 0.5942008933164462,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2512,
+      "step": 5234
+    },
+    {
+      "epoch": 0.5943144204263653,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2504,
+      "step": 5235
+    },
+    {
+      "epoch": 0.5944279475362844,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2473,
+      "step": 5236
+    },
+    {
+      "epoch": 0.5945414746462034,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.2347,
+      "step": 5237
+    },
+    {
+      "epoch": 0.5946550017561225,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.2628,
+      "step": 5238
+    },
+    {
+      "epoch": 0.5947685288660416,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.2607,
+      "step": 5239
+    },
+    {
+      "epoch": 0.5948820559759607,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.2365,
+      "step": 5240
+    },
+    {
+      "epoch": 0.5949955830858797,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2566,
+      "step": 5241
+    },
+    {
+      "epoch": 0.5951091101957988,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2744,
+      "step": 5242
+    },
+    {
+      "epoch": 0.5952226373057179,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2475,
+      "step": 5243
+    },
+    {
+      "epoch": 0.595336164415637,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2586,
+      "step": 5244
+    },
+    {
+      "epoch": 0.595449691525556,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2646,
+      "step": 5245
+    },
+    {
+      "epoch": 0.5955632186354751,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2392,
+      "step": 5246
+    },
+    {
+      "epoch": 0.5956767457453942,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2544,
+      "step": 5247
+    },
+    {
+      "epoch": 0.5957902728553133,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.259,
+      "step": 5248
+    },
+    {
+      "epoch": 0.5959037999652324,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2751,
+      "step": 5249
+    },
+    {
+      "epoch": 0.5960173270751514,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2451,
+      "step": 5250
+    },
+    {
+      "epoch": 0.5961308541850705,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2434,
+      "step": 5251
+    },
+    {
+      "epoch": 0.5962443812949896,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2707,
+      "step": 5252
+    },
+    {
+      "epoch": 0.5963579084049087,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2563,
+      "step": 5253
+    },
+    {
+      "epoch": 0.5964714355148277,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2741,
+      "step": 5254
+    },
+    {
+      "epoch": 0.5965849626247468,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2614,
+      "step": 5255
+    },
+    {
+      "epoch": 0.5966984897346659,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2671,
+      "step": 5256
+    },
+    {
+      "epoch": 0.596812016844585,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2739,
+      "step": 5257
+    },
+    {
+      "epoch": 0.596925543954504,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.235,
+      "step": 5258
+    },
+    {
+      "epoch": 0.5970390710644231,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2428,
+      "step": 5259
+    },
+    {
+      "epoch": 0.5971525981743422,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2545,
+      "step": 5260
+    },
+    {
+      "epoch": 0.5972661252842613,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2599,
+      "step": 5261
+    },
+    {
+      "epoch": 0.5973796523941803,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2645,
+      "step": 5262
+    },
+    {
+      "epoch": 0.5974931795040994,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2762,
+      "step": 5263
+    },
+    {
+      "epoch": 0.5976067066140185,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2699,
+      "step": 5264
+    },
+    {
+      "epoch": 0.5977202337239376,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2585,
+      "step": 5265
+    },
+    {
+      "epoch": 0.5978337608338566,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2605,
+      "step": 5266
+    },
+    {
+      "epoch": 0.5979472879437757,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2698,
+      "step": 5267
+    },
+    {
+      "epoch": 0.5980608150536948,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.247,
+      "step": 5268
+    },
+    {
+      "epoch": 0.5981743421636139,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2558,
+      "step": 5269
+    },
+    {
+      "epoch": 0.598287869273533,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2721,
+      "step": 5270
+    },
+    {
+      "epoch": 0.598401396383452,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2664,
+      "step": 5271
+    },
+    {
+      "epoch": 0.5985149234933711,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2497,
+      "step": 5272
+    },
+    {
+      "epoch": 0.5986284506032902,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2392,
+      "step": 5273
+    },
+    {
+      "epoch": 0.5987419777132093,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2861,
+      "step": 5274
+    },
+    {
+      "epoch": 0.5988555048231283,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 5.2516,
+      "step": 5275
+    },
+    {
+      "epoch": 0.5989690319330474,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.2648,
+      "step": 5276
+    },
+    {
+      "epoch": 0.5990825590429665,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2501,
+      "step": 5277
+    },
+    {
+      "epoch": 0.5991960861528856,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2558,
+      "step": 5278
+    },
+    {
+      "epoch": 0.5993096132628046,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2599,
+      "step": 5279
+    },
+    {
+      "epoch": 0.5994231403727237,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2443,
+      "step": 5280
+    },
+    {
+      "epoch": 0.5995366674826428,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2477,
+      "step": 5281
+    },
+    {
+      "epoch": 0.5996501945925619,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2688,
+      "step": 5282
+    },
+    {
+      "epoch": 0.5997637217024809,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.2672,
+      "step": 5283
+    },
+    {
+      "epoch": 0.5998772488124,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.002,
+      "loss": 5.2474,
+      "step": 5284
+    },
+    {
+      "epoch": 0.5999907759223191,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.002,
+      "loss": 5.2648,
+      "step": 5285
+    },
+    {
+      "epoch": 0.6001043030322382,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.2432,
+      "step": 5286
+    },
+    {
+      "epoch": 0.6002178301421572,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.267,
+      "step": 5287
+    },
+    {
+      "epoch": 0.6003313572520763,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2419,
+      "step": 5288
+    },
+    {
+      "epoch": 0.6004448843619954,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2454,
+      "step": 5289
+    },
+    {
+      "epoch": 0.6005584114719145,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2602,
+      "step": 5290
+    },
+    {
+      "epoch": 0.6006719385818335,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2598,
+      "step": 5291
+    },
+    {
+      "epoch": 0.6007854656917526,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2727,
+      "step": 5292
+    },
+    {
+      "epoch": 0.6008989928016717,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2549,
+      "step": 5293
+    },
+    {
+      "epoch": 0.6010125199115908,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.2785,
+      "step": 5294
+    },
+    {
+      "epoch": 0.6011260470215098,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2637,
+      "step": 5295
+    },
+    {
+      "epoch": 0.6012395741314289,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2356,
+      "step": 5296
+    },
+    {
+      "epoch": 0.601353101241348,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2638,
+      "step": 5297
+    },
+    {
+      "epoch": 0.6014666283512671,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2754,
+      "step": 5298
+    },
+    {
+      "epoch": 0.6015801554611862,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2486,
+      "step": 5299
+    },
+    {
+      "epoch": 0.6016936825711052,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2396,
+      "step": 5300
+    },
+    {
+      "epoch": 0.6018072096810243,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2538,
+      "step": 5301
+    },
+    {
+      "epoch": 0.6019207367909434,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.271,
+      "step": 5302
+    },
+    {
+      "epoch": 0.6020342639008625,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2687,
+      "step": 5303
+    },
+    {
+      "epoch": 0.6021477910107815,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2515,
+      "step": 5304
+    },
+    {
+      "epoch": 0.6022613181207006,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2502,
+      "step": 5305
+    },
+    {
+      "epoch": 0.6023748452306197,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 5.2708,
+      "step": 5306
+    },
+    {
+      "epoch": 0.6024883723405388,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2513,
+      "step": 5307
+    },
+    {
+      "epoch": 0.6026018994504578,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2462,
+      "step": 5308
+    },
+    {
+      "epoch": 0.6027154265603769,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.26,
+      "step": 5309
+    },
+    {
+      "epoch": 0.602828953670296,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2584,
+      "step": 5310
+    },
+    {
+      "epoch": 0.6029424807802151,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2566,
+      "step": 5311
+    },
+    {
+      "epoch": 0.6030560078901341,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.273,
+      "step": 5312
+    },
+    {
+      "epoch": 0.6031695350000532,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2666,
+      "step": 5313
+    },
+    {
+      "epoch": 0.6032830621099723,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2355,
+      "step": 5314
+    },
+    {
+      "epoch": 0.6033965892198914,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2795,
+      "step": 5315
+    },
+    {
+      "epoch": 0.6035101163298104,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.2415,
+      "step": 5316
+    },
+    {
+      "epoch": 0.6036236434397295,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2623,
+      "step": 5317
+    },
+    {
+      "epoch": 0.6037371705496486,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.244,
+      "step": 5318
+    },
+    {
+      "epoch": 0.6038506976595677,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2521,
+      "step": 5319
+    },
+    {
+      "epoch": 0.6039642247694867,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2543,
+      "step": 5320
+    },
+    {
+      "epoch": 0.6040777518794058,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.264,
+      "step": 5321
+    },
+    {
+      "epoch": 0.6041912789893249,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.2584,
+      "step": 5322
+    },
+    {
+      "epoch": 0.604304806099244,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2482,
+      "step": 5323
+    },
+    {
+      "epoch": 0.604418333209163,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2522,
+      "step": 5324
+    },
+    {
+      "epoch": 0.6045318603190821,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.242,
+      "step": 5325
+    },
+    {
+      "epoch": 0.6046453874290012,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2637,
+      "step": 5326
+    },
+    {
+      "epoch": 0.6047589145389203,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.2611,
+      "step": 5327
+    },
+    {
+      "epoch": 0.6048724416488394,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.2394,
+      "step": 5328
+    },
+    {
+      "epoch": 0.6049859687587584,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.2666,
+      "step": 5329
+    },
+    {
+      "epoch": 0.6050994958686775,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2464,
+      "step": 5330
+    },
+    {
+      "epoch": 0.6052130229785966,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.2435,
+      "step": 5331
+    },
+    {
+      "epoch": 0.6053265500885157,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2676,
+      "step": 5332
+    },
+    {
+      "epoch": 0.6054400771984347,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2704,
+      "step": 5333
+    },
+    {
+      "epoch": 0.6055536043083538,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2548,
+      "step": 5334
+    },
+    {
+      "epoch": 0.6056671314182729,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2484,
+      "step": 5335
+    },
+    {
+      "epoch": 0.605780658528192,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2498,
+      "step": 5336
+    },
+    {
+      "epoch": 0.605894185638111,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2606,
+      "step": 5337
+    },
+    {
+      "epoch": 0.6060077127480301,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.242,
+      "step": 5338
+    },
+    {
+      "epoch": 0.6061212398579492,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2569,
+      "step": 5339
+    },
+    {
+      "epoch": 0.6062347669678683,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2472,
+      "step": 5340
+    },
+    {
+      "epoch": 0.6063482940777873,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2459,
+      "step": 5341
+    },
+    {
+      "epoch": 0.6064618211877064,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.251,
+      "step": 5342
+    },
+    {
+      "epoch": 0.6065753482976255,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.245,
+      "step": 5343
+    },
+    {
+      "epoch": 0.6066888754075446,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2367,
+      "step": 5344
+    },
+    {
+      "epoch": 0.6068024025174636,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.27,
+      "step": 5345
+    },
+    {
+      "epoch": 0.6069159296273827,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2629,
+      "step": 5346
+    },
+    {
+      "epoch": 0.6070294567373018,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2746,
+      "step": 5347
+    },
+    {
+      "epoch": 0.6071429838472209,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.233,
+      "step": 5348
+    },
+    {
+      "epoch": 0.60725651095714,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2547,
+      "step": 5349
+    },
+    {
+      "epoch": 0.607370038067059,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2669,
+      "step": 5350
+    },
+    {
+      "epoch": 0.6074835651769781,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2702,
+      "step": 5351
+    },
+    {
+      "epoch": 0.6075970922868972,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2626,
+      "step": 5352
+    },
+    {
+      "epoch": 0.6077106193968163,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2409,
+      "step": 5353
+    },
+    {
+      "epoch": 0.6078241465067353,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2521,
+      "step": 5354
+    },
+    {
+      "epoch": 0.6079376736166544,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2363,
+      "step": 5355
+    },
+    {
+      "epoch": 0.6080512007265735,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2493,
+      "step": 5356
+    },
+    {
+      "epoch": 0.6081647278364926,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2609,
+      "step": 5357
+    },
+    {
+      "epoch": 0.6082782549464116,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2474,
+      "step": 5358
+    },
+    {
+      "epoch": 0.6083917820563307,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.232,
+      "step": 5359
+    },
+    {
+      "epoch": 0.6085053091662498,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2526,
+      "step": 5360
+    },
+    {
+      "epoch": 0.6086188362761689,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2606,
+      "step": 5361
+    },
+    {
+      "epoch": 0.6087323633860879,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2512,
+      "step": 5362
+    },
+    {
+      "epoch": 0.608845890496007,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2506,
+      "step": 5363
+    },
+    {
+      "epoch": 0.6089594176059261,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2475,
+      "step": 5364
+    },
+    {
+      "epoch": 0.6090729447158452,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2536,
+      "step": 5365
+    },
+    {
+      "epoch": 0.6091864718257642,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2398,
+      "step": 5366
+    },
+    {
+      "epoch": 0.6092999989356833,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2631,
+      "step": 5367
+    },
+    {
+      "epoch": 0.6094135260456024,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2641,
+      "step": 5368
+    },
+    {
+      "epoch": 0.6095270531555215,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2621,
+      "step": 5369
+    },
+    {
+      "epoch": 0.6096405802654405,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2609,
+      "step": 5370
+    },
+    {
+      "epoch": 0.6097541073753596,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.2591,
+      "step": 5371
+    },
+    {
+      "epoch": 0.6098676344852787,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 5.2626,
+      "step": 5372
+    },
+    {
+      "epoch": 0.6099811615951978,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2579,
+      "step": 5373
+    },
+    {
+      "epoch": 0.6100946887051168,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 5.2467,
+      "step": 5374
+    },
+    {
+      "epoch": 0.6102082158150359,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2537,
+      "step": 5375
+    },
+    {
+      "epoch": 0.610321742924955,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2357,
+      "step": 5376
+    },
+    {
+      "epoch": 0.6104352700348741,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2715,
+      "step": 5377
+    },
+    {
+      "epoch": 0.6105487971447932,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2627,
+      "step": 5378
+    },
+    {
+      "epoch": 0.6106623242547122,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2396,
+      "step": 5379
+    },
+    {
+      "epoch": 0.6107758513646313,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.2629,
+      "step": 5380
+    },
+    {
+      "epoch": 0.6108893784745504,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.2644,
+      "step": 5381
+    },
+    {
+      "epoch": 0.6110029055844695,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2613,
+      "step": 5382
+    },
+    {
+      "epoch": 0.6111164326943885,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.292,
+      "step": 5383
+    },
+    {
+      "epoch": 0.6112299598043076,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2515,
+      "step": 5384
+    },
+    {
+      "epoch": 0.6113434869142267,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2357,
+      "step": 5385
+    },
+    {
+      "epoch": 0.6114570140241458,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2526,
+      "step": 5386
+    },
+    {
+      "epoch": 0.6115705411340648,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2487,
+      "step": 5387
+    },
+    {
+      "epoch": 0.6116840682439839,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2498,
+      "step": 5388
+    },
+    {
+      "epoch": 0.611797595353903,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2431,
+      "step": 5389
+    },
+    {
+      "epoch": 0.6119111224638221,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2743,
+      "step": 5390
+    },
+    {
+      "epoch": 0.6120246495737411,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2472,
+      "step": 5391
+    },
+    {
+      "epoch": 0.6121381766836602,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2494,
+      "step": 5392
+    },
+    {
+      "epoch": 0.6122517037935793,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2495,
+      "step": 5393
+    },
+    {
+      "epoch": 0.6123652309034984,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2873,
+      "step": 5394
+    },
+    {
+      "epoch": 0.6124787580134174,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.246,
+      "step": 5395
+    },
+    {
+      "epoch": 0.6125922851233365,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2416,
+      "step": 5396
+    },
+    {
+      "epoch": 0.6127058122332556,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2757,
+      "step": 5397
+    },
+    {
+      "epoch": 0.6128193393431747,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2772,
+      "step": 5398
+    },
+    {
+      "epoch": 0.6129328664530937,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2584,
+      "step": 5399
+    },
+    {
+      "epoch": 0.6130463935630128,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2593,
+      "step": 5400
+    },
+    {
+      "epoch": 0.6131599206729319,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2691,
+      "step": 5401
+    },
+    {
+      "epoch": 0.613273447782851,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 5.2611,
+      "step": 5402
+    },
+    {
+      "epoch": 0.61338697489277,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.2766,
+      "step": 5403
+    },
+    {
+      "epoch": 0.6135005020026891,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.2555,
+      "step": 5404
+    },
+    {
+      "epoch": 0.6136140291126082,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.2561,
+      "step": 5405
+    },
+    {
+      "epoch": 0.6137275562225273,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2548,
+      "step": 5406
+    },
+    {
+      "epoch": 0.6138410833324464,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2744,
+      "step": 5407
+    },
+    {
+      "epoch": 0.6139546104423654,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2686,
+      "step": 5408
+    },
+    {
+      "epoch": 0.6140681375522845,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2413,
+      "step": 5409
+    },
+    {
+      "epoch": 0.6141816646622036,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2812,
+      "step": 5410
+    },
+    {
+      "epoch": 0.6142951917721227,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2492,
+      "step": 5411
+    },
+    {
+      "epoch": 0.6144087188820417,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2491,
+      "step": 5412
+    },
+    {
+      "epoch": 0.6145222459919608,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2778,
+      "step": 5413
+    },
+    {
+      "epoch": 0.6146357731018799,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2514,
+      "step": 5414
+    },
+    {
+      "epoch": 0.614749300211799,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2692,
+      "step": 5415
+    },
+    {
+      "epoch": 0.614862827321718,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2439,
+      "step": 5416
+    },
+    {
+      "epoch": 0.6149763544316371,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2646,
+      "step": 5417
+    },
+    {
+      "epoch": 0.6150898815415562,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2623,
+      "step": 5418
+    },
+    {
+      "epoch": 0.6152034086514753,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2853,
+      "step": 5419
+    },
+    {
+      "epoch": 0.6153169357613943,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2627,
+      "step": 5420
+    },
+    {
+      "epoch": 0.6154304628713134,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2472,
+      "step": 5421
+    },
+    {
+      "epoch": 0.6155439899812325,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2636,
+      "step": 5422
+    },
+    {
+      "epoch": 0.6156575170911516,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2643,
+      "step": 5423
+    },
+    {
+      "epoch": 0.6157710442010706,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2554,
+      "step": 5424
+    },
+    {
+      "epoch": 0.6158845713109897,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2655,
+      "step": 5425
+    },
+    {
+      "epoch": 0.6159980984209088,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2571,
+      "step": 5426
+    },
+    {
+      "epoch": 0.6161116255308279,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2751,
+      "step": 5427
+    },
+    {
+      "epoch": 0.616225152640747,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2598,
+      "step": 5428
+    },
+    {
+      "epoch": 0.616338679750666,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2603,
+      "step": 5429
+    },
+    {
+      "epoch": 0.6164522068605852,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2542,
+      "step": 5430
+    },
+    {
+      "epoch": 0.6165657339705043,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2505,
+      "step": 5431
+    },
+    {
+      "epoch": 0.6166792610804234,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2404,
+      "step": 5432
+    },
+    {
+      "epoch": 0.6167927881903424,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2635,
+      "step": 5433
+    },
+    {
+      "epoch": 0.6169063153002615,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2748,
+      "step": 5434
+    },
+    {
+      "epoch": 0.6170198424101806,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2609,
+      "step": 5435
+    },
+    {
+      "epoch": 0.6171333695200997,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2477,
+      "step": 5436
+    },
+    {
+      "epoch": 0.6172468966300187,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2531,
+      "step": 5437
+    },
+    {
+      "epoch": 0.6173604237399378,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2567,
+      "step": 5438
+    },
+    {
+      "epoch": 0.6174739508498569,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2669,
+      "step": 5439
+    },
+    {
+      "epoch": 0.617587477959776,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2463,
+      "step": 5440
+    },
+    {
+      "epoch": 0.617701005069695,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2529,
+      "step": 5441
+    },
+    {
+      "epoch": 0.6178145321796141,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2395,
+      "step": 5442
+    },
+    {
+      "epoch": 0.6179280592895332,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2444,
+      "step": 5443
+    },
+    {
+      "epoch": 0.6180415863994523,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2543,
+      "step": 5444
+    },
+    {
+      "epoch": 0.6181551135093714,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2493,
+      "step": 5445
+    },
+    {
+      "epoch": 0.6182686406192904,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.25,
+      "step": 5446
+    },
+    {
+      "epoch": 0.6183821677292095,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2624,
+      "step": 5447
+    },
+    {
+      "epoch": 0.6184956948391286,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2647,
+      "step": 5448
+    },
+    {
+      "epoch": 0.6186092219490477,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2596,
+      "step": 5449
+    },
+    {
+      "epoch": 0.6187227490589667,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2409,
+      "step": 5450
+    },
+    {
+      "epoch": 0.6188362761688858,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2474,
+      "step": 5451
+    },
+    {
+      "epoch": 0.6189498032788049,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2943,
+      "step": 5452
+    },
+    {
+      "epoch": 0.619063330388724,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.268,
+      "step": 5453
+    },
+    {
+      "epoch": 0.619176857498643,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2343,
+      "step": 5454
+    },
+    {
+      "epoch": 0.6192903846085621,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.244,
+      "step": 5455
+    },
+    {
+      "epoch": 0.6194039117184812,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2566,
+      "step": 5456
+    },
+    {
+      "epoch": 0.6195174388284003,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.2622,
+      "step": 5457
+    },
+    {
+      "epoch": 0.6196309659383193,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2345,
+      "step": 5458
+    },
+    {
+      "epoch": 0.6197444930482384,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.225,
+      "step": 5459
+    },
+    {
+      "epoch": 0.6198580201581575,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2564,
+      "step": 5460
+    },
+    {
+      "epoch": 0.6199715472680766,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2531,
+      "step": 5461
+    },
+    {
+      "epoch": 0.6200850743779956,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2493,
+      "step": 5462
+    },
+    {
+      "epoch": 0.6201986014879147,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2716,
+      "step": 5463
+    },
+    {
+      "epoch": 0.6203121285978338,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2576,
+      "step": 5464
+    },
+    {
+      "epoch": 0.6204256557077529,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2286,
+      "step": 5465
+    },
+    {
+      "epoch": 0.620539182817672,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2555,
+      "step": 5466
+    },
+    {
+      "epoch": 0.620652709927591,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2641,
+      "step": 5467
+    },
+    {
+      "epoch": 0.6207662370375101,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2656,
+      "step": 5468
+    },
+    {
+      "epoch": 0.6208797641474292,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2712,
+      "step": 5469
+    },
+    {
+      "epoch": 0.6209932912573483,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2571,
+      "step": 5470
+    },
+    {
+      "epoch": 0.6211068183672673,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.267,
+      "step": 5471
+    },
+    {
+      "epoch": 0.6212203454771864,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2469,
+      "step": 5472
+    },
+    {
+      "epoch": 0.6213338725871055,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2388,
+      "step": 5473
+    },
+    {
+      "epoch": 0.6214473996970246,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2681,
+      "step": 5474
+    },
+    {
+      "epoch": 0.6215609268069436,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2476,
+      "step": 5475
+    },
+    {
+      "epoch": 0.6216744539168627,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2382,
+      "step": 5476
+    },
+    {
+      "epoch": 0.6217879810267818,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2475,
+      "step": 5477
+    },
+    {
+      "epoch": 0.6219015081367009,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2408,
+      "step": 5478
+    },
+    {
+      "epoch": 0.6220150352466199,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2556,
+      "step": 5479
+    },
+    {
+      "epoch": 0.622128562356539,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2779,
+      "step": 5480
+    },
+    {
+      "epoch": 0.6222420894664581,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2565,
+      "step": 5481
+    },
+    {
+      "epoch": 0.6223556165763772,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2546,
+      "step": 5482
+    },
+    {
+      "epoch": 0.6224691436862962,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2408,
+      "step": 5483
+    },
+    {
+      "epoch": 0.6225826707962153,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2502,
+      "step": 5484
+    },
+    {
+      "epoch": 0.6226961979061344,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2628,
+      "step": 5485
+    },
+    {
+      "epoch": 0.6228097250160535,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2394,
+      "step": 5486
+    },
+    {
+      "epoch": 0.6229232521259725,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2688,
+      "step": 5487
+    },
+    {
+      "epoch": 0.6230367792358916,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2547,
+      "step": 5488
+    },
+    {
+      "epoch": 0.6231503063458107,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2511,
+      "step": 5489
+    },
+    {
+      "epoch": 0.6232638334557298,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2576,
+      "step": 5490
+    },
+    {
+      "epoch": 0.6233773605656489,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2314,
+      "step": 5491
+    },
+    {
+      "epoch": 0.6234908876755679,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.267,
+      "step": 5492
+    },
+    {
+      "epoch": 0.623604414785487,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.253,
+      "step": 5493
+    },
+    {
+      "epoch": 0.6237179418954061,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.002,
+      "loss": 5.25,
+      "step": 5494
+    },
+    {
+      "epoch": 0.6238314690053252,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.2679,
+      "step": 5495
+    },
+    {
+      "epoch": 0.6239449961152442,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2804,
+      "step": 5496
+    },
+    {
+      "epoch": 0.6240585232251633,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2516,
+      "step": 5497
+    },
+    {
+      "epoch": 0.6241720503350824,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.236,
+      "step": 5498
+    },
+    {
+      "epoch": 0.6242855774450015,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2427,
+      "step": 5499
+    },
+    {
+      "epoch": 0.6243991045549205,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2578,
+      "step": 5500
+    },
+    {
+      "epoch": 0.6245126316648396,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2594,
+      "step": 5501
+    },
+    {
+      "epoch": 0.6246261587747587,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2399,
+      "step": 5502
+    },
+    {
+      "epoch": 0.6247396858846778,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2344,
+      "step": 5503
+    },
+    {
+      "epoch": 0.6248532129945968,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2543,
+      "step": 5504
+    },
+    {
+      "epoch": 0.6249667401045159,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2599,
+      "step": 5505
+    },
+    {
+      "epoch": 0.625080267214435,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2656,
+      "step": 5506
+    },
+    {
+      "epoch": 0.6251937943243541,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2475,
+      "step": 5507
+    },
+    {
+      "epoch": 0.6253073214342731,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2548,
+      "step": 5508
+    },
+    {
+      "epoch": 0.6254208485441922,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2381,
+      "step": 5509
+    },
+    {
+      "epoch": 0.6255343756541113,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2639,
+      "step": 5510
+    },
+    {
+      "epoch": 0.6256479027640304,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2541,
+      "step": 5511
+    },
+    {
+      "epoch": 0.6257614298739494,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2394,
+      "step": 5512
+    },
+    {
+      "epoch": 0.6258749569838685,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.251,
+      "step": 5513
+    },
+    {
+      "epoch": 0.6259884840937876,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.2513,
+      "step": 5514
+    },
+    {
+      "epoch": 0.6261020112037067,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2463,
+      "step": 5515
+    },
+    {
+      "epoch": 0.6262155383136258,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2692,
+      "step": 5516
+    },
+    {
+      "epoch": 0.6263290654235448,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2495,
+      "step": 5517
+    },
+    {
+      "epoch": 0.6264425925334639,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2588,
+      "step": 5518
+    },
+    {
+      "epoch": 0.626556119643383,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2438,
+      "step": 5519
+    },
+    {
+      "epoch": 0.6266696467533021,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2592,
+      "step": 5520
+    },
+    {
+      "epoch": 0.6267831738632211,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2738,
+      "step": 5521
+    },
+    {
+      "epoch": 0.6268967009731402,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2529,
+      "step": 5522
+    },
+    {
+      "epoch": 0.6270102280830593,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2675,
+      "step": 5523
+    },
+    {
+      "epoch": 0.6271237551929784,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2409,
+      "step": 5524
+    },
+    {
+      "epoch": 0.6272372823028974,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2557,
+      "step": 5525
+    },
+    {
+      "epoch": 0.6273508094128165,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2528,
+      "step": 5526
+    },
+    {
+      "epoch": 0.6274643365227356,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2369,
+      "step": 5527
+    },
+    {
+      "epoch": 0.6275778636326547,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2477,
+      "step": 5528
+    },
+    {
+      "epoch": 0.6276913907425737,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2557,
+      "step": 5529
+    },
+    {
+      "epoch": 0.6278049178524928,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2609,
+      "step": 5530
+    },
+    {
+      "epoch": 0.6279184449624119,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2338,
+      "step": 5531
+    },
+    {
+      "epoch": 0.628031972072331,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2351,
+      "step": 5532
+    },
+    {
+      "epoch": 0.62814549918225,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.245,
+      "step": 5533
+    },
+    {
+      "epoch": 0.6282590262921691,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2501,
+      "step": 5534
+    },
+    {
+      "epoch": 0.6283725534020882,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.26,
+      "step": 5535
+    },
+    {
+      "epoch": 0.6284860805120073,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2596,
+      "step": 5536
+    },
+    {
+      "epoch": 0.6285996076219263,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2456,
+      "step": 5537
+    },
+    {
+      "epoch": 0.6287131347318454,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2576,
+      "step": 5538
+    },
+    {
+      "epoch": 0.6288266618417645,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2544,
+      "step": 5539
+    },
+    {
+      "epoch": 0.6289401889516836,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.252,
+      "step": 5540
+    },
+    {
+      "epoch": 0.6290537160616027,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2464,
+      "step": 5541
+    },
+    {
+      "epoch": 0.6291672431715217,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2549,
+      "step": 5542
+    },
+    {
+      "epoch": 0.6292807702814408,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.2724,
+      "step": 5543
+    },
+    {
+      "epoch": 0.6293942973913599,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2313,
+      "step": 5544
+    },
+    {
+      "epoch": 0.629507824501279,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2623,
+      "step": 5545
+    },
+    {
+      "epoch": 0.629621351611198,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2549,
+      "step": 5546
+    },
+    {
+      "epoch": 0.6297348787211171,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2753,
+      "step": 5547
+    },
+    {
+      "epoch": 0.6298484058310362,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2642,
+      "step": 5548
+    },
+    {
+      "epoch": 0.6299619329409553,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2473,
+      "step": 5549
+    },
+    {
+      "epoch": 0.6300754600508743,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2174,
+      "step": 5550
+    },
+    {
+      "epoch": 0.6301889871607934,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2524,
+      "step": 5551
+    },
+    {
+      "epoch": 0.6303025142707125,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2636,
+      "step": 5552
+    },
+    {
+      "epoch": 0.6304160413806316,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2374,
+      "step": 5553
+    },
+    {
+      "epoch": 0.6305295684905506,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.2524,
+      "step": 5554
+    },
+    {
+      "epoch": 0.6306430956004697,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.249,
+      "step": 5555
+    },
+    {
+      "epoch": 0.6307566227103888,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.2418,
+      "step": 5556
+    },
+    {
+      "epoch": 0.6308701498203079,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2524,
+      "step": 5557
+    },
+    {
+      "epoch": 0.6309836769302269,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2499,
+      "step": 5558
+    },
+    {
+      "epoch": 0.631097204040146,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2548,
+      "step": 5559
+    },
+    {
+      "epoch": 0.6312107311500651,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2683,
+      "step": 5560
+    },
+    {
+      "epoch": 0.6313242582599842,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2556,
+      "step": 5561
+    },
+    {
+      "epoch": 0.6314377853699032,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2494,
+      "step": 5562
+    },
+    {
+      "epoch": 0.6315513124798223,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2292,
+      "step": 5563
+    },
+    {
+      "epoch": 0.6316648395897414,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2361,
+      "step": 5564
+    },
+    {
+      "epoch": 0.6317783666996605,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2515,
+      "step": 5565
+    },
+    {
+      "epoch": 0.6318918938095796,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.267,
+      "step": 5566
+    },
+    {
+      "epoch": 0.6320054209194986,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.254,
+      "step": 5567
+    },
+    {
+      "epoch": 0.6321189480294177,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2403,
+      "step": 5568
+    },
+    {
+      "epoch": 0.6322324751393368,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2366,
+      "step": 5569
+    },
+    {
+      "epoch": 0.6323460022492559,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2506,
+      "step": 5570
+    },
+    {
+      "epoch": 0.6324595293591749,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.26,
+      "step": 5571
+    },
+    {
+      "epoch": 0.632573056469094,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2518,
+      "step": 5572
+    },
+    {
+      "epoch": 0.6326865835790131,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2383,
+      "step": 5573
+    },
+    {
+      "epoch": 0.6328001106889322,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2505,
+      "step": 5574
+    },
+    {
+      "epoch": 0.6329136377988512,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2609,
+      "step": 5575
+    },
+    {
+      "epoch": 0.6330271649087703,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.235,
+      "step": 5576
+    },
+    {
+      "epoch": 0.6331406920186894,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2423,
+      "step": 5577
+    },
+    {
+      "epoch": 0.6332542191286085,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2496,
+      "step": 5578
+    },
+    {
+      "epoch": 0.6333677462385275,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.2495,
+      "step": 5579
+    },
+    {
+      "epoch": 0.6334812733484466,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.2522,
+      "step": 5580
+    },
+    {
+      "epoch": 0.6335948004583657,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2457,
+      "step": 5581
+    },
+    {
+      "epoch": 0.6337083275682848,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2686,
+      "step": 5582
+    },
+    {
+      "epoch": 0.6338218546782038,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2627,
+      "step": 5583
+    },
+    {
+      "epoch": 0.6339353817881229,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2568,
+      "step": 5584
+    },
+    {
+      "epoch": 0.634048908898042,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2353,
+      "step": 5585
+    },
+    {
+      "epoch": 0.6341624360079611,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2608,
+      "step": 5586
+    },
+    {
+      "epoch": 0.6342759631178801,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.242,
+      "step": 5587
+    },
+    {
+      "epoch": 0.6343894902277992,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2517,
+      "step": 5588
+    },
+    {
+      "epoch": 0.6345030173377183,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2244,
+      "step": 5589
+    },
+    {
+      "epoch": 0.6346165444476374,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2546,
+      "step": 5590
+    },
+    {
+      "epoch": 0.6347300715575565,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2615,
+      "step": 5591
+    },
+    {
+      "epoch": 0.6348435986674755,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2606,
+      "step": 5592
+    },
+    {
+      "epoch": 0.6349571257773946,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2555,
+      "step": 5593
+    },
+    {
+      "epoch": 0.6350706528873137,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2448,
+      "step": 5594
+    },
+    {
+      "epoch": 0.6351841799972328,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2369,
+      "step": 5595
+    },
+    {
+      "epoch": 0.6352977071071518,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.223,
+      "step": 5596
+    },
+    {
+      "epoch": 0.6354112342170709,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 5.2474,
+      "step": 5597
+    },
+    {
+      "epoch": 0.63552476132699,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2539,
+      "step": 5598
+    },
+    {
+      "epoch": 0.6356382884369091,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 5.2579,
+      "step": 5599
+    },
+    {
+      "epoch": 0.6357518155468281,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2614,
+      "step": 5600
+    },
+    {
+      "epoch": 0.6358653426567472,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.25,
+      "step": 5601
+    },
+    {
+      "epoch": 0.6359788697666663,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2399,
+      "step": 5602
+    },
+    {
+      "epoch": 0.6360923968765854,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2465,
+      "step": 5603
+    },
+    {
+      "epoch": 0.6362059239865044,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2382,
+      "step": 5604
+    },
+    {
+      "epoch": 0.6363194510964235,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2258,
+      "step": 5605
+    },
+    {
+      "epoch": 0.6364329782063426,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2418,
+      "step": 5606
+    },
+    {
+      "epoch": 0.6365465053162617,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2521,
+      "step": 5607
+    },
+    {
+      "epoch": 0.6366600324261807,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2518,
+      "step": 5608
+    },
+    {
+      "epoch": 0.6367735595360998,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2547,
+      "step": 5609
+    },
+    {
+      "epoch": 0.6368870866460189,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2609,
+      "step": 5610
+    },
+    {
+      "epoch": 0.637000613755938,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2567,
+      "step": 5611
+    },
+    {
+      "epoch": 0.637114140865857,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2522,
+      "step": 5612
+    },
+    {
+      "epoch": 0.6372276679757761,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2504,
+      "step": 5613
+    },
+    {
+      "epoch": 0.6373411950856952,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2562,
+      "step": 5614
+    },
+    {
+      "epoch": 0.6374547221956143,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2664,
+      "step": 5615
+    },
+    {
+      "epoch": 0.6375682493055334,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2328,
+      "step": 5616
+    },
+    {
+      "epoch": 0.6376817764154524,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2663,
+      "step": 5617
+    },
+    {
+      "epoch": 0.6377953035253715,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.2565,
+      "step": 5618
+    },
+    {
+      "epoch": 0.6379088306352906,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2438,
+      "step": 5619
+    },
+    {
+      "epoch": 0.6380223577452097,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.002,
+      "loss": 5.235,
+      "step": 5620
+    },
+    {
+      "epoch": 0.6381358848551287,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2359,
+      "step": 5621
+    },
+    {
+      "epoch": 0.6382494119650478,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2668,
+      "step": 5622
+    },
+    {
+      "epoch": 0.6383629390749669,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2582,
+      "step": 5623
+    },
+    {
+      "epoch": 0.638476466184886,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.265,
+      "step": 5624
+    },
+    {
+      "epoch": 0.638589993294805,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2511,
+      "step": 5625
+    },
+    {
+      "epoch": 0.6387035204047241,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.222,
+      "step": 5626
+    },
+    {
+      "epoch": 0.6388170475146432,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2548,
+      "step": 5627
+    },
+    {
+      "epoch": 0.6389305746245623,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2438,
+      "step": 5628
+    },
+    {
+      "epoch": 0.6390441017344813,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2332,
+      "step": 5629
+    },
+    {
+      "epoch": 0.6391576288444004,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2459,
+      "step": 5630
+    },
+    {
+      "epoch": 0.6392711559543195,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2318,
+      "step": 5631
+    },
+    {
+      "epoch": 0.6393846830642386,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2634,
+      "step": 5632
+    },
+    {
+      "epoch": 0.6394982101741576,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2417,
+      "step": 5633
+    },
+    {
+      "epoch": 0.6396117372840767,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2905,
+      "step": 5634
+    },
+    {
+      "epoch": 0.6397252643939958,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2594,
+      "step": 5635
+    },
+    {
+      "epoch": 0.6398387915039149,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2458,
+      "step": 5636
+    },
+    {
+      "epoch": 0.639952318613834,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2654,
+      "step": 5637
+    },
+    {
+      "epoch": 0.640065845723753,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2493,
+      "step": 5638
+    },
+    {
+      "epoch": 0.6401793728336721,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2461,
+      "step": 5639
+    },
+    {
+      "epoch": 0.6402928999435912,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.255,
+      "step": 5640
+    },
+    {
+      "epoch": 0.6404064270535103,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.2404,
+      "step": 5641
+    },
+    {
+      "epoch": 0.6405199541634293,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.267,
+      "step": 5642
+    },
+    {
+      "epoch": 0.6406334812733484,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.2537,
+      "step": 5643
+    },
+    {
+      "epoch": 0.6407470083832675,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2537,
+      "step": 5644
+    },
+    {
+      "epoch": 0.6408605354931866,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.2737,
+      "step": 5645
+    },
+    {
+      "epoch": 0.6409740626031056,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2442,
+      "step": 5646
+    },
+    {
+      "epoch": 0.6410875897130247,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.243,
+      "step": 5647
+    },
+    {
+      "epoch": 0.6412011168229438,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2641,
+      "step": 5648
+    },
+    {
+      "epoch": 0.6413146439328629,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2508,
+      "step": 5649
+    },
+    {
+      "epoch": 0.6414281710427819,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2665,
+      "step": 5650
+    },
+    {
+      "epoch": 0.641541698152701,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2499,
+      "step": 5651
+    },
+    {
+      "epoch": 0.6416552252626201,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2418,
+      "step": 5652
+    },
+    {
+      "epoch": 0.6417687523725392,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2427,
+      "step": 5653
+    },
+    {
+      "epoch": 0.6418822794824582,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2678,
+      "step": 5654
+    },
+    {
+      "epoch": 0.6419958065923773,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2542,
+      "step": 5655
+    },
+    {
+      "epoch": 0.6421093337022964,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2591,
+      "step": 5656
+    },
+    {
+      "epoch": 0.6422228608122155,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2515,
+      "step": 5657
+    },
+    {
+      "epoch": 0.6423363879221345,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2469,
+      "step": 5658
+    },
+    {
+      "epoch": 0.6424499150320536,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2467,
+      "step": 5659
+    },
+    {
+      "epoch": 0.6425634421419727,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2342,
+      "step": 5660
+    },
+    {
+      "epoch": 0.6426769692518918,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2533,
+      "step": 5661
+    },
+    {
+      "epoch": 0.6427904963618108,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2626,
+      "step": 5662
+    },
+    {
+      "epoch": 0.6429040234717299,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2479,
+      "step": 5663
+    },
+    {
+      "epoch": 0.643017550581649,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2365,
+      "step": 5664
+    },
+    {
+      "epoch": 0.6431310776915681,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2519,
+      "step": 5665
+    },
+    {
+      "epoch": 0.6432446048014872,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2558,
+      "step": 5666
+    },
+    {
+      "epoch": 0.6433581319114062,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2557,
+      "step": 5667
+    },
+    {
+      "epoch": 0.6434716590213253,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2617,
+      "step": 5668
+    },
+    {
+      "epoch": 0.6435851861312444,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2617,
+      "step": 5669
+    },
+    {
+      "epoch": 0.6436987132411635,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2603,
+      "step": 5670
+    },
+    {
+      "epoch": 0.6438122403510825,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2355,
+      "step": 5671
+    },
+    {
+      "epoch": 0.6439257674610017,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2527,
+      "step": 5672
+    },
+    {
+      "epoch": 0.6440392945709208,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2576,
+      "step": 5673
+    },
+    {
+      "epoch": 0.6441528216808399,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.229,
+      "step": 5674
+    },
+    {
+      "epoch": 0.644266348790759,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2531,
+      "step": 5675
+    },
+    {
+      "epoch": 0.644379875900678,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2536,
+      "step": 5676
+    },
+    {
+      "epoch": 0.6444934030105971,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2422,
+      "step": 5677
+    },
+    {
+      "epoch": 0.6446069301205162,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2489,
+      "step": 5678
+    },
+    {
+      "epoch": 0.6447204572304353,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2607,
+      "step": 5679
+    },
+    {
+      "epoch": 0.6448339843403543,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2389,
+      "step": 5680
+    },
+    {
+      "epoch": 0.6449475114502734,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.002,
+      "loss": 5.2634,
+      "step": 5681
+    },
+    {
+      "epoch": 0.6450610385601925,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.002,
+      "loss": 5.2509,
+      "step": 5682
+    },
+    {
+      "epoch": 0.6451745656701116,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.2313,
+      "step": 5683
+    },
+    {
+      "epoch": 0.6452880927800306,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.002,
+      "loss": 5.2576,
+      "step": 5684
+    },
+    {
+      "epoch": 0.6454016198899497,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.2456,
+      "step": 5685
+    },
+    {
+      "epoch": 0.6455151469998688,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2338,
+      "step": 5686
+    },
+    {
+      "epoch": 0.6456286741097879,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2339,
+      "step": 5687
+    },
+    {
+      "epoch": 0.6457422012197069,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2463,
+      "step": 5688
+    },
+    {
+      "epoch": 0.645855728329626,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2319,
+      "step": 5689
+    },
+    {
+      "epoch": 0.6459692554395451,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2573,
+      "step": 5690
+    },
+    {
+      "epoch": 0.6460827825494642,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2331,
+      "step": 5691
+    },
+    {
+      "epoch": 0.6461963096593832,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2396,
+      "step": 5692
+    },
+    {
+      "epoch": 0.6463098367693023,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2397,
+      "step": 5693
+    },
+    {
+      "epoch": 0.6464233638792214,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2427,
+      "step": 5694
+    },
+    {
+      "epoch": 0.6465368909891405,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2612,
+      "step": 5695
+    },
+    {
+      "epoch": 0.6466504180990595,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2435,
+      "step": 5696
+    },
+    {
+      "epoch": 0.6467639452089786,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2546,
+      "step": 5697
+    },
+    {
+      "epoch": 0.6468774723188977,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2511,
+      "step": 5698
+    },
+    {
+      "epoch": 0.6469909994288168,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2284,
+      "step": 5699
+    },
+    {
+      "epoch": 0.6471045265387358,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2534,
+      "step": 5700
+    },
+    {
+      "epoch": 0.6472180536486549,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2725,
+      "step": 5701
+    },
+    {
+      "epoch": 0.647331580758574,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.244,
+      "step": 5702
+    },
+    {
+      "epoch": 0.6474451078684931,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2709,
+      "step": 5703
+    },
+    {
+      "epoch": 0.6475586349784122,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2625,
+      "step": 5704
+    },
+    {
+      "epoch": 0.6476721620883312,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2444,
+      "step": 5705
+    },
+    {
+      "epoch": 0.6477856891982503,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2418,
+      "step": 5706
+    },
+    {
+      "epoch": 0.6478992163081694,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2516,
+      "step": 5707
+    },
+    {
+      "epoch": 0.6480127434180885,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2276,
+      "step": 5708
+    },
+    {
+      "epoch": 0.6481262705280075,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2341,
+      "step": 5709
+    },
+    {
+      "epoch": 0.6482397976379266,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2489,
+      "step": 5710
+    },
+    {
+      "epoch": 0.6483533247478457,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2395,
+      "step": 5711
+    },
+    {
+      "epoch": 0.6484668518577648,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2449,
+      "step": 5712
+    },
+    {
+      "epoch": 0.6485803789676838,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2309,
+      "step": 5713
+    },
+    {
+      "epoch": 0.6486939060776029,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2451,
+      "step": 5714
+    },
+    {
+      "epoch": 0.648807433187522,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.261,
+      "step": 5715
+    },
+    {
+      "epoch": 0.6489209602974411,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2472,
+      "step": 5716
+    },
+    {
+      "epoch": 0.6490344874073601,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.2522,
+      "step": 5717
+    },
+    {
+      "epoch": 0.6491480145172792,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2614,
+      "step": 5718
+    },
+    {
+      "epoch": 0.6492615416271983,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2594,
+      "step": 5719
+    },
+    {
+      "epoch": 0.6493750687371174,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2448,
+      "step": 5720
+    },
+    {
+      "epoch": 0.6494885958470364,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2521,
+      "step": 5721
+    },
+    {
+      "epoch": 0.6496021229569555,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2552,
+      "step": 5722
+    },
+    {
+      "epoch": 0.6497156500668746,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2395,
+      "step": 5723
+    },
+    {
+      "epoch": 0.6498291771767937,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2493,
+      "step": 5724
+    },
+    {
+      "epoch": 0.6499427042867127,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2377,
+      "step": 5725
+    },
+    {
+      "epoch": 0.6500562313966318,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2389,
+      "step": 5726
+    },
+    {
+      "epoch": 0.6501697585065509,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2565,
+      "step": 5727
+    },
+    {
+      "epoch": 0.65028328561647,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2521,
+      "step": 5728
+    },
+    {
+      "epoch": 0.650396812726389,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2386,
+      "step": 5729
+    },
+    {
+      "epoch": 0.6505103398363081,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2743,
+      "step": 5730
+    },
+    {
+      "epoch": 0.6506238669462272,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2373,
+      "step": 5731
+    },
+    {
+      "epoch": 0.6507373940561463,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2528,
+      "step": 5732
+    },
+    {
+      "epoch": 0.6508509211660654,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2502,
+      "step": 5733
+    },
+    {
+      "epoch": 0.6509644482759844,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.002,
+      "loss": 5.2644,
+      "step": 5734
+    },
+    {
+      "epoch": 0.6510779753859035,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.002,
+      "loss": 5.2405,
+      "step": 5735
+    },
+    {
+      "epoch": 0.6511915024958226,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 5.2586,
+      "step": 5736
+    },
+    {
+      "epoch": 0.6513050296057417,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.002,
+      "loss": 5.2561,
+      "step": 5737
+    },
+    {
+      "epoch": 0.6514185567156607,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2477,
+      "step": 5738
+    },
+    {
+      "epoch": 0.6515320838255798,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.26,
+      "step": 5739
+    },
+    {
+      "epoch": 0.6516456109354989,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2371,
+      "step": 5740
+    },
+    {
+      "epoch": 0.651759138045418,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.2549,
+      "step": 5741
+    },
+    {
+      "epoch": 0.651872665155337,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2639,
+      "step": 5742
+    },
+    {
+      "epoch": 0.6519861922652561,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2486,
+      "step": 5743
+    },
+    {
+      "epoch": 0.6520997193751752,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2493,
+      "step": 5744
+    },
+    {
+      "epoch": 0.6522132464850943,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2502,
+      "step": 5745
+    },
+    {
+      "epoch": 0.6523267735950133,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2675,
+      "step": 5746
+    },
+    {
+      "epoch": 0.6524403007049324,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.236,
+      "step": 5747
+    },
+    {
+      "epoch": 0.6525538278148515,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2457,
+      "step": 5748
+    },
+    {
+      "epoch": 0.6526673549247706,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2436,
+      "step": 5749
+    },
+    {
+      "epoch": 0.6527808820346896,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2464,
+      "step": 5750
+    },
+    {
+      "epoch": 0.6528944091446087,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2498,
+      "step": 5751
+    },
+    {
+      "epoch": 0.6530079362545278,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2416,
+      "step": 5752
+    },
+    {
+      "epoch": 0.6531214633644469,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2426,
+      "step": 5753
+    },
+    {
+      "epoch": 0.653234990474366,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2726,
+      "step": 5754
+    },
+    {
+      "epoch": 0.653348517584285,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.269,
+      "step": 5755
+    },
+    {
+      "epoch": 0.6534620446942041,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.248,
+      "step": 5756
+    },
+    {
+      "epoch": 0.6535755718041232,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2545,
+      "step": 5757
+    },
+    {
+      "epoch": 0.6536890989140423,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2515,
+      "step": 5758
+    },
+    {
+      "epoch": 0.6538026260239613,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2563,
+      "step": 5759
+    },
+    {
+      "epoch": 0.6539161531338804,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2698,
+      "step": 5760
+    },
+    {
+      "epoch": 0.6540296802437995,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.2411,
+      "step": 5761
+    },
+    {
+      "epoch": 0.6541432073537186,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2692,
+      "step": 5762
+    },
+    {
+      "epoch": 0.6542567344636376,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2513,
+      "step": 5763
+    },
+    {
+      "epoch": 0.6543702615735567,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2468,
+      "step": 5764
+    },
+    {
+      "epoch": 0.6544837886834758,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2363,
+      "step": 5765
+    },
+    {
+      "epoch": 0.6545973157933949,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.254,
+      "step": 5766
+    },
+    {
+      "epoch": 0.6547108429033139,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2767,
+      "step": 5767
+    },
+    {
+      "epoch": 0.654824370013233,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2168,
+      "step": 5768
+    },
+    {
+      "epoch": 0.6549378971231521,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2608,
+      "step": 5769
+    },
+    {
+      "epoch": 0.6550514242330712,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2596,
+      "step": 5770
+    },
+    {
+      "epoch": 0.6551649513429902,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2526,
+      "step": 5771
+    },
+    {
+      "epoch": 0.6552784784529093,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2665,
+      "step": 5772
+    },
+    {
+      "epoch": 0.6553920055628284,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2638,
+      "step": 5773
+    },
+    {
+      "epoch": 0.6555055326727475,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2483,
+      "step": 5774
+    },
+    {
+      "epoch": 0.6556190597826665,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2619,
+      "step": 5775
+    },
+    {
+      "epoch": 0.6557325868925856,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2591,
+      "step": 5776
+    },
+    {
+      "epoch": 0.6558461140025047,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2255,
+      "step": 5777
+    },
+    {
+      "epoch": 0.6559596411124238,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2452,
+      "step": 5778
+    },
+    {
+      "epoch": 0.6560731682223429,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2466,
+      "step": 5779
+    },
+    {
+      "epoch": 0.6561866953322619,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2522,
+      "step": 5780
+    },
+    {
+      "epoch": 0.656300222442181,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2565,
+      "step": 5781
+    },
+    {
+      "epoch": 0.6564137495521001,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2356,
+      "step": 5782
+    },
+    {
+      "epoch": 0.6565272766620192,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2484,
+      "step": 5783
+    },
+    {
+      "epoch": 0.6566408037719382,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2446,
+      "step": 5784
+    },
+    {
+      "epoch": 0.6567543308818573,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2522,
+      "step": 5785
+    },
+    {
+      "epoch": 0.6568678579917764,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2751,
+      "step": 5786
+    },
+    {
+      "epoch": 0.6569813851016955,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2499,
+      "step": 5787
+    },
+    {
+      "epoch": 0.6570949122116145,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.257,
+      "step": 5788
+    },
+    {
+      "epoch": 0.6572084393215336,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.275,
+      "step": 5789
+    },
+    {
+      "epoch": 0.6573219664314527,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2461,
+      "step": 5790
+    },
+    {
+      "epoch": 0.6574354935413718,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2315,
+      "step": 5791
+    },
+    {
+      "epoch": 0.6575490206512908,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2415,
+      "step": 5792
+    },
+    {
+      "epoch": 0.6576625477612099,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2316,
+      "step": 5793
+    },
+    {
+      "epoch": 0.657776074871129,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2361,
+      "step": 5794
+    },
+    {
+      "epoch": 0.6578896019810481,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2401,
+      "step": 5795
+    },
+    {
+      "epoch": 0.6580031290909671,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2528,
+      "step": 5796
+    },
+    {
+      "epoch": 0.6581166562008862,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2531,
+      "step": 5797
+    },
+    {
+      "epoch": 0.6582301833108053,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2496,
+      "step": 5798
+    },
+    {
+      "epoch": 0.6583437104207244,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2559,
+      "step": 5799
+    },
+    {
+      "epoch": 0.6584572375306434,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2464,
+      "step": 5800
+    },
+    {
+      "epoch": 0.6585707646405625,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2596,
+      "step": 5801
+    },
+    {
+      "epoch": 0.6586842917504816,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.285,
+      "step": 5802
+    },
+    {
+      "epoch": 0.6587978188604007,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.245,
+      "step": 5803
+    },
+    {
+      "epoch": 0.6589113459703198,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2498,
+      "step": 5804
+    },
+    {
+      "epoch": 0.6590248730802388,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.2349,
+      "step": 5805
+    },
+    {
+      "epoch": 0.6591384001901579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.2382,
+      "step": 5806
+    },
+    {
+      "epoch": 0.659251927300077,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.2642,
+      "step": 5807
+    },
+    {
+      "epoch": 0.659365454409996,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.2397,
+      "step": 5808
+    },
+    {
+      "epoch": 0.6594789815199151,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.24,
+      "step": 5809
+    },
+    {
+      "epoch": 0.6595925086298342,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2341,
+      "step": 5810
+    },
+    {
+      "epoch": 0.6597060357397533,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2494,
+      "step": 5811
+    },
+    {
+      "epoch": 0.6598195628496724,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2617,
+      "step": 5812
+    },
+    {
+      "epoch": 0.6599330899595914,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2566,
+      "step": 5813
+    },
+    {
+      "epoch": 0.6600466170695105,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2492,
+      "step": 5814
+    },
+    {
+      "epoch": 0.6601601441794296,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2523,
+      "step": 5815
+    },
+    {
+      "epoch": 0.6602736712893487,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2357,
+      "step": 5816
+    },
+    {
+      "epoch": 0.6603871983992677,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.234,
+      "step": 5817
+    },
+    {
+      "epoch": 0.6605007255091868,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2466,
+      "step": 5818
+    },
+    {
+      "epoch": 0.6606142526191059,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2188,
+      "step": 5819
+    },
+    {
+      "epoch": 0.660727779729025,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2533,
+      "step": 5820
+    },
+    {
+      "epoch": 0.660841306838944,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2456,
+      "step": 5821
+    },
+    {
+      "epoch": 0.6609548339488631,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2541,
+      "step": 5822
+    },
+    {
+      "epoch": 0.6610683610587822,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.258,
+      "step": 5823
+    },
+    {
+      "epoch": 0.6611818881687013,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2659,
+      "step": 5824
+    },
+    {
+      "epoch": 0.6612954152786203,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2447,
+      "step": 5825
+    },
+    {
+      "epoch": 0.6614089423885394,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2445,
+      "step": 5826
+    },
+    {
+      "epoch": 0.6615224694984585,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2328,
+      "step": 5827
+    },
+    {
+      "epoch": 0.6616359966083776,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2688,
+      "step": 5828
+    },
+    {
+      "epoch": 0.6617495237182967,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2465,
+      "step": 5829
+    },
+    {
+      "epoch": 0.6618630508282157,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2442,
+      "step": 5830
+    },
+    {
+      "epoch": 0.6619765779381348,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2527,
+      "step": 5831
+    },
+    {
+      "epoch": 0.6620901050480539,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2277,
+      "step": 5832
+    },
+    {
+      "epoch": 0.662203632157973,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2274,
+      "step": 5833
+    },
+    {
+      "epoch": 0.662317159267892,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2492,
+      "step": 5834
+    },
+    {
+      "epoch": 0.6624306863778111,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2514,
+      "step": 5835
+    },
+    {
+      "epoch": 0.6625442134877302,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2623,
+      "step": 5836
+    },
+    {
+      "epoch": 0.6626577405976493,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2227,
+      "step": 5837
+    },
+    {
+      "epoch": 0.6627712677075683,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2543,
+      "step": 5838
+    },
+    {
+      "epoch": 0.6628847948174874,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2362,
+      "step": 5839
+    },
+    {
+      "epoch": 0.6629983219274065,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2457,
+      "step": 5840
+    },
+    {
+      "epoch": 0.6631118490373256,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2402,
+      "step": 5841
+    },
+    {
+      "epoch": 0.6632253761472446,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2534,
+      "step": 5842
+    },
+    {
+      "epoch": 0.6633389032571637,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2417,
+      "step": 5843
+    },
+    {
+      "epoch": 0.6634524303670828,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.261,
+      "step": 5844
+    },
+    {
+      "epoch": 0.6635659574770019,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2506,
+      "step": 5845
+    },
+    {
+      "epoch": 0.6636794845869209,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2303,
+      "step": 5846
+    },
+    {
+      "epoch": 0.66379301169684,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2469,
+      "step": 5847
+    },
+    {
+      "epoch": 0.6639065388067591,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2291,
+      "step": 5848
+    },
+    {
+      "epoch": 0.6640200659166782,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2097,
+      "step": 5849
+    },
+    {
+      "epoch": 0.6641335930265972,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2439,
+      "step": 5850
+    },
+    {
+      "epoch": 0.6642471201365163,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2527,
+      "step": 5851
+    },
+    {
+      "epoch": 0.6643606472464354,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2313,
+      "step": 5852
+    },
+    {
+      "epoch": 0.6644741743563545,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2552,
+      "step": 5853
+    },
+    {
+      "epoch": 0.6645877014662736,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.261,
+      "step": 5854
+    },
+    {
+      "epoch": 0.6647012285761926,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.002,
+      "loss": 5.2332,
+      "step": 5855
+    },
+    {
+      "epoch": 0.6648147556861117,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.2473,
+      "step": 5856
+    },
+    {
+      "epoch": 0.6649282827960308,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.2228,
+      "step": 5857
+    },
+    {
+      "epoch": 0.6650418099059499,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.2657,
+      "step": 5858
+    },
+    {
+      "epoch": 0.6651553370158689,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2418,
+      "step": 5859
+    },
+    {
+      "epoch": 0.665268864125788,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2578,
+      "step": 5860
+    },
+    {
+      "epoch": 0.6653823912357071,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2593,
+      "step": 5861
+    },
+    {
+      "epoch": 0.6654959183456262,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2296,
+      "step": 5862
+    },
+    {
+      "epoch": 0.6656094454555452,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2287,
+      "step": 5863
+    },
+    {
+      "epoch": 0.6657229725654643,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2352,
+      "step": 5864
+    },
+    {
+      "epoch": 0.6658364996753834,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2523,
+      "step": 5865
+    },
+    {
+      "epoch": 0.6659500267853025,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2488,
+      "step": 5866
+    },
+    {
+      "epoch": 0.6660635538952215,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2259,
+      "step": 5867
+    },
+    {
+      "epoch": 0.6661770810051406,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2569,
+      "step": 5868
+    },
+    {
+      "epoch": 0.6662906081150597,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2565,
+      "step": 5869
+    },
+    {
+      "epoch": 0.6664041352249788,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2087,
+      "step": 5870
+    },
+    {
+      "epoch": 0.6665176623348978,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2486,
+      "step": 5871
+    },
+    {
+      "epoch": 0.6666311894448169,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2596,
+      "step": 5872
+    },
+    {
+      "epoch": 0.666744716554736,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2466,
+      "step": 5873
+    },
+    {
+      "epoch": 0.6668582436646551,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2561,
+      "step": 5874
+    },
+    {
+      "epoch": 0.6669717707745741,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2672,
+      "step": 5875
+    },
+    {
+      "epoch": 0.6670852978844932,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.002,
+      "loss": 5.2672,
+      "step": 5876
+    },
+    {
+      "epoch": 0.6671988249944123,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.233,
+      "step": 5877
+    },
+    {
+      "epoch": 0.6673123521043314,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2522,
+      "step": 5878
+    },
+    {
+      "epoch": 0.6674258792142505,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2337,
+      "step": 5879
+    },
+    {
+      "epoch": 0.6675394063241695,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.237,
+      "step": 5880
+    },
+    {
+      "epoch": 0.6676529334340886,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2511,
+      "step": 5881
+    },
+    {
+      "epoch": 0.6677664605440077,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2521,
+      "step": 5882
+    },
+    {
+      "epoch": 0.6678799876539268,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2497,
+      "step": 5883
+    },
+    {
+      "epoch": 0.6679935147638458,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2291,
+      "step": 5884
+    },
+    {
+      "epoch": 0.6681070418737649,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2697,
+      "step": 5885
+    },
+    {
+      "epoch": 0.668220568983684,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2636,
+      "step": 5886
+    },
+    {
+      "epoch": 0.6683340960936031,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2371,
+      "step": 5887
+    },
+    {
+      "epoch": 0.6684476232035221,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2477,
+      "step": 5888
+    },
+    {
+      "epoch": 0.6685611503134412,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2298,
+      "step": 5889
+    },
+    {
+      "epoch": 0.6686746774233603,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.248,
+      "step": 5890
+    },
+    {
+      "epoch": 0.6687882045332794,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2366,
+      "step": 5891
+    },
+    {
+      "epoch": 0.6689017316431984,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2533,
+      "step": 5892
+    },
+    {
+      "epoch": 0.6690152587531175,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2462,
+      "step": 5893
+    },
+    {
+      "epoch": 0.6691287858630366,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.234,
+      "step": 5894
+    },
+    {
+      "epoch": 0.6692423129729557,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2532,
+      "step": 5895
+    },
+    {
+      "epoch": 0.6693558400828747,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2467,
+      "step": 5896
+    },
+    {
+      "epoch": 0.6694693671927938,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.2415,
+      "step": 5897
+    },
+    {
+      "epoch": 0.6695828943027129,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.235,
+      "step": 5898
+    },
+    {
+      "epoch": 0.669696421412632,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2393,
+      "step": 5899
+    },
+    {
+      "epoch": 0.669809948522551,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2381,
+      "step": 5900
+    },
+    {
+      "epoch": 0.6699234756324701,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2339,
+      "step": 5901
+    },
+    {
+      "epoch": 0.6700370027423892,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2339,
+      "step": 5902
+    },
+    {
+      "epoch": 0.6701505298523083,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2733,
+      "step": 5903
+    },
+    {
+      "epoch": 0.6702640569622274,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2526,
+      "step": 5904
+    },
+    {
+      "epoch": 0.6703775840721464,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2446,
+      "step": 5905
+    },
+    {
+      "epoch": 0.6704911111820655,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2448,
+      "step": 5906
+    },
+    {
+      "epoch": 0.6706046382919846,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2374,
+      "step": 5907
+    },
+    {
+      "epoch": 0.6707181654019037,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2521,
+      "step": 5908
+    },
+    {
+      "epoch": 0.6708316925118227,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2458,
+      "step": 5909
+    },
+    {
+      "epoch": 0.6709452196217418,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2434,
+      "step": 5910
+    },
+    {
+      "epoch": 0.6710587467316609,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2533,
+      "step": 5911
+    },
+    {
+      "epoch": 0.67117227384158,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2413,
+      "step": 5912
+    },
+    {
+      "epoch": 0.6712858009514991,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2272,
+      "step": 5913
+    },
+    {
+      "epoch": 0.6713993280614182,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.239,
+      "step": 5914
+    },
+    {
+      "epoch": 0.6715128551713373,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2312,
+      "step": 5915
+    },
+    {
+      "epoch": 0.6716263822812564,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.2384,
+      "step": 5916
+    },
+    {
+      "epoch": 0.6717399093911754,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2516,
+      "step": 5917
+    },
+    {
+      "epoch": 0.6718534365010945,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.002,
+      "loss": 5.2672,
+      "step": 5918
+    },
+    {
+      "epoch": 0.6719669636110136,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.002,
+      "loss": 5.2504,
+      "step": 5919
+    },
+    {
+      "epoch": 0.6720804907209327,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2494,
+      "step": 5920
+    },
+    {
+      "epoch": 0.6721940178308518,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002,
+      "loss": 5.2597,
+      "step": 5921
+    },
+    {
+      "epoch": 0.6723075449407708,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2467,
+      "step": 5922
+    },
+    {
+      "epoch": 0.6724210720506899,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.2595,
+      "step": 5923
+    },
+    {
+      "epoch": 0.672534599160609,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2413,
+      "step": 5924
+    },
+    {
+      "epoch": 0.6726481262705281,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2274,
+      "step": 5925
+    },
+    {
+      "epoch": 0.6727616533804471,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2361,
+      "step": 5926
+    },
+    {
+      "epoch": 0.6728751804903662,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 5.2595,
+      "step": 5927
+    },
+    {
+      "epoch": 0.6729887076002853,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.002,
+      "loss": 5.2586,
+      "step": 5928
+    },
+    {
+      "epoch": 0.6731022347102044,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2615,
+      "step": 5929
+    },
+    {
+      "epoch": 0.6732157618201234,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2372,
+      "step": 5930
+    },
+    {
+      "epoch": 0.6733292889300425,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2443,
+      "step": 5931
+    },
+    {
+      "epoch": 0.6734428160399616,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2524,
+      "step": 5932
+    },
+    {
+      "epoch": 0.6735563431498807,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2573,
+      "step": 5933
+    },
+    {
+      "epoch": 0.6736698702597997,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2465,
+      "step": 5934
+    },
+    {
+      "epoch": 0.6737833973697188,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2429,
+      "step": 5935
+    },
+    {
+      "epoch": 0.6738969244796379,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2555,
+      "step": 5936
+    },
+    {
+      "epoch": 0.674010451589557,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2504,
+      "step": 5937
+    },
+    {
+      "epoch": 0.674123978699476,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2389,
+      "step": 5938
+    },
+    {
+      "epoch": 0.6742375058093951,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.245,
+      "step": 5939
+    },
+    {
+      "epoch": 0.6743510329193142,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2555,
+      "step": 5940
+    },
+    {
+      "epoch": 0.6744645600292333,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2303,
+      "step": 5941
+    },
+    {
+      "epoch": 0.6745780871391523,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2276,
+      "step": 5942
+    },
+    {
+      "epoch": 0.6746916142490714,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2482,
+      "step": 5943
+    },
+    {
+      "epoch": 0.6748051413589905,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.2511,
+      "step": 5944
+    },
+    {
+      "epoch": 0.6749186684689096,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.002,
+      "loss": 5.2159,
+      "step": 5945
+    },
+    {
+      "epoch": 0.6750321955788287,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.002,
+      "loss": 5.2413,
+      "step": 5946
+    },
+    {
+      "epoch": 0.6751457226887477,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2412,
+      "step": 5947
+    },
+    {
+      "epoch": 0.6752592497986668,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2137,
+      "step": 5948
+    },
+    {
+      "epoch": 0.6753727769085859,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2364,
+      "step": 5949
+    },
+    {
+      "epoch": 0.675486304018505,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2222,
+      "step": 5950
+    },
+    {
+      "epoch": 0.675599831128424,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.237,
+      "step": 5951
+    },
+    {
+      "epoch": 0.6757133582383431,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2289,
+      "step": 5952
+    },
+    {
+      "epoch": 0.6758268853482622,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2329,
+      "step": 5953
+    },
+    {
+      "epoch": 0.6759404124581813,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.245,
+      "step": 5954
+    },
+    {
+      "epoch": 0.6760539395681003,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2435,
+      "step": 5955
+    },
+    {
+      "epoch": 0.6761674666780194,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2424,
+      "step": 5956
+    },
+    {
+      "epoch": 0.6762809937879385,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2476,
+      "step": 5957
+    },
+    {
+      "epoch": 0.6763945208978576,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.229,
+      "step": 5958
+    },
+    {
+      "epoch": 0.6765080480077766,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2421,
+      "step": 5959
+    },
+    {
+      "epoch": 0.6766215751176957,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2506,
+      "step": 5960
+    },
+    {
+      "epoch": 0.6767351022276148,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2437,
+      "step": 5961
+    },
+    {
+      "epoch": 0.6768486293375339,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2622,
+      "step": 5962
+    },
+    {
+      "epoch": 0.676962156447453,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2488,
+      "step": 5963
+    },
+    {
+      "epoch": 0.677075683557372,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2457,
+      "step": 5964
+    },
+    {
+      "epoch": 0.6771892106672911,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2546,
+      "step": 5965
+    },
+    {
+      "epoch": 0.6773027377772102,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2479,
+      "step": 5966
+    },
+    {
+      "epoch": 0.6774162648871292,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.237,
+      "step": 5967
+    },
+    {
+      "epoch": 0.6775297919970483,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2383,
+      "step": 5968
+    },
+    {
+      "epoch": 0.6776433191069674,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2344,
+      "step": 5969
+    },
+    {
+      "epoch": 0.6777568462168865,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2657,
+      "step": 5970
+    },
+    {
+      "epoch": 0.6778703733268056,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2456,
+      "step": 5971
+    },
+    {
+      "epoch": 0.6779839004367246,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2556,
+      "step": 5972
+    },
+    {
+      "epoch": 0.6780974275466437,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2474,
+      "step": 5973
+    },
+    {
+      "epoch": 0.6782109546565628,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2349,
+      "step": 5974
+    },
+    {
+      "epoch": 0.6783244817664819,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2473,
+      "step": 5975
+    },
+    {
+      "epoch": 0.6784380088764009,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2307,
+      "step": 5976
+    },
+    {
+      "epoch": 0.67855153598632,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2411,
+      "step": 5977
+    },
+    {
+      "epoch": 0.6786650630962391,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2479,
+      "step": 5978
+    },
+    {
+      "epoch": 0.6787785902061582,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2395,
+      "step": 5979
+    },
+    {
+      "epoch": 0.6788921173160772,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2386,
+      "step": 5980
+    },
+    {
+      "epoch": 0.6790056444259963,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2367,
+      "step": 5981
+    },
+    {
+      "epoch": 0.6791191715359154,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2321,
+      "step": 5982
+    },
+    {
+      "epoch": 0.6792326986458345,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.252,
+      "step": 5983
+    },
+    {
+      "epoch": 0.6793462257557535,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2381,
+      "step": 5984
+    },
+    {
+      "epoch": 0.6794597528656726,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2544,
+      "step": 5985
+    },
+    {
+      "epoch": 0.6795732799755917,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2316,
+      "step": 5986
+    },
+    {
+      "epoch": 0.6796868070855108,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2453,
+      "step": 5987
+    },
+    {
+      "epoch": 0.6798003341954298,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2529,
+      "step": 5988
+    },
+    {
+      "epoch": 0.6799138613053489,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2416,
+      "step": 5989
+    },
+    {
+      "epoch": 0.680027388415268,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2154,
+      "step": 5990
+    },
+    {
+      "epoch": 0.6801409155251871,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2423,
+      "step": 5991
+    },
+    {
+      "epoch": 0.6802544426351061,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2402,
+      "step": 5992
+    },
+    {
+      "epoch": 0.6803679697450252,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2468,
+      "step": 5993
+    },
+    {
+      "epoch": 0.6804814968549443,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2356,
+      "step": 5994
+    },
+    {
+      "epoch": 0.6805950239648634,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2426,
+      "step": 5995
+    },
+    {
+      "epoch": 0.6807085510747825,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2732,
+      "step": 5996
+    },
+    {
+      "epoch": 0.6808220781847015,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.2705,
+      "step": 5997
+    },
+    {
+      "epoch": 0.6809356052946206,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2548,
+      "step": 5998
+    },
+    {
+      "epoch": 0.6810491324045397,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2282,
+      "step": 5999
+    },
+    {
+      "epoch": 0.6811626595144588,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2515,
+      "step": 6000
+    },
+    {
+      "epoch": 0.6812761866243778,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2479,
+      "step": 6001
+    },
+    {
+      "epoch": 0.6813897137342969,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2339,
+      "step": 6002
+    },
+    {
+      "epoch": 0.681503240844216,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.248,
+      "step": 6003
+    },
+    {
+      "epoch": 0.6816167679541351,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2485,
+      "step": 6004
+    },
+    {
+      "epoch": 0.6817302950640541,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2493,
+      "step": 6005
+    },
+    {
+      "epoch": 0.6818438221739732,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2382,
+      "step": 6006
+    },
+    {
+      "epoch": 0.6819573492838923,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2447,
+      "step": 6007
+    },
+    {
+      "epoch": 0.6820708763938114,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2375,
+      "step": 6008
+    },
+    {
+      "epoch": 0.6821844035037304,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2495,
+      "step": 6009
+    },
+    {
+      "epoch": 0.6822979306136495,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2572,
+      "step": 6010
+    },
+    {
+      "epoch": 0.6824114577235686,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.235,
+      "step": 6011
+    },
+    {
+      "epoch": 0.6825249848334877,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2647,
+      "step": 6012
+    },
+    {
+      "epoch": 0.6826385119434067,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2426,
+      "step": 6013
+    },
+    {
+      "epoch": 0.6827520390533258,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2274,
+      "step": 6014
+    },
+    {
+      "epoch": 0.6828655661632449,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2429,
+      "step": 6015
+    },
+    {
+      "epoch": 0.682979093273164,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2224,
+      "step": 6016
+    },
+    {
+      "epoch": 0.683092620383083,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.229,
+      "step": 6017
+    },
+    {
+      "epoch": 0.6832061474930021,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2476,
+      "step": 6018
+    },
+    {
+      "epoch": 0.6833196746029212,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2536,
+      "step": 6019
+    },
+    {
+      "epoch": 0.6834332017128403,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.261,
+      "step": 6020
+    },
+    {
+      "epoch": 0.6835467288227594,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.252,
+      "step": 6021
+    },
+    {
+      "epoch": 0.6836602559326784,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 5.2217,
+      "step": 6022
+    },
+    {
+      "epoch": 0.6837737830425975,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 5.2421,
+      "step": 6023
+    },
+    {
+      "epoch": 0.6838873101525166,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 5.2475,
+      "step": 6024
+    },
+    {
+      "epoch": 0.6840008372624357,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2412,
+      "step": 6025
+    },
+    {
+      "epoch": 0.6841143643723547,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2385,
+      "step": 6026
+    },
+    {
+      "epoch": 0.6842278914822738,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2638,
+      "step": 6027
+    },
+    {
+      "epoch": 0.6843414185921929,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2516,
+      "step": 6028
+    },
+    {
+      "epoch": 0.684454945702112,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2472,
+      "step": 6029
+    },
+    {
+      "epoch": 0.684568472812031,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2436,
+      "step": 6030
+    },
+    {
+      "epoch": 0.6846819999219501,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2438,
+      "step": 6031
+    },
+    {
+      "epoch": 0.6847955270318692,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2309,
+      "step": 6032
+    },
+    {
+      "epoch": 0.6849090541417883,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2346,
+      "step": 6033
+    },
+    {
+      "epoch": 0.6850225812517073,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2245,
+      "step": 6034
+    },
+    {
+      "epoch": 0.6851361083616264,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2426,
+      "step": 6035
+    },
+    {
+      "epoch": 0.6852496354715455,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2252,
+      "step": 6036
+    },
+    {
+      "epoch": 0.6853631625814646,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2452,
+      "step": 6037
+    },
+    {
+      "epoch": 0.6854766896913836,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2472,
+      "step": 6038
+    },
+    {
+      "epoch": 0.6855902168013027,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2423,
+      "step": 6039
+    },
+    {
+      "epoch": 0.6857037439112218,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2294,
+      "step": 6040
+    },
+    {
+      "epoch": 0.6858172710211409,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.2474,
+      "step": 6041
+    },
+    {
+      "epoch": 0.68593079813106,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2509,
+      "step": 6042
+    },
+    {
+      "epoch": 0.686044325240979,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2411,
+      "step": 6043
+    },
+    {
+      "epoch": 0.6861578523508981,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2347,
+      "step": 6044
+    },
+    {
+      "epoch": 0.6862713794608172,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2391,
+      "step": 6045
+    },
+    {
+      "epoch": 0.6863849065707363,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.2634,
+      "step": 6046
+    },
+    {
+      "epoch": 0.6864984336806553,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2518,
+      "step": 6047
+    },
+    {
+      "epoch": 0.6866119607905744,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.2406,
+      "step": 6048
+    },
+    {
+      "epoch": 0.6867254879004935,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2374,
+      "step": 6049
+    },
+    {
+      "epoch": 0.6868390150104126,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2472,
+      "step": 6050
+    },
+    {
+      "epoch": 0.6869525421203316,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2462,
+      "step": 6051
+    },
+    {
+      "epoch": 0.6870660692302507,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2577,
+      "step": 6052
+    },
+    {
+      "epoch": 0.6871795963401698,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2708,
+      "step": 6053
+    },
+    {
+      "epoch": 0.6872931234500889,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.223,
+      "step": 6054
+    },
+    {
+      "epoch": 0.6874066505600079,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2549,
+      "step": 6055
+    },
+    {
+      "epoch": 0.687520177669927,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2377,
+      "step": 6056
+    },
+    {
+      "epoch": 0.6876337047798461,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2208,
+      "step": 6057
+    },
+    {
+      "epoch": 0.6877472318897652,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2376,
+      "step": 6058
+    },
+    {
+      "epoch": 0.6878607589996842,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2575,
+      "step": 6059
+    },
+    {
+      "epoch": 0.6879742861096033,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2225,
+      "step": 6060
+    },
+    {
+      "epoch": 0.6880878132195224,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2431,
+      "step": 6061
+    },
+    {
+      "epoch": 0.6882013403294415,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2485,
+      "step": 6062
+    },
+    {
+      "epoch": 0.6883148674393605,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2608,
+      "step": 6063
+    },
+    {
+      "epoch": 0.6884283945492796,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2388,
+      "step": 6064
+    },
+    {
+      "epoch": 0.6885419216591987,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2441,
+      "step": 6065
+    },
+    {
+      "epoch": 0.6886554487691178,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.242,
+      "step": 6066
+    },
+    {
+      "epoch": 0.6887689758790368,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2494,
+      "step": 6067
+    },
+    {
+      "epoch": 0.6888825029889559,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2346,
+      "step": 6068
+    },
+    {
+      "epoch": 0.688996030098875,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2521,
+      "step": 6069
+    },
+    {
+      "epoch": 0.6891095572087941,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2533,
+      "step": 6070
+    },
+    {
+      "epoch": 0.6892230843187132,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2626,
+      "step": 6071
+    },
+    {
+      "epoch": 0.6893366114286322,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2497,
+      "step": 6072
+    },
+    {
+      "epoch": 0.6894501385385513,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2533,
+      "step": 6073
+    },
+    {
+      "epoch": 0.6895636656484704,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2295,
+      "step": 6074
+    },
+    {
+      "epoch": 0.6896771927583895,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2321,
+      "step": 6075
+    },
+    {
+      "epoch": 0.6897907198683085,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2544,
+      "step": 6076
+    },
+    {
+      "epoch": 0.6899042469782276,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2312,
+      "step": 6077
+    },
+    {
+      "epoch": 0.6900177740881467,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2409,
+      "step": 6078
+    },
+    {
+      "epoch": 0.6901313011980658,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2494,
+      "step": 6079
+    },
+    {
+      "epoch": 0.6902448283079848,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2417,
+      "step": 6080
+    },
+    {
+      "epoch": 0.6903583554179039,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2411,
+      "step": 6081
+    },
+    {
+      "epoch": 0.690471882527823,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2486,
+      "step": 6082
+    },
+    {
+      "epoch": 0.6905854096377421,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2362,
+      "step": 6083
+    },
+    {
+      "epoch": 0.6906989367476611,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2228,
+      "step": 6084
+    },
+    {
+      "epoch": 0.6908124638575802,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2461,
+      "step": 6085
+    },
+    {
+      "epoch": 0.6909259909674993,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2488,
+      "step": 6086
+    },
+    {
+      "epoch": 0.6910395180774184,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2303,
+      "step": 6087
+    },
+    {
+      "epoch": 0.6911530451873374,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.25,
+      "step": 6088
+    },
+    {
+      "epoch": 0.6912665722972565,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2485,
+      "step": 6089
+    },
+    {
+      "epoch": 0.6913800994071756,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2409,
+      "step": 6090
+    },
+    {
+      "epoch": 0.6914936265170947,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2462,
+      "step": 6091
+    },
+    {
+      "epoch": 0.6916071536270137,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.242,
+      "step": 6092
+    },
+    {
+      "epoch": 0.6917206807369328,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.245,
+      "step": 6093
+    },
+    {
+      "epoch": 0.6918342078468519,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2544,
+      "step": 6094
+    },
+    {
+      "epoch": 0.691947734956771,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2309,
+      "step": 6095
+    },
+    {
+      "epoch": 0.69206126206669,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2487,
+      "step": 6096
+    },
+    {
+      "epoch": 0.6921747891766091,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.002,
+      "loss": 5.2367,
+      "step": 6097
+    },
+    {
+      "epoch": 0.6922883162865282,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.002,
+      "loss": 5.2588,
+      "step": 6098
+    },
+    {
+      "epoch": 0.6924018433964473,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.2361,
+      "step": 6099
+    },
+    {
+      "epoch": 0.6925153705063664,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.002,
+      "loss": 5.2485,
+      "step": 6100
+    },
+    {
+      "epoch": 0.6926288976162854,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.2512,
+      "step": 6101
+    },
+    {
+      "epoch": 0.6927424247262045,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.2459,
+      "step": 6102
+    },
+    {
+      "epoch": 0.6928559518361236,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2693,
+      "step": 6103
+    },
+    {
+      "epoch": 0.6929694789460427,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2527,
+      "step": 6104
+    },
+    {
+      "epoch": 0.6930830060559617,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2467,
+      "step": 6105
+    },
+    {
+      "epoch": 0.6931965331658808,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2445,
+      "step": 6106
+    },
+    {
+      "epoch": 0.6933100602757999,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2548,
+      "step": 6107
+    },
+    {
+      "epoch": 0.693423587385719,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2422,
+      "step": 6108
+    },
+    {
+      "epoch": 0.693537114495638,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2399,
+      "step": 6109
+    },
+    {
+      "epoch": 0.6936506416055571,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2513,
+      "step": 6110
+    },
+    {
+      "epoch": 0.6937641687154762,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2447,
+      "step": 6111
+    },
+    {
+      "epoch": 0.6938776958253953,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2426,
+      "step": 6112
+    },
+    {
+      "epoch": 0.6939912229353143,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2534,
+      "step": 6113
+    },
+    {
+      "epoch": 0.6941047500452334,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2575,
+      "step": 6114
+    },
+    {
+      "epoch": 0.6942182771551525,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2382,
+      "step": 6115
+    },
+    {
+      "epoch": 0.6943318042650716,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2273,
+      "step": 6116
+    },
+    {
+      "epoch": 0.6944453313749906,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2584,
+      "step": 6117
+    },
+    {
+      "epoch": 0.6945588584849097,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2405,
+      "step": 6118
+    },
+    {
+      "epoch": 0.6946723855948288,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2277,
+      "step": 6119
+    },
+    {
+      "epoch": 0.6947859127047479,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2442,
+      "step": 6120
+    },
+    {
+      "epoch": 0.694899439814667,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2261,
+      "step": 6121
+    },
+    {
+      "epoch": 0.695012966924586,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2474,
+      "step": 6122
+    },
+    {
+      "epoch": 0.6951264940345051,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2591,
+      "step": 6123
+    },
+    {
+      "epoch": 0.6952400211444242,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2372,
+      "step": 6124
+    },
+    {
+      "epoch": 0.6953535482543433,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2371,
+      "step": 6125
+    },
+    {
+      "epoch": 0.6954670753642623,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2758,
+      "step": 6126
+    },
+    {
+      "epoch": 0.6955806024741814,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2541,
+      "step": 6127
+    },
+    {
+      "epoch": 0.6956941295841005,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2535,
+      "step": 6128
+    },
+    {
+      "epoch": 0.6958076566940196,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2459,
+      "step": 6129
+    },
+    {
+      "epoch": 0.6959211838039386,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2402,
+      "step": 6130
+    },
+    {
+      "epoch": 0.6960347109138577,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2125,
+      "step": 6131
+    },
+    {
+      "epoch": 0.6961482380237768,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2431,
+      "step": 6132
+    },
+    {
+      "epoch": 0.6962617651336959,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2259,
+      "step": 6133
+    },
+    {
+      "epoch": 0.6963752922436149,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2648,
+      "step": 6134
+    },
+    {
+      "epoch": 0.696488819353534,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.249,
+      "step": 6135
+    },
+    {
+      "epoch": 0.6966023464634531,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2362,
+      "step": 6136
+    },
+    {
+      "epoch": 0.6967158735733722,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2544,
+      "step": 6137
+    },
+    {
+      "epoch": 0.6968294006832912,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.252,
+      "step": 6138
+    },
+    {
+      "epoch": 0.6969429277932103,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2261,
+      "step": 6139
+    },
+    {
+      "epoch": 0.6970564549031294,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2624,
+      "step": 6140
+    },
+    {
+      "epoch": 0.6971699820130485,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2722,
+      "step": 6141
+    },
+    {
+      "epoch": 0.6972835091229675,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2489,
+      "step": 6142
+    },
+    {
+      "epoch": 0.6973970362328866,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2629,
+      "step": 6143
+    },
+    {
+      "epoch": 0.6975105633428057,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2544,
+      "step": 6144
+    },
+    {
+      "epoch": 0.6976240904527248,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2501,
+      "step": 6145
+    },
+    {
+      "epoch": 0.6977376175626439,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2579,
+      "step": 6146
+    },
+    {
+      "epoch": 0.6978511446725629,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2549,
+      "step": 6147
+    },
+    {
+      "epoch": 0.697964671782482,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2363,
+      "step": 6148
+    },
+    {
+      "epoch": 0.6980781988924011,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2774,
+      "step": 6149
+    },
+    {
+      "epoch": 0.6981917260023202,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2517,
+      "step": 6150
+    },
+    {
+      "epoch": 0.6983052531122392,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2218,
+      "step": 6151
+    },
+    {
+      "epoch": 0.6984187802221583,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2258,
+      "step": 6152
+    },
+    {
+      "epoch": 0.6985323073320774,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.2366,
+      "step": 6153
+    },
+    {
+      "epoch": 0.6986458344419966,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.002,
+      "loss": 5.2476,
+      "step": 6154
+    },
+    {
+      "epoch": 0.6987593615519156,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.002,
+      "loss": 5.2622,
+      "step": 6155
+    },
+    {
+      "epoch": 0.6988728886618347,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.2498,
+      "step": 6156
+    },
+    {
+      "epoch": 0.6989864157717538,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.2578,
+      "step": 6157
+    },
+    {
+      "epoch": 0.6990999428816729,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.2555,
+      "step": 6158
+    },
+    {
+      "epoch": 0.699213469991592,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2556,
+      "step": 6159
+    },
+    {
+      "epoch": 0.699326997101511,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2488,
+      "step": 6160
+    },
+    {
+      "epoch": 0.6994405242114301,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2494,
+      "step": 6161
+    },
+    {
+      "epoch": 0.6995540513213492,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2434,
+      "step": 6162
+    },
+    {
+      "epoch": 0.6996675784312683,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2354,
+      "step": 6163
+    },
+    {
+      "epoch": 0.6997811055411873,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2331,
+      "step": 6164
+    },
+    {
+      "epoch": 0.6998946326511064,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2576,
+      "step": 6165
+    },
+    {
+      "epoch": 0.7000081597610255,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2407,
+      "step": 6166
+    },
+    {
+      "epoch": 0.7001216868709446,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2423,
+      "step": 6167
+    },
+    {
+      "epoch": 0.7002352139808636,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2476,
+      "step": 6168
+    },
+    {
+      "epoch": 0.7003487410907827,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2273,
+      "step": 6169
+    },
+    {
+      "epoch": 0.7004622682007018,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2359,
+      "step": 6170
+    },
+    {
+      "epoch": 0.7005757953106209,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2311,
+      "step": 6171
+    },
+    {
+      "epoch": 0.7006893224205399,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2383,
+      "step": 6172
+    },
+    {
+      "epoch": 0.700802849530459,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2472,
+      "step": 6173
+    },
+    {
+      "epoch": 0.7009163766403781,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2446,
+      "step": 6174
+    },
+    {
+      "epoch": 0.7010299037502972,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2264,
+      "step": 6175
+    },
+    {
+      "epoch": 0.7011434308602162,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2312,
+      "step": 6176
+    },
+    {
+      "epoch": 0.7012569579701353,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2238,
+      "step": 6177
+    },
+    {
+      "epoch": 0.7013704850800544,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2237,
+      "step": 6178
+    },
+    {
+      "epoch": 0.7014840121899735,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2507,
+      "step": 6179
+    },
+    {
+      "epoch": 0.7015975392998925,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2243,
+      "step": 6180
+    },
+    {
+      "epoch": 0.7017110664098116,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.2432,
+      "step": 6181
+    },
+    {
+      "epoch": 0.7018245935197307,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2354,
+      "step": 6182
+    },
+    {
+      "epoch": 0.7019381206296498,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.249,
+      "step": 6183
+    },
+    {
+      "epoch": 0.7020516477395689,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2265,
+      "step": 6184
+    },
+    {
+      "epoch": 0.7021651748494879,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2604,
+      "step": 6185
+    },
+    {
+      "epoch": 0.702278701959407,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2491,
+      "step": 6186
+    },
+    {
+      "epoch": 0.7023922290693261,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2426,
+      "step": 6187
+    },
+    {
+      "epoch": 0.7025057561792452,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2493,
+      "step": 6188
+    },
+    {
+      "epoch": 0.7026192832891642,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.229,
+      "step": 6189
+    },
+    {
+      "epoch": 0.7027328103990833,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2454,
+      "step": 6190
+    },
+    {
+      "epoch": 0.7028463375090024,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2199,
+      "step": 6191
+    },
+    {
+      "epoch": 0.7029598646189215,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2349,
+      "step": 6192
+    },
+    {
+      "epoch": 0.7030733917288405,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2488,
+      "step": 6193
+    },
+    {
+      "epoch": 0.7031869188387596,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2423,
+      "step": 6194
+    },
+    {
+      "epoch": 0.7033004459486787,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2295,
+      "step": 6195
+    },
+    {
+      "epoch": 0.7034139730585978,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2474,
+      "step": 6196
+    },
+    {
+      "epoch": 0.7035275001685168,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2513,
+      "step": 6197
+    },
+    {
+      "epoch": 0.7036410272784359,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2557,
+      "step": 6198
+    },
+    {
+      "epoch": 0.703754554388355,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2479,
+      "step": 6199
+    },
+    {
+      "epoch": 0.7038680814982741,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2192,
+      "step": 6200
+    },
+    {
+      "epoch": 0.7039816086081931,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2396,
+      "step": 6201
+    },
+    {
+      "epoch": 0.7040951357181122,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2607,
+      "step": 6202
+    },
+    {
+      "epoch": 0.7042086628280313,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2301,
+      "step": 6203
+    },
+    {
+      "epoch": 0.7043221899379504,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.2418,
+      "step": 6204
+    },
+    {
+      "epoch": 0.7044357170478694,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2536,
+      "step": 6205
+    },
+    {
+      "epoch": 0.7045492441577885,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.2456,
+      "step": 6206
+    },
+    {
+      "epoch": 0.7046627712677076,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.002,
+      "loss": 5.2339,
+      "step": 6207
+    },
+    {
+      "epoch": 0.7047762983776267,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.002,
+      "loss": 5.2302,
+      "step": 6208
+    },
+    {
+      "epoch": 0.7048898254875458,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.002,
+      "loss": 5.2312,
+      "step": 6209
+    },
+    {
+      "epoch": 0.7050033525974648,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.2403,
+      "step": 6210
+    },
+    {
+      "epoch": 0.7051168797073839,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2503,
+      "step": 6211
+    },
+    {
+      "epoch": 0.705230406817303,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.234,
+      "step": 6212
+    },
+    {
+      "epoch": 0.705343933927222,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2318,
+      "step": 6213
+    },
+    {
+      "epoch": 0.7054574610371411,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2595,
+      "step": 6214
+    },
+    {
+      "epoch": 0.7055709881470602,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.239,
+      "step": 6215
+    },
+    {
+      "epoch": 0.7056845152569793,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2301,
+      "step": 6216
+    },
+    {
+      "epoch": 0.7057980423668984,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2209,
+      "step": 6217
+    },
+    {
+      "epoch": 0.7059115694768174,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2533,
+      "step": 6218
+    },
+    {
+      "epoch": 0.7060250965867365,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2491,
+      "step": 6219
+    },
+    {
+      "epoch": 0.7061386236966556,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2531,
+      "step": 6220
+    },
+    {
+      "epoch": 0.7062521508065747,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2298,
+      "step": 6221
+    },
+    {
+      "epoch": 0.7063656779164937,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2367,
+      "step": 6222
+    },
+    {
+      "epoch": 0.7064792050264128,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2367,
+      "step": 6223
+    },
+    {
+      "epoch": 0.7065927321363319,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2328,
+      "step": 6224
+    },
+    {
+      "epoch": 0.706706259246251,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2607,
+      "step": 6225
+    },
+    {
+      "epoch": 0.70681978635617,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2371,
+      "step": 6226
+    },
+    {
+      "epoch": 0.7069333134660891,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2399,
+      "step": 6227
+    },
+    {
+      "epoch": 0.7070468405760082,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2497,
+      "step": 6228
+    },
+    {
+      "epoch": 0.7071603676859273,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2283,
+      "step": 6229
+    },
+    {
+      "epoch": 0.7072738947958463,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2552,
+      "step": 6230
+    },
+    {
+      "epoch": 0.7073874219057654,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2496,
+      "step": 6231
+    },
+    {
+      "epoch": 0.7075009490156845,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2524,
+      "step": 6232
+    },
+    {
+      "epoch": 0.7076144761256036,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2322,
+      "step": 6233
+    },
+    {
+      "epoch": 0.7077280032355227,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2416,
+      "step": 6234
+    },
+    {
+      "epoch": 0.7078415303454417,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2494,
+      "step": 6235
+    },
+    {
+      "epoch": 0.7079550574553608,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2507,
+      "step": 6236
+    },
+    {
+      "epoch": 0.7080685845652799,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2275,
+      "step": 6237
+    },
+    {
+      "epoch": 0.708182111675199,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2449,
+      "step": 6238
+    },
+    {
+      "epoch": 0.708295638785118,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2474,
+      "step": 6239
+    },
+    {
+      "epoch": 0.7084091658950371,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2493,
+      "step": 6240
+    },
+    {
+      "epoch": 0.7085226930049562,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2378,
+      "step": 6241
+    },
+    {
+      "epoch": 0.7086362201148753,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2356,
+      "step": 6242
+    },
+    {
+      "epoch": 0.7087497472247943,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2261,
+      "step": 6243
+    },
+    {
+      "epoch": 0.7088632743347134,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2489,
+      "step": 6244
+    },
+    {
+      "epoch": 0.7089768014446325,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2545,
+      "step": 6245
+    },
+    {
+      "epoch": 0.7090903285545516,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2424,
+      "step": 6246
+    },
+    {
+      "epoch": 0.7092038556644706,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2223,
+      "step": 6247
+    },
+    {
+      "epoch": 0.7093173827743897,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2622,
+      "step": 6248
+    },
+    {
+      "epoch": 0.7094309098843088,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2345,
+      "step": 6249
+    },
+    {
+      "epoch": 0.7095444369942279,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2589,
+      "step": 6250
+    },
+    {
+      "epoch": 0.7096579641041469,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2574,
+      "step": 6251
+    },
+    {
+      "epoch": 0.709771491214066,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2437,
+      "step": 6252
+    },
+    {
+      "epoch": 0.7098850183239851,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.231,
+      "step": 6253
+    },
+    {
+      "epoch": 0.7099985454339042,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2562,
+      "step": 6254
+    },
+    {
+      "epoch": 0.7101120725438232,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2438,
+      "step": 6255
+    },
+    {
+      "epoch": 0.7102255996537423,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2579,
+      "step": 6256
+    },
+    {
+      "epoch": 0.7103391267636614,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2265,
+      "step": 6257
+    },
+    {
+      "epoch": 0.7104526538735805,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2682,
+      "step": 6258
+    },
+    {
+      "epoch": 0.7105661809834996,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2451,
+      "step": 6259
+    },
+    {
+      "epoch": 0.7106797080934186,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2502,
+      "step": 6260
+    },
+    {
+      "epoch": 0.7107932352033377,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2373,
+      "step": 6261
+    },
+    {
+      "epoch": 0.7109067623132568,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.2753,
+      "step": 6262
+    },
+    {
+      "epoch": 0.7110202894231759,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.217,
+      "step": 6263
+    },
+    {
+      "epoch": 0.7111338165330949,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.2446,
+      "step": 6264
+    },
+    {
+      "epoch": 0.711247343643014,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.2324,
+      "step": 6265
+    },
+    {
+      "epoch": 0.7113608707529331,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2599,
+      "step": 6266
+    },
+    {
+      "epoch": 0.7114743978628522,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.217,
+      "step": 6267
+    },
+    {
+      "epoch": 0.7115879249727712,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2671,
+      "step": 6268
+    },
+    {
+      "epoch": 0.7117014520826903,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2372,
+      "step": 6269
+    },
+    {
+      "epoch": 0.7118149791926094,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.244,
+      "step": 6270
+    },
+    {
+      "epoch": 0.7119285063025285,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2523,
+      "step": 6271
+    },
+    {
+      "epoch": 0.7120420334124475,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2354,
+      "step": 6272
+    },
+    {
+      "epoch": 0.7121555605223666,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2386,
+      "step": 6273
+    },
+    {
+      "epoch": 0.7122690876322857,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2426,
+      "step": 6274
+    },
+    {
+      "epoch": 0.7123826147422048,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2166,
+      "step": 6275
+    },
+    {
+      "epoch": 0.7124961418521238,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2319,
+      "step": 6276
+    },
+    {
+      "epoch": 0.7126096689620429,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.25,
+      "step": 6277
+    },
+    {
+      "epoch": 0.712723196071962,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2544,
+      "step": 6278
+    },
+    {
+      "epoch": 0.7128367231818811,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2268,
+      "step": 6279
+    },
+    {
+      "epoch": 0.7129502502918001,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2417,
+      "step": 6280
+    },
+    {
+      "epoch": 0.7130637774017192,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2317,
+      "step": 6281
+    },
+    {
+      "epoch": 0.7131773045116383,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2315,
+      "step": 6282
+    },
+    {
+      "epoch": 0.7132908316215574,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2487,
+      "step": 6283
+    },
+    {
+      "epoch": 0.7134043587314765,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2278,
+      "step": 6284
+    },
+    {
+      "epoch": 0.7135178858413955,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2564,
+      "step": 6285
+    },
+    {
+      "epoch": 0.7136314129513146,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.254,
+      "step": 6286
+    },
+    {
+      "epoch": 0.7137449400612337,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.237,
+      "step": 6287
+    },
+    {
+      "epoch": 0.7138584671711528,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2461,
+      "step": 6288
+    },
+    {
+      "epoch": 0.7139719942810718,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2695,
+      "step": 6289
+    },
+    {
+      "epoch": 0.7140855213909909,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2339,
+      "step": 6290
+    },
+    {
+      "epoch": 0.71419904850091,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2607,
+      "step": 6291
+    },
+    {
+      "epoch": 0.7143125756108291,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2145,
+      "step": 6292
+    },
+    {
+      "epoch": 0.7144261027207481,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2186,
+      "step": 6293
+    },
+    {
+      "epoch": 0.7145396298306672,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2618,
+      "step": 6294
+    },
+    {
+      "epoch": 0.7146531569405863,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2191,
+      "step": 6295
+    },
+    {
+      "epoch": 0.7147666840505054,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2542,
+      "step": 6296
+    },
+    {
+      "epoch": 0.7148802111604244,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2508,
+      "step": 6297
+    },
+    {
+      "epoch": 0.7149937382703435,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2608,
+      "step": 6298
+    },
+    {
+      "epoch": 0.7151072653802626,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2516,
+      "step": 6299
+    },
+    {
+      "epoch": 0.7152207924901817,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2422,
+      "step": 6300
+    },
+    {
+      "epoch": 0.7153343196001007,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2375,
+      "step": 6301
+    },
+    {
+      "epoch": 0.7154478467100198,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.2288,
+      "step": 6302
+    },
+    {
+      "epoch": 0.7155613738199389,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.2386,
+      "step": 6303
+    },
+    {
+      "epoch": 0.715674900929858,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.002,
+      "loss": 5.2415,
+      "step": 6304
+    },
+    {
+      "epoch": 0.715788428039777,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.002,
+      "loss": 5.2467,
+      "step": 6305
+    },
+    {
+      "epoch": 0.7159019551496961,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.2426,
+      "step": 6306
+    },
+    {
+      "epoch": 0.7160154822596152,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2535,
+      "step": 6307
+    },
+    {
+      "epoch": 0.7161290093695343,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.247,
+      "step": 6308
+    },
+    {
+      "epoch": 0.7162425364794534,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2234,
+      "step": 6309
+    },
+    {
+      "epoch": 0.7163560635893724,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.2215,
+      "step": 6310
+    },
+    {
+      "epoch": 0.7164695906992915,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2295,
+      "step": 6311
+    },
+    {
+      "epoch": 0.7165831178092106,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.2615,
+      "step": 6312
+    },
+    {
+      "epoch": 0.7166966449191297,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2268,
+      "step": 6313
+    },
+    {
+      "epoch": 0.7168101720290487,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2612,
+      "step": 6314
+    },
+    {
+      "epoch": 0.7169236991389678,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2545,
+      "step": 6315
+    },
+    {
+      "epoch": 0.7170372262488869,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2335,
+      "step": 6316
+    },
+    {
+      "epoch": 0.717150753358806,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2281,
+      "step": 6317
+    },
+    {
+      "epoch": 0.717264280468725,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.209,
+      "step": 6318
+    },
+    {
+      "epoch": 0.7173778075786441,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2326,
+      "step": 6319
+    },
+    {
+      "epoch": 0.7174913346885632,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.254,
+      "step": 6320
+    },
+    {
+      "epoch": 0.7176048617984823,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.256,
+      "step": 6321
+    },
+    {
+      "epoch": 0.7177183889084013,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.2453,
+      "step": 6322
+    },
+    {
+      "epoch": 0.7178319160183204,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2529,
+      "step": 6323
+    },
+    {
+      "epoch": 0.7179454431282395,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2234,
+      "step": 6324
+    },
+    {
+      "epoch": 0.7180589702381586,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2272,
+      "step": 6325
+    },
+    {
+      "epoch": 0.7181724973480776,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2367,
+      "step": 6326
+    },
+    {
+      "epoch": 0.7182860244579967,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2425,
+      "step": 6327
+    },
+    {
+      "epoch": 0.7183995515679158,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.243,
+      "step": 6328
+    },
+    {
+      "epoch": 0.7185130786778349,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2307,
+      "step": 6329
+    },
+    {
+      "epoch": 0.718626605787754,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2472,
+      "step": 6330
+    },
+    {
+      "epoch": 0.718740132897673,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2287,
+      "step": 6331
+    },
+    {
+      "epoch": 0.7188536600075921,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2418,
+      "step": 6332
+    },
+    {
+      "epoch": 0.7189671871175112,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2604,
+      "step": 6333
+    },
+    {
+      "epoch": 0.7190807142274303,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2445,
+      "step": 6334
+    },
+    {
+      "epoch": 0.7191942413373493,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2534,
+      "step": 6335
+    },
+    {
+      "epoch": 0.7193077684472684,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2262,
+      "step": 6336
+    },
+    {
+      "epoch": 0.7194212955571875,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2634,
+      "step": 6337
+    },
+    {
+      "epoch": 0.7195348226671066,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2402,
+      "step": 6338
+    },
+    {
+      "epoch": 0.7196483497770256,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2427,
+      "step": 6339
+    },
+    {
+      "epoch": 0.7197618768869447,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.25,
+      "step": 6340
+    },
+    {
+      "epoch": 0.7198754039968638,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2438,
+      "step": 6341
+    },
+    {
+      "epoch": 0.7199889311067829,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2458,
+      "step": 6342
+    },
+    {
+      "epoch": 0.7201024582167019,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2153,
+      "step": 6343
+    },
+    {
+      "epoch": 0.720215985326621,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2486,
+      "step": 6344
+    },
+    {
+      "epoch": 0.7203295124365401,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2198,
+      "step": 6345
+    },
+    {
+      "epoch": 0.7204430395464592,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2228,
+      "step": 6346
+    },
+    {
+      "epoch": 0.7205565666563782,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2262,
+      "step": 6347
+    },
+    {
+      "epoch": 0.7206700937662973,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2327,
+      "step": 6348
+    },
+    {
+      "epoch": 0.7207836208762164,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2519,
+      "step": 6349
+    },
+    {
+      "epoch": 0.7208971479861355,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2547,
+      "step": 6350
+    },
+    {
+      "epoch": 0.7210106750960545,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2459,
+      "step": 6351
+    },
+    {
+      "epoch": 0.7211242022059736,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2511,
+      "step": 6352
+    },
+    {
+      "epoch": 0.7212377293158927,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2552,
+      "step": 6353
+    },
+    {
+      "epoch": 0.7213512564258118,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2613,
+      "step": 6354
+    },
+    {
+      "epoch": 0.7214647835357308,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2413,
+      "step": 6355
+    },
+    {
+      "epoch": 0.7215783106456499,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.251,
+      "step": 6356
+    },
+    {
+      "epoch": 0.721691837755569,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2418,
+      "step": 6357
+    },
+    {
+      "epoch": 0.7218053648654881,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2294,
+      "step": 6358
+    },
+    {
+      "epoch": 0.7219188919754072,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2562,
+      "step": 6359
+    },
+    {
+      "epoch": 0.7220324190853262,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.002,
+      "loss": 5.2279,
+      "step": 6360
+    },
+    {
+      "epoch": 0.7221459461952453,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.2671,
+      "step": 6361
+    },
+    {
+      "epoch": 0.7222594733051644,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.23,
+      "step": 6362
+    },
+    {
+      "epoch": 0.7223730004150835,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2403,
+      "step": 6363
+    },
+    {
+      "epoch": 0.7224865275250025,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2304,
+      "step": 6364
+    },
+    {
+      "epoch": 0.7226000546349216,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.253,
+      "step": 6365
+    },
+    {
+      "epoch": 0.7227135817448407,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2448,
+      "step": 6366
+    },
+    {
+      "epoch": 0.7228271088547598,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2385,
+      "step": 6367
+    },
+    {
+      "epoch": 0.7229406359646788,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2347,
+      "step": 6368
+    },
+    {
+      "epoch": 0.7230541630745979,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2655,
+      "step": 6369
+    },
+    {
+      "epoch": 0.723167690184517,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2436,
+      "step": 6370
+    },
+    {
+      "epoch": 0.7232812172944361,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2494,
+      "step": 6371
+    },
+    {
+      "epoch": 0.7233947444043551,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.2626,
+      "step": 6372
+    },
+    {
+      "epoch": 0.7235082715142742,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.002,
+      "loss": 5.2181,
+      "step": 6373
+    },
+    {
+      "epoch": 0.7236217986241933,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2354,
+      "step": 6374
+    },
+    {
+      "epoch": 0.7237353257341124,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.2315,
+      "step": 6375
+    },
+    {
+      "epoch": 0.7238488528440314,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2571,
+      "step": 6376
+    },
+    {
+      "epoch": 0.7239623799539505,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2371,
+      "step": 6377
+    },
+    {
+      "epoch": 0.7240759070638696,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2275,
+      "step": 6378
+    },
+    {
+      "epoch": 0.7241894341737887,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2283,
+      "step": 6379
+    },
+    {
+      "epoch": 0.7243029612837077,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2403,
+      "step": 6380
+    },
+    {
+      "epoch": 0.7244164883936268,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2546,
+      "step": 6381
+    },
+    {
+      "epoch": 0.7245300155035459,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2404,
+      "step": 6382
+    },
+    {
+      "epoch": 0.724643542613465,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2525,
+      "step": 6383
+    },
+    {
+      "epoch": 0.724757069723384,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2326,
+      "step": 6384
+    },
+    {
+      "epoch": 0.7248705968333031,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2494,
+      "step": 6385
+    },
+    {
+      "epoch": 0.7249841239432222,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2328,
+      "step": 6386
+    },
+    {
+      "epoch": 0.7250976510531413,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2292,
+      "step": 6387
+    },
+    {
+      "epoch": 0.7252111781630604,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2334,
+      "step": 6388
+    },
+    {
+      "epoch": 0.7253247052729794,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2517,
+      "step": 6389
+    },
+    {
+      "epoch": 0.7254382323828985,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.265,
+      "step": 6390
+    },
+    {
+      "epoch": 0.7255517594928176,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2655,
+      "step": 6391
+    },
+    {
+      "epoch": 0.7256652866027367,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2321,
+      "step": 6392
+    },
+    {
+      "epoch": 0.7257788137126557,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2326,
+      "step": 6393
+    },
+    {
+      "epoch": 0.7258923408225748,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2581,
+      "step": 6394
+    },
+    {
+      "epoch": 0.7260058679324939,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.245,
+      "step": 6395
+    },
+    {
+      "epoch": 0.7261193950424131,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.254,
+      "step": 6396
+    },
+    {
+      "epoch": 0.7262329221523321,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.249,
+      "step": 6397
+    },
+    {
+      "epoch": 0.7263464492622512,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2758,
+      "step": 6398
+    },
+    {
+      "epoch": 0.7264599763721703,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2435,
+      "step": 6399
+    },
+    {
+      "epoch": 0.7265735034820894,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2326,
+      "step": 6400
+    },
+    {
+      "epoch": 0.7266870305920085,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.002,
+      "loss": 5.2434,
+      "step": 6401
+    },
+    {
+      "epoch": 0.7268005577019275,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.002,
+      "loss": 5.2529,
+      "step": 6402
+    },
+    {
+      "epoch": 0.7269140848118466,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.002,
+      "loss": 5.2284,
+      "step": 6403
+    },
+    {
+      "epoch": 0.7270276119217657,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.002,
+      "loss": 5.2504,
+      "step": 6404
+    },
+    {
+      "epoch": 0.7271411390316848,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.002,
+      "loss": 5.2474,
+      "step": 6405
+    },
+    {
+      "epoch": 0.7272546661416038,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2369,
+      "step": 6406
+    },
+    {
+      "epoch": 0.7273681932515229,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.236,
+      "step": 6407
+    },
+    {
+      "epoch": 0.727481720361442,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2436,
+      "step": 6408
+    },
+    {
+      "epoch": 0.7275952474713611,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2522,
+      "step": 6409
+    },
+    {
+      "epoch": 0.7277087745812801,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2544,
+      "step": 6410
+    },
+    {
+      "epoch": 0.7278223016911992,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.223,
+      "step": 6411
+    },
+    {
+      "epoch": 0.7279358288011183,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2486,
+      "step": 6412
+    },
+    {
+      "epoch": 0.7280493559110374,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2143,
+      "step": 6413
+    },
+    {
+      "epoch": 0.7281628830209564,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2352,
+      "step": 6414
+    },
+    {
+      "epoch": 0.7282764101308755,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2504,
+      "step": 6415
+    },
+    {
+      "epoch": 0.7283899372407946,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2362,
+      "step": 6416
+    },
+    {
+      "epoch": 0.7285034643507137,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.2259,
+      "step": 6417
+    },
+    {
+      "epoch": 0.7286169914606327,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.2635,
+      "step": 6418
+    },
+    {
+      "epoch": 0.7287305185705518,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2481,
+      "step": 6419
+    },
+    {
+      "epoch": 0.7288440456804709,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2477,
+      "step": 6420
+    },
+    {
+      "epoch": 0.72895757279039,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.2511,
+      "step": 6421
+    },
+    {
+      "epoch": 0.729071099900309,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2473,
+      "step": 6422
+    },
+    {
+      "epoch": 0.7291846270102281,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2467,
+      "step": 6423
+    },
+    {
+      "epoch": 0.7292981541201472,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2284,
+      "step": 6424
+    },
+    {
+      "epoch": 0.7294116812300663,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2597,
+      "step": 6425
+    },
+    {
+      "epoch": 0.7295252083399854,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2468,
+      "step": 6426
+    },
+    {
+      "epoch": 0.7296387354499044,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2568,
+      "step": 6427
+    },
+    {
+      "epoch": 0.7297522625598235,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.248,
+      "step": 6428
+    },
+    {
+      "epoch": 0.7298657896697426,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2468,
+      "step": 6429
+    },
+    {
+      "epoch": 0.7299793167796617,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2374,
+      "step": 6430
+    },
+    {
+      "epoch": 0.7300928438895807,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2322,
+      "step": 6431
+    },
+    {
+      "epoch": 0.7302063709994998,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.249,
+      "step": 6432
+    },
+    {
+      "epoch": 0.7303198981094189,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2433,
+      "step": 6433
+    },
+    {
+      "epoch": 0.730433425219338,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2502,
+      "step": 6434
+    },
+    {
+      "epoch": 0.730546952329257,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2283,
+      "step": 6435
+    },
+    {
+      "epoch": 0.7306604794391761,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2358,
+      "step": 6436
+    },
+    {
+      "epoch": 0.7307740065490952,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2377,
+      "step": 6437
+    },
+    {
+      "epoch": 0.7308875336590143,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2598,
+      "step": 6438
+    },
+    {
+      "epoch": 0.7310010607689333,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2473,
+      "step": 6439
+    },
+    {
+      "epoch": 0.7311145878788524,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2511,
+      "step": 6440
+    },
+    {
+      "epoch": 0.7312281149887715,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2519,
+      "step": 6441
+    },
+    {
+      "epoch": 0.7313416420986906,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2497,
+      "step": 6442
+    },
+    {
+      "epoch": 0.7314551692086096,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2443,
+      "step": 6443
+    },
+    {
+      "epoch": 0.7315686963185287,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2332,
+      "step": 6444
+    },
+    {
+      "epoch": 0.7316822234284478,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2344,
+      "step": 6445
+    },
+    {
+      "epoch": 0.7317957505383669,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2467,
+      "step": 6446
+    },
+    {
+      "epoch": 0.731909277648286,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2576,
+      "step": 6447
+    },
+    {
+      "epoch": 0.732022804758205,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2264,
+      "step": 6448
+    },
+    {
+      "epoch": 0.7321363318681241,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2429,
+      "step": 6449
+    },
+    {
+      "epoch": 0.7322498589780432,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.256,
+      "step": 6450
+    },
+    {
+      "epoch": 0.7323633860879623,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.2203,
+      "step": 6451
+    },
+    {
+      "epoch": 0.7324769131978813,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.236,
+      "step": 6452
+    },
+    {
+      "epoch": 0.7325904403078004,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2515,
+      "step": 6453
+    },
+    {
+      "epoch": 0.7327039674177195,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2408,
+      "step": 6454
+    },
+    {
+      "epoch": 0.7328174945276386,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2276,
+      "step": 6455
+    },
+    {
+      "epoch": 0.7329310216375576,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2239,
+      "step": 6456
+    },
+    {
+      "epoch": 0.7330445487474767,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2579,
+      "step": 6457
+    },
+    {
+      "epoch": 0.7331580758573958,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2555,
+      "step": 6458
+    },
+    {
+      "epoch": 0.7332716029673149,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2549,
+      "step": 6459
+    },
+    {
+      "epoch": 0.7333851300772339,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.2386,
+      "step": 6460
+    },
+    {
+      "epoch": 0.733498657187153,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2336,
+      "step": 6461
+    },
+    {
+      "epoch": 0.7336121842970721,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2418,
+      "step": 6462
+    },
+    {
+      "epoch": 0.7337257114069912,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2433,
+      "step": 6463
+    },
+    {
+      "epoch": 0.7338392385169102,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2411,
+      "step": 6464
+    },
+    {
+      "epoch": 0.7339527656268293,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2435,
+      "step": 6465
+    },
+    {
+      "epoch": 0.7340662927367484,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.235,
+      "step": 6466
+    },
+    {
+      "epoch": 0.7341798198466675,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.237,
+      "step": 6467
+    },
+    {
+      "epoch": 0.7342933469565865,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2561,
+      "step": 6468
+    },
+    {
+      "epoch": 0.7344068740665056,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2398,
+      "step": 6469
+    },
+    {
+      "epoch": 0.7345204011764247,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.231,
+      "step": 6470
+    },
+    {
+      "epoch": 0.7346339282863438,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2583,
+      "step": 6471
+    },
+    {
+      "epoch": 0.7347474553962628,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2392,
+      "step": 6472
+    },
+    {
+      "epoch": 0.7348609825061819,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2508,
+      "step": 6473
+    },
+    {
+      "epoch": 0.734974509616101,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2592,
+      "step": 6474
+    },
+    {
+      "epoch": 0.7350880367260201,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2527,
+      "step": 6475
+    },
+    {
+      "epoch": 0.7352015638359392,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2305,
+      "step": 6476
+    },
+    {
+      "epoch": 0.7353150909458582,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.246,
+      "step": 6477
+    },
+    {
+      "epoch": 0.7354286180557773,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2484,
+      "step": 6478
+    },
+    {
+      "epoch": 0.7355421451656964,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.234,
+      "step": 6479
+    },
+    {
+      "epoch": 0.7356556722756155,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2496,
+      "step": 6480
+    },
+    {
+      "epoch": 0.7357691993855345,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.002,
+      "loss": 5.2584,
+      "step": 6481
+    },
+    {
+      "epoch": 0.7358827264954536,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2369,
+      "step": 6482
+    },
+    {
+      "epoch": 0.7359962536053727,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2481,
+      "step": 6483
+    },
+    {
+      "epoch": 0.7361097807152918,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.262,
+      "step": 6484
+    },
+    {
+      "epoch": 0.7362233078252108,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2611,
+      "step": 6485
+    },
+    {
+      "epoch": 0.7363368349351299,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2396,
+      "step": 6486
+    },
+    {
+      "epoch": 0.736450362045049,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2505,
+      "step": 6487
+    },
+    {
+      "epoch": 0.7365638891549681,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2574,
+      "step": 6488
+    },
+    {
+      "epoch": 0.7366774162648871,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2449,
+      "step": 6489
+    },
+    {
+      "epoch": 0.7367909433748062,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2528,
+      "step": 6490
+    },
+    {
+      "epoch": 0.7369044704847253,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2301,
+      "step": 6491
+    },
+    {
+      "epoch": 0.7370179975946444,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2427,
+      "step": 6492
+    },
+    {
+      "epoch": 0.7371315247045634,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2409,
+      "step": 6493
+    },
+    {
+      "epoch": 0.7372450518144825,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2462,
+      "step": 6494
+    },
+    {
+      "epoch": 0.7373585789244016,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2261,
+      "step": 6495
+    },
+    {
+      "epoch": 0.7374721060343207,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2608,
+      "step": 6496
+    },
+    {
+      "epoch": 0.7375856331442397,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2577,
+      "step": 6497
+    },
+    {
+      "epoch": 0.7376991602541588,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2387,
+      "step": 6498
+    },
+    {
+      "epoch": 0.7378126873640779,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2269,
+      "step": 6499
+    },
+    {
+      "epoch": 0.737926214473997,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2523,
+      "step": 6500
+    },
+    {
+      "epoch": 0.738039741583916,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2218,
+      "step": 6501
+    },
+    {
+      "epoch": 0.7381532686938351,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2441,
+      "step": 6502
+    },
+    {
+      "epoch": 0.7382667958037542,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2375,
+      "step": 6503
+    },
+    {
+      "epoch": 0.7383803229136733,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2305,
+      "step": 6504
+    },
+    {
+      "epoch": 0.7384938500235924,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2231,
+      "step": 6505
+    },
+    {
+      "epoch": 0.7386073771335114,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.002,
+      "loss": 5.2292,
+      "step": 6506
+    },
+    {
+      "epoch": 0.7387209042434305,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2048,
+      "step": 6507
+    },
+    {
+      "epoch": 0.7388344313533496,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2132,
+      "step": 6508
+    },
+    {
+      "epoch": 0.7389479584632687,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2315,
+      "step": 6509
+    },
+    {
+      "epoch": 0.7390614855731877,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2397,
+      "step": 6510
+    },
+    {
+      "epoch": 0.7391750126831068,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2287,
+      "step": 6511
+    },
+    {
+      "epoch": 0.7392885397930259,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2405,
+      "step": 6512
+    },
+    {
+      "epoch": 0.739402066902945,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2423,
+      "step": 6513
+    },
+    {
+      "epoch": 0.739515594012864,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2387,
+      "step": 6514
+    },
+    {
+      "epoch": 0.7396291211227831,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2253,
+      "step": 6515
+    },
+    {
+      "epoch": 0.7397426482327022,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2197,
+      "step": 6516
+    },
+    {
+      "epoch": 0.7398561753426213,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2267,
+      "step": 6517
+    },
+    {
+      "epoch": 0.7399697024525403,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2124,
+      "step": 6518
+    },
+    {
+      "epoch": 0.7400832295624594,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2582,
+      "step": 6519
+    },
+    {
+      "epoch": 0.7401967566723785,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2414,
+      "step": 6520
+    },
+    {
+      "epoch": 0.7403102837822976,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.235,
+      "step": 6521
+    },
+    {
+      "epoch": 0.7404238108922166,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2377,
+      "step": 6522
+    },
+    {
+      "epoch": 0.7405373380021357,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2338,
+      "step": 6523
+    },
+    {
+      "epoch": 0.7406508651120548,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2497,
+      "step": 6524
+    },
+    {
+      "epoch": 0.7407643922219739,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2183,
+      "step": 6525
+    },
+    {
+      "epoch": 0.740877919331893,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2492,
+      "step": 6526
+    },
+    {
+      "epoch": 0.740991446441812,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2358,
+      "step": 6527
+    },
+    {
+      "epoch": 0.7411049735517311,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2461,
+      "step": 6528
+    },
+    {
+      "epoch": 0.7412185006616502,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2402,
+      "step": 6529
+    },
+    {
+      "epoch": 0.7413320277715693,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2408,
+      "step": 6530
+    },
+    {
+      "epoch": 0.7414455548814883,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2309,
+      "step": 6531
+    },
+    {
+      "epoch": 0.7415590819914074,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2614,
+      "step": 6532
+    },
+    {
+      "epoch": 0.7416726091013265,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2371,
+      "step": 6533
+    },
+    {
+      "epoch": 0.7417861362112456,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2419,
+      "step": 6534
+    },
+    {
+      "epoch": 0.7418996633211646,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2457,
+      "step": 6535
+    },
+    {
+      "epoch": 0.7420131904310837,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2353,
+      "step": 6536
+    },
+    {
+      "epoch": 0.7421267175410028,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2268,
+      "step": 6537
+    },
+    {
+      "epoch": 0.7422402446509219,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2537,
+      "step": 6538
+    },
+    {
+      "epoch": 0.7423537717608409,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2176,
+      "step": 6539
+    },
+    {
+      "epoch": 0.74246729887076,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2293,
+      "step": 6540
+    },
+    {
+      "epoch": 0.7425808259806791,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2393,
+      "step": 6541
+    },
+    {
+      "epoch": 0.7426943530905982,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.2599,
+      "step": 6542
+    },
+    {
+      "epoch": 0.7428078802005172,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 5.2475,
+      "step": 6543
+    },
+    {
+      "epoch": 0.7429214073104363,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.2291,
+      "step": 6544
+    },
+    {
+      "epoch": 0.7430349344203554,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.002,
+      "loss": 5.244,
+      "step": 6545
+    },
+    {
+      "epoch": 0.7431484615302745,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2286,
+      "step": 6546
+    },
+    {
+      "epoch": 0.7432619886401935,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.2353,
+      "step": 6547
+    },
+    {
+      "epoch": 0.7433755157501126,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2229,
+      "step": 6548
+    },
+    {
+      "epoch": 0.7434890428600317,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2331,
+      "step": 6549
+    },
+    {
+      "epoch": 0.7436025699699508,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2244,
+      "step": 6550
+    },
+    {
+      "epoch": 0.7437160970798699,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2461,
+      "step": 6551
+    },
+    {
+      "epoch": 0.7438296241897889,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 5.2289,
+      "step": 6552
+    },
+    {
+      "epoch": 0.743943151299708,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.2444,
+      "step": 6553
+    },
+    {
+      "epoch": 0.7440566784096271,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.2608,
+      "step": 6554
+    },
+    {
+      "epoch": 0.7441702055195462,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2246,
+      "step": 6555
+    },
+    {
+      "epoch": 0.7442837326294652,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.002,
+      "loss": 5.2412,
+      "step": 6556
+    },
+    {
+      "epoch": 0.7443972597393843,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2227,
+      "step": 6557
+    },
+    {
+      "epoch": 0.7445107868493034,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2438,
+      "step": 6558
+    },
+    {
+      "epoch": 0.7446243139592225,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.002,
+      "loss": 5.2306,
+      "step": 6559
+    },
+    {
+      "epoch": 0.7447378410691415,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2453,
+      "step": 6560
+    },
+    {
+      "epoch": 0.7448513681790606,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.229,
+      "step": 6561
+    },
+    {
+      "epoch": 0.7449648952889797,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2272,
+      "step": 6562
+    },
+    {
+      "epoch": 0.7450784223988988,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2371,
+      "step": 6563
+    },
+    {
+      "epoch": 0.7451919495088178,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2373,
+      "step": 6564
+    },
+    {
+      "epoch": 0.7453054766187369,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2289,
+      "step": 6565
+    },
+    {
+      "epoch": 0.745419003728656,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.224,
+      "step": 6566
+    },
+    {
+      "epoch": 0.7455325308385751,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2315,
+      "step": 6567
+    },
+    {
+      "epoch": 0.7456460579484941,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2401,
+      "step": 6568
+    },
+    {
+      "epoch": 0.7457595850584132,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.241,
+      "step": 6569
+    },
+    {
+      "epoch": 0.7458731121683323,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2505,
+      "step": 6570
+    },
+    {
+      "epoch": 0.7459866392782514,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2421,
+      "step": 6571
+    },
+    {
+      "epoch": 0.7461001663881704,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2558,
+      "step": 6572
+    },
+    {
+      "epoch": 0.7462136934980895,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2396,
+      "step": 6573
+    },
+    {
+      "epoch": 0.7463272206080086,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2588,
+      "step": 6574
+    },
+    {
+      "epoch": 0.7464407477179277,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2467,
+      "step": 6575
+    },
+    {
+      "epoch": 0.7465542748278468,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.002,
+      "loss": 5.2517,
+      "step": 6576
+    },
+    {
+      "epoch": 0.7466678019377658,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.002,
+      "loss": 5.2457,
+      "step": 6577
+    },
+    {
+      "epoch": 0.7467813290476849,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.002,
+      "loss": 5.2204,
+      "step": 6578
+    },
+    {
+      "epoch": 0.746894856157604,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.002,
+      "loss": 5.24,
+      "step": 6579
+    },
+    {
+      "epoch": 0.7470083832675231,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.2324,
+      "step": 6580
+    },
+    {
+      "epoch": 0.7471219103774421,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.002,
+      "loss": 5.2342,
+      "step": 6581
+    },
+    {
+      "epoch": 0.7472354374873612,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2397,
+      "step": 6582
+    },
+    {
+      "epoch": 0.7473489645972803,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2179,
+      "step": 6583
+    },
+    {
+      "epoch": 0.7474624917071994,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2565,
+      "step": 6584
+    },
+    {
+      "epoch": 0.7475760188171184,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2235,
+      "step": 6585
+    },
+    {
+      "epoch": 0.7476895459270375,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2388,
+      "step": 6586
+    },
+    {
+      "epoch": 0.7478030730369566,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.232,
+      "step": 6587
+    },
+    {
+      "epoch": 0.7479166001468757,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2668,
+      "step": 6588
+    },
+    {
+      "epoch": 0.7480301272567947,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.2487,
+      "step": 6589
+    },
+    {
+      "epoch": 0.7481436543667138,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 5.2537,
+      "step": 6590
+    },
+    {
+      "epoch": 0.7482571814766329,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.2132,
+      "step": 6591
+    },
+    {
+      "epoch": 0.748370708586552,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2297,
+      "step": 6592
+    },
+    {
+      "epoch": 0.748484235696471,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2187,
+      "step": 6593
+    },
+    {
+      "epoch": 0.7485977628063901,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2564,
+      "step": 6594
+    },
+    {
+      "epoch": 0.7487112899163092,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2328,
+      "step": 6595
+    },
+    {
+      "epoch": 0.7488248170262283,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2347,
+      "step": 6596
+    },
+    {
+      "epoch": 0.7489383441361473,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.002,
+      "loss": 5.2322,
+      "step": 6597
+    },
+    {
+      "epoch": 0.7490518712460664,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.224,
+      "step": 6598
+    },
+    {
+      "epoch": 0.7491653983559855,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2275,
+      "step": 6599
+    },
+    {
+      "epoch": 0.7492789254659046,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.242,
+      "step": 6600
+    },
+    {
+      "epoch": 0.7493924525758237,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2326,
+      "step": 6601
+    },
+    {
+      "epoch": 0.7495059796857427,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2316,
+      "step": 6602
+    },
+    {
+      "epoch": 0.7496195067956618,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2438,
+      "step": 6603
+    },
+    {
+      "epoch": 0.7497330339055809,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.2464,
+      "step": 6604
+    },
+    {
+      "epoch": 0.7498465610155,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2058,
+      "step": 6605
+    },
+    {
+      "epoch": 0.749960088125419,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2333,
+      "step": 6606
+    },
+    {
+      "epoch": 0.7500736152353381,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2363,
+      "step": 6607
+    },
+    {
+      "epoch": 0.7501871423452572,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2124,
+      "step": 6608
+    },
+    {
+      "epoch": 0.7503006694551763,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2463,
+      "step": 6609
+    },
+    {
+      "epoch": 0.7504141965650953,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2565,
+      "step": 6610
+    },
+    {
+      "epoch": 0.7505277236750144,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2413,
+      "step": 6611
+    },
+    {
+      "epoch": 0.7506412507849335,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2392,
+      "step": 6612
+    },
+    {
+      "epoch": 0.7507547778948526,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2116,
+      "step": 6613
+    },
+    {
+      "epoch": 0.7508683050047716,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2397,
+      "step": 6614
+    },
+    {
+      "epoch": 0.7509818321146907,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2337,
+      "step": 6615
+    },
+    {
+      "epoch": 0.7510953592246098,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2508,
+      "step": 6616
+    },
+    {
+      "epoch": 0.7512088863345289,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2494,
+      "step": 6617
+    },
+    {
+      "epoch": 0.751322413444448,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 5.2424,
+      "step": 6618
+    },
+    {
+      "epoch": 0.751435940554367,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2054,
+      "step": 6619
+    },
+    {
+      "epoch": 0.7515494676642861,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2499,
+      "step": 6620
+    },
+    {
+      "epoch": 0.7516629947742052,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2385,
+      "step": 6621
+    },
+    {
+      "epoch": 0.7517765218841242,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2374,
+      "step": 6622
+    },
+    {
+      "epoch": 0.7518900489940433,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2377,
+      "step": 6623
+    },
+    {
+      "epoch": 0.7520035761039624,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.2351,
+      "step": 6624
+    },
+    {
+      "epoch": 0.7521171032138815,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.227,
+      "step": 6625
+    },
+    {
+      "epoch": 0.7522306303238006,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2301,
+      "step": 6626
+    },
+    {
+      "epoch": 0.7523441574337196,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.237,
+      "step": 6627
+    },
+    {
+      "epoch": 0.7524576845436387,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2243,
+      "step": 6628
+    },
+    {
+      "epoch": 0.7525712116535578,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.257,
+      "step": 6629
+    },
+    {
+      "epoch": 0.7526847387634769,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2434,
+      "step": 6630
+    },
+    {
+      "epoch": 0.7527982658733959,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2383,
+      "step": 6631
+    },
+    {
+      "epoch": 0.752911792983315,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2206,
+      "step": 6632
+    },
+    {
+      "epoch": 0.7530253200932341,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2558,
+      "step": 6633
+    },
+    {
+      "epoch": 0.7531388472031532,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2568,
+      "step": 6634
+    },
+    {
+      "epoch": 0.7532523743130722,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2483,
+      "step": 6635
+    },
+    {
+      "epoch": 0.7533659014229913,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2293,
+      "step": 6636
+    },
+    {
+      "epoch": 0.7534794285329105,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2456,
+      "step": 6637
+    },
+    {
+      "epoch": 0.7535929556428296,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2091,
+      "step": 6638
+    },
+    {
+      "epoch": 0.7537064827527487,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2541,
+      "step": 6639
+    },
+    {
+      "epoch": 0.7538200098626677,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2416,
+      "step": 6640
+    },
+    {
+      "epoch": 0.7539335369725868,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2353,
+      "step": 6641
+    },
+    {
+      "epoch": 0.7540470640825059,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2472,
+      "step": 6642
+    },
+    {
+      "epoch": 0.754160591192425,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2533,
+      "step": 6643
+    },
+    {
+      "epoch": 0.754274118302344,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 5.2283,
+      "step": 6644
+    },
+    {
+      "epoch": 0.7543876454122631,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2423,
+      "step": 6645
+    },
+    {
+      "epoch": 0.7545011725221822,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.002,
+      "loss": 5.2412,
+      "step": 6646
+    },
+    {
+      "epoch": 0.7546146996321013,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.002,
+      "loss": 5.2557,
+      "step": 6647
+    },
+    {
+      "epoch": 0.7547282267420203,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.229,
+      "step": 6648
+    },
+    {
+      "epoch": 0.7548417538519394,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2573,
+      "step": 6649
+    },
+    {
+      "epoch": 0.7549552809618585,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2189,
+      "step": 6650
+    },
+    {
+      "epoch": 0.7550688080717776,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.002,
+      "loss": 5.2383,
+      "step": 6651
+    },
+    {
+      "epoch": 0.7551823351816966,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.2595,
+      "step": 6652
+    },
+    {
+      "epoch": 0.7552958622916157,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2155,
+      "step": 6653
+    },
+    {
+      "epoch": 0.7554093894015348,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2134,
+      "step": 6654
+    },
+    {
+      "epoch": 0.7555229165114539,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2423,
+      "step": 6655
+    },
+    {
+      "epoch": 0.7556364436213729,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2403,
+      "step": 6656
+    },
+    {
+      "epoch": 0.755749970731292,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2458,
+      "step": 6657
+    },
+    {
+      "epoch": 0.7558634978412111,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2019,
+      "step": 6658
+    },
+    {
+      "epoch": 0.7559770249511302,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2393,
+      "step": 6659
+    },
+    {
+      "epoch": 0.7560905520610492,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2406,
+      "step": 6660
+    },
+    {
+      "epoch": 0.7562040791709683,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2323,
+      "step": 6661
+    },
+    {
+      "epoch": 0.7563176062808874,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2272,
+      "step": 6662
+    },
+    {
+      "epoch": 0.7564311333908065,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2457,
+      "step": 6663
+    },
+    {
+      "epoch": 0.7565446605007256,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2305,
+      "step": 6664
+    },
+    {
+      "epoch": 0.7566581876106446,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.002,
+      "loss": 5.2457,
+      "step": 6665
+    },
+    {
+      "epoch": 0.7567717147205637,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.24,
+      "step": 6666
+    },
+    {
+      "epoch": 0.7568852418304828,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.002,
+      "loss": 5.2461,
+      "step": 6667
+    },
+    {
+      "epoch": 0.7569987689404019,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2423,
+      "step": 6668
+    },
+    {
+      "epoch": 0.7571122960503209,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2406,
+      "step": 6669
+    },
+    {
+      "epoch": 0.75722582316024,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2573,
+      "step": 6670
+    },
+    {
+      "epoch": 0.7573393502701591,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2423,
+      "step": 6671
+    },
+    {
+      "epoch": 0.7574528773800782,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2429,
+      "step": 6672
+    },
+    {
+      "epoch": 0.7575664044899972,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2299,
+      "step": 6673
+    },
+    {
+      "epoch": 0.7576799315999163,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2371,
+      "step": 6674
+    },
+    {
+      "epoch": 0.7577934587098354,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2248,
+      "step": 6675
+    },
+    {
+      "epoch": 0.7579069858197545,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2463,
+      "step": 6676
+    },
+    {
+      "epoch": 0.7580205129296735,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.248,
+      "step": 6677
+    },
+    {
+      "epoch": 0.7581340400395926,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2227,
+      "step": 6678
+    },
+    {
+      "epoch": 0.7582475671495117,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2248,
+      "step": 6679
+    },
+    {
+      "epoch": 0.7583610942594308,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.229,
+      "step": 6680
+    },
+    {
+      "epoch": 0.7584746213693498,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.2353,
+      "step": 6681
+    },
+    {
+      "epoch": 0.7585881484792689,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2508,
+      "step": 6682
+    },
+    {
+      "epoch": 0.758701675589188,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2359,
+      "step": 6683
+    },
+    {
+      "epoch": 0.7588152026991071,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2327,
+      "step": 6684
+    },
+    {
+      "epoch": 0.7589287298090261,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2302,
+      "step": 6685
+    },
+    {
+      "epoch": 0.7590422569189452,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2441,
+      "step": 6686
+    },
+    {
+      "epoch": 0.7591557840288643,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.231,
+      "step": 6687
+    },
+    {
+      "epoch": 0.7592693111387834,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2483,
+      "step": 6688
+    },
+    {
+      "epoch": 0.7593828382487025,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.231,
+      "step": 6689
+    },
+    {
+      "epoch": 0.7594963653586215,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2369,
+      "step": 6690
+    },
+    {
+      "epoch": 0.7596098924685406,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2421,
+      "step": 6691
+    },
+    {
+      "epoch": 0.7597234195784597,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.238,
+      "step": 6692
+    },
+    {
+      "epoch": 0.7598369466883788,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2282,
+      "step": 6693
+    },
+    {
+      "epoch": 0.7599504737982978,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2171,
+      "step": 6694
+    },
+    {
+      "epoch": 0.7600640009082169,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2314,
+      "step": 6695
+    },
+    {
+      "epoch": 0.760177528018136,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2506,
+      "step": 6696
+    },
+    {
+      "epoch": 0.7602910551280551,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2679,
+      "step": 6697
+    },
+    {
+      "epoch": 0.7604045822379741,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2272,
+      "step": 6698
+    },
+    {
+      "epoch": 0.7605181093478932,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2184,
+      "step": 6699
+    },
+    {
+      "epoch": 0.7606316364578123,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2413,
+      "step": 6700
+    },
+    {
+      "epoch": 0.7607451635677314,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2358,
+      "step": 6701
+    },
+    {
+      "epoch": 0.7608586906776504,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2341,
+      "step": 6702
+    },
+    {
+      "epoch": 0.7609722177875695,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2514,
+      "step": 6703
+    },
+    {
+      "epoch": 0.7610857448974886,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2317,
+      "step": 6704
+    },
+    {
+      "epoch": 0.7611992720074077,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.002,
+      "loss": 5.2383,
+      "step": 6705
+    },
+    {
+      "epoch": 0.7613127991173267,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.002,
+      "loss": 5.2513,
+      "step": 6706
+    },
+    {
+      "epoch": 0.7614263262272458,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.002,
+      "loss": 5.2426,
+      "step": 6707
+    },
+    {
+      "epoch": 0.7615398533371649,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.2501,
+      "step": 6708
+    },
+    {
+      "epoch": 0.761653380447084,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.2264,
+      "step": 6709
+    },
+    {
+      "epoch": 0.761766907557003,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2307,
+      "step": 6710
+    },
+    {
+      "epoch": 0.7618804346669221,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2472,
+      "step": 6711
+    },
+    {
+      "epoch": 0.7619939617768412,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2241,
+      "step": 6712
+    },
+    {
+      "epoch": 0.7621074888867603,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2492,
+      "step": 6713
+    },
+    {
+      "epoch": 0.7622210159966794,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2182,
+      "step": 6714
+    },
+    {
+      "epoch": 0.7623345431065984,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2345,
+      "step": 6715
+    },
+    {
+      "epoch": 0.7624480702165175,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.235,
+      "step": 6716
+    },
+    {
+      "epoch": 0.7625615973264366,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2307,
+      "step": 6717
+    },
+    {
+      "epoch": 0.7626751244363557,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2234,
+      "step": 6718
+    },
+    {
+      "epoch": 0.7627886515462747,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2426,
+      "step": 6719
+    },
+    {
+      "epoch": 0.7629021786561938,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.237,
+      "step": 6720
+    },
+    {
+      "epoch": 0.7630157057661129,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2426,
+      "step": 6721
+    },
+    {
+      "epoch": 0.763129232876032,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2178,
+      "step": 6722
+    },
+    {
+      "epoch": 0.763242759985951,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.223,
+      "step": 6723
+    },
+    {
+      "epoch": 0.7633562870958701,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2444,
+      "step": 6724
+    },
+    {
+      "epoch": 0.7634698142057892,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2394,
+      "step": 6725
+    },
+    {
+      "epoch": 0.7635833413157083,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.234,
+      "step": 6726
+    },
+    {
+      "epoch": 0.7636968684256273,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2378,
+      "step": 6727
+    },
+    {
+      "epoch": 0.7638103955355464,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2362,
+      "step": 6728
+    },
+    {
+      "epoch": 0.7639239226454655,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.218,
+      "step": 6729
+    },
+    {
+      "epoch": 0.7640374497553846,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2459,
+      "step": 6730
+    },
+    {
+      "epoch": 0.7641509768653036,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2241,
+      "step": 6731
+    },
+    {
+      "epoch": 0.7642645039752227,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2279,
+      "step": 6732
+    },
+    {
+      "epoch": 0.7643780310851418,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2338,
+      "step": 6733
+    },
+    {
+      "epoch": 0.7644915581950609,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2399,
+      "step": 6734
+    },
+    {
+      "epoch": 0.76460508530498,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2254,
+      "step": 6735
+    },
+    {
+      "epoch": 0.764718612414899,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2402,
+      "step": 6736
+    },
+    {
+      "epoch": 0.7648321395248181,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.2512,
+      "step": 6737
+    },
+    {
+      "epoch": 0.7649456666347372,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2268,
+      "step": 6738
+    },
+    {
+      "epoch": 0.7650591937446563,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2291,
+      "step": 6739
+    },
+    {
+      "epoch": 0.7651727208545753,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2577,
+      "step": 6740
+    },
+    {
+      "epoch": 0.7652862479644944,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2588,
+      "step": 6741
+    },
+    {
+      "epoch": 0.7653997750744135,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2401,
+      "step": 6742
+    },
+    {
+      "epoch": 0.7655133021843326,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2435,
+      "step": 6743
+    },
+    {
+      "epoch": 0.7656268292942516,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2362,
+      "step": 6744
+    },
+    {
+      "epoch": 0.7657403564041707,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2474,
+      "step": 6745
+    },
+    {
+      "epoch": 0.7658538835140898,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.002,
+      "loss": 5.2224,
+      "step": 6746
+    },
+    {
+      "epoch": 0.7659674106240089,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2369,
+      "step": 6747
+    },
+    {
+      "epoch": 0.7660809377339279,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2333,
+      "step": 6748
+    },
+    {
+      "epoch": 0.766194464843847,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2431,
+      "step": 6749
+    },
+    {
+      "epoch": 0.7663079919537661,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2459,
+      "step": 6750
+    },
+    {
+      "epoch": 0.7664215190636852,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2325,
+      "step": 6751
+    },
+    {
+      "epoch": 0.7665350461736042,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2352,
+      "step": 6752
+    },
+    {
+      "epoch": 0.7666485732835233,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2107,
+      "step": 6753
+    },
+    {
+      "epoch": 0.7667621003934424,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2366,
+      "step": 6754
+    },
+    {
+      "epoch": 0.7668756275033615,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2143,
+      "step": 6755
+    },
+    {
+      "epoch": 0.7669891546132805,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.231,
+      "step": 6756
+    },
+    {
+      "epoch": 0.7671026817231996,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2416,
+      "step": 6757
+    },
+    {
+      "epoch": 0.7672162088331187,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.229,
+      "step": 6758
+    },
+    {
+      "epoch": 0.7673297359430378,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2533,
+      "step": 6759
+    },
+    {
+      "epoch": 0.7674432630529568,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2351,
+      "step": 6760
+    },
+    {
+      "epoch": 0.7675567901628759,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2277,
+      "step": 6761
+    },
+    {
+      "epoch": 0.767670317272795,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2395,
+      "step": 6762
+    },
+    {
+      "epoch": 0.7677838443827141,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.242,
+      "step": 6763
+    },
+    {
+      "epoch": 0.7678973714926332,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2243,
+      "step": 6764
+    },
+    {
+      "epoch": 0.7680108986025522,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.2223,
+      "step": 6765
+    },
+    {
+      "epoch": 0.7681244257124713,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2535,
+      "step": 6766
+    },
+    {
+      "epoch": 0.7682379528223904,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2203,
+      "step": 6767
+    },
+    {
+      "epoch": 0.7683514799323095,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.223,
+      "step": 6768
+    },
+    {
+      "epoch": 0.7684650070422285,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2298,
+      "step": 6769
+    },
+    {
+      "epoch": 0.7685785341521476,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.235,
+      "step": 6770
+    },
+    {
+      "epoch": 0.7686920612620667,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2343,
+      "step": 6771
+    },
+    {
+      "epoch": 0.7688055883719858,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.222,
+      "step": 6772
+    },
+    {
+      "epoch": 0.7689191154819048,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2294,
+      "step": 6773
+    },
+    {
+      "epoch": 0.7690326425918239,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2549,
+      "step": 6774
+    },
+    {
+      "epoch": 0.769146169701743,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2303,
+      "step": 6775
+    },
+    {
+      "epoch": 0.7692596968116621,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2419,
+      "step": 6776
+    },
+    {
+      "epoch": 0.7693732239215811,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2335,
+      "step": 6777
+    },
+    {
+      "epoch": 0.7694867510315002,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.241,
+      "step": 6778
+    },
+    {
+      "epoch": 0.7696002781414193,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2332,
+      "step": 6779
+    },
+    {
+      "epoch": 0.7697138052513384,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.237,
+      "step": 6780
+    },
+    {
+      "epoch": 0.7698273323612574,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2282,
+      "step": 6781
+    },
+    {
+      "epoch": 0.7699408594711765,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2345,
+      "step": 6782
+    },
+    {
+      "epoch": 0.7700543865810956,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2493,
+      "step": 6783
+    },
+    {
+      "epoch": 0.7701679136910147,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2382,
+      "step": 6784
+    },
+    {
+      "epoch": 0.7702814408009337,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2788,
+      "step": 6785
+    },
+    {
+      "epoch": 0.7703949679108528,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.238,
+      "step": 6786
+    },
+    {
+      "epoch": 0.7705084950207719,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.244,
+      "step": 6787
+    },
+    {
+      "epoch": 0.770622022130691,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2475,
+      "step": 6788
+    },
+    {
+      "epoch": 0.77073554924061,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2095,
+      "step": 6789
+    },
+    {
+      "epoch": 0.7708490763505291,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.002,
+      "loss": 5.2081,
+      "step": 6790
+    },
+    {
+      "epoch": 0.7709626034604482,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2202,
+      "step": 6791
+    },
+    {
+      "epoch": 0.7710761305703673,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2479,
+      "step": 6792
+    },
+    {
+      "epoch": 0.7711896576802864,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2461,
+      "step": 6793
+    },
+    {
+      "epoch": 0.7713031847902054,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2277,
+      "step": 6794
+    },
+    {
+      "epoch": 0.7714167119001245,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.234,
+      "step": 6795
+    },
+    {
+      "epoch": 0.7715302390100436,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2404,
+      "step": 6796
+    },
+    {
+      "epoch": 0.7716437661199627,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2266,
+      "step": 6797
+    },
+    {
+      "epoch": 0.7717572932298817,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2344,
+      "step": 6798
+    },
+    {
+      "epoch": 0.7718708203398008,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2377,
+      "step": 6799
+    },
+    {
+      "epoch": 0.7719843474497199,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2376,
+      "step": 6800
+    },
+    {
+      "epoch": 0.772097874559639,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2333,
+      "step": 6801
+    },
+    {
+      "epoch": 0.772211401669558,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2518,
+      "step": 6802
+    },
+    {
+      "epoch": 0.7723249287794771,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2407,
+      "step": 6803
+    },
+    {
+      "epoch": 0.7724384558893962,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2413,
+      "step": 6804
+    },
+    {
+      "epoch": 0.7725519829993153,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2209,
+      "step": 6805
+    },
+    {
+      "epoch": 0.7726655101092343,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2597,
+      "step": 6806
+    },
+    {
+      "epoch": 0.7727790372191534,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2248,
+      "step": 6807
+    },
+    {
+      "epoch": 0.7728925643290725,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2286,
+      "step": 6808
+    },
+    {
+      "epoch": 0.7730060914389916,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2331,
+      "step": 6809
+    },
+    {
+      "epoch": 0.7731196185489106,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2193,
+      "step": 6810
+    },
+    {
+      "epoch": 0.7732331456588297,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2528,
+      "step": 6811
+    },
+    {
+      "epoch": 0.7733466727687488,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2491,
+      "step": 6812
+    },
+    {
+      "epoch": 0.7734601998786679,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2199,
+      "step": 6813
+    },
+    {
+      "epoch": 0.773573726988587,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.002,
+      "loss": 5.2408,
+      "step": 6814
+    },
+    {
+      "epoch": 0.773687254098506,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2329,
+      "step": 6815
+    },
+    {
+      "epoch": 0.7738007812084251,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2513,
+      "step": 6816
+    },
+    {
+      "epoch": 0.7739143083183442,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2356,
+      "step": 6817
+    },
+    {
+      "epoch": 0.7740278354282633,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2374,
+      "step": 6818
+    },
+    {
+      "epoch": 0.7741413625381823,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.2206,
+      "step": 6819
+    },
+    {
+      "epoch": 0.7742548896481014,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2301,
+      "step": 6820
+    },
+    {
+      "epoch": 0.7743684167580205,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.2602,
+      "step": 6821
+    },
+    {
+      "epoch": 0.7744819438679396,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2196,
+      "step": 6822
+    },
+    {
+      "epoch": 0.7745954709778586,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2355,
+      "step": 6823
+    },
+    {
+      "epoch": 0.7747089980877777,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2368,
+      "step": 6824
+    },
+    {
+      "epoch": 0.7748225251976968,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2678,
+      "step": 6825
+    },
+    {
+      "epoch": 0.7749360523076159,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2182,
+      "step": 6826
+    },
+    {
+      "epoch": 0.7750495794175349,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.2361,
+      "step": 6827
+    },
+    {
+      "epoch": 0.775163106527454,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.232,
+      "step": 6828
+    },
+    {
+      "epoch": 0.7752766336373731,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.002,
+      "loss": 5.2196,
+      "step": 6829
+    },
+    {
+      "epoch": 0.7753901607472922,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.2364,
+      "step": 6830
+    },
+    {
+      "epoch": 0.7755036878572112,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.258,
+      "step": 6831
+    },
+    {
+      "epoch": 0.7756172149671303,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2222,
+      "step": 6832
+    },
+    {
+      "epoch": 0.7757307420770494,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2422,
+      "step": 6833
+    },
+    {
+      "epoch": 0.7758442691869685,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2332,
+      "step": 6834
+    },
+    {
+      "epoch": 0.7759577962968875,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2299,
+      "step": 6835
+    },
+    {
+      "epoch": 0.7760713234068066,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2432,
+      "step": 6836
+    },
+    {
+      "epoch": 0.7761848505167257,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2389,
+      "step": 6837
+    },
+    {
+      "epoch": 0.7762983776266448,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2381,
+      "step": 6838
+    },
+    {
+      "epoch": 0.7764119047365639,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2311,
+      "step": 6839
+    },
+    {
+      "epoch": 0.7765254318464829,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2197,
+      "step": 6840
+    },
+    {
+      "epoch": 0.776638958956402,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2363,
+      "step": 6841
+    },
+    {
+      "epoch": 0.7767524860663211,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.245,
+      "step": 6842
+    },
+    {
+      "epoch": 0.7768660131762402,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.2618,
+      "step": 6843
+    },
+    {
+      "epoch": 0.7769795402861592,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.25,
+      "step": 6844
+    },
+    {
+      "epoch": 0.7770930673960783,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.247,
+      "step": 6845
+    },
+    {
+      "epoch": 0.7772065945059974,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2188,
+      "step": 6846
+    },
+    {
+      "epoch": 0.7773201216159165,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.255,
+      "step": 6847
+    },
+    {
+      "epoch": 0.7774336487258355,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2345,
+      "step": 6848
+    },
+    {
+      "epoch": 0.7775471758357546,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.002,
+      "loss": 5.2509,
+      "step": 6849
+    },
+    {
+      "epoch": 0.7776607029456737,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2479,
+      "step": 6850
+    },
+    {
+      "epoch": 0.7777742300555928,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2289,
+      "step": 6851
+    },
+    {
+      "epoch": 0.7778877571655118,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2379,
+      "step": 6852
+    },
+    {
+      "epoch": 0.7780012842754309,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.2318,
+      "step": 6853
+    },
+    {
+      "epoch": 0.77811481138535,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2342,
+      "step": 6854
+    },
+    {
+      "epoch": 0.7782283384952691,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.271,
+      "step": 6855
+    },
+    {
+      "epoch": 0.7783418656051881,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2582,
+      "step": 6856
+    },
+    {
+      "epoch": 0.7784553927151072,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2279,
+      "step": 6857
+    },
+    {
+      "epoch": 0.7785689198250263,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.002,
+      "loss": 5.2574,
+      "step": 6858
+    },
+    {
+      "epoch": 0.7786824469349454,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.002,
+      "loss": 5.2319,
+      "step": 6859
+    },
+    {
+      "epoch": 0.7787959740448644,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.2146,
+      "step": 6860
+    },
+    {
+      "epoch": 0.7789095011547835,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2272,
+      "step": 6861
+    },
+    {
+      "epoch": 0.7790230282647026,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2288,
+      "step": 6862
+    },
+    {
+      "epoch": 0.7791365553746217,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2431,
+      "step": 6863
+    },
+    {
+      "epoch": 0.7792500824845408,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2463,
+      "step": 6864
+    },
+    {
+      "epoch": 0.7793636095944598,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.002,
+      "loss": 5.2358,
+      "step": 6865
+    },
+    {
+      "epoch": 0.7794771367043789,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.002,
+      "loss": 5.2503,
+      "step": 6866
+    },
+    {
+      "epoch": 0.779590663814298,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2333,
+      "step": 6867
+    },
+    {
+      "epoch": 0.779704190924217,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2463,
+      "step": 6868
+    },
+    {
+      "epoch": 0.7798177180341361,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.002,
+      "loss": 5.232,
+      "step": 6869
+    },
+    {
+      "epoch": 0.7799312451440552,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2399,
+      "step": 6870
+    },
+    {
+      "epoch": 0.7800447722539743,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2346,
+      "step": 6871
+    },
+    {
+      "epoch": 0.7801582993638934,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2398,
+      "step": 6872
+    },
+    {
+      "epoch": 0.7802718264738124,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2493,
+      "step": 6873
+    },
+    {
+      "epoch": 0.7803853535837315,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2278,
+      "step": 6874
+    },
+    {
+      "epoch": 0.7804988806936506,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2305,
+      "step": 6875
+    },
+    {
+      "epoch": 0.7806124078035697,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2213,
+      "step": 6876
+    },
+    {
+      "epoch": 0.7807259349134887,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.002,
+      "loss": 5.243,
+      "step": 6877
+    },
+    {
+      "epoch": 0.7808394620234079,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.002,
+      "loss": 5.2457,
+      "step": 6878
+    },
+    {
+      "epoch": 0.780952989133327,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.002,
+      "loss": 5.2353,
+      "step": 6879
+    },
+    {
+      "epoch": 0.7810665162432461,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.002,
+      "loss": 5.2473,
+      "step": 6880
+    },
+    {
+      "epoch": 0.7811800433531652,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.002,
+      "loss": 5.2453,
+      "step": 6881
+    },
+    {
+      "epoch": 0.7812935704630842,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2442,
+      "step": 6882
+    },
+    {
+      "epoch": 0.7814070975730033,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.002,
+      "loss": 5.2573,
+      "step": 6883
+    },
+    {
+      "epoch": 0.7815206246829224,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2277,
+      "step": 6884
+    },
+    {
+      "epoch": 0.7816341517928415,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2445,
+      "step": 6885
+    },
+    {
+      "epoch": 0.7817476789027605,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2392,
+      "step": 6886
+    },
+    {
+      "epoch": 0.7818612060126796,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2451,
+      "step": 6887
+    },
+    {
+      "epoch": 0.7819747331225987,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2335,
+      "step": 6888
+    },
+    {
+      "epoch": 0.7820882602325178,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2373,
+      "step": 6889
+    },
+    {
+      "epoch": 0.7822017873424368,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2308,
+      "step": 6890
+    },
+    {
+      "epoch": 0.7823153144523559,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2202,
+      "step": 6891
+    },
+    {
+      "epoch": 0.782428841562275,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2322,
+      "step": 6892
+    },
+    {
+      "epoch": 0.7825423686721941,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2481,
+      "step": 6893
+    },
+    {
+      "epoch": 0.7826558957821131,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2461,
+      "step": 6894
+    },
+    {
+      "epoch": 0.7827694228920322,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2435,
+      "step": 6895
+    },
+    {
+      "epoch": 0.7828829500019513,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2343,
+      "step": 6896
+    },
+    {
+      "epoch": 0.7829964771118704,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2486,
+      "step": 6897
+    },
+    {
+      "epoch": 0.7831100042217894,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2181,
+      "step": 6898
+    },
+    {
+      "epoch": 0.7832235313317085,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2498,
+      "step": 6899
+    },
+    {
+      "epoch": 0.7833370584416276,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2332,
+      "step": 6900
+    },
+    {
+      "epoch": 0.7834505855515467,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.002,
+      "loss": 5.2143,
+      "step": 6901
+    },
+    {
+      "epoch": 0.7835641126614658,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2363,
+      "step": 6902
+    },
+    {
+      "epoch": 0.7836776397713848,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.241,
+      "step": 6903
+    },
+    {
+      "epoch": 0.7837911668813039,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2391,
+      "step": 6904
+    },
+    {
+      "epoch": 0.783904693991223,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.245,
+      "step": 6905
+    },
+    {
+      "epoch": 0.784018221101142,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2314,
+      "step": 6906
+    },
+    {
+      "epoch": 0.7841317482110611,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.002,
+      "loss": 5.2241,
+      "step": 6907
+    },
+    {
+      "epoch": 0.7842452753209802,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.002,
+      "loss": 5.2169,
+      "step": 6908
+    },
+    {
+      "epoch": 0.7843588024308993,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.002,
+      "loss": 5.2409,
+      "step": 6909
+    },
+    {
+      "epoch": 0.7844723295408184,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.236,
+      "step": 6910
+    },
+    {
+      "epoch": 0.7845858566507374,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2588,
+      "step": 6911
+    },
+    {
+      "epoch": 0.7846993837606565,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2417,
+      "step": 6912
+    },
+    {
+      "epoch": 0.7848129108705756,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2552,
+      "step": 6913
+    },
+    {
+      "epoch": 0.7849264379804947,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2175,
+      "step": 6914
+    },
+    {
+      "epoch": 0.7850399650904137,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.2256,
+      "step": 6915
+    },
+    {
+      "epoch": 0.7851534922003328,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2154,
+      "step": 6916
+    },
+    {
+      "epoch": 0.7852670193102519,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2367,
+      "step": 6917
+    },
+    {
+      "epoch": 0.785380546420171,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2481,
+      "step": 6918
+    },
+    {
+      "epoch": 0.78549407353009,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2453,
+      "step": 6919
+    },
+    {
+      "epoch": 0.7856076006400091,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2562,
+      "step": 6920
+    },
+    {
+      "epoch": 0.7857211277499282,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2279,
+      "step": 6921
+    },
+    {
+      "epoch": 0.7858346548598473,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2364,
+      "step": 6922
+    },
+    {
+      "epoch": 0.7859481819697663,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2455,
+      "step": 6923
+    },
+    {
+      "epoch": 0.7860617090796854,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.237,
+      "step": 6924
+    },
+    {
+      "epoch": 0.7861752361896045,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2437,
+      "step": 6925
+    },
+    {
+      "epoch": 0.7862887632995236,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2476,
+      "step": 6926
+    },
+    {
+      "epoch": 0.7864022904094427,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2362,
+      "step": 6927
+    },
+    {
+      "epoch": 0.7865158175193617,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.248,
+      "step": 6928
+    },
+    {
+      "epoch": 0.7866293446292808,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2252,
+      "step": 6929
+    },
+    {
+      "epoch": 0.7867428717391999,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.002,
+      "loss": 5.2306,
+      "step": 6930
+    },
+    {
+      "epoch": 0.786856398849119,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.002,
+      "loss": 5.2265,
+      "step": 6931
+    },
+    {
+      "epoch": 0.786969925959038,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2217,
+      "step": 6932
+    },
+    {
+      "epoch": 0.7870834530689571,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2369,
+      "step": 6933
+    },
+    {
+      "epoch": 0.7871969801788762,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2235,
+      "step": 6934
+    },
+    {
+      "epoch": 0.7873105072887953,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2451,
+      "step": 6935
+    },
+    {
+      "epoch": 0.7874240343987143,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.002,
+      "loss": 5.2467,
+      "step": 6936
+    },
+    {
+      "epoch": 0.7875375615086334,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.2405,
+      "step": 6937
+    },
+    {
+      "epoch": 0.7876510886185525,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.2115,
+      "step": 6938
+    },
+    {
+      "epoch": 0.7877646157284716,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2136,
+      "step": 6939
+    },
+    {
+      "epoch": 0.7878781428383906,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2299,
+      "step": 6940
+    },
+    {
+      "epoch": 0.7879916699483097,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2345,
+      "step": 6941
+    },
+    {
+      "epoch": 0.7881051970582288,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2467,
+      "step": 6942
+    },
+    {
+      "epoch": 0.7882187241681479,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2191,
+      "step": 6943
+    },
+    {
+      "epoch": 0.7883322512780669,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2442,
+      "step": 6944
+    },
+    {
+      "epoch": 0.788445778387986,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2374,
+      "step": 6945
+    },
+    {
+      "epoch": 0.7885593054979051,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.002,
+      "loss": 5.2301,
+      "step": 6946
+    },
+    {
+      "epoch": 0.7886728326078242,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2383,
+      "step": 6947
+    },
+    {
+      "epoch": 0.7887863597177432,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2345,
+      "step": 6948
+    },
+    {
+      "epoch": 0.7888998868276623,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2313,
+      "step": 6949
+    },
+    {
+      "epoch": 0.7890134139375814,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.002,
+      "loss": 5.2563,
+      "step": 6950
+    },
+    {
+      "epoch": 0.7891269410475005,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.002,
+      "loss": 5.2591,
+      "step": 6951
+    },
+    {
+      "epoch": 0.7892404681574196,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2393,
+      "step": 6952
+    },
+    {
+      "epoch": 0.7893539952673386,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2363,
+      "step": 6953
+    },
+    {
+      "epoch": 0.7894675223772577,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.002,
+      "loss": 5.2446,
+      "step": 6954
+    },
+    {
+      "epoch": 0.7895810494871768,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2312,
+      "step": 6955
+    },
+    {
+      "epoch": 0.7896945765970959,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.226,
+      "step": 6956
+    },
+    {
+      "epoch": 0.7898081037070149,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2148,
+      "step": 6957
+    },
+    {
+      "epoch": 0.789921630816934,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2506,
+      "step": 6958
+    },
+    {
+      "epoch": 0.7900351579268531,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.234,
+      "step": 6959
+    },
+    {
+      "epoch": 0.7901486850367722,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2247,
+      "step": 6960
+    },
+    {
+      "epoch": 0.7902622121466912,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.2458,
+      "step": 6961
+    },
+    {
+      "epoch": 0.7903757392566103,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2491,
+      "step": 6962
+    },
+    {
+      "epoch": 0.7904892663665294,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.002,
+      "loss": 5.247,
+      "step": 6963
+    },
+    {
+      "epoch": 0.7906027934764485,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2535,
+      "step": 6964
+    },
+    {
+      "epoch": 0.7907163205863675,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.002,
+      "loss": 5.243,
+      "step": 6965
+    },
+    {
+      "epoch": 0.7908298476962866,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2401,
+      "step": 6966
+    },
+    {
+      "epoch": 0.7909433748062057,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2325,
+      "step": 6967
+    },
+    {
+      "epoch": 0.7910569019161248,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.2415,
+      "step": 6968
+    },
+    {
+      "epoch": 0.7911704290260438,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.002,
+      "loss": 5.2203,
+      "step": 6969
+    },
+    {
+      "epoch": 0.7912839561359629,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.2427,
+      "step": 6970
+    },
+    {
+      "epoch": 0.791397483245882,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2402,
+      "step": 6971
+    },
+    {
+      "epoch": 0.7915110103558011,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.002,
+      "loss": 5.2275,
+      "step": 6972
+    },
+    {
+      "epoch": 0.7916245374657201,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.002,
+      "loss": 5.2234,
+      "step": 6973
+    },
+    {
+      "epoch": 0.7917380645756392,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2497,
+      "step": 6974
+    },
+    {
+      "epoch": 0.7918515916855583,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.002,
+      "loss": 5.2208,
+      "step": 6975
+    },
+    {
+      "epoch": 0.7919651187954774,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.236,
+      "step": 6976
+    },
+    {
+      "epoch": 0.7920786459053965,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.002,
+      "loss": 5.2368,
+      "step": 6977
+    },
+    {
+      "epoch": 0.7921921730153155,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.002,
+      "loss": 5.2332,
+      "step": 6978
+    },
+    {
+      "epoch": 0.7923057001252346,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.002,
+      "loss": 5.2314,
+      "step": 6979
+    },
+    {
+      "epoch": 0.7924192272351537,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2378,
+      "step": 6980
+    },
+    {
+      "epoch": 0.7925327543450728,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.2369,
+      "step": 6981
+    },
+    {
+      "epoch": 0.7926462814549918,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2576,
+      "step": 6982
+    },
+    {
+      "epoch": 0.7927598085649109,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.002,
+      "loss": 5.2327,
+      "step": 6983
+    },
+    {
+      "epoch": 0.79287333567483,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.002,
+      "loss": 5.2248,
+      "step": 6984
+    },
+    {
+      "epoch": 0.7929868627847491,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.002,
+      "loss": 5.2396,
+      "step": 6985
+    },
+    {
+      "epoch": 0.7931003898946681,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.2326,
+      "step": 6986
+    },
+    {
+      "epoch": 0.7932139170045872,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.002,
+      "loss": 5.2195,
+      "step": 6987
+    },
+    {
+      "epoch": 0.7933274441145063,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.002,
+      "loss": 5.2251,
+      "step": 6988
+    },
+    {
+      "epoch": 0.7934409712244254,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2463,
+      "step": 6989
+    },
+    {
+      "epoch": 0.7935544983343444,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2387,
+      "step": 6990
+    },
+    {
+      "epoch": 0.7936680254442635,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.262,
+      "step": 6991
+    },
+    {
+      "epoch": 0.7937815525541826,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.002,
+      "loss": 5.2209,
+      "step": 6992
+    },
+    {
+      "epoch": 0.7938950796641017,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.223,
+      "step": 6993
+    },
+    {
+      "epoch": 0.7940086067740207,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2487,
+      "step": 6994
+    },
+    {
+      "epoch": 0.7941221338839398,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.002,
+      "loss": 5.2105,
+      "step": 6995
+    },
+    {
+      "epoch": 0.7942356609938589,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.002,
+      "loss": 5.2564,
+      "step": 6996
+    },
+    {
+      "epoch": 0.794349188103778,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.002,
+      "loss": 5.2414,
+      "step": 6997
+    },
+    {
+      "epoch": 0.794462715213697,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.002,
+      "loss": 5.2365,
+      "step": 6998
+    },
+    {
+      "epoch": 0.7945762423236161,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.002,
+      "loss": 5.2439,
+      "step": 6999
+    },
+    {
+      "epoch": 0.7946897694335352,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2292,
+      "step": 7000
+    },
+    {
+      "epoch": 0.7948032965434543,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2461,
+      "step": 7001
+    },
+    {
+      "epoch": 0.7949168236533734,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.002,
+      "loss": 5.246,
+      "step": 7002
+    },
+    {
+      "epoch": 0.7950303507632924,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2428,
+      "step": 7003
+    },
+    {
+      "epoch": 0.7951438778732115,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2487,
+      "step": 7004
+    },
+    {
+      "epoch": 0.7952574049831306,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.002,
+      "loss": 5.2452,
+      "step": 7005
+    },
+    {
+      "epoch": 0.7953709320930497,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2282,
+      "step": 7006
+    },
+    {
+      "epoch": 0.7954844592029687,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2371,
+      "step": 7007
+    },
+    {
+      "epoch": 0.7955979863128878,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.002,
+      "loss": 5.2197,
+      "step": 7008
+    },
+    {
+      "epoch": 0.7957115134228069,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.002,
+      "loss": 5.216,
+      "step": 7009
+    },
+    {
+      "epoch": 0.795825040532726,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.002,
+      "loss": 5.2231,
+      "step": 7010
+    },
+    {
+      "epoch": 0.795938567642645,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2396,
+      "step": 7011
+    },
+    {
+      "epoch": 0.7960520947525641,
+      "grad_norm": 0.375,
+      "learning_rate": 0.002,
+      "loss": 5.2239,
+      "step": 7012
+    },
+    {
+      "epoch": 0.7961656218624832,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.002,
+      "loss": 5.2562,
+      "step": 7013
+    },
+    {
+      "epoch": 0.7962791489724023,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.002,
+      "loss": 5.248,
+      "step": 7014
+    },
+    {
+      "epoch": 0.7963926760823213,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.002,
+      "loss": 5.2318,
+      "step": 7015
+    },
+    {
+      "epoch": 0.7965062031922404,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.2435,
+      "step": 7016
+    },
+    {
+      "epoch": 0.7966197303021595,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2274,
+      "step": 7017
+    },
+    {
+      "epoch": 0.7967332574120786,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2473,
+      "step": 7018
+    },
+    {
+      "epoch": 0.7968467845219976,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2277,
+      "step": 7019
+    },
+    {
+      "epoch": 0.7969603116319167,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.002,
+      "loss": 5.2257,
+      "step": 7020
+    },
+    {
+      "epoch": 0.7970738387418358,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.002,
+      "loss": 5.2436,
+      "step": 7021
+    },
+    {
+      "epoch": 0.7971873658517549,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.249,
+      "step": 7022
+    },
+    {
+      "epoch": 0.797300892961674,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.002,
+      "loss": 5.251,
+      "step": 7023
+    },
+    {
+      "epoch": 0.797414420071593,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.232,
+      "step": 7024
+    },
+    {
+      "epoch": 0.7975279471815121,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2185,
+      "step": 7025
+    },
+    {
+      "epoch": 0.7976414742914312,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2482,
+      "step": 7026
+    },
+    {
+      "epoch": 0.7977550014013502,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2369,
+      "step": 7027
+    },
+    {
+      "epoch": 0.7978685285112693,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2301,
+      "step": 7028
+    },
+    {
+      "epoch": 0.7979820556211884,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.002,
+      "loss": 5.2426,
+      "step": 7029
+    },
+    {
+      "epoch": 0.7980955827311075,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.002,
+      "loss": 5.2391,
+      "step": 7030
+    },
+    {
+      "epoch": 0.7982091098410266,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.002,
+      "loss": 5.217,
+      "step": 7031
+    },
+    {
+      "epoch": 0.7983226369509456,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.002,
+      "loss": 5.2424,
+      "step": 7032
+    },
+    {
+      "epoch": 0.7984361640608647,
+      "grad_norm": 0.25,
+      "learning_rate": 0.002,
+      "loss": 5.2188,
+      "step": 7033
+    },
+    {
+      "epoch": 0.7985496911707838,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.002,
+      "loss": 5.2291,
+      "step": 7034
+    },
+    {
+      "epoch": 0.7986632182807029,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.002,
+      "loss": 5.2345,
+      "step": 7035
+    },
+    {
+      "epoch": 0.7987767453906219,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.002,
+      "loss": 5.2464,
+      "step": 7036
+    },
+    {
+      "epoch": 0.798890272500541,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.002,
+      "loss": 5.2364,
+      "step": 7037
+    },
+    {
+      "epoch": 0.7990037996104601,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.002,
+      "loss": 5.2175,
+      "step": 7038
+    },
+    {
+      "epoch": 0.7991173267203792,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.002,
+      "loss": 5.2395,
+      "step": 7039
+    },
+    {
+      "epoch": 0.7992308538302982,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.002,
+      "loss": 5.1986,
+      "step": 7040
+    },
+    {
+      "epoch": 0.7993443809402173,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2579,
+      "step": 7041
+    },
+    {
+      "epoch": 0.7994579080501364,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.002,
+      "loss": 5.255,
+      "step": 7042
+    },
+    {
+      "epoch": 0.7995714351600555,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.002,
+      "loss": 5.2412,
+      "step": 7043
+    },
+    {
+      "epoch": 0.7996849622699745,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.002,
+      "loss": 5.2219,
+      "step": 7044
+    },
+    {
+      "epoch": 0.7997984893798936,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.002,
+      "loss": 5.232,
+      "step": 7045
+    },
+    {
+      "epoch": 0.7999120164898127,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.002,
+      "loss": 5.2227,
+      "step": 7046
+    },
+    {
+      "epoch": 0.8000255435997318,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.002,
+      "loss": 5.2288,
+      "step": 7047
+    },
+    {
+      "epoch": 0.8001390707096508,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.002,
+      "loss": 5.2299,
+      "step": 7048
+    },
+    {
+      "epoch": 0.8002525978195699,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.002,
+      "loss": 5.2249,
+      "step": 7049
+    },
+    {
+      "epoch": 0.800366124929489,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.002,
+      "loss": 5.2269,
+      "step": 7050
+    },
+    {
+      "epoch": 0.8004796520394081,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2385,
+      "step": 7051
+    },
+    {
+      "epoch": 0.8005931791493271,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.002,
+      "loss": 5.2385,
+      "step": 7052
+    },
+    {
+      "epoch": 0.8007067062592462,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.002,
+      "loss": 5.2264,
+      "step": 7053
+    },
+    {
+      "epoch": 0.8008202333691653,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.002,
+      "loss": 5.2482,
+      "step": 7054
+    },
+    {
+      "epoch": 0.8009337604790844,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.002,
+      "loss": 5.2317,
+      "step": 7055
+    },
+    {
+      "epoch": 0.8010472875890035,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.002,
+      "loss": 5.249,
+      "step": 7056
+    },
+    {
+      "epoch": 0.8011608146989225,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.002,
+      "loss": 5.2467,
+      "step": 7057
+    },
+    {
+      "epoch": 0.8012743418088416,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.002,
+      "loss": 5.2354,
+      "step": 7058
+    },
+    {
+      "epoch": 0.8013878689187607,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0019999985497728043,
+      "loss": 5.2593,
+      "step": 7059
+    },
+    {
+      "epoch": 0.8015013960286798,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0019999941990958897,
+      "loss": 5.2228,
+      "step": 7060
+    },
+    {
+      "epoch": 0.8016149231385988,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0019999869479832783,
+      "loss": 5.2192,
+      "step": 7061
+    },
+    {
+      "epoch": 0.8017284502485179,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0019999767964583377,
+      "loss": 5.2464,
+      "step": 7062
+    },
+    {
+      "epoch": 0.801841977358437,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.001999963744553784,
+      "loss": 5.235,
+      "step": 7063
+    },
+    {
+      "epoch": 0.8019555044683561,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00199994779231168,
+      "loss": 5.2344,
+      "step": 7064
+    },
+    {
+      "epoch": 0.8020690315782751,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0019999289397834344,
+      "loss": 5.251,
+      "step": 7065
+    },
+    {
+      "epoch": 0.8021825586881942,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0019999071870298053,
+      "loss": 5.2342,
+      "step": 7066
+    },
+    {
+      "epoch": 0.8022960857981133,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0019998825341208943,
+      "loss": 5.2311,
+      "step": 7067
+    },
+    {
+      "epoch": 0.8024096129080324,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0019998549811361525,
+      "loss": 5.2038,
+      "step": 7068
+    },
+    {
+      "epoch": 0.8025231400179514,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0019998245281643745,
+      "loss": 5.2382,
+      "step": 7069
+    },
+    {
+      "epoch": 0.8026366671278705,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0019997911753037024,
+      "loss": 5.2386,
+      "step": 7070
+    },
+    {
+      "epoch": 0.8027501942377896,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0019997549226616236,
+      "loss": 5.208,
+      "step": 7071
+    },
+    {
+      "epoch": 0.8028637213477087,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00199971577035497,
+      "loss": 5.2324,
+      "step": 7072
+    },
+    {
+      "epoch": 0.8029772484576277,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.001999673718509919,
+      "loss": 5.2286,
+      "step": 7073
+    },
+    {
+      "epoch": 0.8030907755675468,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0019996287672619925,
+      "loss": 5.2365,
+      "step": 7074
+    },
+    {
+      "epoch": 0.8032043026774659,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.001999580916756055,
+      "loss": 5.2468,
+      "step": 7075
+    },
+    {
+      "epoch": 0.803317829787385,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0019995301671463174,
+      "loss": 5.2252,
+      "step": 7076
+    },
+    {
+      "epoch": 0.803431356897304,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0019994765185963304,
+      "loss": 5.216,
+      "step": 7077
+    },
+    {
+      "epoch": 0.8035448840072231,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.001999419971278989,
+      "loss": 5.2266,
+      "step": 7078
+    },
+    {
+      "epoch": 0.8036584111171422,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0019993605253765304,
+      "loss": 5.2387,
+      "step": 7079
+    },
+    {
+      "epoch": 0.8037719382270613,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0019992981810805317,
+      "loss": 5.205,
+      "step": 7080
+    },
+    {
+      "epoch": 0.8038854653369804,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0019992329385919117,
+      "loss": 5.2228,
+      "step": 7081
+    },
+    {
+      "epoch": 0.8039989924468994,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.00199916479812093,
+      "loss": 5.2332,
+      "step": 7082
+    },
+    {
+      "epoch": 0.8041125195568185,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.001999093759887183,
+      "loss": 5.2142,
+      "step": 7083
+    },
+    {
+      "epoch": 0.8042260466667376,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0019990198241196092,
+      "loss": 5.2272,
+      "step": 7084
+    },
+    {
+      "epoch": 0.8043395737766567,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0019989429910564826,
+      "loss": 5.248,
+      "step": 7085
+    },
+    {
+      "epoch": 0.8044531008865757,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.001998863260945416,
+      "loss": 5.2242,
+      "step": 7086
+    },
+    {
+      "epoch": 0.8045666279964948,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0019987806340433564,
+      "loss": 5.2396,
+      "step": 7087
+    },
+    {
+      "epoch": 0.8046801551064139,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.001998695110616589,
+      "loss": 5.2022,
+      "step": 7088
+    },
+    {
+      "epoch": 0.804793682216333,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0019986066909407317,
+      "loss": 5.2301,
+      "step": 7089
+    },
+    {
+      "epoch": 0.804907209326252,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0019985153753007373,
+      "loss": 5.213,
+      "step": 7090
+    },
+    {
+      "epoch": 0.8050207364361711,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0019984211639908914,
+      "loss": 5.2323,
+      "step": 7091
+    },
+    {
+      "epoch": 0.8051342635460902,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.001998324057314811,
+      "loss": 5.25,
+      "step": 7092
+    },
+    {
+      "epoch": 0.8052477906560093,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0019982240555854445,
+      "loss": 5.2288,
+      "step": 7093
+    },
+    {
+      "epoch": 0.8053613177659283,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00199812115912507,
+      "loss": 5.2443,
+      "step": 7094
+    },
+    {
+      "epoch": 0.8054748448758474,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.001998015368265295,
+      "loss": 5.2326,
+      "step": 7095
+    },
+    {
+      "epoch": 0.8055883719857665,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.001997906683347055,
+      "loss": 5.2554,
+      "step": 7096
+    },
+    {
+      "epoch": 0.8057018990956856,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0019977951047206108,
+      "loss": 5.2406,
+      "step": 7097
+    },
+    {
+      "epoch": 0.8058154262056046,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0019976806327455508,
+      "loss": 5.2463,
+      "step": 7098
+    },
+    {
+      "epoch": 0.8059289533155237,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.001997563267790786,
+      "loss": 5.2427,
+      "step": 7099
+    },
+    {
+      "epoch": 0.8060424804254428,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0019974430102345526,
+      "loss": 5.2501,
+      "step": 7100
+    },
+    {
+      "epoch": 0.8061560075353619,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.001997319860464407,
+      "loss": 5.2372,
+      "step": 7101
+    },
+    {
+      "epoch": 0.806269534645281,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.001997193818877228,
+      "loss": 5.2439,
+      "step": 7102
+    },
+    {
+      "epoch": 0.8063830617552,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.001997064885879213,
+      "loss": 5.2292,
+      "step": 7103
+    },
+    {
+      "epoch": 0.8064965888651191,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0019969330618858777,
+      "loss": 5.2216,
+      "step": 7104
+    },
+    {
+      "epoch": 0.8066101159750382,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0019967983473220554,
+      "loss": 5.2187,
+      "step": 7105
+    },
+    {
+      "epoch": 0.8067236430849573,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0019966607426218936,
+      "loss": 5.2221,
+      "step": 7106
+    },
+    {
+      "epoch": 0.8068371701948763,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0019965202482288553,
+      "loss": 5.2452,
+      "step": 7107
+    },
+    {
+      "epoch": 0.8069506973047954,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.001996376864595715,
+      "loss": 5.2446,
+      "step": 7108
+    },
+    {
+      "epoch": 0.8070642244147145,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0019962305921845595,
+      "loss": 5.2246,
+      "step": 7109
+    },
+    {
+      "epoch": 0.8071777515246336,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.001996081431466785,
+      "loss": 5.2344,
+      "step": 7110
+    },
+    {
+      "epoch": 0.8072912786345526,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.001995929382923095,
+      "loss": 5.2461,
+      "step": 7111
+    },
+    {
+      "epoch": 0.8074048057444717,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0019957744470435017,
+      "loss": 5.2463,
+      "step": 7112
+    },
+    {
+      "epoch": 0.8075183328543908,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0019956166243273203,
+      "loss": 5.2233,
+      "step": 7113
+    },
+    {
+      "epoch": 0.8076318599643099,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0019954559152831706,
+      "loss": 5.2364,
+      "step": 7114
+    },
+    {
+      "epoch": 0.8077453870742289,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.001995292320428973,
+      "loss": 5.2483,
+      "step": 7115
+    },
+    {
+      "epoch": 0.807858914184148,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.001995125840291951,
+      "loss": 5.2371,
+      "step": 7116
+    },
+    {
+      "epoch": 0.8079724412940671,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.001994956475408623,
+      "loss": 5.2271,
+      "step": 7117
+    },
+    {
+      "epoch": 0.8080859684039862,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0019947842263248064,
+      "loss": 5.2429,
+      "step": 7118
+    },
+    {
+      "epoch": 0.8081994955139052,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.001994609093595613,
+      "loss": 5.2098,
+      "step": 7119
+    },
+    {
+      "epoch": 0.8083130226238244,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.001994431077785448,
+      "loss": 5.2377,
+      "step": 7120
+    },
+    {
+      "epoch": 0.8084265497337435,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0019942501794680077,
+      "loss": 5.2504,
+      "step": 7121
+    },
+    {
+      "epoch": 0.8085400768436626,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0019940663992262775,
+      "loss": 5.2359,
+      "step": 7122
+    },
+    {
+      "epoch": 0.8086536039535817,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0019938797376525316,
+      "loss": 5.2292,
+      "step": 7123
+    },
+    {
+      "epoch": 0.8087671310635007,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.001993690195348329,
+      "loss": 5.2436,
+      "step": 7124
+    },
+    {
+      "epoch": 0.8088806581734198,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.001993497772924513,
+      "loss": 5.226,
+      "step": 7125
+    },
+    {
+      "epoch": 0.8089941852833389,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0019933024710012083,
+      "loss": 5.2094,
+      "step": 7126
+    },
+    {
+      "epoch": 0.809107712393258,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00199310429020782,
+      "loss": 5.2277,
+      "step": 7127
+    },
+    {
+      "epoch": 0.809221239503177,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00199290323118303,
+      "loss": 5.2299,
+      "step": 7128
+    },
+    {
+      "epoch": 0.8093347666130961,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.001992699294574798,
+      "loss": 5.2419,
+      "step": 7129
+    },
+    {
+      "epoch": 0.8094482937230152,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0019924924810403545,
+      "loss": 5.2246,
+      "step": 7130
+    },
+    {
+      "epoch": 0.8095618208329343,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0019922827912462046,
+      "loss": 5.2314,
+      "step": 7131
+    },
+    {
+      "epoch": 0.8096753479428533,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0019920702258681196,
+      "loss": 5.2273,
+      "step": 7132
+    },
+    {
+      "epoch": 0.8097888750527724,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.001991854785591141,
+      "loss": 5.2327,
+      "step": 7133
+    },
+    {
+      "epoch": 0.8099024021626915,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0019916364711095743,
+      "loss": 5.2324,
+      "step": 7134
+    },
+    {
+      "epoch": 0.8100159292726106,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.001991415283126986,
+      "loss": 5.2381,
+      "step": 7135
+    },
+    {
+      "epoch": 0.8101294563825296,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0019911912223562057,
+      "loss": 5.2377,
+      "step": 7136
+    },
+    {
+      "epoch": 0.8102429834924487,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00199096428951932,
+      "loss": 5.2477,
+      "step": 7137
+    },
+    {
+      "epoch": 0.8103565106023678,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0019907344853476714,
+      "loss": 5.2349,
+      "step": 7138
+    },
+    {
+      "epoch": 0.8104700377122869,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.001990501810581856,
+      "loss": 5.2307,
+      "step": 7139
+    },
+    {
+      "epoch": 0.810583564822206,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.001990266265971721,
+      "loss": 5.2297,
+      "step": 7140
+    },
+    {
+      "epoch": 0.810697091932125,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0019900278522763624,
+      "loss": 5.2133,
+      "step": 7141
+    },
+    {
+      "epoch": 0.8108106190420441,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0019897865702641227,
+      "loss": 5.244,
+      "step": 7142
+    },
+    {
+      "epoch": 0.8109241461519632,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.001989542420712588,
+      "loss": 5.2312,
+      "step": 7143
+    },
+    {
+      "epoch": 0.8110376732618823,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.001989295404408585,
+      "loss": 5.2455,
+      "step": 7144
+    },
+    {
+      "epoch": 0.8111512003718013,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0019890455221481806,
+      "loss": 5.2388,
+      "step": 7145
+    },
+    {
+      "epoch": 0.8112647274817204,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.001988792774736677,
+      "loss": 5.2303,
+      "step": 7146
+    },
+    {
+      "epoch": 0.8113782545916395,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.00198853716298861,
+      "loss": 5.2365,
+      "step": 7147
+    },
+    {
+      "epoch": 0.8114917817015586,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0019882786877277463,
+      "loss": 5.2388,
+      "step": 7148
+    },
+    {
+      "epoch": 0.8116053088114776,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.001988017349787081,
+      "loss": 5.2471,
+      "step": 7149
+    },
+    {
+      "epoch": 0.8117188359213967,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.001987753150008836,
+      "loss": 5.2396,
+      "step": 7150
+    },
+    {
+      "epoch": 0.8118323630313158,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0019874860892444544,
+      "loss": 5.2562,
+      "step": 7151
+    },
+    {
+      "epoch": 0.8119458901412349,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0019872161683545998,
+      "loss": 5.2307,
+      "step": 7152
+    },
+    {
+      "epoch": 0.8120594172511539,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.001986943388209154,
+      "loss": 5.2181,
+      "step": 7153
+    },
+    {
+      "epoch": 0.812172944361073,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.001986667749687213,
+      "loss": 5.2511,
+      "step": 7154
+    },
+    {
+      "epoch": 0.8122864714709921,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0019863892536770844,
+      "loss": 5.2599,
+      "step": 7155
+    },
+    {
+      "epoch": 0.8123999985809112,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.001986107901076285,
+      "loss": 5.2153,
+      "step": 7156
+    },
+    {
+      "epoch": 0.8125135256908302,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.001985823692791537,
+      "loss": 5.2369,
+      "step": 7157
+    },
+    {
+      "epoch": 0.8126270528007493,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0019855366297387668,
+      "loss": 5.2278,
+      "step": 7158
+    },
+    {
+      "epoch": 0.8127405799106684,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0019852467128430996,
+      "loss": 5.2292,
+      "step": 7159
+    },
+    {
+      "epoch": 0.8128541070205875,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.001984953943038859,
+      "loss": 5.2276,
+      "step": 7160
+    },
+    {
+      "epoch": 0.8129676341305065,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.001984658321269562,
+      "loss": 5.258,
+      "step": 7161
+    },
+    {
+      "epoch": 0.8130811612404256,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.001984359848487917,
+      "loss": 5.2062,
+      "step": 7162
+    },
+    {
+      "epoch": 0.8131946883503447,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00198405852565582,
+      "loss": 5.2148,
+      "step": 7163
+    },
+    {
+      "epoch": 0.8133082154602638,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0019837543537443523,
+      "loss": 5.2288,
+      "step": 7164
+    },
+    {
+      "epoch": 0.8134217425701828,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0019834473337337773,
+      "loss": 5.243,
+      "step": 7165
+    },
+    {
+      "epoch": 0.8135352696801019,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0019831374666135363,
+      "loss": 5.2008,
+      "step": 7166
+    },
+    {
+      "epoch": 0.813648796790021,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0019828247533822466,
+      "loss": 5.215,
+      "step": 7167
+    },
+    {
+      "epoch": 0.8137623238999401,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.001982509195047698,
+      "loss": 5.2385,
+      "step": 7168
+    },
+    {
+      "epoch": 0.8138758510098592,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0019821907926268487,
+      "loss": 5.2461,
+      "step": 7169
+    },
+    {
+      "epoch": 0.8139893781197782,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.001981869547145822,
+      "loss": 5.2426,
+      "step": 7170
+    },
+    {
+      "epoch": 0.8141029052296973,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.001981545459639906,
+      "loss": 5.2326,
+      "step": 7171
+    },
+    {
+      "epoch": 0.8142164323396164,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0019812185311535446,
+      "loss": 5.2196,
+      "step": 7172
+    },
+    {
+      "epoch": 0.8143299594495355,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0019808887627403406,
+      "loss": 5.2386,
+      "step": 7173
+    },
+    {
+      "epoch": 0.8144434865594545,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.001980556155463046,
+      "loss": 5.256,
+      "step": 7174
+    },
+    {
+      "epoch": 0.8145570136693736,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0019802207103935647,
+      "loss": 5.2262,
+      "step": 7175
+    },
+    {
+      "epoch": 0.8146705407792927,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0019798824286129443,
+      "loss": 5.2309,
+      "step": 7176
+    },
+    {
+      "epoch": 0.8147840678892118,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0019795413112113744,
+      "loss": 5.2336,
+      "step": 7177
+    },
+    {
+      "epoch": 0.8148975949991308,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.001979197359288183,
+      "loss": 5.2155,
+      "step": 7178
+    },
+    {
+      "epoch": 0.8150111221090499,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.001978850573951834,
+      "loss": 5.2059,
+      "step": 7179
+    },
+    {
+      "epoch": 0.815124649218969,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.001978500956319921,
+      "loss": 5.246,
+      "step": 7180
+    },
+    {
+      "epoch": 0.8152381763288881,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0019781485075191668,
+      "loss": 5.2322,
+      "step": 7181
+    },
+    {
+      "epoch": 0.8153517034388071,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0019777932286854185,
+      "loss": 5.2339,
+      "step": 7182
+    },
+    {
+      "epoch": 0.8154652305487262,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.001977435120963641,
+      "loss": 5.2343,
+      "step": 7183
+    },
+    {
+      "epoch": 0.8155787576586453,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0019770741855079195,
+      "loss": 5.2315,
+      "step": 7184
+    },
+    {
+      "epoch": 0.8156922847685644,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.001976710423481449,
+      "loss": 5.2266,
+      "step": 7185
+    },
+    {
+      "epoch": 0.8158058118784834,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0019763438360565367,
+      "loss": 5.238,
+      "step": 7186
+    },
+    {
+      "epoch": 0.8159193389884025,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0019759744244145926,
+      "loss": 5.2134,
+      "step": 7187
+    },
+    {
+      "epoch": 0.8160328660983216,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0019756021897461306,
+      "loss": 5.2253,
+      "step": 7188
+    },
+    {
+      "epoch": 0.8161463932082407,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0019752271332507607,
+      "loss": 5.2367,
+      "step": 7189
+    },
+    {
+      "epoch": 0.8162599203181597,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0019748492561371877,
+      "loss": 5.246,
+      "step": 7190
+    },
+    {
+      "epoch": 0.8163734474280788,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0019744685596232068,
+      "loss": 5.213,
+      "step": 7191
+    },
+    {
+      "epoch": 0.8164869745379979,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0019740850449356984,
+      "loss": 5.2335,
+      "step": 7192
+    },
+    {
+      "epoch": 0.816600501647917,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.001973698713310626,
+      "loss": 5.2367,
+      "step": 7193
+    },
+    {
+      "epoch": 0.816714028757836,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.001973309565993032,
+      "loss": 5.2401,
+      "step": 7194
+    },
+    {
+      "epoch": 0.8168275558677551,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00197291760423703,
+      "loss": 5.2281,
+      "step": 7195
+    },
+    {
+      "epoch": 0.8169410829776742,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0019725228293058076,
+      "loss": 5.2098,
+      "step": 7196
+    },
+    {
+      "epoch": 0.8170546100875933,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.001972125242471615,
+      "loss": 5.2306,
+      "step": 7197
+    },
+    {
+      "epoch": 0.8171681371975124,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.001971724845015768,
+      "loss": 5.2341,
+      "step": 7198
+    },
+    {
+      "epoch": 0.8172816643074314,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.001971321638228637,
+      "loss": 5.2291,
+      "step": 7199
+    },
+    {
+      "epoch": 0.8173951914173505,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.001970915623409648,
+      "loss": 5.2319,
+      "step": 7200
+    },
+    {
+      "epoch": 0.8175087185272696,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0019705068018672747,
+      "loss": 5.2428,
+      "step": 7201
+    },
+    {
+      "epoch": 0.8176222456371887,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.001970095174919039,
+      "loss": 5.2308,
+      "step": 7202
+    },
+    {
+      "epoch": 0.8177357727471077,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0019696807438915015,
+      "loss": 5.2295,
+      "step": 7203
+    },
+    {
+      "epoch": 0.8178492998570268,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0019692635101202604,
+      "loss": 5.2513,
+      "step": 7204
+    },
+    {
+      "epoch": 0.8179628269669459,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0019688434749499466,
+      "loss": 5.2363,
+      "step": 7205
+    },
+    {
+      "epoch": 0.818076354076865,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.001968420639734218,
+      "loss": 5.2003,
+      "step": 7206
+    },
+    {
+      "epoch": 0.818189881186784,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0019679950058357583,
+      "loss": 5.2274,
+      "step": 7207
+    },
+    {
+      "epoch": 0.8183034082967031,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.001967566574626268,
+      "loss": 5.2399,
+      "step": 7208
+    },
+    {
+      "epoch": 0.8184169354066222,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.001967135347486465,
+      "loss": 5.2159,
+      "step": 7209
+    },
+    {
+      "epoch": 0.8185304625165413,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.001966701325806076,
+      "loss": 5.2317,
+      "step": 7210
+    },
+    {
+      "epoch": 0.8186439896264603,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0019662645109838352,
+      "loss": 5.2416,
+      "step": 7211
+    },
+    {
+      "epoch": 0.8187575167363794,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.001965824904427477,
+      "loss": 5.2259,
+      "step": 7212
+    },
+    {
+      "epoch": 0.8188710438462985,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.001965382507553734,
+      "loss": 5.2457,
+      "step": 7213
+    },
+    {
+      "epoch": 0.8189845709562176,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00196493732178833,
+      "loss": 5.2565,
+      "step": 7214
+    },
+    {
+      "epoch": 0.8190980980661366,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.001964489348565977,
+      "loss": 5.2399,
+      "step": 7215
+    },
+    {
+      "epoch": 0.8192116251760557,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0019640385893303705,
+      "loss": 5.2222,
+      "step": 7216
+    },
+    {
+      "epoch": 0.8193251522859748,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0019635850455341857,
+      "loss": 5.228,
+      "step": 7217
+    },
+    {
+      "epoch": 0.8194386793958939,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.001963128718639069,
+      "loss": 5.229,
+      "step": 7218
+    },
+    {
+      "epoch": 0.819552206505813,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0019626696101156384,
+      "loss": 5.2476,
+      "step": 7219
+    },
+    {
+      "epoch": 0.819665733615732,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.001962207721443475,
+      "loss": 5.2394,
+      "step": 7220
+    },
+    {
+      "epoch": 0.8197792607256511,
+      "grad_norm": 0.25,
+      "learning_rate": 0.00196174305411112,
+      "loss": 5.2235,
+      "step": 7221
+    },
+    {
+      "epoch": 0.8198927878355702,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0019612756096160687,
+      "loss": 5.2381,
+      "step": 7222
+    },
+    {
+      "epoch": 0.8200063149454893,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0019608053894647685,
+      "loss": 5.2386,
+      "step": 7223
+    },
+    {
+      "epoch": 0.8201198420554083,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00196033239517261,
+      "loss": 5.2171,
+      "step": 7224
+    },
+    {
+      "epoch": 0.8202333691653274,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0019598566282639242,
+      "loss": 5.253,
+      "step": 7225
+    },
+    {
+      "epoch": 0.8203468962752465,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.001959378090271979,
+      "loss": 5.2352,
+      "step": 7226
+    },
+    {
+      "epoch": 0.8204604233851656,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0019588967827389703,
+      "loss": 5.233,
+      "step": 7227
+    },
+    {
+      "epoch": 0.8205739504950846,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.001958412707216023,
+      "loss": 5.2228,
+      "step": 7228
+    },
+    {
+      "epoch": 0.8206874776050037,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.001957925865263179,
+      "loss": 5.2065,
+      "step": 7229
+    },
+    {
+      "epoch": 0.8208010047149228,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0019574362584493968,
+      "loss": 5.2348,
+      "step": 7230
+    },
+    {
+      "epoch": 0.8209145318248419,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0019569438883525465,
+      "loss": 5.2241,
+      "step": 7231
+    },
+    {
+      "epoch": 0.8210280589347609,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.001956448756559402,
+      "loss": 5.2316,
+      "step": 7232
+    },
+    {
+      "epoch": 0.82114158604468,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0019559508646656383,
+      "loss": 5.2187,
+      "step": 7233
+    },
+    {
+      "epoch": 0.8212551131545991,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0019554502142758246,
+      "loss": 5.242,
+      "step": 7234
+    },
+    {
+      "epoch": 0.8213686402645182,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.001954946807003421,
+      "loss": 5.2443,
+      "step": 7235
+    },
+    {
+      "epoch": 0.8214821673744372,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0019544406444707715,
+      "loss": 5.2319,
+      "step": 7236
+    },
+    {
+      "epoch": 0.8215956944843563,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0019539317283091,
+      "loss": 5.2334,
+      "step": 7237
+    },
+    {
+      "epoch": 0.8217092215942754,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0019534200601585036,
+      "loss": 5.2381,
+      "step": 7238
+    },
+    {
+      "epoch": 0.8218227487041945,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0019529056416679506,
+      "loss": 5.243,
+      "step": 7239
+    },
+    {
+      "epoch": 0.8219362758141135,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0019523884744952701,
+      "loss": 5.2346,
+      "step": 7240
+    },
+    {
+      "epoch": 0.8220498029240326,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0019518685603071512,
+      "loss": 5.2412,
+      "step": 7241
+    },
+    {
+      "epoch": 0.8221633300339517,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0019513459007791355,
+      "loss": 5.2575,
+      "step": 7242
+    },
+    {
+      "epoch": 0.8222768571438708,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0019508204975956119,
+      "loss": 5.2204,
+      "step": 7243
+    },
+    {
+      "epoch": 0.8223903842537899,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.001950292352449812,
+      "loss": 5.2482,
+      "step": 7244
+    },
+    {
+      "epoch": 0.8225039113637089,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0019497614670438027,
+      "loss": 5.2401,
+      "step": 7245
+    },
+    {
+      "epoch": 0.822617438473628,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0019492278430884833,
+      "loss": 5.2485,
+      "step": 7246
+    },
+    {
+      "epoch": 0.8227309655835471,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0019486914823035782,
+      "loss": 5.2242,
+      "step": 7247
+    },
+    {
+      "epoch": 0.8228444926934662,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0019481523864176313,
+      "loss": 5.2347,
+      "step": 7248
+    },
+    {
+      "epoch": 0.8229580198033852,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.001947610557168002,
+      "loss": 5.2267,
+      "step": 7249
+    },
+    {
+      "epoch": 0.8230715469133043,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0019470659963008582,
+      "loss": 5.231,
+      "step": 7250
+    },
+    {
+      "epoch": 0.8231850740232234,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0019465187055711707,
+      "loss": 5.2222,
+      "step": 7251
+    },
+    {
+      "epoch": 0.8232986011331425,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.001945968686742708,
+      "loss": 5.2463,
+      "step": 7252
+    },
+    {
+      "epoch": 0.8234121282430615,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0019454159415880313,
+      "loss": 5.2299,
+      "step": 7253
+    },
+    {
+      "epoch": 0.8235256553529806,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0019448604718884867,
+      "loss": 5.2249,
+      "step": 7254
+    },
+    {
+      "epoch": 0.8236391824628997,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0019443022794342019,
+      "loss": 5.2376,
+      "step": 7255
+    },
+    {
+      "epoch": 0.8237527095728188,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0019437413660240786,
+      "loss": 5.2104,
+      "step": 7256
+    },
+    {
+      "epoch": 0.8238662366827378,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.001943177733465788,
+      "loss": 5.2189,
+      "step": 7257
+    },
+    {
+      "epoch": 0.8239797637926569,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0019426113835757635,
+      "loss": 5.2393,
+      "step": 7258
+    },
+    {
+      "epoch": 0.824093290902576,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0019420423181791967,
+      "loss": 5.2351,
+      "step": 7259
+    },
+    {
+      "epoch": 0.8242068180124951,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0019414705391100299,
+      "loss": 5.2582,
+      "step": 7260
+    },
+    {
+      "epoch": 0.8243203451224141,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.001940896048210951,
+      "loss": 5.2146,
+      "step": 7261
+    },
+    {
+      "epoch": 0.8244338722323332,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0019403188473333872,
+      "loss": 5.2421,
+      "step": 7262
+    },
+    {
+      "epoch": 0.8245473993422523,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0019397389383375,
+      "loss": 5.2155,
+      "step": 7263
+    },
+    {
+      "epoch": 0.8246609264521714,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.001939156323092177,
+      "loss": 5.2283,
+      "step": 7264
+    },
+    {
+      "epoch": 0.8247744535620904,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0019385710034750287,
+      "loss": 5.2023,
+      "step": 7265
+    },
+    {
+      "epoch": 0.8248879806720095,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0019379829813723807,
+      "loss": 5.2374,
+      "step": 7266
+    },
+    {
+      "epoch": 0.8250015077819286,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0019373922586792673,
+      "loss": 5.2174,
+      "step": 7267
+    },
+    {
+      "epoch": 0.8251150348918477,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0019367988372994264,
+      "loss": 5.2171,
+      "step": 7268
+    },
+    {
+      "epoch": 0.8252285620017668,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0019362027191452936,
+      "loss": 5.2411,
+      "step": 7269
+    },
+    {
+      "epoch": 0.8253420891116858,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0019356039061379949,
+      "loss": 5.2315,
+      "step": 7270
+    },
+    {
+      "epoch": 0.8254556162216049,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0019350024002073411,
+      "loss": 5.2372,
+      "step": 7271
+    },
+    {
+      "epoch": 0.825569143331524,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0019343982032918217,
+      "loss": 5.2191,
+      "step": 7272
+    },
+    {
+      "epoch": 0.8256826704414431,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0019337913173385988,
+      "loss": 5.2386,
+      "step": 7273
+    },
+    {
+      "epoch": 0.8257961975513621,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0019331817443034999,
+      "loss": 5.2264,
+      "step": 7274
+    },
+    {
+      "epoch": 0.8259097246612812,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0019325694861510121,
+      "loss": 5.2239,
+      "step": 7275
+    },
+    {
+      "epoch": 0.8260232517712003,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0019319545448542772,
+      "loss": 5.2328,
+      "step": 7276
+    },
+    {
+      "epoch": 0.8261367788811194,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.001931336922395083,
+      "loss": 5.228,
+      "step": 7277
+    },
+    {
+      "epoch": 0.8262503059910384,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0019307166207638577,
+      "loss": 5.2125,
+      "step": 7278
+    },
+    {
+      "epoch": 0.8263638331009575,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0019300936419596644,
+      "loss": 5.2516,
+      "step": 7279
+    },
+    {
+      "epoch": 0.8264773602108766,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0019294679879901938,
+      "loss": 5.2344,
+      "step": 7280
+    },
+    {
+      "epoch": 0.8265908873207957,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0019288396608717578,
+      "loss": 5.2307,
+      "step": 7281
+    },
+    {
+      "epoch": 0.8267044144307147,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0019282086626292833,
+      "loss": 5.2337,
+      "step": 7282
+    },
+    {
+      "epoch": 0.8268179415406338,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0019275749952963057,
+      "loss": 5.2255,
+      "step": 7283
+    },
+    {
+      "epoch": 0.8269314686505529,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0019269386609149612,
+      "loss": 5.2403,
+      "step": 7284
+    },
+    {
+      "epoch": 0.827044995760472,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.001926299661535983,
+      "loss": 5.2156,
+      "step": 7285
+    },
+    {
+      "epoch": 0.827158522870391,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0019256579992186905,
+      "loss": 5.2405,
+      "step": 7286
+    },
+    {
+      "epoch": 0.8272720499803101,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0019250136760309868,
+      "loss": 5.2254,
+      "step": 7287
+    },
+    {
+      "epoch": 0.8273855770902292,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0019243666940493497,
+      "loss": 5.2372,
+      "step": 7288
+    },
+    {
+      "epoch": 0.8274991042001483,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.001923717055358826,
+      "loss": 5.2192,
+      "step": 7289
+    },
+    {
+      "epoch": 0.8276126313100673,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0019230647620530228,
+      "loss": 5.2281,
+      "step": 7290
+    },
+    {
+      "epoch": 0.8277261584199864,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0019224098162341043,
+      "loss": 5.2192,
+      "step": 7291
+    },
+    {
+      "epoch": 0.8278396855299055,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0019217522200127822,
+      "loss": 5.2353,
+      "step": 7292
+    },
+    {
+      "epoch": 0.8279532126398246,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0019210919755083089,
+      "loss": 5.2337,
+      "step": 7293
+    },
+    {
+      "epoch": 0.8280667397497437,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0019204290848484728,
+      "loss": 5.236,
+      "step": 7294
+    },
+    {
+      "epoch": 0.8281802668596627,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.00191976355016959,
+      "loss": 5.2139,
+      "step": 7295
+    },
+    {
+      "epoch": 0.8282937939695818,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0019190953736164963,
+      "loss": 5.2483,
+      "step": 7296
+    },
+    {
+      "epoch": 0.8284073210795009,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0019184245573425425,
+      "loss": 5.2305,
+      "step": 7297
+    },
+    {
+      "epoch": 0.82852084818942,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.001917751103509587,
+      "loss": 5.2192,
+      "step": 7298
+    },
+    {
+      "epoch": 0.828634375299339,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0019170750142879873,
+      "loss": 5.203,
+      "step": 7299
+    },
+    {
+      "epoch": 0.8287479024092581,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0019163962918565946,
+      "loss": 5.223,
+      "step": 7300
+    },
+    {
+      "epoch": 0.8288614295191772,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0019157149384027457,
+      "loss": 5.2191,
+      "step": 7301
+    },
+    {
+      "epoch": 0.8289749566290963,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0019150309561222574,
+      "loss": 5.2087,
+      "step": 7302
+    },
+    {
+      "epoch": 0.8290884837390153,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0019143443472194178,
+      "loss": 5.2119,
+      "step": 7303
+    },
+    {
+      "epoch": 0.8292020108489344,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0019136551139069793,
+      "loss": 5.2168,
+      "step": 7304
+    },
+    {
+      "epoch": 0.8293155379588535,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0019129632584061538,
+      "loss": 5.2104,
+      "step": 7305
+    },
+    {
+      "epoch": 0.8294290650687726,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0019122687829466023,
+      "loss": 5.2178,
+      "step": 7306
+    },
+    {
+      "epoch": 0.8295425921786916,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0019115716897664299,
+      "loss": 5.2493,
+      "step": 7307
+    },
+    {
+      "epoch": 0.8296561192886107,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0019108719811121771,
+      "loss": 5.2356,
+      "step": 7308
+    },
+    {
+      "epoch": 0.8297696463985298,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0019101696592388148,
+      "loss": 5.201,
+      "step": 7309
+    },
+    {
+      "epoch": 0.8298831735084489,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.001909464726409734,
+      "loss": 5.219,
+      "step": 7310
+    },
+    {
+      "epoch": 0.8299967006183679,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.001908757184896741,
+      "loss": 5.2147,
+      "step": 7311
+    },
+    {
+      "epoch": 0.830110227728287,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0019080470369800494,
+      "loss": 5.2397,
+      "step": 7312
+    },
+    {
+      "epoch": 0.8302237548382061,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0019073342849482717,
+      "loss": 5.2212,
+      "step": 7313
+    },
+    {
+      "epoch": 0.8303372819481252,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0019066189310984135,
+      "loss": 5.2484,
+      "step": 7314
+    },
+    {
+      "epoch": 0.8304508090580442,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.001905900977735865,
+      "loss": 5.2294,
+      "step": 7315
+    },
+    {
+      "epoch": 0.8305643361679633,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0019051804271743935,
+      "loss": 5.2219,
+      "step": 7316
+    },
+    {
+      "epoch": 0.8306778632778824,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0019044572817361375,
+      "loss": 5.2292,
+      "step": 7317
+    },
+    {
+      "epoch": 0.8307913903878015,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.001903731543751597,
+      "loss": 5.2066,
+      "step": 7318
+    },
+    {
+      "epoch": 0.8309049174977206,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0019030032155596276,
+      "loss": 5.2262,
+      "step": 7319
+    },
+    {
+      "epoch": 0.8310184446076396,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0019022722995074319,
+      "loss": 5.2266,
+      "step": 7320
+    },
+    {
+      "epoch": 0.8311319717175587,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0019015387979505536,
+      "loss": 5.245,
+      "step": 7321
+    },
+    {
+      "epoch": 0.8312454988274778,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0019008027132528676,
+      "loss": 5.2294,
+      "step": 7322
+    },
+    {
+      "epoch": 0.8313590259373969,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0019000640477865739,
+      "loss": 5.2244,
+      "step": 7323
+    },
+    {
+      "epoch": 0.8314725530473159,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00189932280393219,
+      "loss": 5.239,
+      "step": 7324
+    },
+    {
+      "epoch": 0.831586080157235,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0018985789840785426,
+      "loss": 5.2362,
+      "step": 7325
+    },
+    {
+      "epoch": 0.8316996072671541,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0018978325906227598,
+      "loss": 5.2193,
+      "step": 7326
+    },
+    {
+      "epoch": 0.8318131343770732,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0018970836259702643,
+      "loss": 5.2269,
+      "step": 7327
+    },
+    {
+      "epoch": 0.8319266614869922,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0018963320925347647,
+      "loss": 5.2273,
+      "step": 7328
+    },
+    {
+      "epoch": 0.8320401885969113,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.001895577992738248,
+      "loss": 5.2182,
+      "step": 7329
+    },
+    {
+      "epoch": 0.8321537157068304,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0018948213290109722,
+      "loss": 5.2334,
+      "step": 7330
+    },
+    {
+      "epoch": 0.8322672428167495,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.001894062103791458,
+      "loss": 5.225,
+      "step": 7331
+    },
+    {
+      "epoch": 0.8323807699266685,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.001893300319526481,
+      "loss": 5.2291,
+      "step": 7332
+    },
+    {
+      "epoch": 0.8324942970365876,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0018925359786710642,
+      "loss": 5.2212,
+      "step": 7333
+    },
+    {
+      "epoch": 0.8326078241465067,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0018917690836884693,
+      "loss": 5.2243,
+      "step": 7334
+    },
+    {
+      "epoch": 0.8327213512564258,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0018909996370501901,
+      "loss": 5.2375,
+      "step": 7335
+    },
+    {
+      "epoch": 0.8328348783663448,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0018902276412359423,
+      "loss": 5.2317,
+      "step": 7336
+    },
+    {
+      "epoch": 0.8329484054762639,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0018894530987336585,
+      "loss": 5.2565,
+      "step": 7337
+    },
+    {
+      "epoch": 0.833061932586183,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0018886760120394772,
+      "loss": 5.2575,
+      "step": 7338
+    },
+    {
+      "epoch": 0.8331754596961021,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0018878963836577373,
+      "loss": 5.2332,
+      "step": 7339
+    },
+    {
+      "epoch": 0.8332889868060211,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0018871142161009677,
+      "loss": 5.2165,
+      "step": 7340
+    },
+    {
+      "epoch": 0.8334025139159402,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0018863295118898816,
+      "loss": 5.2251,
+      "step": 7341
+    },
+    {
+      "epoch": 0.8335160410258593,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0018855422735533662,
+      "loss": 5.2341,
+      "step": 7342
+    },
+    {
+      "epoch": 0.8336295681357784,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0018847525036284761,
+      "loss": 5.2221,
+      "step": 7343
+    },
+    {
+      "epoch": 0.8337430952456975,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0018839602046604239,
+      "loss": 5.2277,
+      "step": 7344
+    },
+    {
+      "epoch": 0.8338566223556165,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0018831653792025732,
+      "loss": 5.2105,
+      "step": 7345
+    },
+    {
+      "epoch": 0.8339701494655356,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0018823680298164294,
+      "loss": 5.2287,
+      "step": 7346
+    },
+    {
+      "epoch": 0.8340836765754547,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0018815681590716319,
+      "loss": 5.2431,
+      "step": 7347
+    },
+    {
+      "epoch": 0.8341972036853738,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.001880765769545946,
+      "loss": 5.2252,
+      "step": 7348
+    },
+    {
+      "epoch": 0.8343107307952928,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.001879960863825254,
+      "loss": 5.2205,
+      "step": 7349
+    },
+    {
+      "epoch": 0.8344242579052119,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0018791534445035472,
+      "loss": 5.232,
+      "step": 7350
+    },
+    {
+      "epoch": 0.834537785015131,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0018783435141829187,
+      "loss": 5.2107,
+      "step": 7351
+    },
+    {
+      "epoch": 0.8346513121250501,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0018775310754735517,
+      "loss": 5.201,
+      "step": 7352
+    },
+    {
+      "epoch": 0.8347648392349691,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0018767161309937153,
+      "loss": 5.2314,
+      "step": 7353
+    },
+    {
+      "epoch": 0.8348783663448882,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0018758986833697527,
+      "loss": 5.2384,
+      "step": 7354
+    },
+    {
+      "epoch": 0.8349918934548073,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0018750787352360746,
+      "loss": 5.2412,
+      "step": 7355
+    },
+    {
+      "epoch": 0.8351054205647264,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0018742562892351503,
+      "loss": 5.2106,
+      "step": 7356
+    },
+    {
+      "epoch": 0.8352189476746454,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0018734313480174983,
+      "loss": 5.2153,
+      "step": 7357
+    },
+    {
+      "epoch": 0.8353324747845645,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0018726039142416796,
+      "loss": 5.2318,
+      "step": 7358
+    },
+    {
+      "epoch": 0.8354460018944836,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0018717739905742873,
+      "loss": 5.2403,
+      "step": 7359
+    },
+    {
+      "epoch": 0.8355595290044027,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0018709415796899383,
+      "loss": 5.2338,
+      "step": 7360
+    },
+    {
+      "epoch": 0.8356730561143219,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0018701066842712667,
+      "loss": 5.2164,
+      "step": 7361
+    },
+    {
+      "epoch": 0.8357865832242409,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.001869269307008912,
+      "loss": 5.2091,
+      "step": 7362
+    },
+    {
+      "epoch": 0.83590011033416,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0018684294506015125,
+      "loss": 5.2272,
+      "step": 7363
+    },
+    {
+      "epoch": 0.8360136374440791,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0018675871177556967,
+      "loss": 5.2254,
+      "step": 7364
+    },
+    {
+      "epoch": 0.8361271645539982,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.001866742311186073,
+      "loss": 5.2338,
+      "step": 7365
+    },
+    {
+      "epoch": 0.8362406916639172,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0018658950336152228,
+      "loss": 5.2252,
+      "step": 7366
+    },
+    {
+      "epoch": 0.8363542187738363,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0018650452877736901,
+      "loss": 5.2021,
+      "step": 7367
+    },
+    {
+      "epoch": 0.8364677458837554,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0018641930763999743,
+      "loss": 5.2147,
+      "step": 7368
+    },
+    {
+      "epoch": 0.8365812729936745,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0018633384022405197,
+      "loss": 5.2393,
+      "step": 7369
+    },
+    {
+      "epoch": 0.8366948001035935,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0018624812680497081,
+      "loss": 5.2228,
+      "step": 7370
+    },
+    {
+      "epoch": 0.8368083272135126,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.001861621676589849,
+      "loss": 5.2268,
+      "step": 7371
+    },
+    {
+      "epoch": 0.8369218543234317,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0018607596306311712,
+      "loss": 5.2217,
+      "step": 7372
+    },
+    {
+      "epoch": 0.8370353814333508,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0018598951329518137,
+      "loss": 5.2215,
+      "step": 7373
+    },
+    {
+      "epoch": 0.8371489085432698,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0018590281863378163,
+      "loss": 5.2059,
+      "step": 7374
+    },
+    {
+      "epoch": 0.8372624356531889,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0018581587935831113,
+      "loss": 5.2297,
+      "step": 7375
+    },
+    {
+      "epoch": 0.837375962763108,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0018572869574895144,
+      "loss": 5.2313,
+      "step": 7376
+    },
+    {
+      "epoch": 0.8374894898730271,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0018564126808667156,
+      "loss": 5.2385,
+      "step": 7377
+    },
+    {
+      "epoch": 0.8376030169829461,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0018555359665322695,
+      "loss": 5.2205,
+      "step": 7378
+    },
+    {
+      "epoch": 0.8377165440928652,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0018546568173115874,
+      "loss": 5.239,
+      "step": 7379
+    },
+    {
+      "epoch": 0.8378300712027843,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0018537752360379277,
+      "loss": 5.225,
+      "step": 7380
+    },
+    {
+      "epoch": 0.8379435983127034,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0018528912255523855,
+      "loss": 5.2082,
+      "step": 7381
+    },
+    {
+      "epoch": 0.8380571254226225,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0018520047887038858,
+      "loss": 5.2277,
+      "step": 7382
+    },
+    {
+      "epoch": 0.8381706525325415,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0018511159283491722,
+      "loss": 5.2383,
+      "step": 7383
+    },
+    {
+      "epoch": 0.8382841796424606,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0018502246473528,
+      "loss": 5.2243,
+      "step": 7384
+    },
+    {
+      "epoch": 0.8383977067523797,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.001849330948587124,
+      "loss": 5.2148,
+      "step": 7385
+    },
+    {
+      "epoch": 0.8385112338622988,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.001848434834932291,
+      "loss": 5.2117,
+      "step": 7386
+    },
+    {
+      "epoch": 0.8386247609722178,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0018475363092762315,
+      "loss": 5.2222,
+      "step": 7387
+    },
+    {
+      "epoch": 0.8387382880821369,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0018466353745146481,
+      "loss": 5.2113,
+      "step": 7388
+    },
+    {
+      "epoch": 0.838851815192056,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0018457320335510078,
+      "loss": 5.2312,
+      "step": 7389
+    },
+    {
+      "epoch": 0.8389653423019751,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0018448262892965322,
+      "loss": 5.2219,
+      "step": 7390
+    },
+    {
+      "epoch": 0.8390788694118941,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0018439181446701875,
+      "loss": 5.2092,
+      "step": 7391
+    },
+    {
+      "epoch": 0.8391923965218132,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0018430076025986763,
+      "loss": 5.217,
+      "step": 7392
+    },
+    {
+      "epoch": 0.8393059236317323,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0018420946660164268,
+      "loss": 5.2323,
+      "step": 7393
+    },
+    {
+      "epoch": 0.8394194507416514,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0018411793378655845,
+      "loss": 5.2479,
+      "step": 7394
+    },
+    {
+      "epoch": 0.8395329778515704,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0018402616210960031,
+      "loss": 5.1887,
+      "step": 7395
+    },
+    {
+      "epoch": 0.8396465049614895,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0018393415186652323,
+      "loss": 5.2388,
+      "step": 7396
+    },
+    {
+      "epoch": 0.8397600320714086,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0018384190335385116,
+      "loss": 5.2259,
+      "step": 7397
+    },
+    {
+      "epoch": 0.8398735591813277,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0018374941686887588,
+      "loss": 5.2321,
+      "step": 7398
+    },
+    {
+      "epoch": 0.8399870862912467,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0018365669270965607,
+      "loss": 5.2157,
+      "step": 7399
+    },
+    {
+      "epoch": 0.8401006134011658,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.001835637311750164,
+      "loss": 5.2224,
+      "step": 7400
+    },
+    {
+      "epoch": 0.8402141405110849,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0018347053256454657,
+      "loss": 5.2302,
+      "step": 7401
+    },
+    {
+      "epoch": 0.840327667621004,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0018337709717860021,
+      "loss": 5.2177,
+      "step": 7402
+    },
+    {
+      "epoch": 0.840441194730923,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0018328342531829407,
+      "loss": 5.227,
+      "step": 7403
+    },
+    {
+      "epoch": 0.8405547218408421,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0018318951728550706,
+      "loss": 5.2082,
+      "step": 7404
+    },
+    {
+      "epoch": 0.8406682489507612,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0018309537338287904,
+      "loss": 5.2236,
+      "step": 7405
+    },
+    {
+      "epoch": 0.8407817760606803,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0018300099391381022,
+      "loss": 5.2196,
+      "step": 7406
+    },
+    {
+      "epoch": 0.8408953031705994,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.001829063791824598,
+      "loss": 5.2117,
+      "step": 7407
+    },
+    {
+      "epoch": 0.8410088302805184,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0018281152949374527,
+      "loss": 5.2316,
+      "step": 7408
+    },
+    {
+      "epoch": 0.8411223573904375,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.001827164451533413,
+      "loss": 5.2309,
+      "step": 7409
+    },
+    {
+      "epoch": 0.8412358845003566,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0018262112646767875,
+      "loss": 5.2474,
+      "step": 7410
+    },
+    {
+      "epoch": 0.8413494116102757,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0018252557374394374,
+      "loss": 5.2165,
+      "step": 7411
+    },
+    {
+      "epoch": 0.8414629387201947,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.001824297872900766,
+      "loss": 5.2043,
+      "step": 7412
+    },
+    {
+      "epoch": 0.8415764658301138,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0018233376741477098,
+      "loss": 5.2403,
+      "step": 7413
+    },
+    {
+      "epoch": 0.8416899929400329,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0018223751442747271,
+      "loss": 5.2441,
+      "step": 7414
+    },
+    {
+      "epoch": 0.841803520049952,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.001821410286383789,
+      "loss": 5.2222,
+      "step": 7415
+    },
+    {
+      "epoch": 0.841917047159871,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.001820443103584369,
+      "loss": 5.2222,
+      "step": 7416
+    },
+    {
+      "epoch": 0.8420305742697901,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0018194735989934337,
+      "loss": 5.2188,
+      "step": 7417
+    },
+    {
+      "epoch": 0.8421441013797092,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0018185017757354313,
+      "loss": 5.2177,
+      "step": 7418
+    },
+    {
+      "epoch": 0.8422576284896283,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0018175276369422832,
+      "loss": 5.2055,
+      "step": 7419
+    },
+    {
+      "epoch": 0.8423711555995473,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0018165511857533736,
+      "loss": 5.1936,
+      "step": 7420
+    },
+    {
+      "epoch": 0.8424846827094664,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0018155724253155368,
+      "loss": 5.2188,
+      "step": 7421
+    },
+    {
+      "epoch": 0.8425982098193855,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0018145913587830518,
+      "loss": 5.2243,
+      "step": 7422
+    },
+    {
+      "epoch": 0.8427117369293046,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0018136079893176279,
+      "loss": 5.2046,
+      "step": 7423
+    },
+    {
+      "epoch": 0.8428252640392236,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.001812622320088396,
+      "loss": 5.2183,
+      "step": 7424
+    },
+    {
+      "epoch": 0.8429387911491427,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0018116343542719001,
+      "loss": 5.1986,
+      "step": 7425
+    },
+    {
+      "epoch": 0.8430523182590618,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0018106440950520835,
+      "loss": 5.2239,
+      "step": 7426
+    },
+    {
+      "epoch": 0.8431658453689809,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0018096515456202818,
+      "loss": 5.2253,
+      "step": 7427
+    },
+    {
+      "epoch": 0.8432793724789,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0018086567091752108,
+      "loss": 5.2263,
+      "step": 7428
+    },
+    {
+      "epoch": 0.843392899588819,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.001807659588922957,
+      "loss": 5.2151,
+      "step": 7429
+    },
+    {
+      "epoch": 0.8435064266987381,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.001806660188076967,
+      "loss": 5.2337,
+      "step": 7430
+    },
+    {
+      "epoch": 0.8436199538086572,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.001805658509858037,
+      "loss": 5.2317,
+      "step": 7431
+    },
+    {
+      "epoch": 0.8437334809185763,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0018046545574943022,
+      "loss": 5.238,
+      "step": 7432
+    },
+    {
+      "epoch": 0.8438470080284953,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0018036483342212268,
+      "loss": 5.2044,
+      "step": 7433
+    },
+    {
+      "epoch": 0.8439605351384144,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0018026398432815947,
+      "loss": 5.2297,
+      "step": 7434
+    },
+    {
+      "epoch": 0.8440740622483335,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0018016290879254966,
+      "loss": 5.2301,
+      "step": 7435
+    },
+    {
+      "epoch": 0.8441875893582526,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0018006160714103213,
+      "loss": 5.2368,
+      "step": 7436
+    },
+    {
+      "epoch": 0.8443011164681716,
+      "grad_norm": 0.375,
+      "learning_rate": 0.001799600797000744,
+      "loss": 5.2175,
+      "step": 7437
+    },
+    {
+      "epoch": 0.8444146435780907,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.001798583267968718,
+      "loss": 5.2314,
+      "step": 7438
+    },
+    {
+      "epoch": 0.8445281706880098,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0017975634875934605,
+      "loss": 5.205,
+      "step": 7439
+    },
+    {
+      "epoch": 0.8446416977979289,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0017965414591614465,
+      "loss": 5.2173,
+      "step": 7440
+    },
+    {
+      "epoch": 0.8447552249078479,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0017955171859663941,
+      "loss": 5.2231,
+      "step": 7441
+    },
+    {
+      "epoch": 0.844868752017767,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0017944906713092566,
+      "loss": 5.2094,
+      "step": 7442
+    },
+    {
+      "epoch": 0.8449822791276861,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0017934619184982103,
+      "loss": 5.2393,
+      "step": 7443
+    },
+    {
+      "epoch": 0.8450958062376052,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.001792430930848645,
+      "loss": 5.2216,
+      "step": 7444
+    },
+    {
+      "epoch": 0.8452093333475242,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0017913977116831524,
+      "loss": 5.2167,
+      "step": 7445
+    },
+    {
+      "epoch": 0.8453228604574433,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0017903622643315159,
+      "loss": 5.2135,
+      "step": 7446
+    },
+    {
+      "epoch": 0.8454363875673624,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0017893245921307,
+      "loss": 5.2207,
+      "step": 7447
+    },
+    {
+      "epoch": 0.8455499146772815,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0017882846984248386,
+      "loss": 5.2202,
+      "step": 7448
+    },
+    {
+      "epoch": 0.8456634417872005,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0017872425865652257,
+      "loss": 5.203,
+      "step": 7449
+    },
+    {
+      "epoch": 0.8457769688971196,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.001786198259910303,
+      "loss": 5.2125,
+      "step": 7450
+    },
+    {
+      "epoch": 0.8458904960070387,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.001785151721825651,
+      "loss": 5.2425,
+      "step": 7451
+    },
+    {
+      "epoch": 0.8460040231169578,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0017841029756839758,
+      "loss": 5.2119,
+      "step": 7452
+    },
+    {
+      "epoch": 0.8461175502268768,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0017830520248651,
+      "loss": 5.2316,
+      "step": 7453
+    },
+    {
+      "epoch": 0.8462310773367959,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0017819988727559514,
+      "loss": 5.2465,
+      "step": 7454
+    },
+    {
+      "epoch": 0.846344604446715,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0017809435227505522,
+      "loss": 5.2205,
+      "step": 7455
+    },
+    {
+      "epoch": 0.8464581315566341,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.001779885978250007,
+      "loss": 5.2195,
+      "step": 7456
+    },
+    {
+      "epoch": 0.8465716586665532,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0017788262426624935,
+      "loss": 5.2111,
+      "step": 7457
+    },
+    {
+      "epoch": 0.8466851857764722,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00177776431940325,
+      "loss": 5.2225,
+      "step": 7458
+    },
+    {
+      "epoch": 0.8467987128863913,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0017767002118945652,
+      "loss": 5.219,
+      "step": 7459
+    },
+    {
+      "epoch": 0.8469122399963104,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0017756339235657677,
+      "loss": 5.22,
+      "step": 7460
+    },
+    {
+      "epoch": 0.8470257671062295,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0017745654578532133,
+      "loss": 5.2136,
+      "step": 7461
+    },
+    {
+      "epoch": 0.8471392942161485,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0017734948182002762,
+      "loss": 5.2108,
+      "step": 7462
+    },
+    {
+      "epoch": 0.8472528213260676,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0017724220080573348,
+      "loss": 5.2412,
+      "step": 7463
+    },
+    {
+      "epoch": 0.8473663484359867,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.001771347030881764,
+      "loss": 5.2288,
+      "step": 7464
+    },
+    {
+      "epoch": 0.8474798755459058,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0017702698901379218,
+      "loss": 5.2324,
+      "step": 7465
+    },
+    {
+      "epoch": 0.8475934026558248,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.001769190589297139,
+      "loss": 5.2034,
+      "step": 7466
+    },
+    {
+      "epoch": 0.8477069297657439,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0017681091318377073,
+      "loss": 5.2271,
+      "step": 7467
+    },
+    {
+      "epoch": 0.847820456875663,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0017670255212448693,
+      "loss": 5.2168,
+      "step": 7468
+    },
+    {
+      "epoch": 0.8479339839855821,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.001765939761010806,
+      "loss": 5.2402,
+      "step": 7469
+    },
+    {
+      "epoch": 0.8480475110955011,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.001764851854634627,
+      "loss": 5.2227,
+      "step": 7470
+    },
+    {
+      "epoch": 0.8481610382054202,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0017637618056223566,
+      "loss": 5.2304,
+      "step": 7471
+    },
+    {
+      "epoch": 0.8482745653153393,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0017626696174869259,
+      "loss": 5.2111,
+      "step": 7472
+    },
+    {
+      "epoch": 0.8483880924252584,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.001761575293748159,
+      "loss": 5.2403,
+      "step": 7473
+    },
+    {
+      "epoch": 0.8485016195351774,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.001760478837932763,
+      "loss": 5.2097,
+      "step": 7474
+    },
+    {
+      "epoch": 0.8486151466450965,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0017593802535743155,
+      "loss": 5.2026,
+      "step": 7475
+    },
+    {
+      "epoch": 0.8487286737550156,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0017582795442132543,
+      "loss": 5.2282,
+      "step": 7476
+    },
+    {
+      "epoch": 0.8488422008649347,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.001757176713396865,
+      "loss": 5.2219,
+      "step": 7477
+    },
+    {
+      "epoch": 0.8489557279748537,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0017560717646792703,
+      "loss": 5.2334,
+      "step": 7478
+    },
+    {
+      "epoch": 0.8490692550847728,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0017549647016214186,
+      "loss": 5.2092,
+      "step": 7479
+    },
+    {
+      "epoch": 0.8491827821946919,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0017538555277910725,
+      "loss": 5.2101,
+      "step": 7480
+    },
+    {
+      "epoch": 0.849296309304611,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0017527442467627954,
+      "loss": 5.2334,
+      "step": 7481
+    },
+    {
+      "epoch": 0.84940983641453,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0017516308621179438,
+      "loss": 5.2092,
+      "step": 7482
+    },
+    {
+      "epoch": 0.8495233635244491,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0017505153774446526,
+      "loss": 5.219,
+      "step": 7483
+    },
+    {
+      "epoch": 0.8496368906343682,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0017493977963378233,
+      "loss": 5.2115,
+      "step": 7484
+    },
+    {
+      "epoch": 0.8497504177442873,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0017482781223991163,
+      "loss": 5.1933,
+      "step": 7485
+    },
+    {
+      "epoch": 0.8498639448542064,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0017471563592369342,
+      "loss": 5.2172,
+      "step": 7486
+    },
+    {
+      "epoch": 0.8499774719641254,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0017460325104664137,
+      "loss": 5.2017,
+      "step": 7487
+    },
+    {
+      "epoch": 0.8500909990740445,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0017449065797094131,
+      "loss": 5.2281,
+      "step": 7488
+    },
+    {
+      "epoch": 0.8502045261839636,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0017437785705944992,
+      "loss": 5.194,
+      "step": 7489
+    },
+    {
+      "epoch": 0.8503180532938827,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0017426484867569389,
+      "loss": 5.2077,
+      "step": 7490
+    },
+    {
+      "epoch": 0.8504315804038017,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0017415163318386822,
+      "loss": 5.2195,
+      "step": 7491
+    },
+    {
+      "epoch": 0.8505451075137208,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.001740382109488357,
+      "loss": 5.2201,
+      "step": 7492
+    },
+    {
+      "epoch": 0.8506586346236399,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0017392458233612513,
+      "loss": 5.2055,
+      "step": 7493
+    },
+    {
+      "epoch": 0.850772161733559,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0017381074771193059,
+      "loss": 5.217,
+      "step": 7494
+    },
+    {
+      "epoch": 0.850885688843478,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0017369670744310998,
+      "loss": 5.2222,
+      "step": 7495
+    },
+    {
+      "epoch": 0.8509992159533971,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00173582461897184,
+      "loss": 5.202,
+      "step": 7496
+    },
+    {
+      "epoch": 0.8511127430633162,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0017346801144233483,
+      "loss": 5.2117,
+      "step": 7497
+    },
+    {
+      "epoch": 0.8512262701732353,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0017335335644740504,
+      "loss": 5.2163,
+      "step": 7498
+    },
+    {
+      "epoch": 0.8513397972831543,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0017323849728189645,
+      "loss": 5.2031,
+      "step": 7499
+    },
+    {
+      "epoch": 0.8514533243930734,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0017312343431596874,
+      "loss": 5.2112,
+      "step": 7500
+    },
+    {
+      "epoch": 0.8515668515029925,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0017300816792043849,
+      "loss": 5.2134,
+      "step": 7501
+    },
+    {
+      "epoch": 0.8516803786129116,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0017289269846677779,
+      "loss": 5.1887,
+      "step": 7502
+    },
+    {
+      "epoch": 0.8517939057228306,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0017277702632711323,
+      "loss": 5.2175,
+      "step": 7503
+    },
+    {
+      "epoch": 0.8519074328327497,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0017266115187422452,
+      "loss": 5.2027,
+      "step": 7504
+    },
+    {
+      "epoch": 0.8520209599426688,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.001725450754815434,
+      "loss": 5.2239,
+      "step": 7505
+    },
+    {
+      "epoch": 0.8521344870525879,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0017242879752315244,
+      "loss": 5.2257,
+      "step": 7506
+    },
+    {
+      "epoch": 0.852248014162507,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0017231231837378377,
+      "loss": 5.2303,
+      "step": 7507
+    },
+    {
+      "epoch": 0.852361541272426,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0017219563840881783,
+      "loss": 5.1914,
+      "step": 7508
+    },
+    {
+      "epoch": 0.8524750683823451,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0017207875800428235,
+      "loss": 5.2349,
+      "step": 7509
+    },
+    {
+      "epoch": 0.8525885954922642,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.00171961677536851,
+      "loss": 5.2074,
+      "step": 7510
+    },
+    {
+      "epoch": 0.8527021226021833,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0017184439738384214,
+      "loss": 5.203,
+      "step": 7511
+    },
+    {
+      "epoch": 0.8528156497121023,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0017172691792321773,
+      "loss": 5.219,
+      "step": 7512
+    },
+    {
+      "epoch": 0.8529291768220214,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0017160923953358199,
+      "loss": 5.2137,
+      "step": 7513
+    },
+    {
+      "epoch": 0.8530427039319405,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0017149136259418025,
+      "loss": 5.2038,
+      "step": 7514
+    },
+    {
+      "epoch": 0.8531562310418596,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.001713732874848977,
+      "loss": 5.206,
+      "step": 7515
+    },
+    {
+      "epoch": 0.8532697581517786,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0017125501458625826,
+      "loss": 5.2176,
+      "step": 7516
+    },
+    {
+      "epoch": 0.8533832852616977,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0017113654427942317,
+      "loss": 5.2352,
+      "step": 7517
+    },
+    {
+      "epoch": 0.8534968123716168,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0017101787694618988,
+      "loss": 5.225,
+      "step": 7518
+    },
+    {
+      "epoch": 0.8536103394815359,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0017089901296899084,
+      "loss": 5.1893,
+      "step": 7519
+    },
+    {
+      "epoch": 0.8537238665914549,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0017077995273089219,
+      "loss": 5.2056,
+      "step": 7520
+    },
+    {
+      "epoch": 0.853837393701374,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0017066069661559266,
+      "loss": 5.2175,
+      "step": 7521
+    },
+    {
+      "epoch": 0.8539509208112931,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0017054124500742204,
+      "loss": 5.2128,
+      "step": 7522
+    },
+    {
+      "epoch": 0.8540644479212122,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0017042159829134037,
+      "loss": 5.2205,
+      "step": 7523
+    },
+    {
+      "epoch": 0.8541779750311312,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.001703017568529363,
+      "loss": 5.2286,
+      "step": 7524
+    },
+    {
+      "epoch": 0.8542915021410503,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0017018172107842616,
+      "loss": 5.2186,
+      "step": 7525
+    },
+    {
+      "epoch": 0.8544050292509694,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0017006149135465238,
+      "loss": 5.2357,
+      "step": 7526
+    },
+    {
+      "epoch": 0.8545185563608885,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0016994106806908266,
+      "loss": 5.1987,
+      "step": 7527
+    },
+    {
+      "epoch": 0.8546320834708075,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0016982045160980831,
+      "loss": 5.2158,
+      "step": 7528
+    },
+    {
+      "epoch": 0.8547456105807266,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0016969964236554329,
+      "loss": 5.2177,
+      "step": 7529
+    },
+    {
+      "epoch": 0.8548591376906457,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0016957864072562286,
+      "loss": 5.2184,
+      "step": 7530
+    },
+    {
+      "epoch": 0.8549726648005648,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0016945744708000221,
+      "loss": 5.211,
+      "step": 7531
+    },
+    {
+      "epoch": 0.8550861919104839,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0016933606181925547,
+      "loss": 5.2166,
+      "step": 7532
+    },
+    {
+      "epoch": 0.8551997190204029,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0016921448533457415,
+      "loss": 5.2279,
+      "step": 7533
+    },
+    {
+      "epoch": 0.855313246130322,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0016909271801776608,
+      "loss": 5.2373,
+      "step": 7534
+    },
+    {
+      "epoch": 0.8554267732402411,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0016897076026125414,
+      "loss": 5.2019,
+      "step": 7535
+    },
+    {
+      "epoch": 0.8555403003501602,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0016884861245807486,
+      "loss": 5.2134,
+      "step": 7536
+    },
+    {
+      "epoch": 0.8556538274600792,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0016872627500187727,
+      "loss": 5.2257,
+      "step": 7537
+    },
+    {
+      "epoch": 0.8557673545699983,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0016860374828692162,
+      "loss": 5.2173,
+      "step": 7538
+    },
+    {
+      "epoch": 0.8558808816799174,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0016848103270807808,
+      "loss": 5.2107,
+      "step": 7539
+    },
+    {
+      "epoch": 0.8559944087898365,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0016835812866082547,
+      "loss": 5.2267,
+      "step": 7540
+    },
+    {
+      "epoch": 0.8561079358997555,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0016823503654125002,
+      "loss": 5.2244,
+      "step": 7541
+    },
+    {
+      "epoch": 0.8562214630096746,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0016811175674604395,
+      "loss": 5.2292,
+      "step": 7542
+    },
+    {
+      "epoch": 0.8563349901195937,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.001679882896725045,
+      "loss": 5.2056,
+      "step": 7543
+    },
+    {
+      "epoch": 0.8564485172295128,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.001678646357185323,
+      "loss": 5.2114,
+      "step": 7544
+    },
+    {
+      "epoch": 0.8565620443394318,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.001677407952826303,
+      "loss": 5.2037,
+      "step": 7545
+    },
+    {
+      "epoch": 0.8566755714493509,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0016761676876390246,
+      "loss": 5.2066,
+      "step": 7546
+    },
+    {
+      "epoch": 0.85678909855927,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0016749255656205238,
+      "loss": 5.1757,
+      "step": 7547
+    },
+    {
+      "epoch": 0.8569026256691891,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0016736815907738207,
+      "loss": 5.2106,
+      "step": 7548
+    },
+    {
+      "epoch": 0.8570161527791081,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.001672435767107907,
+      "loss": 5.2196,
+      "step": 7549
+    },
+    {
+      "epoch": 0.8571296798890272,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0016711880986377328,
+      "loss": 5.2458,
+      "step": 7550
+    },
+    {
+      "epoch": 0.8572432069989463,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0016699385893841924,
+      "loss": 5.214,
+      "step": 7551
+    },
+    {
+      "epoch": 0.8573567341088654,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0016686872433741133,
+      "loss": 5.2182,
+      "step": 7552
+    },
+    {
+      "epoch": 0.8574702612187844,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0016674340646402424,
+      "loss": 5.241,
+      "step": 7553
+    },
+    {
+      "epoch": 0.8575837883287035,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0016661790572212328,
+      "loss": 5.1946,
+      "step": 7554
+    },
+    {
+      "epoch": 0.8576973154386226,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0016649222251616305,
+      "loss": 5.2025,
+      "step": 7555
+    },
+    {
+      "epoch": 0.8578108425485417,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0016636635725118627,
+      "loss": 5.2113,
+      "step": 7556
+    },
+    {
+      "epoch": 0.8579243696584608,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0016624031033282232,
+      "loss": 5.2039,
+      "step": 7557
+    },
+    {
+      "epoch": 0.8580378967683798,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0016611408216728604,
+      "loss": 5.2319,
+      "step": 7558
+    },
+    {
+      "epoch": 0.8581514238782989,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0016598767316137633,
+      "loss": 5.2122,
+      "step": 7559
+    },
+    {
+      "epoch": 0.858264950988218,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0016586108372247492,
+      "loss": 5.2349,
+      "step": 7560
+    },
+    {
+      "epoch": 0.858378478098137,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0016573431425854503,
+      "loss": 5.2317,
+      "step": 7561
+    },
+    {
+      "epoch": 0.8584920052080561,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0016560736517813011,
+      "loss": 5.2297,
+      "step": 7562
+    },
+    {
+      "epoch": 0.8586055323179752,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0016548023689035229,
+      "loss": 5.2215,
+      "step": 7563
+    },
+    {
+      "epoch": 0.8587190594278943,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0016535292980491146,
+      "loss": 5.2162,
+      "step": 7564
+    },
+    {
+      "epoch": 0.8588325865378134,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0016522544433208353,
+      "loss": 5.2133,
+      "step": 7565
+    },
+    {
+      "epoch": 0.8589461136477324,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0016509778088271941,
+      "loss": 5.2195,
+      "step": 7566
+    },
+    {
+      "epoch": 0.8590596407576515,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.001649699398682436,
+      "loss": 5.2002,
+      "step": 7567
+    },
+    {
+      "epoch": 0.8591731678675706,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0016484192170065275,
+      "loss": 5.2216,
+      "step": 7568
+    },
+    {
+      "epoch": 0.8592866949774897,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.001647137267925145,
+      "loss": 5.2125,
+      "step": 7569
+    },
+    {
+      "epoch": 0.8594002220874087,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0016458535555696602,
+      "loss": 5.2175,
+      "step": 7570
+    },
+    {
+      "epoch": 0.8595137491973278,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0016445680840771284,
+      "loss": 5.2155,
+      "step": 7571
+    },
+    {
+      "epoch": 0.8596272763072469,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.001643280857590273,
+      "loss": 5.2244,
+      "step": 7572
+    },
+    {
+      "epoch": 0.859740803417166,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.001641991880257474,
+      "loss": 5.2184,
+      "step": 7573
+    },
+    {
+      "epoch": 0.859854330527085,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.001640701156232753,
+      "loss": 5.2109,
+      "step": 7574
+    },
+    {
+      "epoch": 0.8599678576370041,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0016394086896757616,
+      "loss": 5.2202,
+      "step": 7575
+    },
+    {
+      "epoch": 0.8600813847469232,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0016381144847517671,
+      "loss": 5.2075,
+      "step": 7576
+    },
+    {
+      "epoch": 0.8601949118568423,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0016368185456316381,
+      "loss": 5.2054,
+      "step": 7577
+    },
+    {
+      "epoch": 0.8603084389667613,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0016355208764918334,
+      "loss": 5.1963,
+      "step": 7578
+    },
+    {
+      "epoch": 0.8604219660766804,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.001634221481514386,
+      "loss": 5.2361,
+      "step": 7579
+    },
+    {
+      "epoch": 0.8605354931865995,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.001632920364886892,
+      "loss": 5.224,
+      "step": 7580
+    },
+    {
+      "epoch": 0.8606490202965186,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0016316175308024943,
+      "loss": 5.2148,
+      "step": 7581
+    },
+    {
+      "epoch": 0.8607625474064377,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0016303129834598724,
+      "loss": 5.2131,
+      "step": 7582
+    },
+    {
+      "epoch": 0.8608760745163567,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.001629006727063226,
+      "loss": 5.2143,
+      "step": 7583
+    },
+    {
+      "epoch": 0.8609896016262758,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.001627698765822263,
+      "loss": 5.2074,
+      "step": 7584
+    },
+    {
+      "epoch": 0.8611031287361949,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0016263891039521854,
+      "loss": 5.1969,
+      "step": 7585
+    },
+    {
+      "epoch": 0.861216655846114,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0016250777456736762,
+      "loss": 5.207,
+      "step": 7586
+    },
+    {
+      "epoch": 0.861330182956033,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0016237646952128855,
+      "loss": 5.1918,
+      "step": 7587
+    },
+    {
+      "epoch": 0.8614437100659521,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.001622449956801416,
+      "loss": 5.2103,
+      "step": 7588
+    },
+    {
+      "epoch": 0.8615572371758712,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0016211335346763116,
+      "loss": 5.2023,
+      "step": 7589
+    },
+    {
+      "epoch": 0.8616707642857903,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0016198154330800407,
+      "loss": 5.2133,
+      "step": 7590
+    },
+    {
+      "epoch": 0.8617842913957093,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0016184956562604858,
+      "loss": 5.1942,
+      "step": 7591
+    },
+    {
+      "epoch": 0.8618978185056284,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0016171742084709268,
+      "loss": 5.2057,
+      "step": 7592
+    },
+    {
+      "epoch": 0.8620113456155475,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00161585109397003,
+      "loss": 5.2038,
+      "step": 7593
+    },
+    {
+      "epoch": 0.8621248727254666,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0016145263170218318,
+      "loss": 5.2123,
+      "step": 7594
+    },
+    {
+      "epoch": 0.8622383998353856,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0016131998818957269,
+      "loss": 5.1869,
+      "step": 7595
+    },
+    {
+      "epoch": 0.8623519269453047,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0016118717928664537,
+      "loss": 5.1967,
+      "step": 7596
+    },
+    {
+      "epoch": 0.8624654540552238,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0016105420542140809,
+      "loss": 5.2073,
+      "step": 7597
+    },
+    {
+      "epoch": 0.8625789811651429,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.001609210670223993,
+      "loss": 5.2152,
+      "step": 7598
+    },
+    {
+      "epoch": 0.8626925082750619,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0016078776451868774,
+      "loss": 5.2186,
+      "step": 7599
+    },
+    {
+      "epoch": 0.862806035384981,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0016065429833987104,
+      "loss": 5.2069,
+      "step": 7600
+    },
+    {
+      "epoch": 0.8629195624949001,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.001605206689160742,
+      "loss": 5.1865,
+      "step": 7601
+    },
+    {
+      "epoch": 0.8630330896048193,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0016038687667794846,
+      "loss": 5.2122,
+      "step": 7602
+    },
+    {
+      "epoch": 0.8631466167147384,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.001602529220566696,
+      "loss": 5.2025,
+      "step": 7603
+    },
+    {
+      "epoch": 0.8632601438246574,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0016011880548393693,
+      "loss": 5.2108,
+      "step": 7604
+    },
+    {
+      "epoch": 0.8633736709345765,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0015998452739197145,
+      "loss": 5.19,
+      "step": 7605
+    },
+    {
+      "epoch": 0.8634871980444956,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0015985008821351489,
+      "loss": 5.2129,
+      "step": 7606
+    },
+    {
+      "epoch": 0.8636007251544147,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0015971548838182793,
+      "loss": 5.2129,
+      "step": 7607
+    },
+    {
+      "epoch": 0.8637142522643337,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0015958072833068924,
+      "loss": 5.2162,
+      "step": 7608
+    },
+    {
+      "epoch": 0.8638277793742528,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0015944580849439358,
+      "loss": 5.2314,
+      "step": 7609
+    },
+    {
+      "epoch": 0.8639413064841719,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0015931072930775076,
+      "loss": 5.21,
+      "step": 7610
+    },
+    {
+      "epoch": 0.864054833594091,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0015917549120608423,
+      "loss": 5.1931,
+      "step": 7611
+    },
+    {
+      "epoch": 0.86416836070401,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.001590400946252294,
+      "loss": 5.2007,
+      "step": 7612
+    },
+    {
+      "epoch": 0.8642818878139291,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0015890454000153254,
+      "loss": 5.203,
+      "step": 7613
+    },
+    {
+      "epoch": 0.8643954149238482,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0015876882777184915,
+      "loss": 5.2026,
+      "step": 7614
+    },
+    {
+      "epoch": 0.8645089420337673,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0015863295837354279,
+      "loss": 5.2039,
+      "step": 7615
+    },
+    {
+      "epoch": 0.8646224691436863,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.001584969322444834,
+      "loss": 5.2212,
+      "step": 7616
+    },
+    {
+      "epoch": 0.8647359962536054,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0015836074982304613,
+      "loss": 5.2077,
+      "step": 7617
+    },
+    {
+      "epoch": 0.8648495233635245,
+      "grad_norm": 0.25,
+      "learning_rate": 0.001582244115481097,
+      "loss": 5.2058,
+      "step": 7618
+    },
+    {
+      "epoch": 0.8649630504734436,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.001580879178590552,
+      "loss": 5.1931,
+      "step": 7619
+    },
+    {
+      "epoch": 0.8650765775833626,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.001579512691957645,
+      "loss": 5.218,
+      "step": 7620
+    },
+    {
+      "epoch": 0.8651901046932817,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00157814465998619,
+      "loss": 5.2069,
+      "step": 7621
+    },
+    {
+      "epoch": 0.8653036318032008,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0015767750870849804,
+      "loss": 5.2055,
+      "step": 7622
+    },
+    {
+      "epoch": 0.8654171589131199,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0015754039776677761,
+      "loss": 5.1989,
+      "step": 7623
+    },
+    {
+      "epoch": 0.865530686023039,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0015740313361532882,
+      "loss": 5.2105,
+      "step": 7624
+    },
+    {
+      "epoch": 0.865644213132958,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0015726571669651664,
+      "loss": 5.2033,
+      "step": 7625
+    },
+    {
+      "epoch": 0.8657577402428771,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0015712814745319822,
+      "loss": 5.1956,
+      "step": 7626
+    },
+    {
+      "epoch": 0.8658712673527962,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0015699042632872173,
+      "loss": 5.2051,
+      "step": 7627
+    },
+    {
+      "epoch": 0.8659847944627153,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0015685255376692478,
+      "loss": 5.2271,
+      "step": 7628
+    },
+    {
+      "epoch": 0.8660983215726343,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0015671453021213297,
+      "loss": 5.1899,
+      "step": 7629
+    },
+    {
+      "epoch": 0.8662118486825534,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0015657635610915861,
+      "loss": 5.2224,
+      "step": 7630
+    },
+    {
+      "epoch": 0.8663253757924725,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.001564380319032991,
+      "loss": 5.1975,
+      "step": 7631
+    },
+    {
+      "epoch": 0.8664389029023916,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0015629955804033558,
+      "loss": 5.1879,
+      "step": 7632
+    },
+    {
+      "epoch": 0.8665524300123106,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0015616093496653156,
+      "loss": 5.223,
+      "step": 7633
+    },
+    {
+      "epoch": 0.8666659571222297,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0015602216312863135,
+      "loss": 5.2054,
+      "step": 7634
+    },
+    {
+      "epoch": 0.8667794842321488,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0015588324297385872,
+      "loss": 5.2198,
+      "step": 7635
+    },
+    {
+      "epoch": 0.8668930113420679,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0015574417494991545,
+      "loss": 5.2062,
+      "step": 7636
+    },
+    {
+      "epoch": 0.8670065384519869,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0015560495950497977,
+      "loss": 5.1861,
+      "step": 7637
+    },
+    {
+      "epoch": 0.867120065561906,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0015546559708770511,
+      "loss": 5.1992,
+      "step": 7638
+    },
+    {
+      "epoch": 0.8672335926718251,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0015532608814721848,
+      "loss": 5.2023,
+      "step": 7639
+    },
+    {
+      "epoch": 0.8673471197817442,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0015518643313311915,
+      "loss": 5.1679,
+      "step": 7640
+    },
+    {
+      "epoch": 0.8674606468916632,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0015504663249547713,
+      "loss": 5.1957,
+      "step": 7641
+    },
+    {
+      "epoch": 0.8675741740015823,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0015490668668483166,
+      "loss": 5.2118,
+      "step": 7642
+    },
+    {
+      "epoch": 0.8676877011115014,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0015476659615218998,
+      "loss": 5.2051,
+      "step": 7643
+    },
+    {
+      "epoch": 0.8678012282214205,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0015462636134902562,
+      "loss": 5.185,
+      "step": 7644
+    },
+    {
+      "epoch": 0.8679147553313395,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0015448598272727702,
+      "loss": 5.2109,
+      "step": 7645
+    },
+    {
+      "epoch": 0.8680282824412586,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0015434546073934622,
+      "loss": 5.2174,
+      "step": 7646
+    },
+    {
+      "epoch": 0.8681418095511777,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0015420479583809728,
+      "loss": 5.1935,
+      "step": 7647
+    },
+    {
+      "epoch": 0.8682553366610968,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0015406398847685472,
+      "loss": 5.2076,
+      "step": 7648
+    },
+    {
+      "epoch": 0.8683688637710159,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0015392303910940228,
+      "loss": 5.1892,
+      "step": 7649
+    },
+    {
+      "epoch": 0.8684823908809349,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0015378194818998125,
+      "loss": 5.2103,
+      "step": 7650
+    },
+    {
+      "epoch": 0.868595917990854,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0015364071617328923,
+      "loss": 5.2292,
+      "step": 7651
+    },
+    {
+      "epoch": 0.8687094451007731,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0015349934351447845,
+      "loss": 5.2191,
+      "step": 7652
+    },
+    {
+      "epoch": 0.8688229722106922,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0015335783066915435,
+      "loss": 5.2028,
+      "step": 7653
+    },
+    {
+      "epoch": 0.8689364993206112,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0015321617809337432,
+      "loss": 5.1941,
+      "step": 7654
+    },
+    {
+      "epoch": 0.8690500264305303,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0015307438624364588,
+      "loss": 5.222,
+      "step": 7655
+    },
+    {
+      "epoch": 0.8691635535404494,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0015293245557692547,
+      "loss": 5.1929,
+      "step": 7656
+    },
+    {
+      "epoch": 0.8692770806503685,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0015279038655061688,
+      "loss": 5.2246,
+      "step": 7657
+    },
+    {
+      "epoch": 0.8693906077602875,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0015264817962256988,
+      "loss": 5.1833,
+      "step": 7658
+    },
+    {
+      "epoch": 0.8695041348702066,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0015250583525107855,
+      "loss": 5.2075,
+      "step": 7659
+    },
+    {
+      "epoch": 0.8696176619801257,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0015236335389487996,
+      "loss": 5.1948,
+      "step": 7660
+    },
+    {
+      "epoch": 0.8697311890900448,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.001522207360131526,
+      "loss": 5.2117,
+      "step": 7661
+    },
+    {
+      "epoch": 0.8698447161999638,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0015207798206551503,
+      "loss": 5.2014,
+      "step": 7662
+    },
+    {
+      "epoch": 0.8699582433098829,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0015193509251202422,
+      "loss": 5.2132,
+      "step": 7663
+    },
+    {
+      "epoch": 0.870071770419802,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.001517920678131743,
+      "loss": 5.2317,
+      "step": 7664
+    },
+    {
+      "epoch": 0.8701852975297211,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0015164890842989474,
+      "loss": 5.2171,
+      "step": 7665
+    },
+    {
+      "epoch": 0.8702988246396401,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0015150561482354921,
+      "loss": 5.1957,
+      "step": 7666
+    },
+    {
+      "epoch": 0.8704123517495592,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0015136218745593394,
+      "loss": 5.2112,
+      "step": 7667
+    },
+    {
+      "epoch": 0.8705258788594783,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.001512186267892761,
+      "loss": 5.188,
+      "step": 7668
+    },
+    {
+      "epoch": 0.8706394059693974,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0015107493328623258,
+      "loss": 5.2025,
+      "step": 7669
+    },
+    {
+      "epoch": 0.8707529330793164,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0015093110740988837,
+      "loss": 5.2158,
+      "step": 7670
+    },
+    {
+      "epoch": 0.8708664601892355,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0015078714962375498,
+      "loss": 5.1936,
+      "step": 7671
+    },
+    {
+      "epoch": 0.8709799872991546,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0015064306039176906,
+      "loss": 5.2162,
+      "step": 7672
+    },
+    {
+      "epoch": 0.8710935144090737,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0015049884017829089,
+      "loss": 5.2106,
+      "step": 7673
+    },
+    {
+      "epoch": 0.8712070415189928,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0015035448944810293,
+      "loss": 5.187,
+      "step": 7674
+    },
+    {
+      "epoch": 0.8713205686289118,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0015021000866640806,
+      "loss": 5.2236,
+      "step": 7675
+    },
+    {
+      "epoch": 0.8714340957388309,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.001500653982988285,
+      "loss": 5.2045,
+      "step": 7676
+    },
+    {
+      "epoch": 0.87154762284875,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00149920658811404,
+      "loss": 5.2114,
+      "step": 7677
+    },
+    {
+      "epoch": 0.8716611499586691,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.001497757906705904,
+      "loss": 5.186,
+      "step": 7678
+    },
+    {
+      "epoch": 0.8717746770685881,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0014963079434325822,
+      "loss": 5.1795,
+      "step": 7679
+    },
+    {
+      "epoch": 0.8718882041785072,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00149485670296691,
+      "loss": 5.2045,
+      "step": 7680
+    },
+    {
+      "epoch": 0.8720017312884263,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0014934041899858398,
+      "loss": 5.2013,
+      "step": 7681
+    },
+    {
+      "epoch": 0.8721152583983454,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.001491950409170424,
+      "loss": 5.2121,
+      "step": 7682
+    },
+    {
+      "epoch": 0.8722287855082644,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0014904953652058021,
+      "loss": 5.2104,
+      "step": 7683
+    },
+    {
+      "epoch": 0.8723423126181835,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0014890390627811838,
+      "loss": 5.1955,
+      "step": 7684
+    },
+    {
+      "epoch": 0.8724558397281026,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0014875815065898339,
+      "loss": 5.2033,
+      "step": 7685
+    },
+    {
+      "epoch": 0.8725693668380217,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.001486122701329058,
+      "loss": 5.1966,
+      "step": 7686
+    },
+    {
+      "epoch": 0.8726828939479407,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0014846626517001883,
+      "loss": 5.2291,
+      "step": 7687
+    },
+    {
+      "epoch": 0.8727964210578598,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0014832013624085654,
+      "loss": 5.1995,
+      "step": 7688
+    },
+    {
+      "epoch": 0.8729099481677789,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0014817388381635262,
+      "loss": 5.2014,
+      "step": 7689
+    },
+    {
+      "epoch": 0.873023475277698,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0014802750836783877,
+      "loss": 5.2039,
+      "step": 7690
+    },
+    {
+      "epoch": 0.873137002387617,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0014788101036704304,
+      "loss": 5.1871,
+      "step": 7691
+    },
+    {
+      "epoch": 0.8732505294975361,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0014773439028608858,
+      "loss": 5.1953,
+      "step": 7692
+    },
+    {
+      "epoch": 0.8733640566074552,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.001475876485974918,
+      "loss": 5.1917,
+      "step": 7693
+    },
+    {
+      "epoch": 0.8734775837173743,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0014744078577416122,
+      "loss": 5.2164,
+      "step": 7694
+    },
+    {
+      "epoch": 0.8735911108272933,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0014729380228939559,
+      "loss": 5.1981,
+      "step": 7695
+    },
+    {
+      "epoch": 0.8737046379372124,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0014714669861688257,
+      "loss": 5.1998,
+      "step": 7696
+    },
+    {
+      "epoch": 0.8738181650471315,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0014699947523069716,
+      "loss": 5.2217,
+      "step": 7697
+    },
+    {
+      "epoch": 0.8739316921570506,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0014685213260530016,
+      "loss": 5.1956,
+      "step": 7698
+    },
+    {
+      "epoch": 0.8740452192669697,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0014670467121553662,
+      "loss": 5.2216,
+      "step": 7699
+    },
+    {
+      "epoch": 0.8741587463768887,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.001465570915366344,
+      "loss": 5.2032,
+      "step": 7700
+    },
+    {
+      "epoch": 0.8742722734868078,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0014640939404420251,
+      "loss": 5.2171,
+      "step": 7701
+    },
+    {
+      "epoch": 0.8743858005967269,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0014626157921422964,
+      "loss": 5.2001,
+      "step": 7702
+    },
+    {
+      "epoch": 0.874499327706646,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.001461136475230827,
+      "loss": 5.2108,
+      "step": 7703
+    },
+    {
+      "epoch": 0.874612854816565,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0014596559944750507,
+      "loss": 5.2096,
+      "step": 7704
+    },
+    {
+      "epoch": 0.8747263819264841,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.001458174354646154,
+      "loss": 5.1888,
+      "step": 7705
+    },
+    {
+      "epoch": 0.8748399090364032,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0014566915605190571,
+      "loss": 5.1753,
+      "step": 7706
+    },
+    {
+      "epoch": 0.8749534361463223,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0014552076168724013,
+      "loss": 5.2088,
+      "step": 7707
+    },
+    {
+      "epoch": 0.8750669632562413,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0014537225284885313,
+      "loss": 5.2031,
+      "step": 7708
+    },
+    {
+      "epoch": 0.8751804903661604,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0014522363001534823,
+      "loss": 5.1987,
+      "step": 7709
+    },
+    {
+      "epoch": 0.8752940174760795,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0014507489366569625,
+      "loss": 5.1918,
+      "step": 7710
+    },
+    {
+      "epoch": 0.8754075445859986,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0014492604427923381,
+      "loss": 5.2174,
+      "step": 7711
+    },
+    {
+      "epoch": 0.8755210716959176,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0014477708233566191,
+      "loss": 5.2064,
+      "step": 7712
+    },
+    {
+      "epoch": 0.8756345988058367,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0014462800831504426,
+      "loss": 5.2025,
+      "step": 7713
+    },
+    {
+      "epoch": 0.8757481259157558,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.001444788226978057,
+      "loss": 5.2152,
+      "step": 7714
+    },
+    {
+      "epoch": 0.8758616530256749,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0014432952596473074,
+      "loss": 5.2251,
+      "step": 7715
+    },
+    {
+      "epoch": 0.875975180135594,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0014418011859696211,
+      "loss": 5.1927,
+      "step": 7716
+    },
+    {
+      "epoch": 0.876088707245513,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.001440306010759989,
+      "loss": 5.2099,
+      "step": 7717
+    },
+    {
+      "epoch": 0.8762022343554321,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0014388097388369529,
+      "loss": 5.1982,
+      "step": 7718
+    },
+    {
+      "epoch": 0.8763157614653512,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.001437312375022589,
+      "loss": 5.1945,
+      "step": 7719
+    },
+    {
+      "epoch": 0.8764292885752702,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0014358139241424923,
+      "loss": 5.2065,
+      "step": 7720
+    },
+    {
+      "epoch": 0.8765428156851893,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.001434314391025761,
+      "loss": 5.2015,
+      "step": 7721
+    },
+    {
+      "epoch": 0.8766563427951084,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0014328137805049808,
+      "loss": 5.2127,
+      "step": 7722
+    },
+    {
+      "epoch": 0.8767698699050275,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0014313120974162104,
+      "loss": 5.1895,
+      "step": 7723
+    },
+    {
+      "epoch": 0.8768833970149466,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0014298093465989641,
+      "loss": 5.2186,
+      "step": 7724
+    },
+    {
+      "epoch": 0.8769969241248656,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.001428305532896198,
+      "loss": 5.2036,
+      "step": 7725
+    },
+    {
+      "epoch": 0.8771104512347847,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0014268006611542936,
+      "loss": 5.1912,
+      "step": 7726
+    },
+    {
+      "epoch": 0.8772239783447038,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0014252947362230412,
+      "loss": 5.1829,
+      "step": 7727
+    },
+    {
+      "epoch": 0.8773375054546229,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0014237877629556263,
+      "loss": 5.2052,
+      "step": 7728
+    },
+    {
+      "epoch": 0.8774510325645419,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0014222797462086123,
+      "loss": 5.2,
+      "step": 7729
+    },
+    {
+      "epoch": 0.877564559674461,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0014207706908419256,
+      "loss": 5.194,
+      "step": 7730
+    },
+    {
+      "epoch": 0.8776780867843801,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00141926060171884,
+      "loss": 5.204,
+      "step": 7731
+    },
+    {
+      "epoch": 0.8777916138942992,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0014177494837059608,
+      "loss": 5.2019,
+      "step": 7732
+    },
+    {
+      "epoch": 0.8779051410042182,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0014162373416732087,
+      "loss": 5.2007,
+      "step": 7733
+    },
+    {
+      "epoch": 0.8780186681141373,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0014147241804938046,
+      "loss": 5.2203,
+      "step": 7734
+    },
+    {
+      "epoch": 0.8781321952240564,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0014132100050442543,
+      "loss": 5.2002,
+      "step": 7735
+    },
+    {
+      "epoch": 0.8782457223339755,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0014116948202043322,
+      "loss": 5.2191,
+      "step": 7736
+    },
+    {
+      "epoch": 0.8783592494438945,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0014101786308570652,
+      "loss": 5.2027,
+      "step": 7737
+    },
+    {
+      "epoch": 0.8784727765538136,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0014086614418887182,
+      "loss": 5.1952,
+      "step": 7738
+    },
+    {
+      "epoch": 0.8785863036637327,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0014071432581887772,
+      "loss": 5.2017,
+      "step": 7739
+    },
+    {
+      "epoch": 0.8786998307736518,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0014056240846499336,
+      "loss": 5.2191,
+      "step": 7740
+    },
+    {
+      "epoch": 0.8788133578835708,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0014041039261680692,
+      "loss": 5.1863,
+      "step": 7741
+    },
+    {
+      "epoch": 0.8789268849934899,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0014025827876422404,
+      "loss": 5.1991,
+      "step": 7742
+    },
+    {
+      "epoch": 0.879040412103409,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.001401060673974661,
+      "loss": 5.1886,
+      "step": 7743
+    },
+    {
+      "epoch": 0.8791539392133281,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.001399537590070688,
+      "loss": 5.1917,
+      "step": 7744
+    },
+    {
+      "epoch": 0.8792674663232471,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0013980135408388056,
+      "loss": 5.1942,
+      "step": 7745
+    },
+    {
+      "epoch": 0.8793809934331662,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0013964885311906082,
+      "loss": 5.235,
+      "step": 7746
+    },
+    {
+      "epoch": 0.8794945205430853,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0013949625660407859,
+      "loss": 5.1829,
+      "step": 7747
+    },
+    {
+      "epoch": 0.8796080476530044,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0013934356503071078,
+      "loss": 5.2025,
+      "step": 7748
+    },
+    {
+      "epoch": 0.8797215747629235,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0013919077889104066,
+      "loss": 5.2172,
+      "step": 7749
+    },
+    {
+      "epoch": 0.8798351018728425,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.001390378986774563,
+      "loss": 5.1914,
+      "step": 7750
+    },
+    {
+      "epoch": 0.8799486289827616,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0013888492488264887,
+      "loss": 5.1833,
+      "step": 7751
+    },
+    {
+      "epoch": 0.8800621560926807,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0013873185799961115,
+      "loss": 5.2095,
+      "step": 7752
+    },
+    {
+      "epoch": 0.8801756832025998,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00138578698521636,
+      "loss": 5.2073,
+      "step": 7753
+    },
+    {
+      "epoch": 0.8802892103125188,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0013842544694231457,
+      "loss": 5.192,
+      "step": 7754
+    },
+    {
+      "epoch": 0.8804027374224379,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0013827210375553485,
+      "loss": 5.1857,
+      "step": 7755
+    },
+    {
+      "epoch": 0.880516264532357,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0013811866945548017,
+      "loss": 5.1908,
+      "step": 7756
+    },
+    {
+      "epoch": 0.8806297916422761,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0013796514453662733,
+      "loss": 5.2199,
+      "step": 7757
+    },
+    {
+      "epoch": 0.8807433187521951,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0013781152949374528,
+      "loss": 5.2044,
+      "step": 7758
+    },
+    {
+      "epoch": 0.8808568458621142,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0013765782482189337,
+      "loss": 5.2011,
+      "step": 7759
+    },
+    {
+      "epoch": 0.8809703729720333,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0013750403101641983,
+      "loss": 5.1926,
+      "step": 7760
+    },
+    {
+      "epoch": 0.8810839000819524,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.001373501485729601,
+      "loss": 5.2098,
+      "step": 7761
+    },
+    {
+      "epoch": 0.8811974271918714,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0013719617798743531,
+      "loss": 5.1877,
+      "step": 7762
+    },
+    {
+      "epoch": 0.8813109543017905,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0013704211975605067,
+      "loss": 5.1889,
+      "step": 7763
+    },
+    {
+      "epoch": 0.8814244814117096,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0013688797437529376,
+      "loss": 5.202,
+      "step": 7764
+    },
+    {
+      "epoch": 0.8815380085216287,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0013673374234193314,
+      "loss": 5.1935,
+      "step": 7765
+    },
+    {
+      "epoch": 0.8816515356315477,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0013657942415301653,
+      "loss": 5.2271,
+      "step": 7766
+    },
+    {
+      "epoch": 0.8817650627414668,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0013642502030586931,
+      "loss": 5.1846,
+      "step": 7767
+    },
+    {
+      "epoch": 0.8818785898513859,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00136270531298093,
+      "loss": 5.2102,
+      "step": 7768
+    },
+    {
+      "epoch": 0.881992116961305,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0013611595762756349,
+      "loss": 5.185,
+      "step": 7769
+    },
+    {
+      "epoch": 0.882105644071224,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0013596129979242951,
+      "loss": 5.1883,
+      "step": 7770
+    },
+    {
+      "epoch": 0.8822191711811431,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.001358065582911111,
+      "loss": 5.202,
+      "step": 7771
+    },
+    {
+      "epoch": 0.8823326982910622,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0013565173362229785,
+      "loss": 5.1774,
+      "step": 7772
+    },
+    {
+      "epoch": 0.8824462254009813,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0013549682628494744,
+      "loss": 5.1813,
+      "step": 7773
+    },
+    {
+      "epoch": 0.8825597525109004,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0013534183677828397,
+      "loss": 5.1968,
+      "step": 7774
+    },
+    {
+      "epoch": 0.8826732796208194,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0013518676560179627,
+      "loss": 5.2038,
+      "step": 7775
+    },
+    {
+      "epoch": 0.8827868067307385,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0013503161325523645,
+      "loss": 5.1878,
+      "step": 7776
+    },
+    {
+      "epoch": 0.8829003338406576,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0013487638023861818,
+      "loss": 5.1836,
+      "step": 7777
+    },
+    {
+      "epoch": 0.8830138609505767,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0013472106705221512,
+      "loss": 5.1969,
+      "step": 7778
+    },
+    {
+      "epoch": 0.8831273880604957,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0013456567419655927,
+      "loss": 5.1836,
+      "step": 7779
+    },
+    {
+      "epoch": 0.8832409151704148,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0013441020217243942,
+      "loss": 5.2085,
+      "step": 7780
+    },
+    {
+      "epoch": 0.8833544422803339,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0013425465148089945,
+      "loss": 5.2036,
+      "step": 7781
+    },
+    {
+      "epoch": 0.883467969390253,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.001340990226232368,
+      "loss": 5.1866,
+      "step": 7782
+    },
+    {
+      "epoch": 0.883581496500172,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0013394331610100077,
+      "loss": 5.1936,
+      "step": 7783
+    },
+    {
+      "epoch": 0.8836950236100911,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0013378753241599104,
+      "loss": 5.1605,
+      "step": 7784
+    },
+    {
+      "epoch": 0.8838085507200102,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.001336316720702559,
+      "loss": 5.1877,
+      "step": 7785
+    },
+    {
+      "epoch": 0.8839220778299293,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0013347573556609074,
+      "loss": 5.1972,
+      "step": 7786
+    },
+    {
+      "epoch": 0.8840356049398483,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0013331972340603628,
+      "loss": 5.1899,
+      "step": 7787
+    },
+    {
+      "epoch": 0.8841491320497674,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0013316363609287723,
+      "loss": 5.2045,
+      "step": 7788
+    },
+    {
+      "epoch": 0.8842626591596865,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0013300747412964035,
+      "loss": 5.1771,
+      "step": 7789
+    },
+    {
+      "epoch": 0.8843761862696056,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0013285123801959303,
+      "loss": 5.1944,
+      "step": 7790
+    },
+    {
+      "epoch": 0.8844897133795246,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0013269492826624158,
+      "loss": 5.2104,
+      "step": 7791
+    },
+    {
+      "epoch": 0.8846032404894437,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0013253854537332978,
+      "loss": 5.2055,
+      "step": 7792
+    },
+    {
+      "epoch": 0.8847167675993628,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0013238208984483696,
+      "loss": 5.1728,
+      "step": 7793
+    },
+    {
+      "epoch": 0.8848302947092819,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0013222556218497655,
+      "loss": 5.203,
+      "step": 7794
+    },
+    {
+      "epoch": 0.884943821819201,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0013206896289819456,
+      "loss": 5.201,
+      "step": 7795
+    },
+    {
+      "epoch": 0.88505734892912,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0013191229248916772,
+      "loss": 5.2059,
+      "step": 7796
+    },
+    {
+      "epoch": 0.8851708760390391,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0013175555146280199,
+      "loss": 5.1975,
+      "step": 7797
+    },
+    {
+      "epoch": 0.8852844031489582,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0013159874032423094,
+      "loss": 5.197,
+      "step": 7798
+    },
+    {
+      "epoch": 0.8853979302588773,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0013144185957881408,
+      "loss": 5.1856,
+      "step": 7799
+    },
+    {
+      "epoch": 0.8855114573687963,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0013128490973213522,
+      "loss": 5.2092,
+      "step": 7800
+    },
+    {
+      "epoch": 0.8856249844787154,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0013112789129000094,
+      "loss": 5.1998,
+      "step": 7801
+    },
+    {
+      "epoch": 0.8857385115886345,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0013097080475843875,
+      "loss": 5.187,
+      "step": 7802
+    },
+    {
+      "epoch": 0.8858520386985536,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.001308136506436957,
+      "loss": 5.1945,
+      "step": 7803
+    },
+    {
+      "epoch": 0.8859655658084726,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0013065642945223664,
+      "loss": 5.1942,
+      "step": 7804
+    },
+    {
+      "epoch": 0.8860790929183917,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0013049914169074256,
+      "loss": 5.1937,
+      "step": 7805
+    },
+    {
+      "epoch": 0.8861926200283108,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0013034178786610895,
+      "loss": 5.195,
+      "step": 7806
+    },
+    {
+      "epoch": 0.8863061471382299,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0013018436848544431,
+      "loss": 5.2196,
+      "step": 7807
+    },
+    {
+      "epoch": 0.8864196742481489,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.001300268840560683,
+      "loss": 5.2071,
+      "step": 7808
+    },
+    {
+      "epoch": 0.886533201358068,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0012986933508551024,
+      "loss": 5.1843,
+      "step": 7809
+    },
+    {
+      "epoch": 0.8866467284679871,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0012971172208150754,
+      "loss": 5.188,
+      "step": 7810
+    },
+    {
+      "epoch": 0.8867602555779062,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0012955404555200385,
+      "loss": 5.1877,
+      "step": 7811
+    },
+    {
+      "epoch": 0.8868737826878252,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.001293963060051476,
+      "loss": 5.1989,
+      "step": 7812
+    },
+    {
+      "epoch": 0.8869873097977443,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0012923850394929031,
+      "loss": 5.187,
+      "step": 7813
+    },
+    {
+      "epoch": 0.8871008369076634,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0012908063989298493,
+      "loss": 5.1738,
+      "step": 7814
+    },
+    {
+      "epoch": 0.8872143640175825,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0012892271434498423,
+      "loss": 5.1949,
+      "step": 7815
+    },
+    {
+      "epoch": 0.8873278911275015,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0012876472781423916,
+      "loss": 5.1852,
+      "step": 7816
+    },
+    {
+      "epoch": 0.8874414182374206,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0012860668080989723,
+      "loss": 5.1872,
+      "step": 7817
+    },
+    {
+      "epoch": 0.8875549453473397,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0012844857384130075,
+      "loss": 5.1924,
+      "step": 7818
+    },
+    {
+      "epoch": 0.8876684724572588,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0012829040741798535,
+      "loss": 5.1844,
+      "step": 7819
+    },
+    {
+      "epoch": 0.8877819995671778,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0012813218204967823,
+      "loss": 5.1962,
+      "step": 7820
+    },
+    {
+      "epoch": 0.8878955266770969,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0012797389824629661,
+      "loss": 5.1862,
+      "step": 7821
+    },
+    {
+      "epoch": 0.888009053787016,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00127815556517946,
+      "loss": 5.1679,
+      "step": 7822
+    },
+    {
+      "epoch": 0.8881225808969351,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0012765715737491857,
+      "loss": 5.1899,
+      "step": 7823
+    },
+    {
+      "epoch": 0.8882361080068542,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0012749870132769147,
+      "loss": 5.1907,
+      "step": 7824
+    },
+    {
+      "epoch": 0.8883496351167732,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.001273401888869254,
+      "loss": 5.1936,
+      "step": 7825
+    },
+    {
+      "epoch": 0.8884631622266923,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0012718162056346266,
+      "loss": 5.1973,
+      "step": 7826
+    },
+    {
+      "epoch": 0.8885766893366114,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0012702299686832561,
+      "loss": 5.1864,
+      "step": 7827
+    },
+    {
+      "epoch": 0.8886902164465305,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0012686431831271524,
+      "loss": 5.1934,
+      "step": 7828
+    },
+    {
+      "epoch": 0.8888037435564495,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0012670558540800916,
+      "loss": 5.1848,
+      "step": 7829
+    },
+    {
+      "epoch": 0.8889172706663686,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0012654679866576021,
+      "loss": 5.1932,
+      "step": 7830
+    },
+    {
+      "epoch": 0.8890307977762877,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0012638795859769476,
+      "loss": 5.1923,
+      "step": 7831
+    },
+    {
+      "epoch": 0.8891443248862068,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00126229065715711,
+      "loss": 5.2058,
+      "step": 7832
+    },
+    {
+      "epoch": 0.8892578519961258,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.001260701205318773,
+      "loss": 5.191,
+      "step": 7833
+    },
+    {
+      "epoch": 0.8893713791060449,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0012591112355843062,
+      "loss": 5.1946,
+      "step": 7834
+    },
+    {
+      "epoch": 0.889484906215964,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0012575207530777486,
+      "loss": 5.2004,
+      "step": 7835
+    },
+    {
+      "epoch": 0.8895984333258831,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0012559297629247906,
+      "loss": 5.1966,
+      "step": 7836
+    },
+    {
+      "epoch": 0.8897119604358021,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0012543382702527596,
+      "loss": 5.1583,
+      "step": 7837
+    },
+    {
+      "epoch": 0.8898254875457212,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.001252746280190602,
+      "loss": 5.1799,
+      "step": 7838
+    },
+    {
+      "epoch": 0.8899390146556403,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0012511537978688678,
+      "loss": 5.203,
+      "step": 7839
+    },
+    {
+      "epoch": 0.8900525417655594,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0012495608284196925,
+      "loss": 5.1752,
+      "step": 7840
+    },
+    {
+      "epoch": 0.8901660688754784,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0012479673769767818,
+      "loss": 5.1947,
+      "step": 7841
+    },
+    {
+      "epoch": 0.8902795959853975,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0012463734486753953,
+      "loss": 5.1873,
+      "step": 7842
+    },
+    {
+      "epoch": 0.8903931230953166,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0012447790486523288,
+      "loss": 5.1958,
+      "step": 7843
+    },
+    {
+      "epoch": 0.8905066502052358,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0012431841820458981,
+      "loss": 5.1801,
+      "step": 7844
+    },
+    {
+      "epoch": 0.8906201773151549,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0012415888539959233,
+      "loss": 5.1912,
+      "step": 7845
+    },
+    {
+      "epoch": 0.8907337044250739,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0012399930696437114,
+      "loss": 5.1867,
+      "step": 7846
+    },
+    {
+      "epoch": 0.890847231534993,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0012383968341320402,
+      "loss": 5.228,
+      "step": 7847
+    },
+    {
+      "epoch": 0.8909607586449121,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0012368001526051407,
+      "loss": 5.1707,
+      "step": 7848
+    },
+    {
+      "epoch": 0.8910742857548312,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0012352030302086815,
+      "loss": 5.1916,
+      "step": 7849
+    },
+    {
+      "epoch": 0.8911878128647502,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0012336054720897527,
+      "loss": 5.1911,
+      "step": 7850
+    },
+    {
+      "epoch": 0.8913013399746693,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.001232007483396848,
+      "loss": 5.1767,
+      "step": 7851
+    },
+    {
+      "epoch": 0.8914148670845884,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0012304090692798486,
+      "loss": 5.2039,
+      "step": 7852
+    },
+    {
+      "epoch": 0.8915283941945075,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0012288102348900077,
+      "loss": 5.2016,
+      "step": 7853
+    },
+    {
+      "epoch": 0.8916419213044265,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0012272109853799316,
+      "loss": 5.207,
+      "step": 7854
+    },
+    {
+      "epoch": 0.8917554484143456,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0012256113259035651,
+      "loss": 5.2131,
+      "step": 7855
+    },
+    {
+      "epoch": 0.8918689755242647,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0012240112616161744,
+      "loss": 5.1855,
+      "step": 7856
+    },
+    {
+      "epoch": 0.8919825026341838,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0012224107976743297,
+      "loss": 5.1954,
+      "step": 7857
+    },
+    {
+      "epoch": 0.8920960297441028,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00122080993923589,
+      "loss": 5.1868,
+      "step": 7858
+    },
+    {
+      "epoch": 0.8922095568540219,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0012192086914599846,
+      "loss": 5.1843,
+      "step": 7859
+    },
+    {
+      "epoch": 0.892323083963941,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0012176070595069988,
+      "loss": 5.1921,
+      "step": 7860
+    },
+    {
+      "epoch": 0.8924366110738601,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0012160050485385547,
+      "loss": 5.1675,
+      "step": 7861
+    },
+    {
+      "epoch": 0.8925501381837792,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0012144026637174968,
+      "loss": 5.1671,
+      "step": 7862
+    },
+    {
+      "epoch": 0.8926636652936982,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.001212799910207874,
+      "loss": 5.1872,
+      "step": 7863
+    },
+    {
+      "epoch": 0.8927771924036173,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0012111967931749232,
+      "loss": 5.1853,
+      "step": 7864
+    },
+    {
+      "epoch": 0.8928907195135364,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0012095933177850536,
+      "loss": 5.1713,
+      "step": 7865
+    },
+    {
+      "epoch": 0.8930042466234555,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0012079894892058282,
+      "loss": 5.1619,
+      "step": 7866
+    },
+    {
+      "epoch": 0.8931177737333745,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0012063853126059497,
+      "loss": 5.2004,
+      "step": 7867
+    },
+    {
+      "epoch": 0.8932313008432936,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0012047807931552406,
+      "loss": 5.2044,
+      "step": 7868
+    },
+    {
+      "epoch": 0.8933448279532127,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0012031759360246297,
+      "loss": 5.1908,
+      "step": 7869
+    },
+    {
+      "epoch": 0.8934583550631318,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0012015707463861333,
+      "loss": 5.1902,
+      "step": 7870
+    },
+    {
+      "epoch": 0.8935718821730508,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0011999652294128402,
+      "loss": 5.2204,
+      "step": 7871
+    },
+    {
+      "epoch": 0.8936854092829699,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0011983593902788927,
+      "loss": 5.1862,
+      "step": 7872
+    },
+    {
+      "epoch": 0.893798936392889,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0011967532341594727,
+      "loss": 5.1808,
+      "step": 7873
+    },
+    {
+      "epoch": 0.8939124635028081,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0011951467662307822,
+      "loss": 5.1677,
+      "step": 7874
+    },
+    {
+      "epoch": 0.8940259906127271,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0011935399916700297,
+      "loss": 5.1634,
+      "step": 7875
+    },
+    {
+      "epoch": 0.8941395177226462,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0011919329156554112,
+      "loss": 5.1936,
+      "step": 7876
+    },
+    {
+      "epoch": 0.8942530448325653,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0011903255433660933,
+      "loss": 5.1836,
+      "step": 7877
+    },
+    {
+      "epoch": 0.8943665719424844,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0011887178799821992,
+      "loss": 5.1945,
+      "step": 7878
+    },
+    {
+      "epoch": 0.8944800990524034,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0011871099306847886,
+      "loss": 5.205,
+      "step": 7879
+    },
+    {
+      "epoch": 0.8945936261623225,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0011855017006558437,
+      "loss": 5.1798,
+      "step": 7880
+    },
+    {
+      "epoch": 0.8947071532722416,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0011838931950782505,
+      "loss": 5.1999,
+      "step": 7881
+    },
+    {
+      "epoch": 0.8948206803821607,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0011822844191357843,
+      "loss": 5.1847,
+      "step": 7882
+    },
+    {
+      "epoch": 0.8949342074920797,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0011806753780130904,
+      "loss": 5.1886,
+      "step": 7883
+    },
+    {
+      "epoch": 0.8950477346019988,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0011790660768956692,
+      "loss": 5.2067,
+      "step": 7884
+    },
+    {
+      "epoch": 0.8951612617119179,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0011774565209698587,
+      "loss": 5.1954,
+      "step": 7885
+    },
+    {
+      "epoch": 0.895274788821837,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0011758467154228194,
+      "loss": 5.2096,
+      "step": 7886
+    },
+    {
+      "epoch": 0.895388315931756,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0011742366654425144,
+      "loss": 5.1883,
+      "step": 7887
+    },
+    {
+      "epoch": 0.8955018430416751,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0011726263762176956,
+      "loss": 5.1892,
+      "step": 7888
+    },
+    {
+      "epoch": 0.8956153701515942,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.001171015852937886,
+      "loss": 5.1557,
+      "step": 7889
+    },
+    {
+      "epoch": 0.8957288972615133,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0011694051007933627,
+      "loss": 5.1942,
+      "step": 7890
+    },
+    {
+      "epoch": 0.8958424243714324,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0011677941249751396,
+      "loss": 5.1717,
+      "step": 7891
+    },
+    {
+      "epoch": 0.8959559514813514,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0011661829306749525,
+      "loss": 5.179,
+      "step": 7892
+    },
+    {
+      "epoch": 0.8960694785912705,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0011645715230852415,
+      "loss": 5.1974,
+      "step": 7893
+    },
+    {
+      "epoch": 0.8961830057011896,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0011629599073991327,
+      "loss": 5.1684,
+      "step": 7894
+    },
+    {
+      "epoch": 0.8962965328111087,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0011613480888104243,
+      "loss": 5.1968,
+      "step": 7895
+    },
+    {
+      "epoch": 0.8964100599210277,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0011597360725135667,
+      "loss": 5.2139,
+      "step": 7896
+    },
+    {
+      "epoch": 0.8965235870309468,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0011581238637036499,
+      "loss": 5.188,
+      "step": 7897
+    },
+    {
+      "epoch": 0.8966371141408659,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0011565114675763822,
+      "loss": 5.1818,
+      "step": 7898
+    },
+    {
+      "epoch": 0.896750641250785,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0011548988893280761,
+      "loss": 5.1698,
+      "step": 7899
+    },
+    {
+      "epoch": 0.896864168360704,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0011532861341556316,
+      "loss": 5.1732,
+      "step": 7900
+    },
+    {
+      "epoch": 0.8969776954706231,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0011516732072565186,
+      "loss": 5.1746,
+      "step": 7901
+    },
+    {
+      "epoch": 0.8970912225805422,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0011500601138287596,
+      "loss": 5.1823,
+      "step": 7902
+    },
+    {
+      "epoch": 0.8972047496904613,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0011484468590709153,
+      "loss": 5.1689,
+      "step": 7903
+    },
+    {
+      "epoch": 0.8973182768003803,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0011468334481820656,
+      "loss": 5.1704,
+      "step": 7904
+    },
+    {
+      "epoch": 0.8974318039102994,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0011452198863617926,
+      "loss": 5.1858,
+      "step": 7905
+    },
+    {
+      "epoch": 0.8975453310202185,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0011436061788101665,
+      "loss": 5.172,
+      "step": 7906
+    },
+    {
+      "epoch": 0.8976588581301376,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0011419923307277263,
+      "loss": 5.1728,
+      "step": 7907
+    },
+    {
+      "epoch": 0.8977723852400566,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0011403783473154635,
+      "loss": 5.1778,
+      "step": 7908
+    },
+    {
+      "epoch": 0.8978859123499757,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0011387642337748068,
+      "loss": 5.1722,
+      "step": 7909
+    },
+    {
+      "epoch": 0.8979994394598948,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0011371499953076027,
+      "loss": 5.1951,
+      "step": 7910
+    },
+    {
+      "epoch": 0.8981129665698139,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0011355356371161025,
+      "loss": 5.2079,
+      "step": 7911
+    },
+    {
+      "epoch": 0.898226493679733,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0011339211644029413,
+      "loss": 5.1825,
+      "step": 7912
+    },
+    {
+      "epoch": 0.898340020789652,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0011323065823711242,
+      "loss": 5.2076,
+      "step": 7913
+    },
+    {
+      "epoch": 0.8984535478995711,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0011306918962240083,
+      "loss": 5.1994,
+      "step": 7914
+    },
+    {
+      "epoch": 0.8985670750094902,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0011290771111652866,
+      "loss": 5.1968,
+      "step": 7915
+    },
+    {
+      "epoch": 0.8986806021194093,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0011274622323989712,
+      "loss": 5.1793,
+      "step": 7916
+    },
+    {
+      "epoch": 0.8987941292293283,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0011258472651293748,
+      "loss": 5.1857,
+      "step": 7917
+    },
+    {
+      "epoch": 0.8989076563392474,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0011242322145610965,
+      "loss": 5.1803,
+      "step": 7918
+    },
+    {
+      "epoch": 0.8990211834491665,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0011226170858990039,
+      "loss": 5.2064,
+      "step": 7919
+    },
+    {
+      "epoch": 0.8991347105590856,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0011210018843482155,
+      "loss": 5.1755,
+      "step": 7920
+    },
+    {
+      "epoch": 0.8992482376690046,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0011193866151140849,
+      "loss": 5.1773,
+      "step": 7921
+    },
+    {
+      "epoch": 0.8993617647789237,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0011177712834021848,
+      "loss": 5.1708,
+      "step": 7922
+    },
+    {
+      "epoch": 0.8994752918888428,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0011161558944182877,
+      "loss": 5.1751,
+      "step": 7923
+    },
+    {
+      "epoch": 0.8995888189987619,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0011145404533683518,
+      "loss": 5.176,
+      "step": 7924
+    },
+    {
+      "epoch": 0.8997023461086809,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.001112924965458503,
+      "loss": 5.2079,
+      "step": 7925
+    },
+    {
+      "epoch": 0.8998158732186,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0011113094358950176,
+      "loss": 5.1855,
+      "step": 7926
+    },
+    {
+      "epoch": 0.8999294003285191,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0011096938698843064,
+      "loss": 5.1644,
+      "step": 7927
+    },
+    {
+      "epoch": 0.9000429274384382,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0011080782726328982,
+      "loss": 5.1846,
+      "step": 7928
+    },
+    {
+      "epoch": 0.9001564545483572,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0011064626493474219,
+      "loss": 5.1596,
+      "step": 7929
+    },
+    {
+      "epoch": 0.9002699816582763,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0011048470052345905,
+      "loss": 5.1702,
+      "step": 7930
+    },
+    {
+      "epoch": 0.9003835087681954,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0011032313455011837,
+      "loss": 5.1814,
+      "step": 7931
+    },
+    {
+      "epoch": 0.9004970358781145,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.001101615675354032,
+      "loss": 5.1957,
+      "step": 7932
+    },
+    {
+      "epoch": 0.9006105629880335,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0011,
+      "loss": 5.16,
+      "step": 7933
+    },
+    {
+      "epoch": 0.9007240900979526,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.001098384324645968,
+      "loss": 5.1868,
+      "step": 7934
+    },
+    {
+      "epoch": 0.9008376172078717,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0010967686544988166,
+      "loss": 5.1864,
+      "step": 7935
+    },
+    {
+      "epoch": 0.9009511443177908,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0010951529947654103,
+      "loss": 5.1908,
+      "step": 7936
+    },
+    {
+      "epoch": 0.9010646714277099,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.001093537350652578,
+      "loss": 5.1767,
+      "step": 7937
+    },
+    {
+      "epoch": 0.9011781985376289,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.001091921727367102,
+      "loss": 5.166,
+      "step": 7938
+    },
+    {
+      "epoch": 0.901291725647548,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0010903061301156937,
+      "loss": 5.1636,
+      "step": 7939
+    },
+    {
+      "epoch": 0.9014052527574671,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0010886905641049828,
+      "loss": 5.175,
+      "step": 7940
+    },
+    {
+      "epoch": 0.9015187798673862,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0010870750345414973,
+      "loss": 5.1638,
+      "step": 7941
+    },
+    {
+      "epoch": 0.9016323069773052,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0010854595466316484,
+      "loss": 5.1865,
+      "step": 7942
+    },
+    {
+      "epoch": 0.9017458340872243,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0010838441055817127,
+      "loss": 5.1731,
+      "step": 7943
+    },
+    {
+      "epoch": 0.9018593611971434,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0010822287165978156,
+      "loss": 5.1617,
+      "step": 7944
+    },
+    {
+      "epoch": 0.9019728883070625,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.001080613384885915,
+      "loss": 5.1774,
+      "step": 7945
+    },
+    {
+      "epoch": 0.9020864154169815,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0010789981156517847,
+      "loss": 5.179,
+      "step": 7946
+    },
+    {
+      "epoch": 0.9021999425269006,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0010773829141009963,
+      "loss": 5.1648,
+      "step": 7947
+    },
+    {
+      "epoch": 0.9023134696368197,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0010757677854389036,
+      "loss": 5.1832,
+      "step": 7948
+    },
+    {
+      "epoch": 0.9024269967467388,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0010741527348706254,
+      "loss": 5.1857,
+      "step": 7949
+    },
+    {
+      "epoch": 0.9025405238566578,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.001072537767601029,
+      "loss": 5.177,
+      "step": 7950
+    },
+    {
+      "epoch": 0.9026540509665769,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0010709228888347133,
+      "loss": 5.1861,
+      "step": 7951
+    },
+    {
+      "epoch": 0.902767578076496,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0010693081037759916,
+      "loss": 5.1883,
+      "step": 7952
+    },
+    {
+      "epoch": 0.9028811051864151,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.001067693417628876,
+      "loss": 5.1767,
+      "step": 7953
+    },
+    {
+      "epoch": 0.9029946322963341,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.001066078835597059,
+      "loss": 5.1763,
+      "step": 7954
+    },
+    {
+      "epoch": 0.9031081594062532,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0010644643628838976,
+      "loss": 5.1664,
+      "step": 7955
+    },
+    {
+      "epoch": 0.9032216865161723,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0010628500046923975,
+      "loss": 5.1894,
+      "step": 7956
+    },
+    {
+      "epoch": 0.9033352136260914,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0010612357662251938,
+      "loss": 5.2043,
+      "step": 7957
+    },
+    {
+      "epoch": 0.9034487407360104,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.001059621652684537,
+      "loss": 5.1603,
+      "step": 7958
+    },
+    {
+      "epoch": 0.9035622678459295,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0010580076692722738,
+      "loss": 5.1801,
+      "step": 7959
+    },
+    {
+      "epoch": 0.9036757949558486,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0010563938211898337,
+      "loss": 5.1819,
+      "step": 7960
+    },
+    {
+      "epoch": 0.9037893220657677,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0010547801136382075,
+      "loss": 5.1908,
+      "step": 7961
+    },
+    {
+      "epoch": 0.9039028491756868,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0010531665518179348,
+      "loss": 5.1787,
+      "step": 7962
+    },
+    {
+      "epoch": 0.9040163762856058,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0010515531409290846,
+      "loss": 5.1688,
+      "step": 7963
+    },
+    {
+      "epoch": 0.9041299033955249,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0010499398861712405,
+      "loss": 5.1948,
+      "step": 7964
+    },
+    {
+      "epoch": 0.904243430505444,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0010483267927434818,
+      "loss": 5.1724,
+      "step": 7965
+    },
+    {
+      "epoch": 0.904356957615363,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0010467138658443683,
+      "loss": 5.1701,
+      "step": 7966
+    },
+    {
+      "epoch": 0.9044704847252821,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.001045101110671924,
+      "loss": 5.1979,
+      "step": 7967
+    },
+    {
+      "epoch": 0.9045840118352012,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0010434885324236182,
+      "loss": 5.175,
+      "step": 7968
+    },
+    {
+      "epoch": 0.9046975389451203,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0010418761362963502,
+      "loss": 5.1852,
+      "step": 7969
+    },
+    {
+      "epoch": 0.9048110660550394,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0010402639274864334,
+      "loss": 5.1887,
+      "step": 7970
+    },
+    {
+      "epoch": 0.9049245931649584,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0010386519111895763,
+      "loss": 5.1572,
+      "step": 7971
+    },
+    {
+      "epoch": 0.9050381202748775,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0010370400926008677,
+      "loss": 5.1821,
+      "step": 7972
+    },
+    {
+      "epoch": 0.9051516473847966,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0010354284769147587,
+      "loss": 5.1717,
+      "step": 7973
+    },
+    {
+      "epoch": 0.9052651744947157,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0010338170693250475,
+      "loss": 5.1817,
+      "step": 7974
+    },
+    {
+      "epoch": 0.9053787016046347,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0010322058750248605,
+      "loss": 5.1866,
+      "step": 7975
+    },
+    {
+      "epoch": 0.9054922287145538,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0010305948992066377,
+      "loss": 5.1933,
+      "step": 7976
+    },
+    {
+      "epoch": 0.9056057558244729,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0010289841470621142,
+      "loss": 5.1607,
+      "step": 7977
+    },
+    {
+      "epoch": 0.905719282934392,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0010273736237823045,
+      "loss": 5.1882,
+      "step": 7978
+    },
+    {
+      "epoch": 0.905832810044311,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.001025763334557486,
+      "loss": 5.1713,
+      "step": 7979
+    },
+    {
+      "epoch": 0.9059463371542301,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0010241532845771814,
+      "loss": 5.17,
+      "step": 7980
+    },
+    {
+      "epoch": 0.9060598642641492,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0010225434790301414,
+      "loss": 5.1675,
+      "step": 7981
+    },
+    {
+      "epoch": 0.9061733913740683,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0010209339231043314,
+      "loss": 5.1693,
+      "step": 7982
+    },
+    {
+      "epoch": 0.9062869184839873,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00101932462198691,
+      "loss": 5.1671,
+      "step": 7983
+    },
+    {
+      "epoch": 0.9064004455939064,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.001017715580864216,
+      "loss": 5.1734,
+      "step": 7984
+    },
+    {
+      "epoch": 0.9065139727038255,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0010161068049217494,
+      "loss": 5.1768,
+      "step": 7985
+    },
+    {
+      "epoch": 0.9066274998137446,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0010144982993441564,
+      "loss": 5.1915,
+      "step": 7986
+    },
+    {
+      "epoch": 0.9067410269236637,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0010128900693152114,
+      "loss": 5.1684,
+      "step": 7987
+    },
+    {
+      "epoch": 0.9068545540335827,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0010112821200178009,
+      "loss": 5.1797,
+      "step": 7988
+    },
+    {
+      "epoch": 0.9069680811435018,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0010096744566339064,
+      "loss": 5.1773,
+      "step": 7989
+    },
+    {
+      "epoch": 0.9070816082534209,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.001008067084344589,
+      "loss": 5.1864,
+      "step": 7990
+    },
+    {
+      "epoch": 0.90719513536334,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0010064600083299704,
+      "loss": 5.1773,
+      "step": 7991
+    },
+    {
+      "epoch": 0.907308662473259,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.001004853233769218,
+      "loss": 5.1919,
+      "step": 7992
+    },
+    {
+      "epoch": 0.9074221895831781,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0010032467658405277,
+      "loss": 5.1791,
+      "step": 7993
+    },
+    {
+      "epoch": 0.9075357166930972,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0010016406097211074,
+      "loss": 5.162,
+      "step": 7994
+    },
+    {
+      "epoch": 0.9076492438030163,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0010000347705871597,
+      "loss": 5.1776,
+      "step": 7995
+    },
+    {
+      "epoch": 0.9077627709129353,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0009984292536138666,
+      "loss": 5.1733,
+      "step": 7996
+    },
+    {
+      "epoch": 0.9078762980228544,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0009968240639753702,
+      "loss": 5.1716,
+      "step": 7997
+    },
+    {
+      "epoch": 0.9079898251327735,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0009952192068447597,
+      "loss": 5.172,
+      "step": 7998
+    },
+    {
+      "epoch": 0.9081033522426926,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0009936146873940506,
+      "loss": 5.182,
+      "step": 7999
+    },
+    {
+      "epoch": 0.9082168793526116,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0009920105107941717,
+      "loss": 5.1707,
+      "step": 8000
+    },
+    {
+      "epoch": 0.9083304064625307,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0009904066822149467,
+      "loss": 5.175,
+      "step": 8001
+    },
+    {
+      "epoch": 0.9084439335724498,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0009888032068250767,
+      "loss": 5.1907,
+      "step": 8002
+    },
+    {
+      "epoch": 0.9085574606823689,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0009872000897921261,
+      "loss": 5.1779,
+      "step": 8003
+    },
+    {
+      "epoch": 0.9086709877922879,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0009855973362825033,
+      "loss": 5.1776,
+      "step": 8004
+    },
+    {
+      "epoch": 0.908784514902207,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0009839949514614454,
+      "loss": 5.1975,
+      "step": 8005
+    },
+    {
+      "epoch": 0.9088980420121261,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0009823929404930015,
+      "loss": 5.1914,
+      "step": 8006
+    },
+    {
+      "epoch": 0.9090115691220452,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0009807913085400155,
+      "loss": 5.1892,
+      "step": 8007
+    },
+    {
+      "epoch": 0.9091250962319642,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0009791900607641103,
+      "loss": 5.1815,
+      "step": 8008
+    },
+    {
+      "epoch": 0.9092386233418833,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0009775892023256704,
+      "loss": 5.1674,
+      "step": 8009
+    },
+    {
+      "epoch": 0.9093521504518024,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.000975988738383826,
+      "loss": 5.1825,
+      "step": 8010
+    },
+    {
+      "epoch": 0.9094656775617215,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0009743886740964352,
+      "loss": 5.1813,
+      "step": 8011
+    },
+    {
+      "epoch": 0.9095792046716406,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0009727890146200687,
+      "loss": 5.1862,
+      "step": 8012
+    },
+    {
+      "epoch": 0.9096927317815596,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0009711897651099925,
+      "loss": 5.189,
+      "step": 8013
+    },
+    {
+      "epoch": 0.9098062588914787,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0009695909307201513,
+      "loss": 5.1727,
+      "step": 8014
+    },
+    {
+      "epoch": 0.9099197860013978,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0009679925166031525,
+      "loss": 5.1764,
+      "step": 8015
+    },
+    {
+      "epoch": 0.9100333131113169,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0009663945279102477,
+      "loss": 5.1844,
+      "step": 8016
+    },
+    {
+      "epoch": 0.9101468402212359,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0009647969697913186,
+      "loss": 5.1706,
+      "step": 8017
+    },
+    {
+      "epoch": 0.910260367331155,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0009631998473948596,
+      "loss": 5.1549,
+      "step": 8018
+    },
+    {
+      "epoch": 0.9103738944410741,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00096160316586796,
+      "loss": 5.192,
+      "step": 8019
+    },
+    {
+      "epoch": 0.9104874215509932,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0009600069303562887,
+      "loss": 5.1895,
+      "step": 8020
+    },
+    {
+      "epoch": 0.9106009486609122,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0009584111460040767,
+      "loss": 5.195,
+      "step": 8021
+    },
+    {
+      "epoch": 0.9107144757708313,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0009568158179541022,
+      "loss": 5.1845,
+      "step": 8022
+    },
+    {
+      "epoch": 0.9108280028807504,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0009552209513476718,
+      "loss": 5.1668,
+      "step": 8023
+    },
+    {
+      "epoch": 0.9109415299906695,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0009536265513246047,
+      "loss": 5.1755,
+      "step": 8024
+    },
+    {
+      "epoch": 0.9110550571005885,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0009520326230232181,
+      "loss": 5.1808,
+      "step": 8025
+    },
+    {
+      "epoch": 0.9111685842105076,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0009504391715803076,
+      "loss": 5.1793,
+      "step": 8026
+    },
+    {
+      "epoch": 0.9112821113204267,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0009488462021311324,
+      "loss": 5.1889,
+      "step": 8027
+    },
+    {
+      "epoch": 0.9113956384303458,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.000947253719809398,
+      "loss": 5.1645,
+      "step": 8028
+    },
+    {
+      "epoch": 0.9115091655402648,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0009456617297472407,
+      "loss": 5.1742,
+      "step": 8029
+    },
+    {
+      "epoch": 0.9116226926501839,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0009440702370752097,
+      "loss": 5.167,
+      "step": 8030
+    },
+    {
+      "epoch": 0.911736219760103,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0009424792469222516,
+      "loss": 5.158,
+      "step": 8031
+    },
+    {
+      "epoch": 0.9118497468700221,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0009408887644156938,
+      "loss": 5.1752,
+      "step": 8032
+    },
+    {
+      "epoch": 0.9119632739799411,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.000939298794681227,
+      "loss": 5.1599,
+      "step": 8033
+    },
+    {
+      "epoch": 0.9120768010898602,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0009377093428428903,
+      "loss": 5.1586,
+      "step": 8034
+    },
+    {
+      "epoch": 0.9121903281997793,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0009361204140230524,
+      "loss": 5.1629,
+      "step": 8035
+    },
+    {
+      "epoch": 0.9123038553096984,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.000934532013342398,
+      "loss": 5.161,
+      "step": 8036
+    },
+    {
+      "epoch": 0.9124173824196175,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0009329441459199089,
+      "loss": 5.172,
+      "step": 8037
+    },
+    {
+      "epoch": 0.9125309095295365,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0009313568168728477,
+      "loss": 5.17,
+      "step": 8038
+    },
+    {
+      "epoch": 0.9126444366394556,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0009297700313167439,
+      "loss": 5.1936,
+      "step": 8039
+    },
+    {
+      "epoch": 0.9127579637493747,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0009281837943653737,
+      "loss": 5.1874,
+      "step": 8040
+    },
+    {
+      "epoch": 0.9128714908592938,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000926598111130746,
+      "loss": 5.1661,
+      "step": 8041
+    },
+    {
+      "epoch": 0.9129850179692128,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0009250129867230852,
+      "loss": 5.161,
+      "step": 8042
+    },
+    {
+      "epoch": 0.9130985450791319,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0009234284262508146,
+      "loss": 5.1861,
+      "step": 8043
+    },
+    {
+      "epoch": 0.913212072189051,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0009218444348205401,
+      "loss": 5.176,
+      "step": 8044
+    },
+    {
+      "epoch": 0.9133255992989701,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0009202610175370338,
+      "loss": 5.1729,
+      "step": 8045
+    },
+    {
+      "epoch": 0.9134391264088891,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0009186781795032178,
+      "loss": 5.1907,
+      "step": 8046
+    },
+    {
+      "epoch": 0.9135526535188082,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0009170959258201468,
+      "loss": 5.1914,
+      "step": 8047
+    },
+    {
+      "epoch": 0.9136661806287273,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0009155142615869927,
+      "loss": 5.1819,
+      "step": 8048
+    },
+    {
+      "epoch": 0.9137797077386464,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0009139331919010281,
+      "loss": 5.1969,
+      "step": 8049
+    },
+    {
+      "epoch": 0.9138932348485654,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0009123527218576085,
+      "loss": 5.1699,
+      "step": 8050
+    },
+    {
+      "epoch": 0.9140067619584845,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0009107728565501581,
+      "loss": 5.1461,
+      "step": 8051
+    },
+    {
+      "epoch": 0.9141202890684036,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0009091936010701511,
+      "loss": 5.1691,
+      "step": 8052
+    },
+    {
+      "epoch": 0.9142338161783227,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0009076149605070969,
+      "loss": 5.1581,
+      "step": 8053
+    },
+    {
+      "epoch": 0.9143473432882417,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.000906036939948524,
+      "loss": 5.1846,
+      "step": 8054
+    },
+    {
+      "epoch": 0.9144608703981608,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0009044595444799615,
+      "loss": 5.1796,
+      "step": 8055
+    },
+    {
+      "epoch": 0.9145743975080799,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0009028827791849245,
+      "loss": 5.1764,
+      "step": 8056
+    },
+    {
+      "epoch": 0.914687924617999,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0009013066491448976,
+      "loss": 5.1737,
+      "step": 8057
+    },
+    {
+      "epoch": 0.914801451727918,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008997311594393172,
+      "loss": 5.1796,
+      "step": 8058
+    },
+    {
+      "epoch": 0.9149149788378371,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008981563151455572,
+      "loss": 5.1749,
+      "step": 8059
+    },
+    {
+      "epoch": 0.9150285059477562,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0008965821213389105,
+      "loss": 5.1714,
+      "step": 8060
+    },
+    {
+      "epoch": 0.9151420330576753,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008950085830925745,
+      "loss": 5.1859,
+      "step": 8061
+    },
+    {
+      "epoch": 0.9152555601675944,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008934357054776336,
+      "loss": 5.1832,
+      "step": 8062
+    },
+    {
+      "epoch": 0.9153690872775134,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008918634935630432,
+      "loss": 5.1824,
+      "step": 8063
+    },
+    {
+      "epoch": 0.9154826143874325,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008902919524156128,
+      "loss": 5.1731,
+      "step": 8064
+    },
+    {
+      "epoch": 0.9155961414973516,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0008887210870999912,
+      "loss": 5.1876,
+      "step": 8065
+    },
+    {
+      "epoch": 0.9157096686072707,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008871509026786479,
+      "loss": 5.1597,
+      "step": 8066
+    },
+    {
+      "epoch": 0.9158231957171897,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008855814042118593,
+      "loss": 5.1837,
+      "step": 8067
+    },
+    {
+      "epoch": 0.9159367228271088,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0008840125967576907,
+      "loss": 5.1709,
+      "step": 8068
+    },
+    {
+      "epoch": 0.9160502499370279,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008824444853719802,
+      "loss": 5.1555,
+      "step": 8069
+    },
+    {
+      "epoch": 0.916163777046947,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008808770751083231,
+      "loss": 5.15,
+      "step": 8070
+    },
+    {
+      "epoch": 0.916277304156866,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008793103710180545,
+      "loss": 5.1665,
+      "step": 8071
+    },
+    {
+      "epoch": 0.9163908312667851,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008777443781502345,
+      "loss": 5.1557,
+      "step": 8072
+    },
+    {
+      "epoch": 0.9165043583767042,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.000876179101551631,
+      "loss": 5.1617,
+      "step": 8073
+    },
+    {
+      "epoch": 0.9166178854866233,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008746145462667026,
+      "loss": 5.1653,
+      "step": 8074
+    },
+    {
+      "epoch": 0.9167314125965423,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008730507173375841,
+      "loss": 5.1468,
+      "step": 8075
+    },
+    {
+      "epoch": 0.9168449397064614,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008714876198040699,
+      "loss": 5.1362,
+      "step": 8076
+    },
+    {
+      "epoch": 0.9169584668163805,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008699252587035969,
+      "loss": 5.1758,
+      "step": 8077
+    },
+    {
+      "epoch": 0.9170719939262996,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008683636390712282,
+      "loss": 5.1589,
+      "step": 8078
+    },
+    {
+      "epoch": 0.9171855210362186,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008668027659396374,
+      "loss": 5.1656,
+      "step": 8079
+    },
+    {
+      "epoch": 0.9172990481461377,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.000865242644339093,
+      "loss": 5.1928,
+      "step": 8080
+    },
+    {
+      "epoch": 0.9174125752560568,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008636832792974409,
+      "loss": 5.1834,
+      "step": 8081
+    },
+    {
+      "epoch": 0.9175261023659759,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008621246758400898,
+      "loss": 5.1582,
+      "step": 8082
+    },
+    {
+      "epoch": 0.917639629475895,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0008605668389899925,
+      "loss": 5.1827,
+      "step": 8083
+    },
+    {
+      "epoch": 0.917753156585814,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0008590097737676323,
+      "loss": 5.156,
+      "step": 8084
+    },
+    {
+      "epoch": 0.9178666836957332,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008574534851910059,
+      "loss": 5.1729,
+      "step": 8085
+    },
+    {
+      "epoch": 0.9179802108056523,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0008558979782756062,
+      "loss": 5.1755,
+      "step": 8086
+    },
+    {
+      "epoch": 0.9180937379155714,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008543432580344075,
+      "loss": 5.1685,
+      "step": 8087
+    },
+    {
+      "epoch": 0.9182072650254904,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.000852789329477849,
+      "loss": 5.1662,
+      "step": 8088
+    },
+    {
+      "epoch": 0.9183207921354095,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0008512361976138181,
+      "loss": 5.1779,
+      "step": 8089
+    },
+    {
+      "epoch": 0.9184343192453286,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008496838674476355,
+      "loss": 5.1755,
+      "step": 8090
+    },
+    {
+      "epoch": 0.9185478463552477,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0008481323439820374,
+      "loss": 5.1647,
+      "step": 8091
+    },
+    {
+      "epoch": 0.9186613734651667,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008465816322171605,
+      "loss": 5.1856,
+      "step": 8092
+    },
+    {
+      "epoch": 0.9187749005750858,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008450317371505258,
+      "loss": 5.183,
+      "step": 8093
+    },
+    {
+      "epoch": 0.9188884276850049,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008434826637770216,
+      "loss": 5.162,
+      "step": 8094
+    },
+    {
+      "epoch": 0.919001954794924,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008419344170888892,
+      "loss": 5.1651,
+      "step": 8095
+    },
+    {
+      "epoch": 0.919115481904843,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.000840387002075705,
+      "loss": 5.1786,
+      "step": 8096
+    },
+    {
+      "epoch": 0.9192290090147621,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008388404237243652,
+      "loss": 5.1682,
+      "step": 8097
+    },
+    {
+      "epoch": 0.9193425361246812,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00083729468701907,
+      "loss": 5.1688,
+      "step": 8098
+    },
+    {
+      "epoch": 0.9194560632346003,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008357497969413068,
+      "loss": 5.1825,
+      "step": 8099
+    },
+    {
+      "epoch": 0.9195695903445193,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0008342057584698349,
+      "loss": 5.1891,
+      "step": 8100
+    },
+    {
+      "epoch": 0.9196831174544384,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008326625765806688,
+      "loss": 5.1981,
+      "step": 8101
+    },
+    {
+      "epoch": 0.9197966445643575,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008311202562470626,
+      "loss": 5.1757,
+      "step": 8102
+    },
+    {
+      "epoch": 0.9199101716742766,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008295788024394932,
+      "loss": 5.1916,
+      "step": 8103
+    },
+    {
+      "epoch": 0.9200236987841957,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008280382201256469,
+      "loss": 5.1824,
+      "step": 8104
+    },
+    {
+      "epoch": 0.9201372258941147,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008264985142703991,
+      "loss": 5.1651,
+      "step": 8105
+    },
+    {
+      "epoch": 0.9202507530040338,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008249596898358019,
+      "loss": 5.1896,
+      "step": 8106
+    },
+    {
+      "epoch": 0.9203642801139529,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008234217517810663,
+      "loss": 5.1711,
+      "step": 8107
+    },
+    {
+      "epoch": 0.920477807223872,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008218847050625476,
+      "loss": 5.1697,
+      "step": 8108
+    },
+    {
+      "epoch": 0.920591334333791,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.000820348554633727,
+      "loss": 5.1701,
+      "step": 8109
+    },
+    {
+      "epoch": 0.9207048614437101,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008188133054451987,
+      "loss": 5.1778,
+      "step": 8110
+    },
+    {
+      "epoch": 0.9208183885536292,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008172789624446513,
+      "loss": 5.1575,
+      "step": 8111
+    },
+    {
+      "epoch": 0.9209319156635483,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008157455305768544,
+      "loss": 5.1888,
+      "step": 8112
+    },
+    {
+      "epoch": 0.9210454427734673,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008142130147836402,
+      "loss": 5.1734,
+      "step": 8113
+    },
+    {
+      "epoch": 0.9211589698833864,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008126814200038885,
+      "loss": 5.1834,
+      "step": 8114
+    },
+    {
+      "epoch": 0.9212724969933055,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008111507511735117,
+      "loss": 5.1632,
+      "step": 8115
+    },
+    {
+      "epoch": 0.9213860241032246,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008096210132254373,
+      "loss": 5.1709,
+      "step": 8116
+    },
+    {
+      "epoch": 0.9214995512131436,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008080922110895937,
+      "loss": 5.1507,
+      "step": 8117
+    },
+    {
+      "epoch": 0.9216130783230627,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008065643496928925,
+      "loss": 5.1566,
+      "step": 8118
+    },
+    {
+      "epoch": 0.9217266054329818,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008050374339592144,
+      "loss": 5.1587,
+      "step": 8119
+    },
+    {
+      "epoch": 0.9218401325429009,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008035114688093918,
+      "loss": 5.1667,
+      "step": 8120
+    },
+    {
+      "epoch": 0.92195365965282,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008019864591611945,
+      "loss": 5.1718,
+      "step": 8121
+    },
+    {
+      "epoch": 0.922067186762739,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008004624099293121,
+      "loss": 5.1607,
+      "step": 8122
+    },
+    {
+      "epoch": 0.9221807138726581,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0007989393260253392,
+      "loss": 5.1886,
+      "step": 8123
+    },
+    {
+      "epoch": 0.9222942409825772,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007974172123577599,
+      "loss": 5.1664,
+      "step": 8124
+    },
+    {
+      "epoch": 0.9224077680924962,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0007958960738319305,
+      "loss": 5.1883,
+      "step": 8125
+    },
+    {
+      "epoch": 0.9225212952024153,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007943759153500665,
+      "loss": 5.156,
+      "step": 8126
+    },
+    {
+      "epoch": 0.9226348223123344,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007928567418112229,
+      "loss": 5.1808,
+      "step": 8127
+    },
+    {
+      "epoch": 0.9227483494222535,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007913385581112817,
+      "loss": 5.1537,
+      "step": 8128
+    },
+    {
+      "epoch": 0.9228618765321726,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0007898213691429348,
+      "loss": 5.1715,
+      "step": 8129
+    },
+    {
+      "epoch": 0.9229754036420916,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0007883051797956679,
+      "loss": 5.1693,
+      "step": 8130
+    },
+    {
+      "epoch": 0.9230889307520107,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.000786789994955746,
+      "loss": 5.196,
+      "step": 8131
+    },
+    {
+      "epoch": 0.9232024578619298,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007852758195061955,
+      "loss": 5.1696,
+      "step": 8132
+    },
+    {
+      "epoch": 0.9233159849718489,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007837626583267916,
+      "loss": 5.1719,
+      "step": 8133
+    },
+    {
+      "epoch": 0.9234295120817679,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007822505162940392,
+      "loss": 5.1936,
+      "step": 8134
+    },
+    {
+      "epoch": 0.923543039191687,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0007807393982811599,
+      "loss": 5.1906,
+      "step": 8135
+    },
+    {
+      "epoch": 0.9236565663016061,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007792293091580745,
+      "loss": 5.1816,
+      "step": 8136
+    },
+    {
+      "epoch": 0.9237700934115252,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0007777202537913878,
+      "loss": 5.1559,
+      "step": 8137
+    },
+    {
+      "epoch": 0.9238836205214442,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.000776212237044374,
+      "loss": 5.1688,
+      "step": 8138
+    },
+    {
+      "epoch": 0.9239971476313633,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0007747052637769591,
+      "loss": 5.185,
+      "step": 8139
+    },
+    {
+      "epoch": 0.9241106747412824,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007731993388457066,
+      "loss": 5.1774,
+      "step": 8140
+    },
+    {
+      "epoch": 0.9242242018512015,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.000771694467103802,
+      "loss": 5.1767,
+      "step": 8141
+    },
+    {
+      "epoch": 0.9243377289611205,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0007701906534010361,
+      "loss": 5.169,
+      "step": 8142
+    },
+    {
+      "epoch": 0.9244512560710396,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0007686879025837898,
+      "loss": 5.1879,
+      "step": 8143
+    },
+    {
+      "epoch": 0.9245647831809587,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0007671862194950194,
+      "loss": 5.1811,
+      "step": 8144
+    },
+    {
+      "epoch": 0.9246783102908778,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0007656856089742396,
+      "loss": 5.1636,
+      "step": 8145
+    },
+    {
+      "epoch": 0.9247918374007968,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0007641860758575081,
+      "loss": 5.1663,
+      "step": 8146
+    },
+    {
+      "epoch": 0.9249053645107159,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.000762687624977411,
+      "loss": 5.1732,
+      "step": 8147
+    },
+    {
+      "epoch": 0.925018891620635,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0007611902611630473,
+      "loss": 5.142,
+      "step": 8148
+    },
+    {
+      "epoch": 0.9251324187305541,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0007596939892400112,
+      "loss": 5.1859,
+      "step": 8149
+    },
+    {
+      "epoch": 0.9252459458404731,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007581988140303791,
+      "loss": 5.17,
+      "step": 8150
+    },
+    {
+      "epoch": 0.9253594729503922,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0007567047403526925,
+      "loss": 5.1652,
+      "step": 8151
+    },
+    {
+      "epoch": 0.9254730000603113,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0007552117730219434,
+      "loss": 5.1747,
+      "step": 8152
+    },
+    {
+      "epoch": 0.9255865271702304,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0007537199168495577,
+      "loss": 5.1422,
+      "step": 8153
+    },
+    {
+      "epoch": 0.9257000542801495,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0007522291766433809,
+      "loss": 5.1636,
+      "step": 8154
+    },
+    {
+      "epoch": 0.9258135813900685,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0007507395572076621,
+      "loss": 5.1438,
+      "step": 8155
+    },
+    {
+      "epoch": 0.9259271084999876,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0007492510633430378,
+      "loss": 5.164,
+      "step": 8156
+    },
+    {
+      "epoch": 0.9260406356099067,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0007477636998465178,
+      "loss": 5.1579,
+      "step": 8157
+    },
+    {
+      "epoch": 0.9261541627198258,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0007462774715114688,
+      "loss": 5.1659,
+      "step": 8158
+    },
+    {
+      "epoch": 0.9262676898297448,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0007447923831275989,
+      "loss": 5.1565,
+      "step": 8159
+    },
+    {
+      "epoch": 0.9263812169396639,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0007433084394809431,
+      "loss": 5.1695,
+      "step": 8160
+    },
+    {
+      "epoch": 0.926494744049583,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007418256453538459,
+      "loss": 5.1713,
+      "step": 8161
+    },
+    {
+      "epoch": 0.9266082711595021,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007403440055249491,
+      "loss": 5.1808,
+      "step": 8162
+    },
+    {
+      "epoch": 0.9267217982694211,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0007388635247691733,
+      "loss": 5.175,
+      "step": 8163
+    },
+    {
+      "epoch": 0.9268353253793402,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0007373842078577038,
+      "loss": 5.1892,
+      "step": 8164
+    },
+    {
+      "epoch": 0.9269488524892593,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0007359060595579752,
+      "loss": 5.175,
+      "step": 8165
+    },
+    {
+      "epoch": 0.9270623795991784,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0007344290846336561,
+      "loss": 5.1682,
+      "step": 8166
+    },
+    {
+      "epoch": 0.9271759067090974,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0007329532878446339,
+      "loss": 5.173,
+      "step": 8167
+    },
+    {
+      "epoch": 0.9272894338190165,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0007314786739469987,
+      "loss": 5.1673,
+      "step": 8168
+    },
+    {
+      "epoch": 0.9274029609289356,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0007300052476930286,
+      "loss": 5.172,
+      "step": 8169
+    },
+    {
+      "epoch": 0.9275164880388547,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0007285330138311746,
+      "loss": 5.1901,
+      "step": 8170
+    },
+    {
+      "epoch": 0.9276300151487737,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0007270619771060443,
+      "loss": 5.1783,
+      "step": 8171
+    },
+    {
+      "epoch": 0.9277435422586928,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.000725592142258388,
+      "loss": 5.1678,
+      "step": 8172
+    },
+    {
+      "epoch": 0.9278570693686119,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0007241235140250822,
+      "loss": 5.1475,
+      "step": 8173
+    },
+    {
+      "epoch": 0.927970596478531,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0007226560971391145,
+      "loss": 5.1827,
+      "step": 8174
+    },
+    {
+      "epoch": 0.92808412358845,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00072118989632957,
+      "loss": 5.1874,
+      "step": 8175
+    },
+    {
+      "epoch": 0.9281976506983691,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0007197249163216122,
+      "loss": 5.1481,
+      "step": 8176
+    },
+    {
+      "epoch": 0.9283111778082882,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0007182611618364736,
+      "loss": 5.1714,
+      "step": 8177
+    },
+    {
+      "epoch": 0.9284247049182073,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0007167986375914346,
+      "loss": 5.1695,
+      "step": 8178
+    },
+    {
+      "epoch": 0.9285382320281264,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.000715337348299812,
+      "loss": 5.1633,
+      "step": 8179
+    },
+    {
+      "epoch": 0.9286517591380454,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0007138772986709421,
+      "loss": 5.1806,
+      "step": 8180
+    },
+    {
+      "epoch": 0.9287652862479645,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0007124184934101665,
+      "loss": 5.1536,
+      "step": 8181
+    },
+    {
+      "epoch": 0.9288788133578836,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0007109609372188164,
+      "loss": 5.1775,
+      "step": 8182
+    },
+    {
+      "epoch": 0.9289923404678027,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0007095046347941978,
+      "loss": 5.1505,
+      "step": 8183
+    },
+    {
+      "epoch": 0.9291058675777217,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0007080495908295759,
+      "loss": 5.1821,
+      "step": 8184
+    },
+    {
+      "epoch": 0.9292193946876408,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0007065958100141607,
+      "loss": 5.1717,
+      "step": 8185
+    },
+    {
+      "epoch": 0.9293329217975599,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0007051432970330902,
+      "loss": 5.1634,
+      "step": 8186
+    },
+    {
+      "epoch": 0.929446448907479,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0007036920565674181,
+      "loss": 5.1572,
+      "step": 8187
+    },
+    {
+      "epoch": 0.929559976017398,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0007022420932940962,
+      "loss": 5.1538,
+      "step": 8188
+    },
+    {
+      "epoch": 0.9296735031273171,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.00070079341188596,
+      "loss": 5.1414,
+      "step": 8189
+    },
+    {
+      "epoch": 0.9297870302372362,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.000699346017011715,
+      "loss": 5.1603,
+      "step": 8190
+    },
+    {
+      "epoch": 0.9299005573471553,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0006978999133359192,
+      "loss": 5.1521,
+      "step": 8191
+    },
+    {
+      "epoch": 0.9300140844570743,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006964551055189712,
+      "loss": 5.1734,
+      "step": 8192
+    },
+    {
+      "epoch": 0.9301276115669934,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.000695011598217091,
+      "loss": 5.1705,
+      "step": 8193
+    },
+    {
+      "epoch": 0.9302411386769125,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006935693960823097,
+      "loss": 5.1803,
+      "step": 8194
+    },
+    {
+      "epoch": 0.9303546657868316,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0006921285037624507,
+      "loss": 5.176,
+      "step": 8195
+    },
+    {
+      "epoch": 0.9304681928967506,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0006906889259011165,
+      "loss": 5.1605,
+      "step": 8196
+    },
+    {
+      "epoch": 0.9305817200066697,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0006892506671376742,
+      "loss": 5.1814,
+      "step": 8197
+    },
+    {
+      "epoch": 0.9306952471165888,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0006878137321072394,
+      "loss": 5.1576,
+      "step": 8198
+    },
+    {
+      "epoch": 0.9308087742265079,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0006863781254406611,
+      "loss": 5.184,
+      "step": 8199
+    },
+    {
+      "epoch": 0.930922301336427,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0006849438517645077,
+      "loss": 5.1449,
+      "step": 8200
+    },
+    {
+      "epoch": 0.931035828446346,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0006835109157010528,
+      "loss": 5.1922,
+      "step": 8201
+    },
+    {
+      "epoch": 0.9311493555562651,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006820793218682572,
+      "loss": 5.1593,
+      "step": 8202
+    },
+    {
+      "epoch": 0.9312628826661842,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0006806490748797581,
+      "loss": 5.1539,
+      "step": 8203
+    },
+    {
+      "epoch": 0.9313764097761033,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006792201793448501,
+      "loss": 5.1673,
+      "step": 8204
+    },
+    {
+      "epoch": 0.9314899368860223,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0006777926398684743,
+      "loss": 5.1838,
+      "step": 8205
+    },
+    {
+      "epoch": 0.9316034639959414,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0006763664610512007,
+      "loss": 5.1741,
+      "step": 8206
+    },
+    {
+      "epoch": 0.9317169911058605,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0006749416474892148,
+      "loss": 5.161,
+      "step": 8207
+    },
+    {
+      "epoch": 0.9318305182157796,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0006735182037743013,
+      "loss": 5.1653,
+      "step": 8208
+    },
+    {
+      "epoch": 0.9319440453256986,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.000672096134493831,
+      "loss": 5.1721,
+      "step": 8209
+    },
+    {
+      "epoch": 0.9320575724356177,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0006706754442307456,
+      "loss": 5.1812,
+      "step": 8210
+    },
+    {
+      "epoch": 0.9321710995455368,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0006692561375635414,
+      "loss": 5.1738,
+      "step": 8211
+    },
+    {
+      "epoch": 0.9322846266554559,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0006678382190662568,
+      "loss": 5.1649,
+      "step": 8212
+    },
+    {
+      "epoch": 0.9323981537653749,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0006664216933084562,
+      "loss": 5.1495,
+      "step": 8213
+    },
+    {
+      "epoch": 0.932511680875294,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006650065648552158,
+      "loss": 5.16,
+      "step": 8214
+    },
+    {
+      "epoch": 0.9326252079852131,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0006635928382671077,
+      "loss": 5.1602,
+      "step": 8215
+    },
+    {
+      "epoch": 0.9327387350951322,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0006621805181001876,
+      "loss": 5.1514,
+      "step": 8216
+    },
+    {
+      "epoch": 0.9328522622050512,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0006607696089059775,
+      "loss": 5.1794,
+      "step": 8217
+    },
+    {
+      "epoch": 0.9329657893149703,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0006593601152314532,
+      "loss": 5.1523,
+      "step": 8218
+    },
+    {
+      "epoch": 0.9330793164248894,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0006579520416190272,
+      "loss": 5.1533,
+      "step": 8219
+    },
+    {
+      "epoch": 0.9331928435348085,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0006565453926065377,
+      "loss": 5.155,
+      "step": 8220
+    },
+    {
+      "epoch": 0.9333063706447275,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0006551401727272299,
+      "loss": 5.1594,
+      "step": 8221
+    },
+    {
+      "epoch": 0.9334198977546466,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0006537363865097438,
+      "loss": 5.1779,
+      "step": 8222
+    },
+    {
+      "epoch": 0.9335334248645657,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0006523340384781003,
+      "loss": 5.1652,
+      "step": 8223
+    },
+    {
+      "epoch": 0.9336469519744848,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0006509331331516834,
+      "loss": 5.1586,
+      "step": 8224
+    },
+    {
+      "epoch": 0.9337604790844038,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0006495336750452292,
+      "loss": 5.1692,
+      "step": 8225
+    },
+    {
+      "epoch": 0.9338740061943229,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0006481356686688084,
+      "loss": 5.1772,
+      "step": 8226
+    },
+    {
+      "epoch": 0.933987533304242,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0006467391185278153,
+      "loss": 5.1742,
+      "step": 8227
+    },
+    {
+      "epoch": 0.9341010604141611,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0006453440291229491,
+      "loss": 5.1521,
+      "step": 8228
+    },
+    {
+      "epoch": 0.9342145875240802,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0006439504049502026,
+      "loss": 5.1452,
+      "step": 8229
+    },
+    {
+      "epoch": 0.9343281146339992,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0006425582505008459,
+      "loss": 5.1677,
+      "step": 8230
+    },
+    {
+      "epoch": 0.9344416417439183,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0006411675702614131,
+      "loss": 5.1749,
+      "step": 8231
+    },
+    {
+      "epoch": 0.9345551688538374,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0006397783687136868,
+      "loss": 5.1502,
+      "step": 8232
+    },
+    {
+      "epoch": 0.9346686959637565,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0006383906503346846,
+      "loss": 5.1791,
+      "step": 8233
+    },
+    {
+      "epoch": 0.9347822230736755,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0006370044195966443,
+      "loss": 5.152,
+      "step": 8234
+    },
+    {
+      "epoch": 0.9348957501835946,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0006356196809670089,
+      "loss": 5.144,
+      "step": 8235
+    },
+    {
+      "epoch": 0.9350092772935137,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.000634236438908414,
+      "loss": 5.1706,
+      "step": 8236
+    },
+    {
+      "epoch": 0.9351228044034328,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0006328546978786702,
+      "loss": 5.1641,
+      "step": 8237
+    },
+    {
+      "epoch": 0.9352363315133518,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0006314744623307526,
+      "loss": 5.1651,
+      "step": 8238
+    },
+    {
+      "epoch": 0.9353498586232709,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0006300957367127829,
+      "loss": 5.1847,
+      "step": 8239
+    },
+    {
+      "epoch": 0.93546338573319,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0006287185254680183,
+      "loss": 5.1746,
+      "step": 8240
+    },
+    {
+      "epoch": 0.9355769128431091,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0006273428330348338,
+      "loss": 5.16,
+      "step": 8241
+    },
+    {
+      "epoch": 0.9356904399530281,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0006259686638467119,
+      "loss": 5.1647,
+      "step": 8242
+    },
+    {
+      "epoch": 0.9358039670629472,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0006245960223322241,
+      "loss": 5.1594,
+      "step": 8243
+    },
+    {
+      "epoch": 0.9359174941728663,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0006232249129150201,
+      "loss": 5.1901,
+      "step": 8244
+    },
+    {
+      "epoch": 0.9360310212827854,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0006218553400138102,
+      "loss": 5.1631,
+      "step": 8245
+    },
+    {
+      "epoch": 0.9361445483927044,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0006204873080423549,
+      "loss": 5.1782,
+      "step": 8246
+    },
+    {
+      "epoch": 0.9362580755026235,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0006191208214094484,
+      "loss": 5.1767,
+      "step": 8247
+    },
+    {
+      "epoch": 0.9363716026125426,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0006177558845189029,
+      "loss": 5.1477,
+      "step": 8248
+    },
+    {
+      "epoch": 0.9364851297224617,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0006163925017695389,
+      "loss": 5.177,
+      "step": 8249
+    },
+    {
+      "epoch": 0.9365986568323807,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0006150306775551659,
+      "loss": 5.1655,
+      "step": 8250
+    },
+    {
+      "epoch": 0.9367121839422998,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0006136704162645724,
+      "loss": 5.17,
+      "step": 8251
+    },
+    {
+      "epoch": 0.9368257110522189,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0006123117222815085,
+      "loss": 5.1784,
+      "step": 8252
+    },
+    {
+      "epoch": 0.936939238162138,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0006109545999846751,
+      "loss": 5.1483,
+      "step": 8253
+    },
+    {
+      "epoch": 0.937052765272057,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0006095990537477063,
+      "loss": 5.1718,
+      "step": 8254
+    },
+    {
+      "epoch": 0.9371662923819761,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0006082450879391579,
+      "loss": 5.1824,
+      "step": 8255
+    },
+    {
+      "epoch": 0.9372798194918952,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0006068927069224924,
+      "loss": 5.1844,
+      "step": 8256
+    },
+    {
+      "epoch": 0.9373933466018143,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0006055419150560646,
+      "loss": 5.1544,
+      "step": 8257
+    },
+    {
+      "epoch": 0.9375068737117334,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0006041927166931078,
+      "loss": 5.1615,
+      "step": 8258
+    },
+    {
+      "epoch": 0.9376204008216524,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0006028451161817206,
+      "loss": 5.1449,
+      "step": 8259
+    },
+    {
+      "epoch": 0.9377339279315715,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0006014991178648515,
+      "loss": 5.1657,
+      "step": 8260
+    },
+    {
+      "epoch": 0.9378474550414906,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0006001547260802855,
+      "loss": 5.1387,
+      "step": 8261
+    },
+    {
+      "epoch": 0.9379609821514097,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005988119451606312,
+      "loss": 5.1431,
+      "step": 8262
+    },
+    {
+      "epoch": 0.9380745092613287,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0005974707794333036,
+      "loss": 5.1639,
+      "step": 8263
+    },
+    {
+      "epoch": 0.9381880363712478,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005961312332205157,
+      "loss": 5.1696,
+      "step": 8264
+    },
+    {
+      "epoch": 0.9383015634811669,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005947933108392579,
+      "loss": 5.1578,
+      "step": 8265
+    },
+    {
+      "epoch": 0.938415090591086,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0005934570166012898,
+      "loss": 5.1651,
+      "step": 8266
+    },
+    {
+      "epoch": 0.938528617701005,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005921223548131225,
+      "loss": 5.1502,
+      "step": 8267
+    },
+    {
+      "epoch": 0.9386421448109241,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005907893297760069,
+      "loss": 5.1696,
+      "step": 8268
+    },
+    {
+      "epoch": 0.9387556719208432,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005894579457859194,
+      "loss": 5.1743,
+      "step": 8269
+    },
+    {
+      "epoch": 0.9388691990307623,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005881282071335464,
+      "loss": 5.1559,
+      "step": 8270
+    },
+    {
+      "epoch": 0.9389827261406813,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005868001181042733,
+      "loss": 5.1433,
+      "step": 8271
+    },
+    {
+      "epoch": 0.9390962532506004,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0005854736829781681,
+      "loss": 5.1427,
+      "step": 8272
+    },
+    {
+      "epoch": 0.9392097803605195,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005841489060299701,
+      "loss": 5.1766,
+      "step": 8273
+    },
+    {
+      "epoch": 0.9393233074704386,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.000582825791529073,
+      "loss": 5.1475,
+      "step": 8274
+    },
+    {
+      "epoch": 0.9394368345803576,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0005815043437395144,
+      "loss": 5.1636,
+      "step": 8275
+    },
+    {
+      "epoch": 0.9395503616902767,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005801845669199594,
+      "loss": 5.1669,
+      "step": 8276
+    },
+    {
+      "epoch": 0.9396638888001958,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005788664653236886,
+      "loss": 5.1647,
+      "step": 8277
+    },
+    {
+      "epoch": 0.9397774159101149,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005775500431985838,
+      "loss": 5.1584,
+      "step": 8278
+    },
+    {
+      "epoch": 0.939890943020034,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005762353047871148,
+      "loss": 5.1642,
+      "step": 8279
+    },
+    {
+      "epoch": 0.940004470129953,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.000574922254326324,
+      "loss": 5.1611,
+      "step": 8280
+    },
+    {
+      "epoch": 0.9401179972398721,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005736108960478148,
+      "loss": 5.1681,
+      "step": 8281
+    },
+    {
+      "epoch": 0.9402315243497912,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005723012341777374,
+      "loss": 5.1757,
+      "step": 8282
+    },
+    {
+      "epoch": 0.9403450514597103,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005709932729367744,
+      "loss": 5.1648,
+      "step": 8283
+    },
+    {
+      "epoch": 0.9404585785696293,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005696870165401276,
+      "loss": 5.1755,
+      "step": 8284
+    },
+    {
+      "epoch": 0.9405721056795484,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005683824691975056,
+      "loss": 5.166,
+      "step": 8285
+    },
+    {
+      "epoch": 0.9406856327894675,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005670796351131082,
+      "loss": 5.1804,
+      "step": 8286
+    },
+    {
+      "epoch": 0.9407991598993866,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005657785184856138,
+      "loss": 5.1485,
+      "step": 8287
+    },
+    {
+      "epoch": 0.9409126870093056,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005644791235081668,
+      "loss": 5.1624,
+      "step": 8288
+    },
+    {
+      "epoch": 0.9410262141192247,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.000563181454368362,
+      "loss": 5.1624,
+      "step": 8289
+    },
+    {
+      "epoch": 0.9411397412291438,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0005618855152482334,
+      "loss": 5.166,
+      "step": 8290
+    },
+    {
+      "epoch": 0.9412532683390629,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0005605913103242381,
+      "loss": 5.1648,
+      "step": 8291
+    },
+    {
+      "epoch": 0.9413667954489819,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.000559298843767247,
+      "loss": 5.1579,
+      "step": 8292
+    },
+    {
+      "epoch": 0.941480322558901,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.000558008119742526,
+      "loss": 5.166,
+      "step": 8293
+    },
+    {
+      "epoch": 0.9415938496688201,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0005567191424097268,
+      "loss": 5.1716,
+      "step": 8294
+    },
+    {
+      "epoch": 0.9417073767787392,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005554319159228716,
+      "loss": 5.1862,
+      "step": 8295
+    },
+    {
+      "epoch": 0.9418209038886582,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005541464444303397,
+      "loss": 5.1987,
+      "step": 8296
+    },
+    {
+      "epoch": 0.9419344309985773,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005528627320748554,
+      "loss": 5.161,
+      "step": 8297
+    },
+    {
+      "epoch": 0.9420479581084964,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005515807829934728,
+      "loss": 5.1589,
+      "step": 8298
+    },
+    {
+      "epoch": 0.9421614852184155,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005503006013175642,
+      "loss": 5.1627,
+      "step": 8299
+    },
+    {
+      "epoch": 0.9422750123283345,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005490221911728059,
+      "loss": 5.1718,
+      "step": 8300
+    },
+    {
+      "epoch": 0.9423885394382536,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005477455566791649,
+      "loss": 5.1723,
+      "step": 8301
+    },
+    {
+      "epoch": 0.9425020665481727,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005464707019508856,
+      "loss": 5.1596,
+      "step": 8302
+    },
+    {
+      "epoch": 0.9426155936580918,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005451976310964772,
+      "loss": 5.162,
+      "step": 8303
+    },
+    {
+      "epoch": 0.9427291207680109,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005439263482186992,
+      "loss": 5.1533,
+      "step": 8304
+    },
+    {
+      "epoch": 0.9428426478779299,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005426568574145496,
+      "loss": 5.1798,
+      "step": 8305
+    },
+    {
+      "epoch": 0.942956174987849,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0005413891627752509,
+      "loss": 5.1338,
+      "step": 8306
+    },
+    {
+      "epoch": 0.9430697020977681,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005401232683862367,
+      "loss": 5.1738,
+      "step": 8307
+    },
+    {
+      "epoch": 0.9431832292076872,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0005388591783271399,
+      "loss": 5.1531,
+      "step": 8308
+    },
+    {
+      "epoch": 0.9432967563176062,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005375968966717769,
+      "loss": 5.157,
+      "step": 8309
+    },
+    {
+      "epoch": 0.9434102834275253,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005363364274881375,
+      "loss": 5.1576,
+      "step": 8310
+    },
+    {
+      "epoch": 0.9435238105374444,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005350777748383697,
+      "loss": 5.1734,
+      "step": 8311
+    },
+    {
+      "epoch": 0.9436373376473635,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005338209427787677,
+      "loss": 5.1639,
+      "step": 8312
+    },
+    {
+      "epoch": 0.9437508647572825,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0005325659353597575,
+      "loss": 5.1678,
+      "step": 8313
+    },
+    {
+      "epoch": 0.9438643918672016,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0005313127566258869,
+      "loss": 5.1637,
+      "step": 8314
+    },
+    {
+      "epoch": 0.9439779189771207,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005300614106158077,
+      "loss": 5.1608,
+      "step": 8315
+    },
+    {
+      "epoch": 0.9440914460870398,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005288119013622674,
+      "loss": 5.1805,
+      "step": 8316
+    },
+    {
+      "epoch": 0.9442049731969588,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005275642328920929,
+      "loss": 5.1689,
+      "step": 8317
+    },
+    {
+      "epoch": 0.9443185003068779,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0005263184092261792,
+      "loss": 5.1627,
+      "step": 8318
+    },
+    {
+      "epoch": 0.944432027416797,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005250744343794766,
+      "loss": 5.1598,
+      "step": 8319
+    },
+    {
+      "epoch": 0.9445455545267161,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005238323123609753,
+      "loss": 5.1602,
+      "step": 8320
+    },
+    {
+      "epoch": 0.9446590816366351,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.000522592047173697,
+      "loss": 5.1608,
+      "step": 8321
+    },
+    {
+      "epoch": 0.9447726087465542,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.000521353642814677,
+      "loss": 5.1616,
+      "step": 8322
+    },
+    {
+      "epoch": 0.9448861358564733,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005201171032749552,
+      "loss": 5.1671,
+      "step": 8323
+    },
+    {
+      "epoch": 0.9449996629663924,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0005188824325395605,
+      "loss": 5.1789,
+      "step": 8324
+    },
+    {
+      "epoch": 0.9451131900763114,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005176496345875004,
+      "loss": 5.1621,
+      "step": 8325
+    },
+    {
+      "epoch": 0.9452267171862306,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005164187133917455,
+      "loss": 5.1653,
+      "step": 8326
+    },
+    {
+      "epoch": 0.9453402442961497,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005151896729192194,
+      "loss": 5.164,
+      "step": 8327
+    },
+    {
+      "epoch": 0.9454537714060688,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005139625171307838,
+      "loss": 5.1466,
+      "step": 8328
+    },
+    {
+      "epoch": 0.9455672985159879,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005127372499812276,
+      "loss": 5.1636,
+      "step": 8329
+    },
+    {
+      "epoch": 0.9456808256259069,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0005115138754192517,
+      "loss": 5.1528,
+      "step": 8330
+    },
+    {
+      "epoch": 0.945794352735826,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0005102923973874588,
+      "loss": 5.1666,
+      "step": 8331
+    },
+    {
+      "epoch": 0.9459078798457451,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0005090728198223393,
+      "loss": 5.1435,
+      "step": 8332
+    },
+    {
+      "epoch": 0.9460214069556642,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0005078551466542587,
+      "loss": 5.1327,
+      "step": 8333
+    },
+    {
+      "epoch": 0.9461349340655832,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0005066393818074457,
+      "loss": 5.1594,
+      "step": 8334
+    },
+    {
+      "epoch": 0.9462484611755023,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0005054255291999777,
+      "loss": 5.1938,
+      "step": 8335
+    },
+    {
+      "epoch": 0.9463619882854214,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005042135927437716,
+      "loss": 5.1565,
+      "step": 8336
+    },
+    {
+      "epoch": 0.9464755153953405,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.000503003576344567,
+      "loss": 5.1678,
+      "step": 8337
+    },
+    {
+      "epoch": 0.9465890425052595,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0005017954839019172,
+      "loss": 5.1839,
+      "step": 8338
+    },
+    {
+      "epoch": 0.9467025696151786,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0005005893193091736,
+      "loss": 5.168,
+      "step": 8339
+    },
+    {
+      "epoch": 0.9468160967250977,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0004993850864534762,
+      "loss": 5.1828,
+      "step": 8340
+    },
+    {
+      "epoch": 0.9469296238350168,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0004981827892157389,
+      "loss": 5.1729,
+      "step": 8341
+    },
+    {
+      "epoch": 0.9470431509449359,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0004969824314706371,
+      "loss": 5.1595,
+      "step": 8342
+    },
+    {
+      "epoch": 0.9471566780548549,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0004957840170865964,
+      "loss": 5.1748,
+      "step": 8343
+    },
+    {
+      "epoch": 0.947270205164774,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0004945875499257796,
+      "loss": 5.1526,
+      "step": 8344
+    },
+    {
+      "epoch": 0.9473837322746931,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0004933930338440739,
+      "loss": 5.1646,
+      "step": 8345
+    },
+    {
+      "epoch": 0.9474972593846122,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0004922004726910779,
+      "loss": 5.1554,
+      "step": 8346
+    },
+    {
+      "epoch": 0.9476107864945312,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0004910098703100919,
+      "loss": 5.1661,
+      "step": 8347
+    },
+    {
+      "epoch": 0.9477243136044503,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0004898212305381015,
+      "loss": 5.1527,
+      "step": 8348
+    },
+    {
+      "epoch": 0.9478378407143694,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0004886345572057683,
+      "loss": 5.1607,
+      "step": 8349
+    },
+    {
+      "epoch": 0.9479513678242885,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.00048744985413741725,
+      "loss": 5.1639,
+      "step": 8350
+    },
+    {
+      "epoch": 0.9480648949342075,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0004862671251510229,
+      "loss": 5.1745,
+      "step": 8351
+    },
+    {
+      "epoch": 0.9481784220441266,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.00048508637405819763,
+      "loss": 5.159,
+      "step": 8352
+    },
+    {
+      "epoch": 0.9482919491540457,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.00048390760466418016,
+      "loss": 5.1807,
+      "step": 8353
+    },
+    {
+      "epoch": 0.9484054762639648,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0004827308207678229,
+      "loss": 5.1392,
+      "step": 8354
+    },
+    {
+      "epoch": 0.9485190033738838,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0004815560261615786,
+      "loss": 5.1411,
+      "step": 8355
+    },
+    {
+      "epoch": 0.9486325304838029,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.00048038322463149025,
+      "loss": 5.1527,
+      "step": 8356
+    },
+    {
+      "epoch": 0.948746057593722,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0004792124199571763,
+      "loss": 5.1696,
+      "step": 8357
+    },
+    {
+      "epoch": 0.9488595847036411,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0004780436159118218,
+      "loss": 5.1748,
+      "step": 8358
+    },
+    {
+      "epoch": 0.9489731118135601,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0004768768162621625,
+      "loss": 5.1834,
+      "step": 8359
+    },
+    {
+      "epoch": 0.9490866389234792,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00047571202476847567,
+      "loss": 5.167,
+      "step": 8360
+    },
+    {
+      "epoch": 0.9492001660333983,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00047454924518456595,
+      "loss": 5.158,
+      "step": 8361
+    },
+    {
+      "epoch": 0.9493136931433174,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00047338848125775515,
+      "loss": 5.1563,
+      "step": 8362
+    },
+    {
+      "epoch": 0.9494272202532364,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0004722297367288681,
+      "loss": 5.1888,
+      "step": 8363
+    },
+    {
+      "epoch": 0.9495407473631555,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0004710730153322224,
+      "loss": 5.175,
+      "step": 8364
+    },
+    {
+      "epoch": 0.9496542744730746,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00046991832079561554,
+      "loss": 5.1681,
+      "step": 8365
+    },
+    {
+      "epoch": 0.9497678015829937,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0004687656568403127,
+      "loss": 5.1737,
+      "step": 8366
+    },
+    {
+      "epoch": 0.9498813286929128,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00046761502718103585,
+      "loss": 5.1348,
+      "step": 8367
+    },
+    {
+      "epoch": 0.9499948558028318,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0004664664355259496,
+      "loss": 5.141,
+      "step": 8368
+    },
+    {
+      "epoch": 0.9501083829127509,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.000465319885576652,
+      "loss": 5.1691,
+      "step": 8369
+    },
+    {
+      "epoch": 0.95022191002267,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0004641753810281601,
+      "loss": 5.154,
+      "step": 8370
+    },
+    {
+      "epoch": 0.9503354371325891,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00046303292556890005,
+      "loss": 5.1613,
+      "step": 8371
+    },
+    {
+      "epoch": 0.9504489642425081,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0004618925228806939,
+      "loss": 5.1707,
+      "step": 8372
+    },
+    {
+      "epoch": 0.9505624913524272,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0004607541766387488,
+      "loss": 5.1596,
+      "step": 8373
+    },
+    {
+      "epoch": 0.9506760184623463,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0004596178905116432,
+      "loss": 5.1814,
+      "step": 8374
+    },
+    {
+      "epoch": 0.9507895455722654,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00045848366816131803,
+      "loss": 5.1379,
+      "step": 8375
+    },
+    {
+      "epoch": 0.9509030726821844,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00045735151324306154,
+      "loss": 5.1881,
+      "step": 8376
+    },
+    {
+      "epoch": 0.9510165997921035,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0004562214294055007,
+      "loss": 5.1655,
+      "step": 8377
+    },
+    {
+      "epoch": 0.9511301269020226,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.000455093420290587,
+      "loss": 5.1595,
+      "step": 8378
+    },
+    {
+      "epoch": 0.9512436540119417,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00045396748953358614,
+      "loss": 5.1536,
+      "step": 8379
+    },
+    {
+      "epoch": 0.9513571811218607,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.000452843640763066,
+      "loss": 5.1454,
+      "step": 8380
+    },
+    {
+      "epoch": 0.9514707082317798,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0004517218776008838,
+      "loss": 5.173,
+      "step": 8381
+    },
+    {
+      "epoch": 0.9515842353416989,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0004506022036621767,
+      "loss": 5.1873,
+      "step": 8382
+    },
+    {
+      "epoch": 0.951697762451618,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0004494846225553477,
+      "loss": 5.165,
+      "step": 8383
+    },
+    {
+      "epoch": 0.951811289561537,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00044836913788205636,
+      "loss": 5.1473,
+      "step": 8384
+    },
+    {
+      "epoch": 0.9519248166714561,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00044725575323720424,
+      "loss": 5.1566,
+      "step": 8385
+    },
+    {
+      "epoch": 0.9520383437813752,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0004461444722089277,
+      "loss": 5.1857,
+      "step": 8386
+    },
+    {
+      "epoch": 0.9521518708912943,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00044503529837858116,
+      "loss": 5.1842,
+      "step": 8387
+    },
+    {
+      "epoch": 0.9522653980012133,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00044392823532072976,
+      "loss": 5.1414,
+      "step": 8388
+    },
+    {
+      "epoch": 0.9523789251111324,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0004428232866031352,
+      "loss": 5.1723,
+      "step": 8389
+    },
+    {
+      "epoch": 0.9524924522210515,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0004417204557867458,
+      "loss": 5.1782,
+      "step": 8390
+    },
+    {
+      "epoch": 0.9526059793309706,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0004406197464256847,
+      "loss": 5.1488,
+      "step": 8391
+    },
+    {
+      "epoch": 0.9527195064408897,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.000439521162067237,
+      "loss": 5.1583,
+      "step": 8392
+    },
+    {
+      "epoch": 0.9528330335508087,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00043842470625184095,
+      "loss": 5.1665,
+      "step": 8393
+    },
+    {
+      "epoch": 0.9529465606607278,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0004373303825130741,
+      "loss": 5.1728,
+      "step": 8394
+    },
+    {
+      "epoch": 0.9530600877706469,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00043623819437764357,
+      "loss": 5.1462,
+      "step": 8395
+    },
+    {
+      "epoch": 0.953173614880566,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0004351481453653733,
+      "loss": 5.1407,
+      "step": 8396
+    },
+    {
+      "epoch": 0.953287141990485,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00043406023898919405,
+      "loss": 5.1738,
+      "step": 8397
+    },
+    {
+      "epoch": 0.9534006691004041,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.00043297447875513084,
+      "loss": 5.1405,
+      "step": 8398
+    },
+    {
+      "epoch": 0.9535141962103232,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00043189086816229275,
+      "loss": 5.1716,
+      "step": 8399
+    },
+    {
+      "epoch": 0.9536277233202423,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0004308094107028612,
+      "loss": 5.1796,
+      "step": 8400
+    },
+    {
+      "epoch": 0.9537412504301613,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0004297301098620784,
+      "loss": 5.1668,
+      "step": 8401
+    },
+    {
+      "epoch": 0.9538547775400804,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00042865296911823614,
+      "loss": 5.1635,
+      "step": 8402
+    },
+    {
+      "epoch": 0.9539683046499995,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0004275779919426653,
+      "loss": 5.1601,
+      "step": 8403
+    },
+    {
+      "epoch": 0.9540818317599186,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0004265051817997242,
+      "loss": 5.1738,
+      "step": 8404
+    },
+    {
+      "epoch": 0.9541953588698376,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0004254345421467867,
+      "loss": 5.1473,
+      "step": 8405
+    },
+    {
+      "epoch": 0.9543088859797567,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00042436607643423266,
+      "loss": 5.1403,
+      "step": 8406
+    },
+    {
+      "epoch": 0.9544224130896758,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.00042329978810543467,
+      "loss": 5.1473,
+      "step": 8407
+    },
+    {
+      "epoch": 0.9545359401995949,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0004222356805967502,
+      "loss": 5.173,
+      "step": 8408
+    },
+    {
+      "epoch": 0.9546494673095139,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0004211737573375065,
+      "loss": 5.1485,
+      "step": 8409
+    },
+    {
+      "epoch": 0.954762994419433,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.00042011402174999316,
+      "loss": 5.1655,
+      "step": 8410
+    },
+    {
+      "epoch": 0.9548765215293521,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.00041905647724944787,
+      "loss": 5.175,
+      "step": 8411
+    },
+    {
+      "epoch": 0.9549900486392712,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00041800112724404845,
+      "loss": 5.1521,
+      "step": 8412
+    },
+    {
+      "epoch": 0.9551035757491902,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0004169479751349001,
+      "loss": 5.1531,
+      "step": 8413
+    },
+    {
+      "epoch": 0.9552171028591093,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00041589702431602443,
+      "loss": 5.1519,
+      "step": 8414
+    },
+    {
+      "epoch": 0.9553306299690284,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0004148482781743491,
+      "loss": 5.1755,
+      "step": 8415
+    },
+    {
+      "epoch": 0.9554441570789475,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0004138017400896968,
+      "loss": 5.1596,
+      "step": 8416
+    },
+    {
+      "epoch": 0.9555576841888666,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.00041275741343477457,
+      "loss": 5.1418,
+      "step": 8417
+    },
+    {
+      "epoch": 0.9556712112987856,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00041171530157516143,
+      "loss": 5.1324,
+      "step": 8418
+    },
+    {
+      "epoch": 0.9557847384087047,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00041067540786930034,
+      "loss": 5.1526,
+      "step": 8419
+    },
+    {
+      "epoch": 0.9558982655186238,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0004096377356684842,
+      "loss": 5.1527,
+      "step": 8420
+    },
+    {
+      "epoch": 0.9560117926285429,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00040860228831684787,
+      "loss": 5.1771,
+      "step": 8421
+    },
+    {
+      "epoch": 0.9561253197384619,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00040756906915135493,
+      "loss": 5.1582,
+      "step": 8422
+    },
+    {
+      "epoch": 0.956238846848381,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0004065380815017898,
+      "loss": 5.1622,
+      "step": 8423
+    },
+    {
+      "epoch": 0.9563523739583001,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00040550932869074343,
+      "loss": 5.175,
+      "step": 8424
+    },
+    {
+      "epoch": 0.9564659010682192,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00040448281403360583,
+      "loss": 5.1768,
+      "step": 8425
+    },
+    {
+      "epoch": 0.9565794281781382,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0004034585408385536,
+      "loss": 5.1851,
+      "step": 8426
+    },
+    {
+      "epoch": 0.9566929552880573,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.00040243651240653945,
+      "loss": 5.1442,
+      "step": 8427
+    },
+    {
+      "epoch": 0.9568064823979764,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.00040141673203128247,
+      "loss": 5.1728,
+      "step": 8428
+    },
+    {
+      "epoch": 0.9569200095078955,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00040039920299925593,
+      "loss": 5.1554,
+      "step": 8429
+    },
+    {
+      "epoch": 0.9570335366178145,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00039938392858967895,
+      "loss": 5.1575,
+      "step": 8430
+    },
+    {
+      "epoch": 0.9571470637277336,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0003983709120745035,
+      "loss": 5.1694,
+      "step": 8431
+    },
+    {
+      "epoch": 0.9572605908376527,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00039736015671840543,
+      "loss": 5.1523,
+      "step": 8432
+    },
+    {
+      "epoch": 0.9573741179475718,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00039635166577877323,
+      "loss": 5.1434,
+      "step": 8433
+    },
+    {
+      "epoch": 0.9574876450574908,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0003953454425056984,
+      "loss": 5.1476,
+      "step": 8434
+    },
+    {
+      "epoch": 0.9576011721674099,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0003943414901419635,
+      "loss": 5.1356,
+      "step": 8435
+    },
+    {
+      "epoch": 0.957714699277329,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00039333981192303313,
+      "loss": 5.1555,
+      "step": 8436
+    },
+    {
+      "epoch": 0.9578282263872481,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.000392340411077043,
+      "loss": 5.1687,
+      "step": 8437
+    },
+    {
+      "epoch": 0.9579417534971671,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0003913432908247892,
+      "loss": 5.1538,
+      "step": 8438
+    },
+    {
+      "epoch": 0.9580552806070862,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0003903484543797184,
+      "loss": 5.1586,
+      "step": 8439
+    },
+    {
+      "epoch": 0.9581688077170053,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0003893559049479167,
+      "loss": 5.1481,
+      "step": 8440
+    },
+    {
+      "epoch": 0.9582823348269244,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0003883656457281003,
+      "loss": 5.1598,
+      "step": 8441
+    },
+    {
+      "epoch": 0.9583958619368435,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.000387377679911604,
+      "loss": 5.1739,
+      "step": 8442
+    },
+    {
+      "epoch": 0.9585093890467625,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0003863920106823723,
+      "loss": 5.1644,
+      "step": 8443
+    },
+    {
+      "epoch": 0.9586229161566816,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0003854086412169482,
+      "loss": 5.1607,
+      "step": 8444
+    },
+    {
+      "epoch": 0.9587364432666007,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0003844275746844632,
+      "loss": 5.1585,
+      "step": 8445
+    },
+    {
+      "epoch": 0.9588499703765198,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0003834488142466266,
+      "loss": 5.1447,
+      "step": 8446
+    },
+    {
+      "epoch": 0.9589634974864388,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00038247236305771683,
+      "loss": 5.1594,
+      "step": 8447
+    },
+    {
+      "epoch": 0.9590770245963579,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0003814982242645688,
+      "loss": 5.1618,
+      "step": 8448
+    },
+    {
+      "epoch": 0.959190551706277,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0003805264010065665,
+      "loss": 5.1641,
+      "step": 8449
+    },
+    {
+      "epoch": 0.9593040788161961,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0003795568964156312,
+      "loss": 5.1479,
+      "step": 8450
+    },
+    {
+      "epoch": 0.9594176059261151,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00037858971361621097,
+      "loss": 5.1503,
+      "step": 8451
+    },
+    {
+      "epoch": 0.9595311330360342,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00037762485572527294,
+      "loss": 5.1409,
+      "step": 8452
+    },
+    {
+      "epoch": 0.9596446601459533,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00037666232585229013,
+      "loss": 5.1789,
+      "step": 8453
+    },
+    {
+      "epoch": 0.9597581872558724,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.000375702127099234,
+      "loss": 5.1513,
+      "step": 8454
+    },
+    {
+      "epoch": 0.9598717143657914,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0003747442625605628,
+      "loss": 5.155,
+      "step": 8455
+    },
+    {
+      "epoch": 0.9599852414757105,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0003737887353232129,
+      "loss": 5.1621,
+      "step": 8456
+    },
+    {
+      "epoch": 0.9600987685856296,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00037283554846658727,
+      "loss": 5.1607,
+      "step": 8457
+    },
+    {
+      "epoch": 0.9602122956955487,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00037188470506254747,
+      "loss": 5.1743,
+      "step": 8458
+    },
+    {
+      "epoch": 0.9603258228054677,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0003709362081754021,
+      "loss": 5.178,
+      "step": 8459
+    },
+    {
+      "epoch": 0.9604393499153868,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0003699900608618981,
+      "loss": 5.1516,
+      "step": 8460
+    },
+    {
+      "epoch": 0.9605528770253059,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00036904626617120963,
+      "loss": 5.18,
+      "step": 8461
+    },
+    {
+      "epoch": 0.960666404135225,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00036810482714492965,
+      "loss": 5.1616,
+      "step": 8462
+    },
+    {
+      "epoch": 0.960779931245144,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0003671657468170594,
+      "loss": 5.1806,
+      "step": 8463
+    },
+    {
+      "epoch": 0.9608934583550631,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0003662290282139982,
+      "loss": 5.1625,
+      "step": 8464
+    },
+    {
+      "epoch": 0.9610069854649822,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00036529467435453446,
+      "loss": 5.1943,
+      "step": 8465
+    },
+    {
+      "epoch": 0.9611205125749013,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0003643626882498359,
+      "loss": 5.1392,
+      "step": 8466
+    },
+    {
+      "epoch": 0.9612340396848204,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0003634330729034394,
+      "loss": 5.1505,
+      "step": 8467
+    },
+    {
+      "epoch": 0.9613475667947394,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0003625058313112413,
+      "loss": 5.1745,
+      "step": 8468
+    },
+    {
+      "epoch": 0.9614610939046585,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00036158096646148866,
+      "loss": 5.1437,
+      "step": 8469
+    },
+    {
+      "epoch": 0.9615746210145776,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0003606584813347678,
+      "loss": 5.1422,
+      "step": 8470
+    },
+    {
+      "epoch": 0.9616881481244967,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.000359738378903997,
+      "loss": 5.1627,
+      "step": 8471
+    },
+    {
+      "epoch": 0.9618016752344157,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0003588206621344153,
+      "loss": 5.17,
+      "step": 8472
+    },
+    {
+      "epoch": 0.9619152023443348,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0003579053339835735,
+      "loss": 5.1463,
+      "step": 8473
+    },
+    {
+      "epoch": 0.9620287294542539,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.00035699239740132397,
+      "loss": 5.1765,
+      "step": 8474
+    },
+    {
+      "epoch": 0.962142256564173,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00035608185532981263,
+      "loss": 5.1532,
+      "step": 8475
+    },
+    {
+      "epoch": 0.962255783674092,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00035517371070346805,
+      "loss": 5.164,
+      "step": 8476
+    },
+    {
+      "epoch": 0.9623693107840111,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0003542679664489922,
+      "loss": 5.1611,
+      "step": 8477
+    },
+    {
+      "epoch": 0.9624828378939302,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0003533646254853522,
+      "loss": 5.1444,
+      "step": 8478
+    },
+    {
+      "epoch": 0.9625963650038493,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0003524636907237685,
+      "loss": 5.1598,
+      "step": 8479
+    },
+    {
+      "epoch": 0.9627098921137683,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0003515651650677092,
+      "loss": 5.1917,
+      "step": 8480
+    },
+    {
+      "epoch": 0.9628234192236874,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0003506690514128763,
+      "loss": 5.1478,
+      "step": 8481
+    },
+    {
+      "epoch": 0.9629369463336065,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00034977535264720026,
+      "loss": 5.1298,
+      "step": 8482
+    },
+    {
+      "epoch": 0.9630504734435256,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0003488840716508276,
+      "loss": 5.1423,
+      "step": 8483
+    },
+    {
+      "epoch": 0.9631640005534446,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0003479952112961143,
+      "loss": 5.1577,
+      "step": 8484
+    },
+    {
+      "epoch": 0.9632775276633637,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0003471087744476148,
+      "loss": 5.1698,
+      "step": 8485
+    },
+    {
+      "epoch": 0.9633910547732828,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00034622476396207254,
+      "loss": 5.1411,
+      "step": 8486
+    },
+    {
+      "epoch": 0.9635045818832019,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0003453431826884125,
+      "loss": 5.1627,
+      "step": 8487
+    },
+    {
+      "epoch": 0.963618108993121,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0003444640334677305,
+      "loss": 5.1594,
+      "step": 8488
+    },
+    {
+      "epoch": 0.96373163610304,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00034358731913328465,
+      "loss": 5.1723,
+      "step": 8489
+    },
+    {
+      "epoch": 0.9638451632129591,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0003427130425104857,
+      "loss": 5.1604,
+      "step": 8490
+    },
+    {
+      "epoch": 0.9639586903228782,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00034184120641688907,
+      "loss": 5.1615,
+      "step": 8491
+    },
+    {
+      "epoch": 0.9640722174327973,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.000340971813662184,
+      "loss": 5.1714,
+      "step": 8492
+    },
+    {
+      "epoch": 0.9641857445427163,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00034010486704818665,
+      "loss": 5.1526,
+      "step": 8493
+    },
+    {
+      "epoch": 0.9642992716526354,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00033924036936882864,
+      "loss": 5.1567,
+      "step": 8494
+    },
+    {
+      "epoch": 0.9644127987625545,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00033837832341015105,
+      "loss": 5.1515,
+      "step": 8495
+    },
+    {
+      "epoch": 0.9645263258724736,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0003375187319502919,
+      "loss": 5.1505,
+      "step": 8496
+    },
+    {
+      "epoch": 0.9646398529823926,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00033666159775948024,
+      "loss": 5.1558,
+      "step": 8497
+    },
+    {
+      "epoch": 0.9647533800923117,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00033580692360002585,
+      "loss": 5.1429,
+      "step": 8498
+    },
+    {
+      "epoch": 0.9648669072022308,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00033495471222631,
+      "loss": 5.1636,
+      "step": 8499
+    },
+    {
+      "epoch": 0.9649804343121499,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0003341049663847775,
+      "loss": 5.1804,
+      "step": 8500
+    },
+    {
+      "epoch": 0.9650939614220689,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00033325768881392696,
+      "loss": 5.1479,
+      "step": 8501
+    },
+    {
+      "epoch": 0.965207488531988,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0003324128822443035,
+      "loss": 5.1508,
+      "step": 8502
+    },
+    {
+      "epoch": 0.9653210156419071,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0003315705493984875,
+      "loss": 5.1564,
+      "step": 8503
+    },
+    {
+      "epoch": 0.9654345427518262,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0003307306929910883,
+      "loss": 5.1587,
+      "step": 8504
+    },
+    {
+      "epoch": 0.9655480698617452,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00032989331572873335,
+      "loss": 5.1652,
+      "step": 8505
+    },
+    {
+      "epoch": 0.9656615969716643,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00032905842031006184,
+      "loss": 5.138,
+      "step": 8506
+    },
+    {
+      "epoch": 0.9657751240815834,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0003282260094257131,
+      "loss": 5.146,
+      "step": 8507
+    },
+    {
+      "epoch": 0.9658886511915025,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0003273960857583205,
+      "loss": 5.1465,
+      "step": 8508
+    },
+    {
+      "epoch": 0.9660021783014215,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00032656865198250174,
+      "loss": 5.1534,
+      "step": 8509
+    },
+    {
+      "epoch": 0.9661157054113406,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00032574371076484984,
+      "loss": 5.1687,
+      "step": 8510
+    },
+    {
+      "epoch": 0.9662292325212597,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0003249212647639255,
+      "loss": 5.1535,
+      "step": 8511
+    },
+    {
+      "epoch": 0.9663427596311788,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0003241013166302474,
+      "loss": 5.1669,
+      "step": 8512
+    },
+    {
+      "epoch": 0.9664562867410978,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00032328386900628485,
+      "loss": 5.1484,
+      "step": 8513
+    },
+    {
+      "epoch": 0.9665698138510169,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00032246892452644825,
+      "loss": 5.1524,
+      "step": 8514
+    },
+    {
+      "epoch": 0.966683340960936,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00032165648581708163,
+      "loss": 5.131,
+      "step": 8515
+    },
+    {
+      "epoch": 0.9667968680708551,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00032084655549645256,
+      "loss": 5.141,
+      "step": 8516
+    },
+    {
+      "epoch": 0.9669103951807742,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0003200391361747462,
+      "loss": 5.1481,
+      "step": 8517
+    },
+    {
+      "epoch": 0.9670239222906932,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0003192342304540542,
+      "loss": 5.1569,
+      "step": 8518
+    },
+    {
+      "epoch": 0.9671374494006123,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0003184318409283685,
+      "loss": 5.1414,
+      "step": 8519
+    },
+    {
+      "epoch": 0.9672509765105314,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00031763197018357093,
+      "loss": 5.1676,
+      "step": 8520
+    },
+    {
+      "epoch": 0.9673645036204505,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0003168346207974269,
+      "loss": 5.1565,
+      "step": 8521
+    },
+    {
+      "epoch": 0.9674780307303695,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0003160397953395762,
+      "loss": 5.1599,
+      "step": 8522
+    },
+    {
+      "epoch": 0.9675915578402886,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.00031524749637152374,
+      "loss": 5.1333,
+      "step": 8523
+    },
+    {
+      "epoch": 0.9677050849502077,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0003144577264466336,
+      "loss": 5.1722,
+      "step": 8524
+    },
+    {
+      "epoch": 0.9678186120601268,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00031367048811011824,
+      "loss": 5.1816,
+      "step": 8525
+    },
+    {
+      "epoch": 0.9679321391700458,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0003128857838990322,
+      "loss": 5.1324,
+      "step": 8526
+    },
+    {
+      "epoch": 0.9680456662799649,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00031210361634226286,
+      "loss": 5.1443,
+      "step": 8527
+    },
+    {
+      "epoch": 0.968159193389884,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.000311323987960523,
+      "loss": 5.1518,
+      "step": 8528
+    },
+    {
+      "epoch": 0.9682727204998031,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00031054690126634175,
+      "loss": 5.1591,
+      "step": 8529
+    },
+    {
+      "epoch": 0.9683862476097221,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0003097723587640577,
+      "loss": 5.1504,
+      "step": 8530
+    },
+    {
+      "epoch": 0.9684997747196412,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0003090003629498101,
+      "loss": 5.1482,
+      "step": 8531
+    },
+    {
+      "epoch": 0.9686133018295603,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00030823091631153066,
+      "loss": 5.1625,
+      "step": 8532
+    },
+    {
+      "epoch": 0.9687268289394794,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0003074640213289358,
+      "loss": 5.1462,
+      "step": 8533
+    },
+    {
+      "epoch": 0.9688403560493984,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0003066996804735189,
+      "loss": 5.1582,
+      "step": 8534
+    },
+    {
+      "epoch": 0.9689538831593175,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00030593789620854216,
+      "loss": 5.1627,
+      "step": 8535
+    },
+    {
+      "epoch": 0.9690674102692366,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.000305178670989028,
+      "loss": 5.1492,
+      "step": 8536
+    },
+    {
+      "epoch": 0.9691809373791557,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00030442200726175207,
+      "loss": 5.134,
+      "step": 8537
+    },
+    {
+      "epoch": 0.9692944644890747,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00030366790746523537,
+      "loss": 5.1613,
+      "step": 8538
+    },
+    {
+      "epoch": 0.9694079915989938,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00030291637402973576,
+      "loss": 5.1403,
+      "step": 8539
+    },
+    {
+      "epoch": 0.9695215187089129,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00030216740937724016,
+      "loss": 5.1566,
+      "step": 8540
+    },
+    {
+      "epoch": 0.969635045818832,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0003014210159214575,
+      "loss": 5.1305,
+      "step": 8541
+    },
+    {
+      "epoch": 0.969748572928751,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00030067719606780995,
+      "loss": 5.1384,
+      "step": 8542
+    },
+    {
+      "epoch": 0.9698621000386701,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0002999359522134262,
+      "loss": 5.1418,
+      "step": 8543
+    },
+    {
+      "epoch": 0.9699756271485892,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00029919728674713267,
+      "loss": 5.1481,
+      "step": 8544
+    },
+    {
+      "epoch": 0.9700891542585083,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00029846120204944654,
+      "loss": 5.1791,
+      "step": 8545
+    },
+    {
+      "epoch": 0.9702026813684274,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00029772770049256816,
+      "loss": 5.1543,
+      "step": 8546
+    },
+    {
+      "epoch": 0.9703162084783464,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00029699678444037263,
+      "loss": 5.1671,
+      "step": 8547
+    },
+    {
+      "epoch": 0.9704297355882655,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00029626845624840324,
+      "loss": 5.1639,
+      "step": 8548
+    },
+    {
+      "epoch": 0.9705432626981846,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00029554271826386265,
+      "loss": 5.1705,
+      "step": 8549
+    },
+    {
+      "epoch": 0.9706567898081037,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00029481957282560665,
+      "loss": 5.1375,
+      "step": 8550
+    },
+    {
+      "epoch": 0.9707703169180227,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00029409902226413523,
+      "loss": 5.1314,
+      "step": 8551
+    },
+    {
+      "epoch": 0.9708838440279418,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0002933810689015865,
+      "loss": 5.1703,
+      "step": 8552
+    },
+    {
+      "epoch": 0.9709973711378609,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00029266571505172823,
+      "loss": 5.1458,
+      "step": 8553
+    },
+    {
+      "epoch": 0.97111089824778,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0002919529630199507,
+      "loss": 5.1653,
+      "step": 8554
+    },
+    {
+      "epoch": 0.971224425357699,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00029124281510325896,
+      "loss": 5.1452,
+      "step": 8555
+    },
+    {
+      "epoch": 0.9713379524676181,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.000290535273590266,
+      "loss": 5.1518,
+      "step": 8556
+    },
+    {
+      "epoch": 0.9714514795775372,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00028983034076118547,
+      "loss": 5.1601,
+      "step": 8557
+    },
+    {
+      "epoch": 0.9715650066874563,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00028912801888782293,
+      "loss": 5.1352,
+      "step": 8558
+    },
+    {
+      "epoch": 0.9716785337973753,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00028842831023357025,
+      "loss": 5.1599,
+      "step": 8559
+    },
+    {
+      "epoch": 0.9717920609072944,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00028773121705339755,
+      "loss": 5.1502,
+      "step": 8560
+    },
+    {
+      "epoch": 0.9719055880172135,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0002870367415938462,
+      "loss": 5.1485,
+      "step": 8561
+    },
+    {
+      "epoch": 0.9720191151271326,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0002863448860930206,
+      "loss": 5.1572,
+      "step": 8562
+    },
+    {
+      "epoch": 0.9721326422370516,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0002856556527805826,
+      "loss": 5.1339,
+      "step": 8563
+    },
+    {
+      "epoch": 0.9722461693469707,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00028496904387774263,
+      "loss": 5.1607,
+      "step": 8564
+    },
+    {
+      "epoch": 0.9723596964568898,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00028428506159725443,
+      "loss": 5.1588,
+      "step": 8565
+    },
+    {
+      "epoch": 0.9724732235668089,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00028360370814340546,
+      "loss": 5.1873,
+      "step": 8566
+    },
+    {
+      "epoch": 0.972586750676728,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0002829249857120128,
+      "loss": 5.1185,
+      "step": 8567
+    },
+    {
+      "epoch": 0.9727002777866471,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00028224889649041296,
+      "loss": 5.1456,
+      "step": 8568
+    },
+    {
+      "epoch": 0.9728138048965662,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0002815754426574574,
+      "loss": 5.1567,
+      "step": 8569
+    },
+    {
+      "epoch": 0.9729273320064853,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00028090462638350394,
+      "loss": 5.1474,
+      "step": 8570
+    },
+    {
+      "epoch": 0.9730408591164044,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0002802364498304102,
+      "loss": 5.166,
+      "step": 8571
+    },
+    {
+      "epoch": 0.9731543862263234,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0002795709151515272,
+      "loss": 5.1605,
+      "step": 8572
+    },
+    {
+      "epoch": 0.9732679133362425,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0002789080244916909,
+      "loss": 5.1688,
+      "step": 8573
+    },
+    {
+      "epoch": 0.9733814404461616,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00027824777998721806,
+      "loss": 5.1593,
+      "step": 8574
+    },
+    {
+      "epoch": 0.9734949675560807,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0002775901837658956,
+      "loss": 5.1504,
+      "step": 8575
+    },
+    {
+      "epoch": 0.9736084946659997,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0002769352379469773,
+      "loss": 5.1656,
+      "step": 8576
+    },
+    {
+      "epoch": 0.9737220217759188,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00027628294464117424,
+      "loss": 5.1571,
+      "step": 8577
+    },
+    {
+      "epoch": 0.9738355488858379,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00027563330595065024,
+      "loss": 5.1568,
+      "step": 8578
+    },
+    {
+      "epoch": 0.973949075995757,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0002749863239690132,
+      "loss": 5.1562,
+      "step": 8579
+    },
+    {
+      "epoch": 0.974062603105676,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00027434200078130954,
+      "loss": 5.1615,
+      "step": 8580
+    },
+    {
+      "epoch": 0.9741761302155951,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00027370033846401716,
+      "loss": 5.1633,
+      "step": 8581
+    },
+    {
+      "epoch": 0.9742896573255142,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0002730613390850386,
+      "loss": 5.1638,
+      "step": 8582
+    },
+    {
+      "epoch": 0.9744031844354333,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.00027242500470369454,
+      "loss": 5.1587,
+      "step": 8583
+    },
+    {
+      "epoch": 0.9745167115453524,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0002717913373707167,
+      "loss": 5.1436,
+      "step": 8584
+    },
+    {
+      "epoch": 0.9746302386552714,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00027116033912824247,
+      "loss": 5.1434,
+      "step": 8585
+    },
+    {
+      "epoch": 0.9747437657651905,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00027053201200980643,
+      "loss": 5.1793,
+      "step": 8586
+    },
+    {
+      "epoch": 0.9748572928751096,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0002699063580403359,
+      "loss": 5.1637,
+      "step": 8587
+    },
+    {
+      "epoch": 0.9749708199850287,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0002692833792361424,
+      "loss": 5.1565,
+      "step": 8588
+    },
+    {
+      "epoch": 0.9750843470949477,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0002686630776049172,
+      "loss": 5.1466,
+      "step": 8589
+    },
+    {
+      "epoch": 0.9751978742048668,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0002680454551457228,
+      "loss": 5.1508,
+      "step": 8590
+    },
+    {
+      "epoch": 0.9753114013147859,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.00026743051384898786,
+      "loss": 5.1237,
+      "step": 8591
+    },
+    {
+      "epoch": 0.975424928424705,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00026681825569650037,
+      "loss": 5.1547,
+      "step": 8592
+    },
+    {
+      "epoch": 0.975538455534624,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0002662086826614013,
+      "loss": 5.1528,
+      "step": 8593
+    },
+    {
+      "epoch": 0.9756519826445431,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00026560179670817835,
+      "loss": 5.1782,
+      "step": 8594
+    },
+    {
+      "epoch": 0.9757655097544622,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0002649975997926589,
+      "loss": 5.172,
+      "step": 8595
+    },
+    {
+      "epoch": 0.9758790368643813,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0002643960938620053,
+      "loss": 5.1388,
+      "step": 8596
+    },
+    {
+      "epoch": 0.9759925639743003,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0002637972808547065,
+      "loss": 5.1525,
+      "step": 8597
+    },
+    {
+      "epoch": 0.9761060910842194,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00026320116270057384,
+      "loss": 5.1637,
+      "step": 8598
+    },
+    {
+      "epoch": 0.9762196181941385,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00026260774132073314,
+      "loss": 5.1719,
+      "step": 8599
+    },
+    {
+      "epoch": 0.9763331453040576,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00026201701862761965,
+      "loss": 5.1764,
+      "step": 8600
+    },
+    {
+      "epoch": 0.9764466724139766,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00026142899652497145,
+      "loss": 5.1427,
+      "step": 8601
+    },
+    {
+      "epoch": 0.9765601995238957,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0002608436769078231,
+      "loss": 5.1842,
+      "step": 8602
+    },
+    {
+      "epoch": 0.9766737266338148,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.00026026106166250027,
+      "loss": 5.1527,
+      "step": 8603
+    },
+    {
+      "epoch": 0.9767872537437339,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00025968115266661285,
+      "loss": 5.1621,
+      "step": 8604
+    },
+    {
+      "epoch": 0.976900780853653,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0002591039517890492,
+      "loss": 5.1766,
+      "step": 8605
+    },
+    {
+      "epoch": 0.977014307963572,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00025852946088997025,
+      "loss": 5.1638,
+      "step": 8606
+    },
+    {
+      "epoch": 0.9771278350734911,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00025795768182080344,
+      "loss": 5.1427,
+      "step": 8607
+    },
+    {
+      "epoch": 0.9772413621834102,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0002573886164242366,
+      "loss": 5.1831,
+      "step": 8608
+    },
+    {
+      "epoch": 0.9773548892933293,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00025682226653421226,
+      "loss": 5.1355,
+      "step": 8609
+    },
+    {
+      "epoch": 0.9774684164032483,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00025625863397592133,
+      "loss": 5.1669,
+      "step": 8610
+    },
+    {
+      "epoch": 0.9775819435131674,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0002556977205657982,
+      "loss": 5.172,
+      "step": 8611
+    },
+    {
+      "epoch": 0.9776954706230865,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0002551395281115134,
+      "loss": 5.1424,
+      "step": 8612
+    },
+    {
+      "epoch": 0.9778089977330056,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.000254584058411969,
+      "loss": 5.1695,
+      "step": 8613
+    },
+    {
+      "epoch": 0.9779225248429246,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00025403131325729195,
+      "loss": 5.1417,
+      "step": 8614
+    },
+    {
+      "epoch": 0.9780360519528437,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0002534812944288294,
+      "loss": 5.1522,
+      "step": 8615
+    },
+    {
+      "epoch": 0.9781495790627628,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.000252934003699142,
+      "loss": 5.1485,
+      "step": 8616
+    },
+    {
+      "epoch": 0.9782631061726819,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.000252389442831998,
+      "loss": 5.1561,
+      "step": 8617
+    },
+    {
+      "epoch": 0.9783766332826009,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00025184761358236875,
+      "loss": 5.1676,
+      "step": 8618
+    },
+    {
+      "epoch": 0.97849016039252,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0002513085176964219,
+      "loss": 5.1676,
+      "step": 8619
+    },
+    {
+      "epoch": 0.9786036875024391,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0002507721569115168,
+      "loss": 5.1599,
+      "step": 8620
+    },
+    {
+      "epoch": 0.9787172146123582,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00025023853295619725,
+      "loss": 5.1543,
+      "step": 8621
+    },
+    {
+      "epoch": 0.9788307417222772,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00024970764755018817,
+      "loss": 5.1539,
+      "step": 8622
+    },
+    {
+      "epoch": 0.9789442688321963,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00024917950240438804,
+      "loss": 5.1604,
+      "step": 8623
+    },
+    {
+      "epoch": 0.9790577959421154,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0002486540992208646,
+      "loss": 5.1331,
+      "step": 8624
+    },
+    {
+      "epoch": 0.9791713230520345,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00024813143969284893,
+      "loss": 5.1693,
+      "step": 8625
+    },
+    {
+      "epoch": 0.9792848501619535,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0002476115255047302,
+      "loss": 5.1358,
+      "step": 8626
+    },
+    {
+      "epoch": 0.9793983772718726,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0002470943583320497,
+      "loss": 5.1416,
+      "step": 8627
+    },
+    {
+      "epoch": 0.9795119043817917,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0002465799398414962,
+      "loss": 5.1871,
+      "step": 8628
+    },
+    {
+      "epoch": 0.9796254314917108,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0002460682716909004,
+      "loss": 5.1599,
+      "step": 8629
+    },
+    {
+      "epoch": 0.9797389586016299,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0002455593555292287,
+      "loss": 5.1541,
+      "step": 8630
+    },
+    {
+      "epoch": 0.9798524857115489,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.000245053192996579,
+      "loss": 5.1649,
+      "step": 8631
+    },
+    {
+      "epoch": 0.979966012821468,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00024454978572417533,
+      "loss": 5.1633,
+      "step": 8632
+    },
+    {
+      "epoch": 0.9800795399313871,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00024404913533436186,
+      "loss": 5.1495,
+      "step": 8633
+    },
+    {
+      "epoch": 0.9801930670413062,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00024355124344059792,
+      "loss": 5.1779,
+      "step": 8634
+    },
+    {
+      "epoch": 0.9803065941512252,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00024305611164745358,
+      "loss": 5.1705,
+      "step": 8635
+    },
+    {
+      "epoch": 0.9804201212611443,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0002425637415506032,
+      "loss": 5.1446,
+      "step": 8636
+    },
+    {
+      "epoch": 0.9805336483710634,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00024207413473682147,
+      "loss": 5.1645,
+      "step": 8637
+    },
+    {
+      "epoch": 0.9806471754809825,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00024158729278397724,
+      "loss": 5.1464,
+      "step": 8638
+    },
+    {
+      "epoch": 0.9807607025909015,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00024110321726102937,
+      "loss": 5.1403,
+      "step": 8639
+    },
+    {
+      "epoch": 0.9808742297008206,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00024062190972802137,
+      "loss": 5.1649,
+      "step": 8640
+    },
+    {
+      "epoch": 0.9809877568107397,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00024014337173607585,
+      "loss": 5.1627,
+      "step": 8641
+    },
+    {
+      "epoch": 0.9811012839206588,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00023966760482739037,
+      "loss": 5.1667,
+      "step": 8642
+    },
+    {
+      "epoch": 0.9812148110305778,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00023919461053523167,
+      "loss": 5.1573,
+      "step": 8643
+    },
+    {
+      "epoch": 0.9813283381404969,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0002387243903839313,
+      "loss": 5.1371,
+      "step": 8644
+    },
+    {
+      "epoch": 0.981441865250416,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00023825694588888046,
+      "loss": 5.1593,
+      "step": 8645
+    },
+    {
+      "epoch": 0.9815553923603351,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00023779227855652525,
+      "loss": 5.1372,
+      "step": 8646
+    },
+    {
+      "epoch": 0.9816689194702541,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00023733038988436167,
+      "loss": 5.1428,
+      "step": 8647
+    },
+    {
+      "epoch": 0.9817824465801732,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.000236871281360931,
+      "loss": 5.1505,
+      "step": 8648
+    },
+    {
+      "epoch": 0.9818959736900923,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00023641495446581447,
+      "loss": 5.1238,
+      "step": 8649
+    },
+    {
+      "epoch": 0.9820095008000114,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0002359614106696293,
+      "loss": 5.1757,
+      "step": 8650
+    },
+    {
+      "epoch": 0.9821230279099304,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00023551065143402318,
+      "loss": 5.127,
+      "step": 8651
+    },
+    {
+      "epoch": 0.9822365550198495,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00023506267821167033,
+      "loss": 5.1608,
+      "step": 8652
+    },
+    {
+      "epoch": 0.9823500821297686,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00023461749244626604,
+      "loss": 5.1723,
+      "step": 8653
+    },
+    {
+      "epoch": 0.9824636092396877,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00023417509557252267,
+      "loss": 5.1628,
+      "step": 8654
+    },
+    {
+      "epoch": 0.9825771363496068,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00023373548901616464,
+      "loss": 5.162,
+      "step": 8655
+    },
+    {
+      "epoch": 0.9826906634595258,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00023329867419392373,
+      "loss": 5.1505,
+      "step": 8656
+    },
+    {
+      "epoch": 0.9828041905694449,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0002328646525135352,
+      "loss": 5.1481,
+      "step": 8657
+    },
+    {
+      "epoch": 0.982917717679364,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0002324334253737321,
+      "loss": 5.1644,
+      "step": 8658
+    },
+    {
+      "epoch": 0.983031244789283,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0002320049941642421,
+      "loss": 5.1591,
+      "step": 8659
+    },
+    {
+      "epoch": 0.9831447718992021,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00023157936026578185,
+      "loss": 5.1626,
+      "step": 8660
+    },
+    {
+      "epoch": 0.9832582990091212,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00023115652505005353,
+      "loss": 5.1559,
+      "step": 8661
+    },
+    {
+      "epoch": 0.9833718261190403,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00023073648987973942,
+      "loss": 5.1334,
+      "step": 8662
+    },
+    {
+      "epoch": 0.9834853532289594,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00023031925610849852,
+      "loss": 5.1693,
+      "step": 8663
+    },
+    {
+      "epoch": 0.9835988803388784,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00022990482508096103,
+      "loss": 5.1548,
+      "step": 8664
+    },
+    {
+      "epoch": 0.9837124074487975,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0002294931981327252,
+      "loss": 5.1424,
+      "step": 8665
+    },
+    {
+      "epoch": 0.9838259345587166,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0002290843765903526,
+      "loss": 5.1587,
+      "step": 8666
+    },
+    {
+      "epoch": 0.9839394616686357,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00022867836177136312,
+      "loss": 5.1506,
+      "step": 8667
+    },
+    {
+      "epoch": 0.9840529887785547,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00022827515498423205,
+      "loss": 5.1191,
+      "step": 8668
+    },
+    {
+      "epoch": 0.9841665158884738,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00022787475752838448,
+      "loss": 5.1645,
+      "step": 8669
+    },
+    {
+      "epoch": 0.9842800429983929,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00022747717069419255,
+      "loss": 5.1468,
+      "step": 8670
+    },
+    {
+      "epoch": 0.984393570108312,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00022708239576296987,
+      "loss": 5.1568,
+      "step": 8671
+    },
+    {
+      "epoch": 0.984507097218231,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00022669043400696837,
+      "loss": 5.1601,
+      "step": 8672
+    },
+    {
+      "epoch": 0.9846206243281501,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0002263012866893737,
+      "loss": 5.1506,
+      "step": 8673
+    },
+    {
+      "epoch": 0.9847341514380692,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00022591495506430156,
+      "loss": 5.145,
+      "step": 8674
+    },
+    {
+      "epoch": 0.9848476785479883,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00022553144037679344,
+      "loss": 5.1336,
+      "step": 8675
+    },
+    {
+      "epoch": 0.9849612056579073,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00022515074386281257,
+      "loss": 5.1797,
+      "step": 8676
+    },
+    {
+      "epoch": 0.9850747327678264,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0002247728667492396,
+      "loss": 5.1442,
+      "step": 8677
+    },
+    {
+      "epoch": 0.9851882598777455,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00022439781025386957,
+      "loss": 5.1528,
+      "step": 8678
+    },
+    {
+      "epoch": 0.9853017869876646,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00022402557558540726,
+      "loss": 5.1608,
+      "step": 8679
+    },
+    {
+      "epoch": 0.9854153140975837,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00022365616394346333,
+      "loss": 5.1588,
+      "step": 8680
+    },
+    {
+      "epoch": 0.9855288412075027,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0002232895765185508,
+      "loss": 5.1482,
+      "step": 8681
+    },
+    {
+      "epoch": 0.9856423683174218,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00022292581449208048,
+      "loss": 5.1762,
+      "step": 8682
+    },
+    {
+      "epoch": 0.9857558954273409,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00022256487903635875,
+      "loss": 5.1579,
+      "step": 8683
+    },
+    {
+      "epoch": 0.98586942253726,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00022220677131458172,
+      "loss": 5.1438,
+      "step": 8684
+    },
+    {
+      "epoch": 0.985982949647179,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0002218514924808329,
+      "loss": 5.1589,
+      "step": 8685
+    },
+    {
+      "epoch": 0.9860964767570981,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00022149904368007903,
+      "loss": 5.1768,
+      "step": 8686
+    },
+    {
+      "epoch": 0.9862100038670172,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00022114942604816633,
+      "loss": 5.1403,
+      "step": 8687
+    },
+    {
+      "epoch": 0.9863235309769363,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0002208026407118172,
+      "loss": 5.165,
+      "step": 8688
+    },
+    {
+      "epoch": 0.9864370580868553,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0002204586887886259,
+      "loss": 5.1627,
+      "step": 8689
+    },
+    {
+      "epoch": 0.9865505851967744,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0002201175713870558,
+      "loss": 5.171,
+      "step": 8690
+    },
+    {
+      "epoch": 0.9866641123066935,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00021977928960643513,
+      "loss": 5.156,
+      "step": 8691
+    },
+    {
+      "epoch": 0.9867776394166126,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00021944384453695387,
+      "loss": 5.1631,
+      "step": 8692
+    },
+    {
+      "epoch": 0.9868911665265316,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00021911123725965972,
+      "loss": 5.1614,
+      "step": 8693
+    },
+    {
+      "epoch": 0.9870046936364507,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00021878146884645542,
+      "loss": 5.1535,
+      "step": 8694
+    },
+    {
+      "epoch": 0.9871182207463698,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00021845454036009438,
+      "loss": 5.1541,
+      "step": 8695
+    },
+    {
+      "epoch": 0.9872317478562889,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00021813045285417785,
+      "loss": 5.173,
+      "step": 8696
+    },
+    {
+      "epoch": 0.9873452749662079,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00021780920737315152,
+      "loss": 5.1459,
+      "step": 8697
+    },
+    {
+      "epoch": 0.987458802076127,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.000217490804952302,
+      "loss": 5.1545,
+      "step": 8698
+    },
+    {
+      "epoch": 0.9875723291860461,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00021717524661775322,
+      "loss": 5.1579,
+      "step": 8699
+    },
+    {
+      "epoch": 0.9876858562959652,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00021686253338646378,
+      "loss": 5.1641,
+      "step": 8700
+    },
+    {
+      "epoch": 0.9877993834058842,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.00021655266626622295,
+      "loss": 5.1597,
+      "step": 8701
+    },
+    {
+      "epoch": 0.9879129105158033,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00021624564625564782,
+      "loss": 5.1852,
+      "step": 8702
+    },
+    {
+      "epoch": 0.9880264376257224,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00021594147434418027,
+      "loss": 5.159,
+      "step": 8703
+    },
+    {
+      "epoch": 0.9881399647356415,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00021564015151208314,
+      "loss": 5.1526,
+      "step": 8704
+    },
+    {
+      "epoch": 0.9882534918455605,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00021534167873043806,
+      "loss": 5.157,
+      "step": 8705
+    },
+    {
+      "epoch": 0.9883670189554796,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00021504605696114095,
+      "loss": 5.17,
+      "step": 8706
+    },
+    {
+      "epoch": 0.9884805460653987,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0002147532871569004,
+      "loss": 5.1395,
+      "step": 8707
+    },
+    {
+      "epoch": 0.9885940731753178,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0002144633702612333,
+      "loss": 5.1422,
+      "step": 8708
+    },
+    {
+      "epoch": 0.9887076002852369,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00021417630720846305,
+      "loss": 5.152,
+      "step": 8709
+    },
+    {
+      "epoch": 0.9888211273951559,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00021389209892371523,
+      "loss": 5.1682,
+      "step": 8710
+    },
+    {
+      "epoch": 0.988934654505075,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00021361074632291561,
+      "loss": 5.1635,
+      "step": 8711
+    },
+    {
+      "epoch": 0.9890481816149941,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00021333225031278701,
+      "loss": 5.1526,
+      "step": 8712
+    },
+    {
+      "epoch": 0.9891617087249132,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.000213056611790846,
+      "loss": 5.1435,
+      "step": 8713
+    },
+    {
+      "epoch": 0.9892752358348322,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00021278383164540028,
+      "loss": 5.163,
+      "step": 8714
+    },
+    {
+      "epoch": 0.9893887629447513,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00021251391075554583,
+      "loss": 5.1407,
+      "step": 8715
+    },
+    {
+      "epoch": 0.9895022900546704,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00021224684999116407,
+      "loss": 5.1741,
+      "step": 8716
+    },
+    {
+      "epoch": 0.9896158171645895,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0002119826502129188,
+      "loss": 5.1389,
+      "step": 8717
+    },
+    {
+      "epoch": 0.9897293442745085,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0002117213122722539,
+      "loss": 5.1593,
+      "step": 8718
+    },
+    {
+      "epoch": 0.9898428713844276,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00021146283701139028,
+      "loss": 5.1619,
+      "step": 8719
+    },
+    {
+      "epoch": 0.9899563984943467,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00021120722526332328,
+      "loss": 5.15,
+      "step": 8720
+    },
+    {
+      "epoch": 0.9900699256042658,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00021095447785181954,
+      "loss": 5.1518,
+      "step": 8721
+    },
+    {
+      "epoch": 0.9901834527141848,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0002107045955914152,
+      "loss": 5.1544,
+      "step": 8722
+    },
+    {
+      "epoch": 0.9902969798241039,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0002104575792874124,
+      "loss": 5.1629,
+      "step": 8723
+    },
+    {
+      "epoch": 0.990410506934023,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00021021342973587745,
+      "loss": 5.1611,
+      "step": 8724
+    },
+    {
+      "epoch": 0.9905240340439421,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.00020997214772363763,
+      "loss": 5.1418,
+      "step": 8725
+    },
+    {
+      "epoch": 0.9906375611538611,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00020973373402827908,
+      "loss": 5.1578,
+      "step": 8726
+    },
+    {
+      "epoch": 0.9907510882637802,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00020949818941814406,
+      "loss": 5.1534,
+      "step": 8727
+    },
+    {
+      "epoch": 0.9908646153736993,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00020926551465232863,
+      "loss": 5.1665,
+      "step": 8728
+    },
+    {
+      "epoch": 0.9909781424836184,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00020903571048067992,
+      "loss": 5.1685,
+      "step": 8729
+    },
+    {
+      "epoch": 0.9910916695935374,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00020880877764379414,
+      "loss": 5.1565,
+      "step": 8730
+    },
+    {
+      "epoch": 0.9912051967034565,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.00020858471687301393,
+      "loss": 5.1554,
+      "step": 8731
+    },
+    {
+      "epoch": 0.9913187238133756,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.000208363528890426,
+      "loss": 5.1642,
+      "step": 8732
+    },
+    {
+      "epoch": 0.9914322509232947,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00020814521440885877,
+      "loss": 5.1496,
+      "step": 8733
+    },
+    {
+      "epoch": 0.9915457780332138,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0002079297741318803,
+      "loss": 5.1448,
+      "step": 8734
+    },
+    {
+      "epoch": 0.9916593051431328,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00020771720875379586,
+      "loss": 5.134,
+      "step": 8735
+    },
+    {
+      "epoch": 0.9917728322530519,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00020750751895964547,
+      "loss": 5.1615,
+      "step": 8736
+    },
+    {
+      "epoch": 0.991886359362971,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00020730070542520233,
+      "loss": 5.1553,
+      "step": 8737
+    },
+    {
+      "epoch": 0.9919998864728901,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00020709676881697003,
+      "loss": 5.1501,
+      "step": 8738
+    },
+    {
+      "epoch": 0.9921134135828091,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00020689570979218037,
+      "loss": 5.1231,
+      "step": 8739
+    },
+    {
+      "epoch": 0.9922269406927282,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00020669752899879192,
+      "loss": 5.1596,
+      "step": 8740
+    },
+    {
+      "epoch": 0.9923404678026473,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00020650222707548717,
+      "loss": 5.1706,
+      "step": 8741
+    },
+    {
+      "epoch": 0.9924539949125664,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00020630980465167106,
+      "loss": 5.1505,
+      "step": 8742
+    },
+    {
+      "epoch": 0.9925675220224854,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0002061202623474684,
+      "loss": 5.1626,
+      "step": 8743
+    },
+    {
+      "epoch": 0.9926810491324045,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00020593360077372249,
+      "loss": 5.1444,
+      "step": 8744
+    },
+    {
+      "epoch": 0.9927945762423236,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00020574982053199245,
+      "loss": 5.1584,
+      "step": 8745
+    },
+    {
+      "epoch": 0.9929081033522427,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00020556892221455195,
+      "loss": 5.1746,
+      "step": 8746
+    },
+    {
+      "epoch": 0.9930216304621617,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00020539090640438691,
+      "loss": 5.149,
+      "step": 8747
+    },
+    {
+      "epoch": 0.9931351575720808,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00020521577367519373,
+      "loss": 5.1403,
+      "step": 8748
+    },
+    {
+      "epoch": 0.9932486846819999,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0002050435245913772,
+      "loss": 5.1414,
+      "step": 8749
+    },
+    {
+      "epoch": 0.993362211791919,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00020487415970804935,
+      "loss": 5.1603,
+      "step": 8750
+    },
+    {
+      "epoch": 0.993475738901838,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00020470767957102694,
+      "loss": 5.1697,
+      "step": 8751
+    },
+    {
+      "epoch": 0.9935892660117571,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.00020454408471682988,
+      "loss": 5.1719,
+      "step": 8752
+    },
+    {
+      "epoch": 0.9937027931216762,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00020438337567267993,
+      "loss": 5.1562,
+      "step": 8753
+    },
+    {
+      "epoch": 0.9938163202315953,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0002042255529564982,
+      "loss": 5.1772,
+      "step": 8754
+    },
+    {
+      "epoch": 0.9939298473415143,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00020407061707690467,
+      "loss": 5.1478,
+      "step": 8755
+    },
+    {
+      "epoch": 0.9940433744514334,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.000203918568533215,
+      "loss": 5.1444,
+      "step": 8756
+    },
+    {
+      "epoch": 0.9941569015613525,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00020376940781544035,
+      "loss": 5.1625,
+      "step": 8757
+    },
+    {
+      "epoch": 0.9942704286712716,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00020362313540428484,
+      "loss": 5.153,
+      "step": 8758
+    },
+    {
+      "epoch": 0.9943839557811907,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00020347975177114475,
+      "loss": 5.1682,
+      "step": 8759
+    },
+    {
+      "epoch": 0.9944974828911097,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00020333925737810632,
+      "loss": 5.1691,
+      "step": 8760
+    },
+    {
+      "epoch": 0.9946110100010288,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00020320165267794462,
+      "loss": 5.1531,
+      "step": 8761
+    },
+    {
+      "epoch": 0.9947245371109479,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00020306693811412199,
+      "loss": 5.1392,
+      "step": 8762
+    },
+    {
+      "epoch": 0.994838064220867,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.00020293511412078688,
+      "loss": 5.1369,
+      "step": 8763
+    },
+    {
+      "epoch": 0.994951591330786,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0002028061811227718,
+      "loss": 5.1646,
+      "step": 8764
+    },
+    {
+      "epoch": 0.9950651184407051,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00020268013953559277,
+      "loss": 5.1592,
+      "step": 8765
+    },
+    {
+      "epoch": 0.9951786455506242,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00020255698976544747,
+      "loss": 5.1502,
+      "step": 8766
+    },
+    {
+      "epoch": 0.9952921726605433,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00020243673220921393,
+      "loss": 5.142,
+      "step": 8767
+    },
+    {
+      "epoch": 0.9954056997704623,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00020231936725444938,
+      "loss": 5.1435,
+      "step": 8768
+    },
+    {
+      "epoch": 0.9955192268803814,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0002022048952793891,
+      "loss": 5.1484,
+      "step": 8769
+    },
+    {
+      "epoch": 0.9956327539903005,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0002020933166529451,
+      "loss": 5.18,
+      "step": 8770
+    },
+    {
+      "epoch": 0.9957462811002196,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00020198463173470477,
+      "loss": 5.1516,
+      "step": 8771
+    },
+    {
+      "epoch": 0.9958598082101386,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00020187884087492991,
+      "loss": 5.1448,
+      "step": 8772
+    },
+    {
+      "epoch": 0.9959733353200577,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0002017759444145556,
+      "loss": 5.1466,
+      "step": 8773
+    },
+    {
+      "epoch": 0.9960868624299768,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0002016759426851891,
+      "loss": 5.1691,
+      "step": 8774
+    },
+    {
+      "epoch": 0.9962003895398959,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0002015788360091087,
+      "loss": 5.1403,
+      "step": 8775
+    },
+    {
+      "epoch": 0.996313916649815,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00020148462469926265,
+      "loss": 5.1548,
+      "step": 8776
+    },
+    {
+      "epoch": 0.996427443759734,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.00020139330905926835,
+      "loss": 5.1401,
+      "step": 8777
+    },
+    {
+      "epoch": 0.9965409708696531,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00020130488938341118,
+      "loss": 5.1417,
+      "step": 8778
+    },
+    {
+      "epoch": 0.9966544979795722,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00020121936595664358,
+      "loss": 5.1478,
+      "step": 8779
+    },
+    {
+      "epoch": 0.9967680250894912,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0002011367390545843,
+      "loss": 5.1462,
+      "step": 8780
+    },
+    {
+      "epoch": 0.9968815521994103,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00020105700894351716,
+      "loss": 5.1421,
+      "step": 8781
+    },
+    {
+      "epoch": 0.9969950793093294,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0002009801758803907,
+      "loss": 5.1539,
+      "step": 8782
+    },
+    {
+      "epoch": 0.9971086064192485,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00020090624011281688,
+      "loss": 5.1587,
+      "step": 8783
+    },
+    {
+      "epoch": 0.9972221335291676,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00020083520187907033,
+      "loss": 5.1473,
+      "step": 8784
+    },
+    {
+      "epoch": 0.9973356606390866,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00020076706140808814,
+      "loss": 5.149,
+      "step": 8785
+    },
+    {
+      "epoch": 0.9974491877490057,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00020070181891946833,
+      "loss": 5.1587,
+      "step": 8786
+    },
+    {
+      "epoch": 0.9975627148589248,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.00020063947462346975,
+      "loss": 5.1448,
+      "step": 8787
+    },
+    {
+      "epoch": 0.9976762419688439,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00020058002872101077,
+      "loss": 5.1528,
+      "step": 8788
+    },
+    {
+      "epoch": 0.9977897690787629,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0002005234814036696,
+      "loss": 5.1646,
+      "step": 8789
+    },
+    {
+      "epoch": 0.997903296188682,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0002004698328536827,
+      "loss": 5.1574,
+      "step": 8790
+    },
+    {
+      "epoch": 0.9980168232986011,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0002004190832439447,
+      "loss": 5.1553,
+      "step": 8791
+    },
+    {
+      "epoch": 0.9981303504085202,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00020037123273800782,
+      "loss": 5.1813,
+      "step": 8792
+    },
+    {
+      "epoch": 0.9982438775184392,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.000200326281490081,
+      "loss": 5.1492,
+      "step": 8793
+    },
+    {
+      "epoch": 0.9983574046283583,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00020028422964503008,
+      "loss": 5.1408,
+      "step": 8794
+    },
+    {
+      "epoch": 0.9984709317382774,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00020024507733837644,
+      "loss": 5.1743,
+      "step": 8795
+    },
+    {
+      "epoch": 0.9985844588481965,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00020020882469629753,
+      "loss": 5.1757,
+      "step": 8796
+    },
+    {
+      "epoch": 0.9986979859581155,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.00020017547183562553,
+      "loss": 5.1422,
+      "step": 8797
+    },
+    {
+      "epoch": 0.9988115130680346,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.00020014501886384765,
+      "loss": 5.1514,
+      "step": 8798
+    },
+    {
+      "epoch": 0.9989250401779537,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.00020011746587910557,
+      "loss": 5.1403,
+      "step": 8799
+    },
+    {
+      "epoch": 0.9990385672878728,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.00020009281297019498,
+      "loss": 5.1583,
+      "step": 8800
+    },
+    {
+      "epoch": 0.9991520943977918,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0002000710602165655,
+      "loss": 5.1545,
+      "step": 8801
+    },
+    {
+      "epoch": 0.9992656215077109,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0002000522076883203,
+      "loss": 5.1756,
+      "step": 8802
+    },
+    {
+      "epoch": 0.99937914861763,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00020003625544621603,
+      "loss": 5.1673,
+      "step": 8803
+    },
+    {
+      "epoch": 0.9994926757275491,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00020002320354166228,
+      "loss": 5.1647,
+      "step": 8804
+    },
+    {
+      "epoch": 0.9996062028374681,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.00020001305201672182,
+      "loss": 5.1787,
+      "step": 8805
+    },
+    {
+      "epoch": 0.9997197299473872,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.00020000580090411026,
+      "loss": 5.1421,
+      "step": 8806
+    },
+    {
+      "epoch": 0.9998332570573063,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.00020000145022719605,
+      "loss": 5.1743,
+      "step": 8807
+    },
+    {
+      "epoch": 0.9999467841672254,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0002,
+      "loss": 5.143,
+      "step": 8808
+    },
+    {
+      "epoch": 0.9999467841672254,
+      "step": 8808,
+      "total_flos": 4.588406021280372e+20,
+      "train_loss": 5.391583533096487,
+      "train_runtime": 330234.0549,
+      "train_samples_per_second": 23.899,
+      "train_steps_per_second": 0.027
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 8808,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.588406021280372e+20,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}