diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,31534 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.27654867256637167,
+  "eval_steps": 250,
+  "global_step": 4500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 6.145526057030482e-05,
+      "grad_norm": 29.75,
+      "learning_rate": 0.0,
+      "loss": 1.4108,
+      "step": 1
+    },
+    {
+      "epoch": 0.00012291052114060964,
+      "grad_norm": 28.75,
+      "learning_rate": 1.3333333333333334e-08,
+      "loss": 1.4446,
+      "step": 2
+    },
+    {
+      "epoch": 0.00018436578171091445,
+      "grad_norm": 29.875,
+      "learning_rate": 2.6666666666666667e-08,
+      "loss": 1.4226,
+      "step": 3
+    },
+    {
+      "epoch": 0.0002458210422812193,
+      "grad_norm": 28.5,
+      "learning_rate": 4e-08,
+      "loss": 1.4369,
+      "step": 4
+    },
+    {
+      "epoch": 0.00030727630285152407,
+      "grad_norm": 34.0,
+      "learning_rate": 5.3333333333333334e-08,
+      "loss": 1.4434,
+      "step": 5
+    },
+    {
+      "epoch": 0.0003687315634218289,
+      "grad_norm": 35.25,
+      "learning_rate": 6.666666666666668e-08,
+      "loss": 1.4788,
+      "step": 6
+    },
+    {
+      "epoch": 0.00043018682399213374,
+      "grad_norm": 36.25,
+      "learning_rate": 8e-08,
+      "loss": 1.4425,
+      "step": 7
+    },
+    {
+      "epoch": 0.0004916420845624386,
+      "grad_norm": 50.25,
+      "learning_rate": 9.333333333333335e-08,
+      "loss": 1.4805,
+      "step": 8
+    },
+    {
+      "epoch": 0.0005530973451327434,
+      "grad_norm": 33.25,
+      "learning_rate": 1.0666666666666667e-07,
+      "loss": 1.4837,
+      "step": 9
+    },
+    {
+      "epoch": 0.0006145526057030481,
+      "grad_norm": 32.75,
+      "learning_rate": 1.2000000000000002e-07,
+      "loss": 1.472,
+      "step": 10
+    },
+    {
+      "epoch": 0.000676007866273353,
+      "grad_norm": 33.5,
+      "learning_rate": 1.3333333333333336e-07,
+      "loss": 1.5112,
+      "step": 11
+    },
+    {
+      "epoch": 0.0007374631268436578,
+      "grad_norm": 37.5,
+      "learning_rate": 1.4666666666666668e-07,
+      "loss": 1.4344,
+      "step": 12
+    },
+    {
+      "epoch": 0.0007989183874139626,
+      "grad_norm": 41.5,
+      "learning_rate": 1.6e-07,
+      "loss": 1.4732,
+      "step": 13
+    },
+    {
+      "epoch": 0.0008603736479842675,
+      "grad_norm": 30.5,
+      "learning_rate": 1.7333333333333335e-07,
+      "loss": 1.4347,
+      "step": 14
+    },
+    {
+      "epoch": 0.0009218289085545723,
+      "grad_norm": 35.25,
+      "learning_rate": 1.866666666666667e-07,
+      "loss": 1.5008,
+      "step": 15
+    },
+    {
+      "epoch": 0.0009832841691248771,
+      "grad_norm": 33.25,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 1.437,
+      "step": 16
+    },
+    {
+      "epoch": 0.0010447394296951818,
+      "grad_norm": 29.875,
+      "learning_rate": 2.1333333333333334e-07,
+      "loss": 1.5193,
+      "step": 17
+    },
+    {
+      "epoch": 0.0011061946902654867,
+      "grad_norm": 34.0,
+      "learning_rate": 2.266666666666667e-07,
+      "loss": 1.4585,
+      "step": 18
+    },
+    {
+      "epoch": 0.0011676499508357916,
+      "grad_norm": 33.25,
+      "learning_rate": 2.4000000000000003e-07,
+      "loss": 1.5157,
+      "step": 19
+    },
+    {
+      "epoch": 0.0012291052114060963,
+      "grad_norm": 31.875,
+      "learning_rate": 2.533333333333333e-07,
+      "loss": 1.4815,
+      "step": 20
+    },
+    {
+      "epoch": 0.0012905604719764012,
+      "grad_norm": 27.0,
+      "learning_rate": 2.666666666666667e-07,
+      "loss": 1.4444,
+      "step": 21
+    },
+    {
+      "epoch": 0.001352015732546706,
+      "grad_norm": 28.0,
+      "learning_rate": 2.8e-07,
+      "loss": 1.4639,
+      "step": 22
+    },
+    {
+      "epoch": 0.0014134709931170107,
+      "grad_norm": 26.125,
+      "learning_rate": 2.9333333333333337e-07,
+      "loss": 1.4939,
+      "step": 23
+    },
+    {
+      "epoch": 0.0014749262536873156,
+      "grad_norm": 35.5,
+      "learning_rate": 3.0666666666666666e-07,
+      "loss": 1.5076,
+      "step": 24
+    },
+    {
+      "epoch": 0.0015363815142576205,
+      "grad_norm": 32.75,
+      "learning_rate": 3.2e-07,
+      "loss": 1.4532,
+      "step": 25
+    },
+    {
+      "epoch": 0.0015978367748279252,
+      "grad_norm": 27.375,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 1.5085,
+      "step": 26
+    },
+    {
+      "epoch": 0.00165929203539823,
+      "grad_norm": 26.625,
+      "learning_rate": 3.466666666666667e-07,
+      "loss": 1.4565,
+      "step": 27
+    },
+    {
+      "epoch": 0.001720747295968535,
+      "grad_norm": 37.25,
+      "learning_rate": 3.6e-07,
+      "loss": 1.5237,
+      "step": 28
+    },
+    {
+      "epoch": 0.0017822025565388396,
+      "grad_norm": 24.5,
+      "learning_rate": 3.733333333333334e-07,
+      "loss": 1.4745,
+      "step": 29
+    },
+    {
+      "epoch": 0.0018436578171091445,
+      "grad_norm": 30.375,
+      "learning_rate": 3.8666666666666674e-07,
+      "loss": 1.484,
+      "step": 30
+    },
+    {
+      "epoch": 0.0019051130776794494,
+      "grad_norm": 25.25,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 1.4895,
+      "step": 31
+    },
+    {
+      "epoch": 0.0019665683382497543,
+      "grad_norm": 28.0,
+      "learning_rate": 4.133333333333334e-07,
+      "loss": 1.4245,
+      "step": 32
+    },
+    {
+      "epoch": 0.002028023598820059,
+      "grad_norm": 30.875,
+      "learning_rate": 4.266666666666667e-07,
+      "loss": 1.5198,
+      "step": 33
+    },
+    {
+      "epoch": 0.0020894788593903636,
+      "grad_norm": 30.375,
+      "learning_rate": 4.4e-07,
+      "loss": 1.4599,
+      "step": 34
+    },
+    {
+      "epoch": 0.0021509341199606687,
+      "grad_norm": 27.875,
+      "learning_rate": 4.533333333333334e-07,
+      "loss": 1.4397,
+      "step": 35
+    },
+    {
+      "epoch": 0.0022123893805309734,
+      "grad_norm": 27.875,
+      "learning_rate": 4.666666666666667e-07,
+      "loss": 1.4923,
+      "step": 36
+    },
+    {
+      "epoch": 0.002273844641101278,
+      "grad_norm": 24.5,
+      "learning_rate": 4.800000000000001e-07,
+      "loss": 1.468,
+      "step": 37
+    },
+    {
+      "epoch": 0.002335299901671583,
+      "grad_norm": 29.125,
+      "learning_rate": 4.933333333333334e-07,
+      "loss": 1.4585,
+      "step": 38
+    },
+    {
+      "epoch": 0.002396755162241888,
+      "grad_norm": 39.0,
+      "learning_rate": 5.066666666666667e-07,
+      "loss": 1.4864,
+      "step": 39
+    },
+    {
+      "epoch": 0.0024582104228121925,
+      "grad_norm": 28.625,
+      "learning_rate": 5.2e-07,
+      "loss": 1.5024,
+      "step": 40
+    },
+    {
+      "epoch": 0.0025196656833824977,
+      "grad_norm": 29.5,
+      "learning_rate": 5.333333333333335e-07,
+      "loss": 1.4955,
+      "step": 41
+    },
+    {
+      "epoch": 0.0025811209439528023,
+      "grad_norm": 35.25,
+      "learning_rate": 5.466666666666667e-07,
+      "loss": 1.4087,
+      "step": 42
+    },
+    {
+      "epoch": 0.002642576204523107,
+      "grad_norm": 33.5,
+      "learning_rate": 5.6e-07,
+      "loss": 1.4448,
+      "step": 43
+    },
+    {
+      "epoch": 0.002704031465093412,
+      "grad_norm": 32.5,
+      "learning_rate": 5.733333333333334e-07,
+      "loss": 1.4495,
+      "step": 44
+    },
+    {
+      "epoch": 0.0027654867256637168,
+      "grad_norm": 31.75,
+      "learning_rate": 5.866666666666667e-07,
+      "loss": 1.4562,
+      "step": 45
+    },
+    {
+      "epoch": 0.0028269419862340214,
+      "grad_norm": 27.0,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 1.4601,
+      "step": 46
+    },
+    {
+      "epoch": 0.0028883972468043266,
+      "grad_norm": 29.5,
+      "learning_rate": 6.133333333333333e-07,
+      "loss": 1.4944,
+      "step": 47
+    },
+    {
+      "epoch": 0.0029498525073746312,
+      "grad_norm": 27.25,
+      "learning_rate": 6.266666666666667e-07,
+      "loss": 1.4711,
+      "step": 48
+    },
+    {
+      "epoch": 0.003011307767944936,
+      "grad_norm": 29.75,
+      "learning_rate": 6.4e-07,
+      "loss": 1.4145,
+      "step": 49
+    },
+    {
+      "epoch": 0.003072763028515241,
+      "grad_norm": 23.25,
+      "learning_rate": 6.533333333333334e-07,
+      "loss": 1.4167,
+      "step": 50
+    },
+    {
+      "epoch": 0.0031342182890855457,
+      "grad_norm": 26.5,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 1.4101,
+      "step": 51
+    },
+    {
+      "epoch": 0.0031956735496558504,
+      "grad_norm": 34.5,
+      "learning_rate": 6.800000000000001e-07,
+      "loss": 1.4635,
+      "step": 52
+    },
+    {
+      "epoch": 0.0032571288102261555,
+      "grad_norm": 25.25,
+      "learning_rate": 6.933333333333334e-07,
+      "loss": 1.4237,
+      "step": 53
+    },
+    {
+      "epoch": 0.00331858407079646,
+      "grad_norm": 26.375,
+      "learning_rate": 7.066666666666667e-07,
+      "loss": 1.4127,
+      "step": 54
+    },
+    {
+      "epoch": 0.003380039331366765,
+      "grad_norm": 30.5,
+      "learning_rate": 7.2e-07,
+      "loss": 1.4324,
+      "step": 55
+    },
+    {
+      "epoch": 0.00344149459193707,
+      "grad_norm": 23.25,
+      "learning_rate": 7.333333333333334e-07,
+      "loss": 1.4182,
+      "step": 56
+    },
+    {
+      "epoch": 0.0035029498525073746,
+      "grad_norm": 36.75,
+      "learning_rate": 7.466666666666668e-07,
+      "loss": 1.4212,
+      "step": 57
+    },
+    {
+      "epoch": 0.0035644051130776793,
+      "grad_norm": 31.125,
+      "learning_rate": 7.6e-07,
+      "loss": 1.4183,
+      "step": 58
+    },
+    {
+      "epoch": 0.0036258603736479844,
+      "grad_norm": 32.5,
+      "learning_rate": 7.733333333333335e-07,
+      "loss": 1.5004,
+      "step": 59
+    },
+    {
+      "epoch": 0.003687315634218289,
+      "grad_norm": 29.125,
+      "learning_rate": 7.866666666666667e-07,
+      "loss": 1.4667,
+      "step": 60
+    },
+    {
+      "epoch": 0.0037487708947885937,
+      "grad_norm": 30.5,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 1.4404,
+      "step": 61
+    },
+    {
+      "epoch": 0.003810226155358899,
+      "grad_norm": 27.875,
+      "learning_rate": 8.133333333333333e-07,
+      "loss": 1.4657,
+      "step": 62
+    },
+    {
+      "epoch": 0.0038716814159292035,
+      "grad_norm": 40.75,
+      "learning_rate": 8.266666666666668e-07,
+      "loss": 1.4001,
+      "step": 63
+    },
+    {
+      "epoch": 0.003933136676499509,
+      "grad_norm": 27.0,
+      "learning_rate": 8.400000000000001e-07,
+      "loss": 1.4232,
+      "step": 64
+    },
+    {
+      "epoch": 0.003994591937069813,
+      "grad_norm": 27.25,
+      "learning_rate": 8.533333333333334e-07,
+      "loss": 1.3774,
+      "step": 65
+    },
+    {
+      "epoch": 0.004056047197640118,
+      "grad_norm": 29.625,
+      "learning_rate": 8.666666666666668e-07,
+      "loss": 1.3763,
+      "step": 66
+    },
+    {
+      "epoch": 0.004117502458210423,
+      "grad_norm": 27.5,
+      "learning_rate": 8.8e-07,
+      "loss": 1.3875,
+      "step": 67
+    },
+    {
+      "epoch": 0.004178957718780727,
+      "grad_norm": 34.25,
+      "learning_rate": 8.933333333333334e-07,
+      "loss": 1.404,
+      "step": 68
+    },
+    {
+      "epoch": 0.004240412979351032,
+      "grad_norm": 33.5,
+      "learning_rate": 9.066666666666668e-07,
+      "loss": 1.3785,
+      "step": 69
+    },
+    {
+      "epoch": 0.0043018682399213375,
+      "grad_norm": 31.875,
+      "learning_rate": 9.200000000000001e-07,
+      "loss": 1.3663,
+      "step": 70
+    },
+    {
+      "epoch": 0.004363323500491642,
+      "grad_norm": 22.75,
+      "learning_rate": 9.333333333333334e-07,
+      "loss": 1.3843,
+      "step": 71
+    },
+    {
+      "epoch": 0.004424778761061947,
+      "grad_norm": 22.875,
+      "learning_rate": 9.466666666666667e-07,
+      "loss": 1.3844,
+      "step": 72
+    },
+    {
+      "epoch": 0.004486234021632252,
+      "grad_norm": 19.375,
+      "learning_rate": 9.600000000000001e-07,
+      "loss": 1.4336,
+      "step": 73
+    },
+    {
+      "epoch": 0.004547689282202556,
+      "grad_norm": 25.5,
+      "learning_rate": 9.733333333333333e-07,
+      "loss": 1.4111,
+      "step": 74
+    },
+    {
+      "epoch": 0.004609144542772861,
+      "grad_norm": 31.875,
+      "learning_rate": 9.866666666666668e-07,
+      "loss": 1.3456,
+      "step": 75
+    },
+    {
+      "epoch": 0.004670599803343166,
+      "grad_norm": 25.375,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.3526,
+      "step": 76
+    },
+    {
+      "epoch": 0.004732055063913471,
+      "grad_norm": 28.625,
+      "learning_rate": 1.0133333333333333e-06,
+      "loss": 1.3306,
+      "step": 77
+    },
+    {
+      "epoch": 0.004793510324483776,
+      "grad_norm": 22.375,
+      "learning_rate": 1.0266666666666669e-06,
+      "loss": 1.3395,
+      "step": 78
+    },
+    {
+      "epoch": 0.004854965585054081,
+      "grad_norm": 23.125,
+      "learning_rate": 1.04e-06,
+      "loss": 1.3588,
+      "step": 79
+    },
+    {
+      "epoch": 0.004916420845624385,
+      "grad_norm": 22.0,
+      "learning_rate": 1.0533333333333333e-06,
+      "loss": 1.3521,
+      "step": 80
+    },
+    {
+      "epoch": 0.00497787610619469,
+      "grad_norm": 27.25,
+      "learning_rate": 1.066666666666667e-06,
+      "loss": 1.3255,
+      "step": 81
+    },
+    {
+      "epoch": 0.005039331366764995,
+      "grad_norm": 25.375,
+      "learning_rate": 1.08e-06,
+      "loss": 1.3141,
+      "step": 82
+    },
+    {
+      "epoch": 0.0051007866273352995,
+      "grad_norm": 19.0,
+      "learning_rate": 1.0933333333333334e-06,
+      "loss": 1.3346,
+      "step": 83
+    },
+    {
+      "epoch": 0.005162241887905605,
+      "grad_norm": 27.75,
+      "learning_rate": 1.1066666666666667e-06,
+      "loss": 1.2773,
+      "step": 84
+    },
+    {
+      "epoch": 0.00522369714847591,
+      "grad_norm": 26.25,
+      "learning_rate": 1.12e-06,
+      "loss": 1.2768,
+      "step": 85
+    },
+    {
+      "epoch": 0.005285152409046214,
+      "grad_norm": 26.0,
+      "learning_rate": 1.1333333333333334e-06,
+      "loss": 1.3128,
+      "step": 86
+    },
+    {
+      "epoch": 0.005346607669616519,
+      "grad_norm": 19.875,
+      "learning_rate": 1.1466666666666668e-06,
+      "loss": 1.3132,
+      "step": 87
+    },
+    {
+      "epoch": 0.005408062930186824,
+      "grad_norm": 30.875,
+      "learning_rate": 1.1600000000000001e-06,
+      "loss": 1.3141,
+      "step": 88
+    },
+    {
+      "epoch": 0.0054695181907571284,
+      "grad_norm": 22.0,
+      "learning_rate": 1.1733333333333335e-06,
+      "loss": 1.2636,
+      "step": 89
+    },
+    {
+      "epoch": 0.0055309734513274336,
+      "grad_norm": 25.125,
+      "learning_rate": 1.1866666666666668e-06,
+      "loss": 1.2607,
+      "step": 90
+    },
+    {
+      "epoch": 0.005592428711897739,
+      "grad_norm": 17.875,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.2474,
+      "step": 91
+    },
+    {
+      "epoch": 0.005653883972468043,
+      "grad_norm": 21.25,
+      "learning_rate": 1.2133333333333335e-06,
+      "loss": 1.2857,
+      "step": 92
+    },
+    {
+      "epoch": 0.005715339233038348,
+      "grad_norm": 31.125,
+      "learning_rate": 1.2266666666666666e-06,
+      "loss": 1.2651,
+      "step": 93
+    },
+    {
+      "epoch": 0.005776794493608653,
+      "grad_norm": 22.625,
+      "learning_rate": 1.2400000000000002e-06,
+      "loss": 1.2309,
+      "step": 94
+    },
+    {
+      "epoch": 0.005838249754178957,
+      "grad_norm": 26.375,
+      "learning_rate": 1.2533333333333333e-06,
+      "loss": 1.2152,
+      "step": 95
+    },
+    {
+      "epoch": 0.0058997050147492625,
+      "grad_norm": 31.75,
+      "learning_rate": 1.2666666666666669e-06,
+      "loss": 1.2495,
+      "step": 96
+    },
+    {
+      "epoch": 0.0059611602753195676,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.28e-06,
+      "loss": 1.2448,
+      "step": 97
+    },
+    {
+      "epoch": 0.006022615535889872,
+      "grad_norm": 15.875,
+      "learning_rate": 1.2933333333333334e-06,
+      "loss": 1.2347,
+      "step": 98
+    },
+    {
+      "epoch": 0.006084070796460177,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.3066666666666667e-06,
+      "loss": 1.2758,
+      "step": 99
+    },
+    {
+      "epoch": 0.006145526057030482,
+      "grad_norm": 15.25,
+      "learning_rate": 1.32e-06,
+      "loss": 1.3125,
+      "step": 100
+    },
+    {
+      "epoch": 0.006206981317600786,
+      "grad_norm": 28.375,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 1.254,
+      "step": 101
+    },
+    {
+      "epoch": 0.006268436578171091,
+      "grad_norm": 21.0,
+      "learning_rate": 1.3466666666666668e-06,
+      "loss": 1.2636,
+      "step": 102
+    },
+    {
+      "epoch": 0.0063298918387413965,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.3600000000000001e-06,
+      "loss": 1.2687,
+      "step": 103
+    },
+    {
+      "epoch": 0.006391347099311701,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.3733333333333335e-06,
+      "loss": 1.2174,
+      "step": 104
+    },
+    {
+      "epoch": 0.006452802359882006,
+      "grad_norm": 18.75,
+      "learning_rate": 1.3866666666666668e-06,
+      "loss": 1.2783,
+      "step": 105
+    },
+    {
+      "epoch": 0.006514257620452311,
+      "grad_norm": 15.875,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 1.2473,
+      "step": 106
+    },
+    {
+      "epoch": 0.006575712881022615,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.4133333333333335e-06,
+      "loss": 1.2454,
+      "step": 107
+    },
+    {
+      "epoch": 0.00663716814159292,
+      "grad_norm": 41.25,
+      "learning_rate": 1.4266666666666668e-06,
+      "loss": 1.3315,
+      "step": 108
+    },
+    {
+      "epoch": 0.006698623402163225,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.44e-06,
+      "loss": 1.2593,
+      "step": 109
+    },
+    {
+      "epoch": 0.00676007866273353,
+      "grad_norm": 16.75,
+      "learning_rate": 1.4533333333333335e-06,
+      "loss": 1.2575,
+      "step": 110
+    },
+    {
+      "epoch": 0.006821533923303835,
+      "grad_norm": 15.0,
+      "learning_rate": 1.4666666666666669e-06,
+      "loss": 1.265,
+      "step": 111
+    },
+    {
+      "epoch": 0.00688298918387414,
+      "grad_norm": 23.125,
+      "learning_rate": 1.48e-06,
+      "loss": 1.2893,
+      "step": 112
+    },
+    {
+      "epoch": 0.006944444444444444,
+      "grad_norm": 23.0,
+      "learning_rate": 1.4933333333333336e-06,
+      "loss": 1.2425,
+      "step": 113
+    },
+    {
+      "epoch": 0.007005899705014749,
+      "grad_norm": 18.375,
+      "learning_rate": 1.506666666666667e-06,
+      "loss": 1.3032,
+      "step": 114
+    },
+    {
+      "epoch": 0.007067354965585054,
+      "grad_norm": 24.625,
+      "learning_rate": 1.52e-06,
+      "loss": 1.2893,
+      "step": 115
+    },
+    {
+      "epoch": 0.0071288102261553585,
+      "grad_norm": 14.375,
+      "learning_rate": 1.5333333333333334e-06,
+      "loss": 1.1966,
+      "step": 116
+    },
+    {
+      "epoch": 0.007190265486725664,
+      "grad_norm": 21.0,
+      "learning_rate": 1.546666666666667e-06,
+      "loss": 1.23,
+      "step": 117
+    },
+    {
+      "epoch": 0.007251720747295969,
+      "grad_norm": 18.25,
+      "learning_rate": 1.56e-06,
+      "loss": 1.294,
+      "step": 118
+    },
+    {
+      "epoch": 0.007313176007866273,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5733333333333334e-06,
+      "loss": 1.2289,
+      "step": 119
+    },
+    {
+      "epoch": 0.007374631268436578,
+      "grad_norm": 13.875,
+      "learning_rate": 1.586666666666667e-06,
+      "loss": 1.3106,
+      "step": 120
+    },
+    {
+      "epoch": 0.007436086529006883,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.238,
+      "step": 121
+    },
+    {
+      "epoch": 0.007497541789577187,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.6133333333333335e-06,
+      "loss": 1.2299,
+      "step": 122
+    },
+    {
+      "epoch": 0.0075589970501474925,
+      "grad_norm": 13.125,
+      "learning_rate": 1.6266666666666666e-06,
+      "loss": 1.2363,
+      "step": 123
+    },
+    {
+      "epoch": 0.007620452310717798,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.6400000000000002e-06,
+      "loss": 1.1714,
+      "step": 124
+    },
+    {
+      "epoch": 0.007681907571288102,
+      "grad_norm": 19.75,
+      "learning_rate": 1.6533333333333335e-06,
+      "loss": 1.191,
+      "step": 125
+    },
+    {
+      "epoch": 0.007743362831858407,
+      "grad_norm": 18.625,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.1915,
+      "step": 126
+    },
+    {
+      "epoch": 0.007804818092428712,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.6800000000000002e-06,
+      "loss": 1.1701,
+      "step": 127
+    },
+    {
+      "epoch": 0.007866273352999017,
+      "grad_norm": 19.625,
+      "learning_rate": 1.6933333333333336e-06,
+      "loss": 1.173,
+      "step": 128
+    },
+    {
+      "epoch": 0.007927728613569322,
+      "grad_norm": 10.625,
+      "learning_rate": 1.7066666666666667e-06,
+      "loss": 1.2696,
+      "step": 129
+    },
+    {
+      "epoch": 0.007989183874139626,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.72e-06,
+      "loss": 1.2133,
+      "step": 130
+    },
+    {
+      "epoch": 0.00805063913470993,
+      "grad_norm": 13.0,
+      "learning_rate": 1.7333333333333336e-06,
+      "loss": 1.242,
+      "step": 131
+    },
+    {
+      "epoch": 0.008112094395280236,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.7466666666666667e-06,
+      "loss": 1.2084,
+      "step": 132
+    },
+    {
+      "epoch": 0.008173549655850541,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.76e-06,
+      "loss": 1.192,
+      "step": 133
+    },
+    {
+      "epoch": 0.008235004916420846,
+      "grad_norm": 14.625,
+      "learning_rate": 1.7733333333333336e-06,
+      "loss": 1.1556,
+      "step": 134
+    },
+    {
+      "epoch": 0.008296460176991151,
+      "grad_norm": 13.125,
+      "learning_rate": 1.7866666666666668e-06,
+      "loss": 1.2567,
+      "step": 135
+    },
+    {
+      "epoch": 0.008357915437561455,
+      "grad_norm": 23.5,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 1.243,
+      "step": 136
+    },
+    {
+      "epoch": 0.00841937069813176,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.8133333333333337e-06,
+      "loss": 1.1579,
+      "step": 137
+    },
+    {
+      "epoch": 0.008480825958702065,
+      "grad_norm": 10.0,
+      "learning_rate": 1.8266666666666668e-06,
+      "loss": 1.0972,
+      "step": 138
+    },
+    {
+      "epoch": 0.00854228121927237,
+      "grad_norm": 20.875,
+      "learning_rate": 1.8400000000000002e-06,
+      "loss": 1.1399,
+      "step": 139
+    },
+    {
+      "epoch": 0.008603736479842675,
+      "grad_norm": 27.625,
+      "learning_rate": 1.8533333333333333e-06,
+      "loss": 1.1345,
+      "step": 140
+    },
+    {
+      "epoch": 0.00866519174041298,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.8666666666666669e-06,
+      "loss": 1.1324,
+      "step": 141
+    },
+    {
+      "epoch": 0.008726647000983283,
+      "grad_norm": 15.625,
+      "learning_rate": 1.8800000000000002e-06,
+      "loss": 1.1115,
+      "step": 142
+    },
+    {
+      "epoch": 0.008788102261553589,
+      "grad_norm": 21.125,
+      "learning_rate": 1.8933333333333333e-06,
+      "loss": 1.0983,
+      "step": 143
+    },
+    {
+      "epoch": 0.008849557522123894,
+      "grad_norm": 32.0,
+      "learning_rate": 1.906666666666667e-06,
+      "loss": 1.1227,
+      "step": 144
+    },
+    {
+      "epoch": 0.008911012782694199,
+      "grad_norm": 20.125,
+      "learning_rate": 1.9200000000000003e-06,
+      "loss": 1.2015,
+      "step": 145
+    },
+    {
+      "epoch": 0.008972468043264504,
+      "grad_norm": 20.5,
+      "learning_rate": 1.9333333333333336e-06,
+      "loss": 1.2647,
+      "step": 146
+    },
+    {
+      "epoch": 0.009033923303834809,
+      "grad_norm": 16.125,
+      "learning_rate": 1.9466666666666665e-06,
+      "loss": 1.2075,
+      "step": 147
+    },
+    {
+      "epoch": 0.009095378564405112,
+      "grad_norm": 17.0,
+      "learning_rate": 1.9600000000000003e-06,
+      "loss": 1.2263,
+      "step": 148
+    },
+    {
+      "epoch": 0.009156833824975417,
+      "grad_norm": 25.25,
+      "learning_rate": 1.9733333333333336e-06,
+      "loss": 1.313,
+      "step": 149
+    },
+    {
+      "epoch": 0.009218289085545723,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.9866666666666666e-06,
+      "loss": 1.1606,
+      "step": 150
+    },
+    {
+      "epoch": 0.009279744346116028,
+      "grad_norm": 47.5,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.3511,
+      "step": 151
+    },
+    {
+      "epoch": 0.009341199606686333,
+      "grad_norm": 10.6875,
+      "learning_rate": 2.0133333333333337e-06,
+      "loss": 1.2138,
+      "step": 152
+    },
+    {
+      "epoch": 0.009402654867256638,
+      "grad_norm": 14.125,
+      "learning_rate": 2.0266666666666666e-06,
+      "loss": 1.1739,
+      "step": 153
+    },
+    {
+      "epoch": 0.009464110127826941,
+      "grad_norm": 24.0,
+      "learning_rate": 2.04e-06,
+      "loss": 1.1725,
+      "step": 154
+    },
+    {
+      "epoch": 0.009525565388397246,
+      "grad_norm": 21.625,
+      "learning_rate": 2.0533333333333337e-06,
+      "loss": 1.1152,
+      "step": 155
+    },
+    {
+      "epoch": 0.009587020648967551,
+      "grad_norm": 13.375,
+      "learning_rate": 2.0666666666666666e-06,
+      "loss": 1.1109,
+      "step": 156
+    },
+    {
+      "epoch": 0.009648475909537857,
+      "grad_norm": 20.25,
+      "learning_rate": 2.08e-06,
+      "loss": 1.1053,
+      "step": 157
+    },
+    {
+      "epoch": 0.009709931170108162,
+      "grad_norm": 10.375,
+      "learning_rate": 2.0933333333333338e-06,
+      "loss": 1.1225,
+      "step": 158
+    },
+    {
+      "epoch": 0.009771386430678467,
+      "grad_norm": 11.4375,
+      "learning_rate": 2.1066666666666667e-06,
+      "loss": 1.1424,
+      "step": 159
+    },
+    {
+      "epoch": 0.00983284169124877,
+      "grad_norm": 28.5,
+      "learning_rate": 2.12e-06,
+      "loss": 1.1637,
+      "step": 160
+    },
+    {
+      "epoch": 0.009894296951819075,
+      "grad_norm": 19.5,
+      "learning_rate": 2.133333333333334e-06,
+      "loss": 1.1104,
+      "step": 161
+    },
+    {
+      "epoch": 0.00995575221238938,
+      "grad_norm": 30.75,
+      "learning_rate": 2.1466666666666667e-06,
+      "loss": 1.1314,
+      "step": 162
+    },
+    {
+      "epoch": 0.010017207472959685,
+      "grad_norm": 13.625,
+      "learning_rate": 2.16e-06,
+      "loss": 1.0777,
+      "step": 163
+    },
+    {
+      "epoch": 0.01007866273352999,
+      "grad_norm": 14.125,
+      "learning_rate": 2.1733333333333334e-06,
+      "loss": 1.1029,
+      "step": 164
+    },
+    {
+      "epoch": 0.010140117994100296,
+      "grad_norm": 28.625,
+      "learning_rate": 2.1866666666666668e-06,
+      "loss": 1.0845,
+      "step": 165
+    },
+    {
+      "epoch": 0.010201573254670599,
+      "grad_norm": 16.375,
+      "learning_rate": 2.2e-06,
+      "loss": 1.1821,
+      "step": 166
+    },
+    {
+      "epoch": 0.010263028515240904,
+      "grad_norm": 21.25,
+      "learning_rate": 2.2133333333333335e-06,
+      "loss": 1.141,
+      "step": 167
+    },
+    {
+      "epoch": 0.01032448377581121,
+      "grad_norm": 17.0,
+      "learning_rate": 2.226666666666667e-06,
+      "loss": 1.1344,
+      "step": 168
+    },
+    {
+      "epoch": 0.010385939036381514,
+      "grad_norm": 13.125,
+      "learning_rate": 2.24e-06,
+      "loss": 1.1131,
+      "step": 169
+    },
+    {
+      "epoch": 0.01044739429695182,
+      "grad_norm": 16.125,
+      "learning_rate": 2.2533333333333335e-06,
+      "loss": 1.0608,
+      "step": 170
+    },
+    {
+      "epoch": 0.010508849557522125,
+      "grad_norm": 15.0,
+      "learning_rate": 2.266666666666667e-06,
+      "loss": 1.1074,
+      "step": 171
+    },
+    {
+      "epoch": 0.010570304818092428,
+      "grad_norm": 16.75,
+      "learning_rate": 2.28e-06,
+      "loss": 1.1097,
+      "step": 172
+    },
+    {
+      "epoch": 0.010631760078662733,
+      "grad_norm": 12.4375,
+      "learning_rate": 2.2933333333333335e-06,
+      "loss": 1.1263,
+      "step": 173
+    },
+    {
+      "epoch": 0.010693215339233038,
+      "grad_norm": 36.25,
+      "learning_rate": 2.306666666666667e-06,
+      "loss": 1.1327,
+      "step": 174
+    },
+    {
+      "epoch": 0.010754670599803343,
+      "grad_norm": 13.375,
+      "learning_rate": 2.3200000000000002e-06,
+      "loss": 1.1072,
+      "step": 175
+    },
+    {
+      "epoch": 0.010816125860373648,
+      "grad_norm": 16.875,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": 1.1052,
+      "step": 176
+    },
+    {
+      "epoch": 0.010877581120943954,
+      "grad_norm": 12.5625,
+      "learning_rate": 2.346666666666667e-06,
+      "loss": 1.1051,
+      "step": 177
+    },
+    {
+      "epoch": 0.010939036381514257,
+      "grad_norm": 19.0,
+      "learning_rate": 2.3600000000000003e-06,
+      "loss": 1.2418,
+      "step": 178
+    },
+    {
+      "epoch": 0.011000491642084562,
+      "grad_norm": 18.25,
+      "learning_rate": 2.3733333333333336e-06,
+      "loss": 1.0645,
+      "step": 179
+    },
+    {
+      "epoch": 0.011061946902654867,
+      "grad_norm": 9.25,
+      "learning_rate": 2.386666666666667e-06,
+      "loss": 1.1151,
+      "step": 180
+    },
+    {
+      "epoch": 0.011123402163225172,
+      "grad_norm": 34.25,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.1364,
+      "step": 181
+    },
+    {
+      "epoch": 0.011184857423795477,
+      "grad_norm": 21.375,
+      "learning_rate": 2.4133333333333337e-06,
+      "loss": 1.1714,
+      "step": 182
+    },
+    {
+      "epoch": 0.011246312684365782,
+      "grad_norm": 17.875,
+      "learning_rate": 2.426666666666667e-06,
+      "loss": 1.1789,
+      "step": 183
+    },
+    {
+      "epoch": 0.011307767944936086,
+      "grad_norm": 39.0,
+      "learning_rate": 2.4400000000000004e-06,
+      "loss": 1.0766,
+      "step": 184
+    },
+    {
+      "epoch": 0.011369223205506391,
+      "grad_norm": 18.25,
+      "learning_rate": 2.4533333333333333e-06,
+      "loss": 1.2983,
+      "step": 185
+    },
+    {
+      "epoch": 0.011430678466076696,
+      "grad_norm": 12.375,
+      "learning_rate": 2.466666666666667e-06,
+      "loss": 1.1043,
+      "step": 186
+    },
+    {
+      "epoch": 0.011492133726647001,
+      "grad_norm": 22.5,
+      "learning_rate": 2.4800000000000004e-06,
+      "loss": 1.1895,
+      "step": 187
+    },
+    {
+      "epoch": 0.011553588987217306,
+      "grad_norm": 21.75,
+      "learning_rate": 2.4933333333333333e-06,
+      "loss": 1.1812,
+      "step": 188
+    },
+    {
+      "epoch": 0.011615044247787611,
+      "grad_norm": 11.1875,
+      "learning_rate": 2.5066666666666667e-06,
+      "loss": 1.1458,
+      "step": 189
+    },
+    {
+      "epoch": 0.011676499508357915,
+      "grad_norm": 20.625,
+      "learning_rate": 2.52e-06,
+      "loss": 1.3059,
+      "step": 190
+    },
+    {
+      "epoch": 0.01173795476892822,
+      "grad_norm": 16.625,
+      "learning_rate": 2.5333333333333338e-06,
+      "loss": 1.1491,
+      "step": 191
+    },
+    {
+      "epoch": 0.011799410029498525,
+      "grad_norm": 17.25,
+      "learning_rate": 2.5466666666666667e-06,
+      "loss": 1.1945,
+      "step": 192
+    },
+    {
+      "epoch": 0.01186086529006883,
+      "grad_norm": 23.375,
+      "learning_rate": 2.56e-06,
+      "loss": 1.1187,
+      "step": 193
+    },
+    {
+      "epoch": 0.011922320550639135,
+      "grad_norm": 32.5,
+      "learning_rate": 2.573333333333334e-06,
+      "loss": 1.1354,
+      "step": 194
+    },
+    {
+      "epoch": 0.01198377581120944,
+      "grad_norm": 18.875,
+      "learning_rate": 2.5866666666666667e-06,
+      "loss": 1.1791,
+      "step": 195
+    },
+    {
+      "epoch": 0.012045231071779744,
+      "grad_norm": 20.625,
+      "learning_rate": 2.6e-06,
+      "loss": 1.1322,
+      "step": 196
+    },
+    {
+      "epoch": 0.012106686332350049,
+      "grad_norm": 32.0,
+      "learning_rate": 2.6133333333333334e-06,
+      "loss": 1.1313,
+      "step": 197
+    },
+    {
+      "epoch": 0.012168141592920354,
+      "grad_norm": 45.75,
+      "learning_rate": 2.6266666666666668e-06,
+      "loss": 1.1602,
+      "step": 198
+    },
+    {
+      "epoch": 0.012229596853490659,
+      "grad_norm": 14.375,
+      "learning_rate": 2.64e-06,
+      "loss": 1.1752,
+      "step": 199
+    },
+    {
+      "epoch": 0.012291052114060964,
+      "grad_norm": 30.875,
+      "learning_rate": 2.6533333333333335e-06,
+      "loss": 1.2652,
+      "step": 200
+    },
+    {
+      "epoch": 0.012352507374631269,
+      "grad_norm": 23.75,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.1371,
+      "step": 201
+    },
+    {
+      "epoch": 0.012413962635201573,
+      "grad_norm": 21.25,
+      "learning_rate": 2.68e-06,
+      "loss": 1.0729,
+      "step": 202
+    },
+    {
+      "epoch": 0.012475417895771878,
+      "grad_norm": 17.0,
+      "learning_rate": 2.6933333333333335e-06,
+      "loss": 1.0273,
+      "step": 203
+    },
+    {
+      "epoch": 0.012536873156342183,
+      "grad_norm": 20.0,
+      "learning_rate": 2.706666666666667e-06,
+      "loss": 1.1027,
+      "step": 204
+    },
+    {
+      "epoch": 0.012598328416912488,
+      "grad_norm": 21.625,
+      "learning_rate": 2.7200000000000002e-06,
+      "loss": 1.149,
+      "step": 205
+    },
+    {
+      "epoch": 0.012659783677482793,
+      "grad_norm": 19.375,
+      "learning_rate": 2.7333333333333336e-06,
+      "loss": 1.0611,
+      "step": 206
+    },
+    {
+      "epoch": 0.012721238938053098,
+      "grad_norm": 17.125,
+      "learning_rate": 2.746666666666667e-06,
+      "loss": 1.055,
+      "step": 207
+    },
+    {
+      "epoch": 0.012782694198623401,
+      "grad_norm": 9.9375,
+      "learning_rate": 2.7600000000000003e-06,
+      "loss": 1.0826,
+      "step": 208
+    },
+    {
+      "epoch": 0.012844149459193707,
+      "grad_norm": 39.25,
+      "learning_rate": 2.7733333333333336e-06,
+      "loss": 1.0489,
+      "step": 209
+    },
+    {
+      "epoch": 0.012905604719764012,
+      "grad_norm": 11.1875,
+      "learning_rate": 2.786666666666667e-06,
+      "loss": 1.1321,
+      "step": 210
+    },
+    {
+      "epoch": 0.012967059980334317,
+      "grad_norm": 24.75,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 1.1996,
+      "step": 211
+    },
+    {
+      "epoch": 0.013028515240904622,
+      "grad_norm": 21.875,
+      "learning_rate": 2.8133333333333336e-06,
+      "loss": 1.1491,
+      "step": 212
+    },
+    {
+      "epoch": 0.013089970501474927,
+      "grad_norm": 20.375,
+      "learning_rate": 2.826666666666667e-06,
+      "loss": 1.0555,
+      "step": 213
+    },
+    {
+      "epoch": 0.01315142576204523,
+      "grad_norm": 25.125,
+      "learning_rate": 2.84e-06,
+      "loss": 1.1075,
+      "step": 214
+    },
+    {
+      "epoch": 0.013212881022615535,
+      "grad_norm": 10.125,
+      "learning_rate": 2.8533333333333337e-06,
+      "loss": 1.0926,
+      "step": 215
+    },
+    {
+      "epoch": 0.01327433628318584,
+      "grad_norm": 17.375,
+      "learning_rate": 2.866666666666667e-06,
+      "loss": 1.1172,
+      "step": 216
+    },
+    {
+      "epoch": 0.013335791543756146,
+      "grad_norm": 30.875,
+      "learning_rate": 2.88e-06,
+      "loss": 1.1143,
+      "step": 217
+    },
+    {
+      "epoch": 0.01339724680432645,
+      "grad_norm": 14.875,
+      "learning_rate": 2.8933333333333337e-06,
+      "loss": 1.1945,
+      "step": 218
+    },
+    {
+      "epoch": 0.013458702064896756,
+      "grad_norm": 20.0,
+      "learning_rate": 2.906666666666667e-06,
+      "loss": 1.111,
+      "step": 219
+    },
+    {
+      "epoch": 0.01352015732546706,
+      "grad_norm": 12.25,
+      "learning_rate": 2.92e-06,
+      "loss": 1.1282,
+      "step": 220
+    },
+    {
+      "epoch": 0.013581612586037364,
+      "grad_norm": 29.5,
+      "learning_rate": 2.9333333333333338e-06,
+      "loss": 1.0827,
+      "step": 221
+    },
+    {
+      "epoch": 0.01364306784660767,
+      "grad_norm": 10.0625,
+      "learning_rate": 2.946666666666667e-06,
+      "loss": 1.1045,
+      "step": 222
+    },
+    {
+      "epoch": 0.013704523107177975,
+      "grad_norm": 40.5,
+      "learning_rate": 2.96e-06,
+      "loss": 1.1818,
+      "step": 223
+    },
+    {
+      "epoch": 0.01376597836774828,
+      "grad_norm": 38.25,
+      "learning_rate": 2.973333333333334e-06,
+      "loss": 1.1809,
+      "step": 224
+    },
+    {
+      "epoch": 0.013827433628318585,
+      "grad_norm": 20.5,
+      "learning_rate": 2.986666666666667e-06,
+      "loss": 1.0895,
+      "step": 225
+    },
+    {
+      "epoch": 0.013888888888888888,
+      "grad_norm": 13.5,
+      "learning_rate": 3e-06,
+      "loss": 1.0573,
+      "step": 226
+    },
+    {
+      "epoch": 0.013950344149459193,
+      "grad_norm": 22.625,
+      "learning_rate": 3.013333333333334e-06,
+      "loss": 1.1354,
+      "step": 227
+    },
+    {
+      "epoch": 0.014011799410029498,
+      "grad_norm": 21.75,
+      "learning_rate": 3.0266666666666668e-06,
+      "loss": 1.0277,
+      "step": 228
+    },
+    {
+      "epoch": 0.014073254670599803,
+      "grad_norm": 15.375,
+      "learning_rate": 3.04e-06,
+      "loss": 1.0699,
+      "step": 229
+    },
+    {
+      "epoch": 0.014134709931170109,
+      "grad_norm": 28.0,
+      "learning_rate": 3.053333333333334e-06,
+      "loss": 1.0487,
+      "step": 230
+    },
+    {
+      "epoch": 0.014196165191740414,
+      "grad_norm": 43.75,
+      "learning_rate": 3.066666666666667e-06,
+      "loss": 1.0535,
+      "step": 231
+    },
+    {
+      "epoch": 0.014257620452310717,
+      "grad_norm": 9.5,
+      "learning_rate": 3.08e-06,
+      "loss": 1.0433,
+      "step": 232
+    },
+    {
+      "epoch": 0.014319075712881022,
+      "grad_norm": 35.5,
+      "learning_rate": 3.093333333333334e-06,
+      "loss": 1.2087,
+      "step": 233
+    },
+    {
+      "epoch": 0.014380530973451327,
+      "grad_norm": 34.25,
+      "learning_rate": 3.106666666666667e-06,
+      "loss": 1.25,
+      "step": 234
+    },
+    {
+      "epoch": 0.014441986234021632,
+      "grad_norm": 9.75,
+      "learning_rate": 3.12e-06,
+      "loss": 1.1464,
+      "step": 235
+    },
+    {
+      "epoch": 0.014503441494591937,
+      "grad_norm": 12.25,
+      "learning_rate": 3.133333333333334e-06,
+      "loss": 1.1258,
+      "step": 236
+    },
+    {
+      "epoch": 0.014564896755162243,
+      "grad_norm": 13.75,
+      "learning_rate": 3.146666666666667e-06,
+      "loss": 1.2378,
+      "step": 237
+    },
+    {
+      "epoch": 0.014626352015732546,
+      "grad_norm": 52.75,
+      "learning_rate": 3.1600000000000002e-06,
+      "loss": 1.2047,
+      "step": 238
+    },
+    {
+      "epoch": 0.014687807276302851,
+      "grad_norm": 12.25,
+      "learning_rate": 3.173333333333334e-06,
+      "loss": 1.0881,
+      "step": 239
+    },
+    {
+      "epoch": 0.014749262536873156,
+      "grad_norm": 13.3125,
+      "learning_rate": 3.186666666666667e-06,
+      "loss": 1.1176,
+      "step": 240
+    },
+    {
+      "epoch": 0.014810717797443461,
+      "grad_norm": 24.125,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.1777,
+      "step": 241
+    },
+    {
+      "epoch": 0.014872173058013766,
+      "grad_norm": 15.6875,
+      "learning_rate": 3.213333333333334e-06,
+      "loss": 1.2388,
+      "step": 242
+    },
+    {
+      "epoch": 0.014933628318584071,
+      "grad_norm": 37.0,
+      "learning_rate": 3.226666666666667e-06,
+      "loss": 1.1639,
+      "step": 243
+    },
+    {
+      "epoch": 0.014995083579154375,
+      "grad_norm": 26.375,
+      "learning_rate": 3.2400000000000003e-06,
+      "loss": 1.2209,
+      "step": 244
+    },
+    {
+      "epoch": 0.01505653883972468,
+      "grad_norm": 13.0,
+      "learning_rate": 3.2533333333333332e-06,
+      "loss": 1.1995,
+      "step": 245
+    },
+    {
+      "epoch": 0.015117994100294985,
+      "grad_norm": 18.625,
+      "learning_rate": 3.266666666666667e-06,
+      "loss": 1.1579,
+      "step": 246
+    },
+    {
+      "epoch": 0.01517944936086529,
+      "grad_norm": 19.875,
+      "learning_rate": 3.2800000000000004e-06,
+      "loss": 1.1907,
+      "step": 247
+    },
+    {
+      "epoch": 0.015240904621435595,
+      "grad_norm": 18.375,
+      "learning_rate": 3.2933333333333333e-06,
+      "loss": 1.2377,
+      "step": 248
+    },
+    {
+      "epoch": 0.0153023598820059,
+      "grad_norm": 16.5,
+      "learning_rate": 3.306666666666667e-06,
+      "loss": 1.0934,
+      "step": 249
+    },
+    {
+      "epoch": 0.015363815142576204,
+      "grad_norm": 15.0,
+      "learning_rate": 3.3200000000000004e-06,
+      "loss": 1.0654,
+      "step": 250
+    },
+    {
+      "epoch": 0.015425270403146509,
+      "grad_norm": 20.625,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.1563,
+      "step": 251
+    },
+    {
+      "epoch": 0.015486725663716814,
+      "grad_norm": 16.625,
+      "learning_rate": 3.346666666666667e-06,
+      "loss": 1.099,
+      "step": 252
+    },
+    {
+      "epoch": 0.015548180924287119,
+      "grad_norm": 12.6875,
+      "learning_rate": 3.3600000000000004e-06,
+      "loss": 1.0963,
+      "step": 253
+    },
+    {
+      "epoch": 0.015609636184857424,
+      "grad_norm": 18.625,
+      "learning_rate": 3.3733333333333334e-06,
+      "loss": 1.0406,
+      "step": 254
+    },
+    {
+      "epoch": 0.015671091445427728,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.386666666666667e-06,
+      "loss": 1.0701,
+      "step": 255
+    },
+    {
+      "epoch": 0.015732546705998034,
+      "grad_norm": 14.1875,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 1.077,
+      "step": 256
+    },
+    {
+      "epoch": 0.015794001966568338,
+      "grad_norm": 17.75,
+      "learning_rate": 3.4133333333333334e-06,
+      "loss": 1.1132,
+      "step": 257
+    },
+    {
+      "epoch": 0.015855457227138645,
+      "grad_norm": 23.25,
+      "learning_rate": 3.426666666666667e-06,
+      "loss": 1.1191,
+      "step": 258
+    },
+    {
+      "epoch": 0.015916912487708948,
+      "grad_norm": 26.625,
+      "learning_rate": 3.44e-06,
+      "loss": 1.322,
+      "step": 259
+    },
+    {
+      "epoch": 0.01597836774827925,
+      "grad_norm": 14.6875,
+      "learning_rate": 3.4533333333333334e-06,
+      "loss": 1.1019,
+      "step": 260
+    },
+    {
+      "epoch": 0.016039823008849558,
+      "grad_norm": 11.5,
+      "learning_rate": 3.4666666666666672e-06,
+      "loss": 1.1694,
+      "step": 261
+    },
+    {
+      "epoch": 0.01610127826941986,
+      "grad_norm": 17.25,
+      "learning_rate": 3.48e-06,
+      "loss": 1.1177,
+      "step": 262
+    },
+    {
+      "epoch": 0.01616273352999017,
+      "grad_norm": 13.25,
+      "learning_rate": 3.4933333333333335e-06,
+      "loss": 1.0603,
+      "step": 263
+    },
+    {
+      "epoch": 0.016224188790560472,
+      "grad_norm": 23.5,
+      "learning_rate": 3.5066666666666673e-06,
+      "loss": 1.1767,
+      "step": 264
+    },
+    {
+      "epoch": 0.016285644051130775,
+      "grad_norm": 35.0,
+      "learning_rate": 3.52e-06,
+      "loss": 1.2425,
+      "step": 265
+    },
+    {
+      "epoch": 0.016347099311701082,
+      "grad_norm": 10.875,
+      "learning_rate": 3.5333333333333335e-06,
+      "loss": 1.1201,
+      "step": 266
+    },
+    {
+      "epoch": 0.016408554572271385,
+      "grad_norm": 11.0,
+      "learning_rate": 3.5466666666666673e-06,
+      "loss": 1.1221,
+      "step": 267
+    },
+    {
+      "epoch": 0.016470009832841692,
+      "grad_norm": 16.5,
+      "learning_rate": 3.5600000000000002e-06,
+      "loss": 1.27,
+      "step": 268
+    },
+    {
+      "epoch": 0.016531465093411996,
+      "grad_norm": 13.375,
+      "learning_rate": 3.5733333333333336e-06,
+      "loss": 1.1907,
+      "step": 269
+    },
+    {
+      "epoch": 0.016592920353982302,
+      "grad_norm": 11.125,
+      "learning_rate": 3.5866666666666673e-06,
+      "loss": 1.1787,
+      "step": 270
+    },
+    {
+      "epoch": 0.016654375614552606,
+      "grad_norm": 22.25,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 1.232,
+      "step": 271
+    },
+    {
+      "epoch": 0.01671583087512291,
+      "grad_norm": 32.75,
+      "learning_rate": 3.6133333333333336e-06,
+      "loss": 1.1656,
+      "step": 272
+    },
+    {
+      "epoch": 0.016777286135693216,
+      "grad_norm": 17.25,
+      "learning_rate": 3.6266666666666674e-06,
+      "loss": 1.1151,
+      "step": 273
+    },
+    {
+      "epoch": 0.01683874139626352,
+      "grad_norm": 19.0,
+      "learning_rate": 3.6400000000000003e-06,
+      "loss": 1.0429,
+      "step": 274
+    },
+    {
+      "epoch": 0.016900196656833826,
+      "grad_norm": 16.75,
+      "learning_rate": 3.6533333333333336e-06,
+      "loss": 1.1054,
+      "step": 275
+    },
+    {
+      "epoch": 0.01696165191740413,
+      "grad_norm": 16.125,
+      "learning_rate": 3.6666666666666666e-06,
+      "loss": 1.1221,
+      "step": 276
+    },
+    {
+      "epoch": 0.017023107177974433,
+      "grad_norm": 19.375,
+      "learning_rate": 3.6800000000000003e-06,
+      "loss": 1.1401,
+      "step": 277
+    },
+    {
+      "epoch": 0.01708456243854474,
+      "grad_norm": 17.375,
+      "learning_rate": 3.6933333333333337e-06,
+      "loss": 1.1432,
+      "step": 278
+    },
+    {
+      "epoch": 0.017146017699115043,
+      "grad_norm": 25.75,
+      "learning_rate": 3.7066666666666666e-06,
+      "loss": 1.0428,
+      "step": 279
+    },
+    {
+      "epoch": 0.01720747295968535,
+      "grad_norm": 15.6875,
+      "learning_rate": 3.7200000000000004e-06,
+      "loss": 1.0607,
+      "step": 280
+    },
+    {
+      "epoch": 0.017268928220255653,
+      "grad_norm": 10.0,
+      "learning_rate": 3.7333333333333337e-06,
+      "loss": 1.1018,
+      "step": 281
+    },
+    {
+      "epoch": 0.01733038348082596,
+      "grad_norm": 22.875,
+      "learning_rate": 3.7466666666666667e-06,
+      "loss": 1.1253,
+      "step": 282
+    },
+    {
+      "epoch": 0.017391838741396264,
+      "grad_norm": 15.8125,
+      "learning_rate": 3.7600000000000004e-06,
+      "loss": 1.1724,
+      "step": 283
+    },
+    {
+      "epoch": 0.017453294001966567,
+      "grad_norm": 34.25,
+      "learning_rate": 3.7733333333333338e-06,
+      "loss": 1.1102,
+      "step": 284
+    },
+    {
+      "epoch": 0.017514749262536874,
+      "grad_norm": 21.375,
+      "learning_rate": 3.7866666666666667e-06,
+      "loss": 1.1293,
+      "step": 285
+    },
+    {
+      "epoch": 0.017576204523107177,
+      "grad_norm": 29.5,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 1.2461,
+      "step": 286
+    },
+    {
+      "epoch": 0.017637659783677484,
+      "grad_norm": 38.75,
+      "learning_rate": 3.813333333333334e-06,
+      "loss": 1.1235,
+      "step": 287
+    },
+    {
+      "epoch": 0.017699115044247787,
+      "grad_norm": 9.3125,
+      "learning_rate": 3.826666666666667e-06,
+      "loss": 1.1524,
+      "step": 288
+    },
+    {
+      "epoch": 0.01776057030481809,
+      "grad_norm": 21.75,
+      "learning_rate": 3.8400000000000005e-06,
+      "loss": 1.0407,
+      "step": 289
+    },
+    {
+      "epoch": 0.017822025565388398,
+      "grad_norm": 13.8125,
+      "learning_rate": 3.853333333333334e-06,
+      "loss": 1.0842,
+      "step": 290
+    },
+    {
+      "epoch": 0.0178834808259587,
+      "grad_norm": 11.625,
+      "learning_rate": 3.866666666666667e-06,
+      "loss": 1.0728,
+      "step": 291
+    },
+    {
+      "epoch": 0.017944936086529008,
+      "grad_norm": 13.3125,
+      "learning_rate": 3.88e-06,
+      "loss": 1.0725,
+      "step": 292
+    },
+    {
+      "epoch": 0.01800639134709931,
+      "grad_norm": 36.5,
+      "learning_rate": 3.893333333333333e-06,
+      "loss": 1.0338,
+      "step": 293
+    },
+    {
+      "epoch": 0.018067846607669618,
+      "grad_norm": 11.75,
+      "learning_rate": 3.906666666666667e-06,
+      "loss": 1.0135,
+      "step": 294
+    },
+    {
+      "epoch": 0.01812930186823992,
+      "grad_norm": 27.25,
+      "learning_rate": 3.920000000000001e-06,
+      "loss": 1.0667,
+      "step": 295
+    },
+    {
+      "epoch": 0.018190757128810225,
+      "grad_norm": 15.25,
+      "learning_rate": 3.9333333333333335e-06,
+      "loss": 1.068,
+      "step": 296
+    },
+    {
+      "epoch": 0.01825221238938053,
+      "grad_norm": 46.75,
+      "learning_rate": 3.946666666666667e-06,
+      "loss": 1.2074,
+      "step": 297
+    },
+    {
+      "epoch": 0.018313667649950835,
+      "grad_norm": 24.125,
+      "learning_rate": 3.96e-06,
+      "loss": 1.1072,
+      "step": 298
+    },
+    {
+      "epoch": 0.018375122910521142,
+      "grad_norm": 41.75,
+      "learning_rate": 3.973333333333333e-06,
+      "loss": 1.2243,
+      "step": 299
+    },
+    {
+      "epoch": 0.018436578171091445,
+      "grad_norm": 36.0,
+      "learning_rate": 3.986666666666667e-06,
+      "loss": 1.1642,
+      "step": 300
+    },
+    {
+      "epoch": 0.01849803343166175,
+      "grad_norm": 22.875,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.1821,
+      "step": 301
+    },
+    {
+      "epoch": 0.018559488692232055,
+      "grad_norm": 20.625,
+      "learning_rate": 4.013333333333334e-06,
+      "loss": 1.2046,
+      "step": 302
+    },
+    {
+      "epoch": 0.01862094395280236,
+      "grad_norm": 22.625,
+      "learning_rate": 4.026666666666667e-06,
+      "loss": 1.2256,
+      "step": 303
+    },
+    {
+      "epoch": 0.018682399213372666,
+      "grad_norm": 20.0,
+      "learning_rate": 4.04e-06,
+      "loss": 1.2917,
+      "step": 304
+    },
+    {
+      "epoch": 0.01874385447394297,
+      "grad_norm": 34.0,
+      "learning_rate": 4.053333333333333e-06,
+      "loss": 1.1289,
+      "step": 305
+    },
+    {
+      "epoch": 0.018805309734513276,
+      "grad_norm": 37.0,
+      "learning_rate": 4.066666666666667e-06,
+      "loss": 1.1572,
+      "step": 306
+    },
+    {
+      "epoch": 0.01886676499508358,
+      "grad_norm": 24.375,
+      "learning_rate": 4.08e-06,
+      "loss": 1.106,
+      "step": 307
+    },
+    {
+      "epoch": 0.018928220255653883,
+      "grad_norm": 27.0,
+      "learning_rate": 4.093333333333334e-06,
+      "loss": 1.1808,
+      "step": 308
+    },
+    {
+      "epoch": 0.01898967551622419,
+      "grad_norm": 13.5625,
+      "learning_rate": 4.1066666666666674e-06,
+      "loss": 1.1536,
+      "step": 309
+    },
+    {
+      "epoch": 0.019051130776794493,
+      "grad_norm": 25.75,
+      "learning_rate": 4.12e-06,
+      "loss": 1.1452,
+      "step": 310
+    },
+    {
+      "epoch": 0.0191125860373648,
+      "grad_norm": 41.5,
+      "learning_rate": 4.133333333333333e-06,
+      "loss": 1.0979,
+      "step": 311
+    },
+    {
+      "epoch": 0.019174041297935103,
+      "grad_norm": 11.6875,
+      "learning_rate": 4.146666666666667e-06,
+      "loss": 1.1598,
+      "step": 312
+    },
+    {
+      "epoch": 0.019235496558505406,
+      "grad_norm": 18.375,
+      "learning_rate": 4.16e-06,
+      "loss": 1.0453,
+      "step": 313
+    },
+    {
+      "epoch": 0.019296951819075713,
+      "grad_norm": 21.75,
+      "learning_rate": 4.173333333333334e-06,
+      "loss": 1.0437,
+      "step": 314
+    },
+    {
+      "epoch": 0.019358407079646017,
+      "grad_norm": 13.375,
+      "learning_rate": 4.1866666666666675e-06,
+      "loss": 1.0363,
+      "step": 315
+    },
+    {
+      "epoch": 0.019419862340216323,
+      "grad_norm": 8.125,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 1.0984,
+      "step": 316
+    },
+    {
+      "epoch": 0.019481317600786627,
+      "grad_norm": 30.125,
+      "learning_rate": 4.213333333333333e-06,
+      "loss": 1.0599,
+      "step": 317
+    },
+    {
+      "epoch": 0.019542772861356934,
+      "grad_norm": 15.375,
+      "learning_rate": 4.226666666666667e-06,
+      "loss": 1.076,
+      "step": 318
+    },
+    {
+      "epoch": 0.019604228121927237,
+      "grad_norm": 19.375,
+      "learning_rate": 4.24e-06,
+      "loss": 1.0374,
+      "step": 319
+    },
+    {
+      "epoch": 0.01966568338249754,
+      "grad_norm": 11.25,
+      "learning_rate": 4.253333333333334e-06,
+      "loss": 1.0459,
+      "step": 320
+    },
+    {
+      "epoch": 0.019727138643067847,
+      "grad_norm": 8.625,
+      "learning_rate": 4.266666666666668e-06,
+      "loss": 1.0229,
+      "step": 321
+    },
+    {
+      "epoch": 0.01978859390363815,
+      "grad_norm": 19.25,
+      "learning_rate": 4.2800000000000005e-06,
+      "loss": 1.0628,
+      "step": 322
+    },
+    {
+      "epoch": 0.019850049164208457,
+      "grad_norm": 14.4375,
+      "learning_rate": 4.2933333333333334e-06,
+      "loss": 1.0568,
+      "step": 323
+    },
+    {
+      "epoch": 0.01991150442477876,
+      "grad_norm": 9.8125,
+      "learning_rate": 4.306666666666666e-06,
+      "loss": 1.037,
+      "step": 324
+    },
+    {
+      "epoch": 0.019972959685349064,
+      "grad_norm": 26.125,
+      "learning_rate": 4.32e-06,
+      "loss": 1.0578,
+      "step": 325
+    },
+    {
+      "epoch": 0.02003441494591937,
+      "grad_norm": 11.8125,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 1.0025,
+      "step": 326
+    },
+    {
+      "epoch": 0.020095870206489674,
+      "grad_norm": 13.375,
+      "learning_rate": 4.346666666666667e-06,
+      "loss": 1.0442,
+      "step": 327
+    },
+    {
+      "epoch": 0.02015732546705998,
+      "grad_norm": 21.25,
+      "learning_rate": 4.360000000000001e-06,
+      "loss": 0.9953,
+      "step": 328
+    },
+    {
+      "epoch": 0.020218780727630285,
+      "grad_norm": 11.75,
+      "learning_rate": 4.3733333333333335e-06,
+      "loss": 1.1609,
+      "step": 329
+    },
+    {
+      "epoch": 0.02028023598820059,
+      "grad_norm": 23.75,
+      "learning_rate": 4.3866666666666665e-06,
+      "loss": 1.1098,
+      "step": 330
+    },
+    {
+      "epoch": 0.020341691248770895,
+      "grad_norm": 37.25,
+      "learning_rate": 4.4e-06,
+      "loss": 1.0725,
+      "step": 331
+    },
+    {
+      "epoch": 0.020403146509341198,
+      "grad_norm": 32.25,
+      "learning_rate": 4.413333333333334e-06,
+      "loss": 1.0925,
+      "step": 332
+    },
+    {
+      "epoch": 0.020464601769911505,
+      "grad_norm": 14.1875,
+      "learning_rate": 4.426666666666667e-06,
+      "loss": 1.0454,
+      "step": 333
+    },
+    {
+      "epoch": 0.02052605703048181,
+      "grad_norm": 12.375,
+      "learning_rate": 4.440000000000001e-06,
+      "loss": 1.1047,
+      "step": 334
+    },
+    {
+      "epoch": 0.020587512291052115,
+      "grad_norm": 22.125,
+      "learning_rate": 4.453333333333334e-06,
+      "loss": 1.096,
+      "step": 335
+    },
+    {
+      "epoch": 0.02064896755162242,
+      "grad_norm": 14.875,
+      "learning_rate": 4.4666666666666665e-06,
+      "loss": 1.1137,
+      "step": 336
+    },
+    {
+      "epoch": 0.020710422812192722,
+      "grad_norm": 20.75,
+      "learning_rate": 4.48e-06,
+      "loss": 1.121,
+      "step": 337
+    },
+    {
+      "epoch": 0.02077187807276303,
+      "grad_norm": 19.625,
+      "learning_rate": 4.493333333333333e-06,
+      "loss": 1.1034,
+      "step": 338
+    },
+    {
+      "epoch": 0.020833333333333332,
+      "grad_norm": 11.625,
+      "learning_rate": 4.506666666666667e-06,
+      "loss": 1.1099,
+      "step": 339
+    },
+    {
+      "epoch": 0.02089478859390364,
+      "grad_norm": 18.875,
+      "learning_rate": 4.520000000000001e-06,
+      "loss": 1.0467,
+      "step": 340
+    },
+    {
+      "epoch": 0.020956243854473942,
+      "grad_norm": 23.5,
+      "learning_rate": 4.533333333333334e-06,
+      "loss": 1.2196,
+      "step": 341
+    },
+    {
+      "epoch": 0.02101769911504425,
+      "grad_norm": 24.0,
+      "learning_rate": 4.546666666666667e-06,
+      "loss": 1.1509,
+      "step": 342
+    },
+    {
+      "epoch": 0.021079154375614553,
+      "grad_norm": 27.5,
+      "learning_rate": 4.56e-06,
+      "loss": 1.1707,
+      "step": 343
+    },
+    {
+      "epoch": 0.021140609636184856,
+      "grad_norm": 13.375,
+      "learning_rate": 4.573333333333333e-06,
+      "loss": 1.0749,
+      "step": 344
+    },
+    {
+      "epoch": 0.021202064896755163,
+      "grad_norm": 26.75,
+      "learning_rate": 4.586666666666667e-06,
+      "loss": 1.1022,
+      "step": 345
+    },
+    {
+      "epoch": 0.021263520157325466,
+      "grad_norm": 10.1875,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 1.0706,
+      "step": 346
+    },
+    {
+      "epoch": 0.021324975417895773,
+      "grad_norm": 23.0,
+      "learning_rate": 4.613333333333334e-06,
+      "loss": 1.113,
+      "step": 347
+    },
+    {
+      "epoch": 0.021386430678466076,
+      "grad_norm": 26.625,
+      "learning_rate": 4.626666666666667e-06,
+      "loss": 1.0932,
+      "step": 348
+    },
+    {
+      "epoch": 0.021447885939036383,
+      "grad_norm": 43.5,
+      "learning_rate": 4.6400000000000005e-06,
+      "loss": 1.1524,
+      "step": 349
+    },
+    {
+      "epoch": 0.021509341199606687,
+      "grad_norm": 22.125,
+      "learning_rate": 4.653333333333333e-06,
+      "loss": 1.1053,
+      "step": 350
+    },
+    {
+      "epoch": 0.02157079646017699,
+      "grad_norm": 29.875,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 1.1645,
+      "step": 351
+    },
+    {
+      "epoch": 0.021632251720747297,
+      "grad_norm": 17.0,
+      "learning_rate": 4.680000000000001e-06,
+      "loss": 1.0259,
+      "step": 352
+    },
+    {
+      "epoch": 0.0216937069813176,
+      "grad_norm": 47.75,
+      "learning_rate": 4.693333333333334e-06,
+      "loss": 1.1654,
+      "step": 353
+    },
+    {
+      "epoch": 0.021755162241887907,
+      "grad_norm": 21.75,
+      "learning_rate": 4.706666666666667e-06,
+      "loss": 1.1345,
+      "step": 354
+    },
+    {
+      "epoch": 0.02181661750245821,
+      "grad_norm": 21.125,
+      "learning_rate": 4.7200000000000005e-06,
+      "loss": 1.1081,
+      "step": 355
+    },
+    {
+      "epoch": 0.021878072763028514,
+      "grad_norm": 17.25,
+      "learning_rate": 4.7333333333333335e-06,
+      "loss": 1.108,
+      "step": 356
+    },
+    {
+      "epoch": 0.02193952802359882,
+      "grad_norm": 20.875,
+      "learning_rate": 4.746666666666667e-06,
+      "loss": 1.1756,
+      "step": 357
+    },
+    {
+      "epoch": 0.022000983284169124,
+      "grad_norm": 14.5625,
+      "learning_rate": 4.76e-06,
+      "loss": 1.1457,
+      "step": 358
+    },
+    {
+      "epoch": 0.02206243854473943,
+      "grad_norm": 18.375,
+      "learning_rate": 4.773333333333334e-06,
+      "loss": 1.0998,
+      "step": 359
+    },
+    {
+      "epoch": 0.022123893805309734,
+      "grad_norm": 52.25,
+      "learning_rate": 4.786666666666667e-06,
+      "loss": 1.2224,
+      "step": 360
+    },
+    {
+      "epoch": 0.02218534906588004,
+      "grad_norm": 15.6875,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.184,
+      "step": 361
+    },
+    {
+      "epoch": 0.022246804326450344,
+      "grad_norm": 13.25,
+      "learning_rate": 4.8133333333333336e-06,
+      "loss": 1.1848,
+      "step": 362
+    },
+    {
+      "epoch": 0.022308259587020648,
+      "grad_norm": 16.5,
+      "learning_rate": 4.826666666666667e-06,
+      "loss": 1.0546,
+      "step": 363
+    },
+    {
+      "epoch": 0.022369714847590955,
+      "grad_norm": 77.5,
+      "learning_rate": 4.84e-06,
+      "loss": 1.3082,
+      "step": 364
+    },
+    {
+      "epoch": 0.022431170108161258,
+      "grad_norm": 32.25,
+      "learning_rate": 4.853333333333334e-06,
+      "loss": 1.0865,
+      "step": 365
+    },
+    {
+      "epoch": 0.022492625368731565,
+      "grad_norm": 40.5,
+      "learning_rate": 4.866666666666667e-06,
+      "loss": 1.1381,
+      "step": 366
+    },
+    {
+      "epoch": 0.022554080629301868,
+      "grad_norm": 12.125,
+      "learning_rate": 4.880000000000001e-06,
+      "loss": 1.1044,
+      "step": 367
+    },
+    {
+      "epoch": 0.02261553588987217,
+      "grad_norm": 18.625,
+      "learning_rate": 4.893333333333334e-06,
+      "loss": 1.1129,
+      "step": 368
+    },
+    {
+      "epoch": 0.02267699115044248,
+      "grad_norm": 13.1875,
+      "learning_rate": 4.9066666666666666e-06,
+      "loss": 1.168,
+      "step": 369
+    },
+    {
+      "epoch": 0.022738446411012782,
+      "grad_norm": 14.25,
+      "learning_rate": 4.92e-06,
+      "loss": 1.0833,
+      "step": 370
+    },
+    {
+      "epoch": 0.02279990167158309,
+      "grad_norm": 20.625,
+      "learning_rate": 4.933333333333334e-06,
+      "loss": 1.1194,
+      "step": 371
+    },
+    {
+      "epoch": 0.022861356932153392,
+      "grad_norm": 21.25,
+      "learning_rate": 4.946666666666667e-06,
+      "loss": 1.0971,
+      "step": 372
+    },
+    {
+      "epoch": 0.0229228121927237,
+      "grad_norm": 21.25,
+      "learning_rate": 4.960000000000001e-06,
+      "loss": 1.0615,
+      "step": 373
+    },
+    {
+      "epoch": 0.022984267453294002,
+      "grad_norm": 18.25,
+      "learning_rate": 4.973333333333334e-06,
+      "loss": 1.0367,
+      "step": 374
+    },
+    {
+      "epoch": 0.023045722713864306,
+      "grad_norm": 12.375,
+      "learning_rate": 4.986666666666667e-06,
+      "loss": 1.0912,
+      "step": 375
+    },
+    {
+      "epoch": 0.023107177974434612,
+      "grad_norm": 14.375,
+      "learning_rate": 5e-06,
+      "loss": 1.0483,
+      "step": 376
+    },
+    {
+      "epoch": 0.023168633235004916,
+      "grad_norm": 28.0,
+      "learning_rate": 5.013333333333333e-06,
+      "loss": 1.2354,
+      "step": 377
+    },
+    {
+      "epoch": 0.023230088495575223,
+      "grad_norm": 25.625,
+      "learning_rate": 5.026666666666667e-06,
+      "loss": 1.1683,
+      "step": 378
+    },
+    {
+      "epoch": 0.023291543756145526,
+      "grad_norm": 30.0,
+      "learning_rate": 5.04e-06,
+      "loss": 1.2038,
+      "step": 379
+    },
+    {
+      "epoch": 0.02335299901671583,
+      "grad_norm": 18.875,
+      "learning_rate": 5.053333333333334e-06,
+      "loss": 1.1381,
+      "step": 380
+    },
+    {
+      "epoch": 0.023414454277286136,
+      "grad_norm": 19.5,
+      "learning_rate": 5.0666666666666676e-06,
+      "loss": 1.2141,
+      "step": 381
+    },
+    {
+      "epoch": 0.02347590953785644,
+      "grad_norm": 17.5,
+      "learning_rate": 5.0800000000000005e-06,
+      "loss": 1.0728,
+      "step": 382
+    },
+    {
+      "epoch": 0.023537364798426746,
+      "grad_norm": 27.875,
+      "learning_rate": 5.093333333333333e-06,
+      "loss": 1.2395,
+      "step": 383
+    },
+    {
+      "epoch": 0.02359882005899705,
+      "grad_norm": 8.375,
+      "learning_rate": 5.106666666666667e-06,
+      "loss": 1.1278,
+      "step": 384
+    },
+    {
+      "epoch": 0.023660275319567357,
+      "grad_norm": 10.375,
+      "learning_rate": 5.12e-06,
+      "loss": 1.1472,
+      "step": 385
+    },
+    {
+      "epoch": 0.02372173058013766,
+      "grad_norm": 23.75,
+      "learning_rate": 5.133333333333334e-06,
+      "loss": 1.0952,
+      "step": 386
+    },
+    {
+      "epoch": 0.023783185840707963,
+      "grad_norm": 11.375,
+      "learning_rate": 5.146666666666668e-06,
+      "loss": 1.1354,
+      "step": 387
+    },
+    {
+      "epoch": 0.02384464110127827,
+      "grad_norm": 13.5,
+      "learning_rate": 5.1600000000000006e-06,
+      "loss": 1.0963,
+      "step": 388
+    },
+    {
+      "epoch": 0.023906096361848574,
+      "grad_norm": 19.75,
+      "learning_rate": 5.1733333333333335e-06,
+      "loss": 1.1244,
+      "step": 389
+    },
+    {
+      "epoch": 0.02396755162241888,
+      "grad_norm": 15.6875,
+      "learning_rate": 5.186666666666667e-06,
+      "loss": 1.0766,
+      "step": 390
+    },
+    {
+      "epoch": 0.024029006882989184,
+      "grad_norm": 13.375,
+      "learning_rate": 5.2e-06,
+      "loss": 1.1214,
+      "step": 391
+    },
+    {
+      "epoch": 0.024090462143559487,
+      "grad_norm": 14.25,
+      "learning_rate": 5.213333333333334e-06,
+      "loss": 1.0921,
+      "step": 392
+    },
+    {
+      "epoch": 0.024151917404129794,
+      "grad_norm": 10.0625,
+      "learning_rate": 5.226666666666667e-06,
+      "loss": 1.0561,
+      "step": 393
+    },
+    {
+      "epoch": 0.024213372664700097,
+      "grad_norm": 10.375,
+      "learning_rate": 5.240000000000001e-06,
+      "loss": 1.0064,
+      "step": 394
+    },
+    {
+      "epoch": 0.024274827925270404,
+      "grad_norm": 9.875,
+      "learning_rate": 5.2533333333333336e-06,
+      "loss": 1.084,
+      "step": 395
+    },
+    {
+      "epoch": 0.024336283185840708,
+      "grad_norm": 13.375,
+      "learning_rate": 5.2666666666666665e-06,
+      "loss": 1.1051,
+      "step": 396
+    },
+    {
+      "epoch": 0.024397738446411014,
+      "grad_norm": 36.25,
+      "learning_rate": 5.28e-06,
+      "loss": 1.0979,
+      "step": 397
+    },
+    {
+      "epoch": 0.024459193706981318,
+      "grad_norm": 15.6875,
+      "learning_rate": 5.293333333333334e-06,
+      "loss": 1.1049,
+      "step": 398
+    },
+    {
+      "epoch": 0.02452064896755162,
+      "grad_norm": 14.8125,
+      "learning_rate": 5.306666666666667e-06,
+      "loss": 1.1048,
+      "step": 399
+    },
+    {
+      "epoch": 0.024582104228121928,
+      "grad_norm": 23.375,
+      "learning_rate": 5.320000000000001e-06,
+      "loss": 1.0895,
+      "step": 400
+    },
+    {
+      "epoch": 0.02464355948869223,
+      "grad_norm": 17.0,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.032,
+      "step": 401
+    },
+    {
+      "epoch": 0.024705014749262538,
+      "grad_norm": 7.46875,
+      "learning_rate": 5.346666666666667e-06,
+      "loss": 1.0176,
+      "step": 402
+    },
+    {
+      "epoch": 0.02476647000983284,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.36e-06,
+      "loss": 1.1405,
+      "step": 403
+    },
+    {
+      "epoch": 0.024827925270403145,
+      "grad_norm": 26.0,
+      "learning_rate": 5.373333333333334e-06,
+      "loss": 1.0573,
+      "step": 404
+    },
+    {
+      "epoch": 0.024889380530973452,
+      "grad_norm": 10.1875,
+      "learning_rate": 5.386666666666667e-06,
+      "loss": 1.1466,
+      "step": 405
+    },
+    {
+      "epoch": 0.024950835791543755,
+      "grad_norm": 24.25,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 1.1189,
+      "step": 406
+    },
+    {
+      "epoch": 0.025012291052114062,
+      "grad_norm": 19.625,
+      "learning_rate": 5.413333333333334e-06,
+      "loss": 1.0372,
+      "step": 407
+    },
+    {
+      "epoch": 0.025073746312684365,
+      "grad_norm": 36.0,
+      "learning_rate": 5.426666666666667e-06,
+      "loss": 1.0903,
+      "step": 408
+    },
+    {
+      "epoch": 0.025135201573254672,
+      "grad_norm": 41.5,
+      "learning_rate": 5.4400000000000004e-06,
+      "loss": 1.2313,
+      "step": 409
+    },
+    {
+      "epoch": 0.025196656833824976,
+      "grad_norm": 19.0,
+      "learning_rate": 5.453333333333334e-06,
+      "loss": 1.0858,
+      "step": 410
+    },
+    {
+      "epoch": 0.02525811209439528,
+      "grad_norm": 14.9375,
+      "learning_rate": 5.466666666666667e-06,
+      "loss": 1.1539,
+      "step": 411
+    },
+    {
+      "epoch": 0.025319567354965586,
+      "grad_norm": 14.4375,
+      "learning_rate": 5.480000000000001e-06,
+      "loss": 1.0802,
+      "step": 412
+    },
+    {
+      "epoch": 0.02538102261553589,
+      "grad_norm": 26.875,
+      "learning_rate": 5.493333333333334e-06,
+      "loss": 1.0796,
+      "step": 413
+    },
+    {
+      "epoch": 0.025442477876106196,
+      "grad_norm": 17.0,
+      "learning_rate": 5.506666666666667e-06,
+      "loss": 1.0887,
+      "step": 414
+    },
+    {
+      "epoch": 0.0255039331366765,
+      "grad_norm": 22.875,
+      "learning_rate": 5.5200000000000005e-06,
+      "loss": 1.0662,
+      "step": 415
+    },
+    {
+      "epoch": 0.025565388397246803,
+      "grad_norm": 16.5,
+      "learning_rate": 5.533333333333334e-06,
+      "loss": 1.0819,
+      "step": 416
+    },
+    {
+      "epoch": 0.02562684365781711,
+      "grad_norm": 15.5625,
+      "learning_rate": 5.546666666666667e-06,
+      "loss": 1.1185,
+      "step": 417
+    },
+    {
+      "epoch": 0.025688298918387413,
+      "grad_norm": 26.125,
+      "learning_rate": 5.560000000000001e-06,
+      "loss": 1.1053,
+      "step": 418
+    },
+    {
+      "epoch": 0.02574975417895772,
+      "grad_norm": 19.75,
+      "learning_rate": 5.573333333333334e-06,
+      "loss": 1.1059,
+      "step": 419
+    },
+    {
+      "epoch": 0.025811209439528023,
+      "grad_norm": 14.0625,
+      "learning_rate": 5.586666666666667e-06,
+      "loss": 1.0988,
+      "step": 420
+    },
+    {
+      "epoch": 0.02587266470009833,
+      "grad_norm": 9.4375,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.1474,
+      "step": 421
+    },
+    {
+      "epoch": 0.025934119960668633,
+      "grad_norm": 11.0625,
+      "learning_rate": 5.613333333333334e-06,
+      "loss": 1.1135,
+      "step": 422
+    },
+    {
+      "epoch": 0.025995575221238937,
+      "grad_norm": 9.0,
+      "learning_rate": 5.626666666666667e-06,
+      "loss": 1.0204,
+      "step": 423
+    },
+    {
+      "epoch": 0.026057030481809244,
+      "grad_norm": 16.0,
+      "learning_rate": 5.64e-06,
+      "loss": 0.9981,
+      "step": 424
+    },
+    {
+      "epoch": 0.026118485742379547,
+      "grad_norm": 9.0625,
+      "learning_rate": 5.653333333333334e-06,
+      "loss": 1.0163,
+      "step": 425
+    },
+    {
+      "epoch": 0.026179941002949854,
+      "grad_norm": 21.125,
+      "learning_rate": 5.666666666666667e-06,
+      "loss": 1.0589,
+      "step": 426
+    },
+    {
+      "epoch": 0.026241396263520157,
+      "grad_norm": 16.5,
+      "learning_rate": 5.68e-06,
+      "loss": 1.1238,
+      "step": 427
+    },
+    {
+      "epoch": 0.02630285152409046,
+      "grad_norm": 21.25,
+      "learning_rate": 5.6933333333333344e-06,
+      "loss": 1.2123,
+      "step": 428
+    },
+    {
+      "epoch": 0.026364306784660767,
+      "grad_norm": 13.875,
+      "learning_rate": 5.706666666666667e-06,
+      "loss": 1.1196,
+      "step": 429
+    },
+    {
+      "epoch": 0.02642576204523107,
+      "grad_norm": 20.625,
+      "learning_rate": 5.72e-06,
+      "loss": 1.0555,
+      "step": 430
+    },
+    {
+      "epoch": 0.026487217305801378,
+      "grad_norm": 25.375,
+      "learning_rate": 5.733333333333334e-06,
+      "loss": 1.0522,
+      "step": 431
+    },
+    {
+      "epoch": 0.02654867256637168,
+      "grad_norm": 8.5625,
+      "learning_rate": 5.746666666666667e-06,
+      "loss": 1.0864,
+      "step": 432
+    },
+    {
+      "epoch": 0.026610127826941988,
+      "grad_norm": 37.0,
+      "learning_rate": 5.76e-06,
+      "loss": 1.1354,
+      "step": 433
+    },
+    {
+      "epoch": 0.02667158308751229,
+      "grad_norm": 25.125,
+      "learning_rate": 5.7733333333333345e-06,
+      "loss": 1.0033,
+      "step": 434
+    },
+    {
+      "epoch": 0.026733038348082595,
+      "grad_norm": 43.0,
+      "learning_rate": 5.7866666666666674e-06,
+      "loss": 1.047,
+      "step": 435
+    },
+    {
+      "epoch": 0.0267944936086529,
+      "grad_norm": 18.125,
+      "learning_rate": 5.8e-06,
+      "loss": 1.1274,
+      "step": 436
+    },
+    {
+      "epoch": 0.026855948869223205,
+      "grad_norm": 15.4375,
+      "learning_rate": 5.813333333333334e-06,
+      "loss": 1.0806,
+      "step": 437
+    },
+    {
+      "epoch": 0.02691740412979351,
+      "grad_norm": 10.0625,
+      "learning_rate": 5.826666666666667e-06,
+      "loss": 1.0263,
+      "step": 438
+    },
+    {
+      "epoch": 0.026978859390363815,
+      "grad_norm": 16.75,
+      "learning_rate": 5.84e-06,
+      "loss": 1.1131,
+      "step": 439
+    },
+    {
+      "epoch": 0.02704031465093412,
+      "grad_norm": 22.75,
+      "learning_rate": 5.853333333333335e-06,
+      "loss": 1.1802,
+      "step": 440
+    },
+    {
+      "epoch": 0.027101769911504425,
+      "grad_norm": 10.5,
+      "learning_rate": 5.8666666666666675e-06,
+      "loss": 1.0123,
+      "step": 441
+    },
+    {
+      "epoch": 0.02716322517207473,
+      "grad_norm": 13.4375,
+      "learning_rate": 5.8800000000000005e-06,
+      "loss": 1.0406,
+      "step": 442
+    },
+    {
+      "epoch": 0.027224680432645035,
+      "grad_norm": 18.375,
+      "learning_rate": 5.893333333333334e-06,
+      "loss": 1.1492,
+      "step": 443
+    },
+    {
+      "epoch": 0.02728613569321534,
+      "grad_norm": 12.75,
+      "learning_rate": 5.906666666666667e-06,
+      "loss": 1.1121,
+      "step": 444
+    },
+    {
+      "epoch": 0.027347590953785646,
+      "grad_norm": 12.4375,
+      "learning_rate": 5.92e-06,
+      "loss": 1.1758,
+      "step": 445
+    },
+    {
+      "epoch": 0.02740904621435595,
+      "grad_norm": 16.5,
+      "learning_rate": 5.933333333333335e-06,
+      "loss": 1.0911,
+      "step": 446
+    },
+    {
+      "epoch": 0.027470501474926252,
+      "grad_norm": 20.25,
+      "learning_rate": 5.946666666666668e-06,
+      "loss": 1.054,
+      "step": 447
+    },
+    {
+      "epoch": 0.02753195673549656,
+      "grad_norm": 10.4375,
+      "learning_rate": 5.9600000000000005e-06,
+      "loss": 1.0425,
+      "step": 448
+    },
+    {
+      "epoch": 0.027593411996066863,
+      "grad_norm": 17.875,
+      "learning_rate": 5.973333333333334e-06,
+      "loss": 1.224,
+      "step": 449
+    },
+    {
+      "epoch": 0.02765486725663717,
+      "grad_norm": 16.375,
+      "learning_rate": 5.986666666666667e-06,
+      "loss": 1.1029,
+      "step": 450
+    },
+    {
+      "epoch": 0.027716322517207473,
+      "grad_norm": 19.5,
+      "learning_rate": 6e-06,
+      "loss": 1.1007,
+      "step": 451
+    },
+    {
+      "epoch": 0.027777777777777776,
+      "grad_norm": 28.375,
+      "learning_rate": 6.013333333333335e-06,
+      "loss": 1.0575,
+      "step": 452
+    },
+    {
+      "epoch": 0.027839233038348083,
+      "grad_norm": 22.125,
+      "learning_rate": 6.026666666666668e-06,
+      "loss": 1.1137,
+      "step": 453
+    },
+    {
+      "epoch": 0.027900688298918386,
+      "grad_norm": 36.25,
+      "learning_rate": 6.040000000000001e-06,
+      "loss": 1.0707,
+      "step": 454
+    },
+    {
+      "epoch": 0.027962143559488693,
+      "grad_norm": 13.8125,
+      "learning_rate": 6.0533333333333335e-06,
+      "loss": 1.0358,
+      "step": 455
+    },
+    {
+      "epoch": 0.028023598820058997,
+      "grad_norm": 12.5,
+      "learning_rate": 6.066666666666667e-06,
+      "loss": 1.0705,
+      "step": 456
+    },
+    {
+      "epoch": 0.028085054080629304,
+      "grad_norm": 14.75,
+      "learning_rate": 6.08e-06,
+      "loss": 1.0621,
+      "step": 457
+    },
+    {
+      "epoch": 0.028146509341199607,
+      "grad_norm": 14.9375,
+      "learning_rate": 6.093333333333333e-06,
+      "loss": 1.0041,
+      "step": 458
+    },
+    {
+      "epoch": 0.02820796460176991,
+      "grad_norm": 15.6875,
+      "learning_rate": 6.106666666666668e-06,
+      "loss": 0.9826,
+      "step": 459
+    },
+    {
+      "epoch": 0.028269419862340217,
+      "grad_norm": 18.0,
+      "learning_rate": 6.120000000000001e-06,
+      "loss": 1.0493,
+      "step": 460
+    },
+    {
+      "epoch": 0.02833087512291052,
+      "grad_norm": 23.0,
+      "learning_rate": 6.133333333333334e-06,
+      "loss": 0.9997,
+      "step": 461
+    },
+    {
+      "epoch": 0.028392330383480827,
+      "grad_norm": 16.125,
+      "learning_rate": 6.146666666666667e-06,
+      "loss": 1.0271,
+      "step": 462
+    },
+    {
+      "epoch": 0.02845378564405113,
+      "grad_norm": 28.625,
+      "learning_rate": 6.16e-06,
+      "loss": 1.0693,
+      "step": 463
+    },
+    {
+      "epoch": 0.028515240904621434,
+      "grad_norm": 26.25,
+      "learning_rate": 6.173333333333333e-06,
+      "loss": 1.0104,
+      "step": 464
+    },
+    {
+      "epoch": 0.02857669616519174,
+      "grad_norm": 8.9375,
+      "learning_rate": 6.186666666666668e-06,
+      "loss": 0.9769,
+      "step": 465
+    },
+    {
+      "epoch": 0.028638151425762044,
+      "grad_norm": 14.25,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 0.9755,
+      "step": 466
+    },
+    {
+      "epoch": 0.02869960668633235,
+      "grad_norm": 10.6875,
+      "learning_rate": 6.213333333333334e-06,
+      "loss": 1.0195,
+      "step": 467
+    },
+    {
+      "epoch": 0.028761061946902654,
+      "grad_norm": 8.0,
+      "learning_rate": 6.2266666666666675e-06,
+      "loss": 0.9316,
+      "step": 468
+    },
+    {
+      "epoch": 0.02882251720747296,
+      "grad_norm": 10.9375,
+      "learning_rate": 6.24e-06,
+      "loss": 0.8514,
+      "step": 469
+    },
+    {
+      "epoch": 0.028883972468043265,
+      "grad_norm": 24.375,
+      "learning_rate": 6.253333333333333e-06,
+      "loss": 1.0113,
+      "step": 470
+    },
+    {
+      "epoch": 0.028945427728613568,
+      "grad_norm": 20.125,
+      "learning_rate": 6.266666666666668e-06,
+      "loss": 1.0327,
+      "step": 471
+    },
+    {
+      "epoch": 0.029006882989183875,
+      "grad_norm": 35.5,
+      "learning_rate": 6.280000000000001e-06,
+      "loss": 0.885,
+      "step": 472
+    },
+    {
+      "epoch": 0.029068338249754178,
+      "grad_norm": 15.5,
+      "learning_rate": 6.293333333333334e-06,
+      "loss": 1.1404,
+      "step": 473
+    },
+    {
+      "epoch": 0.029129793510324485,
+      "grad_norm": 14.3125,
+      "learning_rate": 6.3066666666666676e-06,
+      "loss": 1.0994,
+      "step": 474
+    },
+    {
+      "epoch": 0.02919124877089479,
+      "grad_norm": 23.5,
+      "learning_rate": 6.3200000000000005e-06,
+      "loss": 1.0879,
+      "step": 475
+    },
+    {
+      "epoch": 0.029252704031465092,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.333333333333333e-06,
+      "loss": 0.8972,
+      "step": 476
+    },
+    {
+      "epoch": 0.0293141592920354,
+      "grad_norm": 11.0,
+      "learning_rate": 6.346666666666668e-06,
+      "loss": 1.0301,
+      "step": 477
+    },
+    {
+      "epoch": 0.029375614552605702,
+      "grad_norm": 28.0,
+      "learning_rate": 6.360000000000001e-06,
+      "loss": 1.0921,
+      "step": 478
+    },
+    {
+      "epoch": 0.02943706981317601,
+      "grad_norm": 16.125,
+      "learning_rate": 6.373333333333334e-06,
+      "loss": 0.9853,
+      "step": 479
+    },
+    {
+      "epoch": 0.029498525073746312,
+      "grad_norm": 11.1875,
+      "learning_rate": 6.386666666666668e-06,
+      "loss": 0.9663,
+      "step": 480
+    },
+    {
+      "epoch": 0.02955998033431662,
+      "grad_norm": 11.75,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 1.0874,
+      "step": 481
+    },
+    {
+      "epoch": 0.029621435594886922,
+      "grad_norm": 12.75,
+      "learning_rate": 6.4133333333333335e-06,
+      "loss": 0.9843,
+      "step": 482
+    },
+    {
+      "epoch": 0.029682890855457226,
+      "grad_norm": 29.625,
+      "learning_rate": 6.426666666666668e-06,
+      "loss": 1.0121,
+      "step": 483
+    },
+    {
+      "epoch": 0.029744346116027533,
+      "grad_norm": 31.625,
+      "learning_rate": 6.440000000000001e-06,
+      "loss": 1.0971,
+      "step": 484
+    },
+    {
+      "epoch": 0.029805801376597836,
+      "grad_norm": 18.0,
+      "learning_rate": 6.453333333333334e-06,
+      "loss": 1.0744,
+      "step": 485
+    },
+    {
+      "epoch": 0.029867256637168143,
+      "grad_norm": 23.125,
+      "learning_rate": 6.466666666666667e-06,
+      "loss": 1.0479,
+      "step": 486
+    },
+    {
+      "epoch": 0.029928711897738446,
+      "grad_norm": 12.6875,
+      "learning_rate": 6.480000000000001e-06,
+      "loss": 0.9901,
+      "step": 487
+    },
+    {
+      "epoch": 0.02999016715830875,
+      "grad_norm": 46.25,
+      "learning_rate": 6.4933333333333336e-06,
+      "loss": 1.038,
+      "step": 488
+    },
+    {
+      "epoch": 0.030051622418879056,
+      "grad_norm": 33.75,
+      "learning_rate": 6.5066666666666665e-06,
+      "loss": 1.0754,
+      "step": 489
+    },
+    {
+      "epoch": 0.03011307767944936,
+      "grad_norm": 14.3125,
+      "learning_rate": 6.520000000000001e-06,
+      "loss": 1.0229,
+      "step": 490
+    },
+    {
+      "epoch": 0.030174532940019667,
+      "grad_norm": 23.75,
+      "learning_rate": 6.533333333333334e-06,
+      "loss": 1.0013,
+      "step": 491
+    },
+    {
+      "epoch": 0.03023598820058997,
+      "grad_norm": 31.75,
+      "learning_rate": 6.546666666666667e-06,
+      "loss": 1.1724,
+      "step": 492
+    },
+    {
+      "epoch": 0.030297443461160277,
+      "grad_norm": 19.5,
+      "learning_rate": 6.560000000000001e-06,
+      "loss": 0.9667,
+      "step": 493
+    },
+    {
+      "epoch": 0.03035889872173058,
+      "grad_norm": 14.125,
+      "learning_rate": 6.573333333333334e-06,
+      "loss": 1.1136,
+      "step": 494
+    },
+    {
+      "epoch": 0.030420353982300884,
+      "grad_norm": 41.25,
+      "learning_rate": 6.5866666666666666e-06,
+      "loss": 1.0007,
+      "step": 495
+    },
+    {
+      "epoch": 0.03048180924287119,
+      "grad_norm": 15.5,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 1.0133,
+      "step": 496
+    },
+    {
+      "epoch": 0.030543264503441494,
+      "grad_norm": 21.0,
+      "learning_rate": 6.613333333333334e-06,
+      "loss": 0.8813,
+      "step": 497
+    },
+    {
+      "epoch": 0.0306047197640118,
+      "grad_norm": 16.375,
+      "learning_rate": 6.626666666666667e-06,
+      "loss": 0.9524,
+      "step": 498
+    },
+    {
+      "epoch": 0.030666175024582104,
+      "grad_norm": 17.75,
+      "learning_rate": 6.640000000000001e-06,
+      "loss": 0.9767,
+      "step": 499
+    },
+    {
+      "epoch": 0.030727630285152407,
+      "grad_norm": 18.125,
+      "learning_rate": 6.653333333333334e-06,
+      "loss": 1.0693,
+      "step": 500
+    },
+    {
+      "epoch": 0.030789085545722714,
+      "grad_norm": 20.125,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.8604,
+      "step": 501
+    },
+    {
+      "epoch": 0.030850540806293018,
+      "grad_norm": 12.0,
+      "learning_rate": 6.680000000000001e-06,
+      "loss": 0.8634,
+      "step": 502
+    },
+    {
+      "epoch": 0.030911996066863325,
+      "grad_norm": 13.5,
+      "learning_rate": 6.693333333333334e-06,
+      "loss": 1.0665,
+      "step": 503
+    },
+    {
+      "epoch": 0.030973451327433628,
+      "grad_norm": 24.375,
+      "learning_rate": 6.706666666666667e-06,
+      "loss": 0.921,
+      "step": 504
+    },
+    {
+      "epoch": 0.031034906588003935,
+      "grad_norm": 28.5,
+      "learning_rate": 6.720000000000001e-06,
+      "loss": 0.935,
+      "step": 505
+    },
+    {
+      "epoch": 0.031096361848574238,
+      "grad_norm": 14.75,
+      "learning_rate": 6.733333333333334e-06,
+      "loss": 0.9524,
+      "step": 506
+    },
+    {
+      "epoch": 0.03115781710914454,
+      "grad_norm": 18.75,
+      "learning_rate": 6.746666666666667e-06,
+      "loss": 0.9773,
+      "step": 507
+    },
+    {
+      "epoch": 0.03121927236971485,
+      "grad_norm": 23.375,
+      "learning_rate": 6.760000000000001e-06,
+      "loss": 0.9076,
+      "step": 508
+    },
+    {
+      "epoch": 0.03128072763028515,
+      "grad_norm": 30.75,
+      "learning_rate": 6.773333333333334e-06,
+      "loss": 0.9671,
+      "step": 509
+    },
+    {
+      "epoch": 0.031342182890855455,
+      "grad_norm": 29.625,
+      "learning_rate": 6.786666666666667e-06,
+      "loss": 0.965,
+      "step": 510
+    },
+    {
+      "epoch": 0.031403638151425765,
+      "grad_norm": 22.5,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.9741,
+      "step": 511
+    },
+    {
+      "epoch": 0.03146509341199607,
+      "grad_norm": 28.0,
+      "learning_rate": 6.813333333333334e-06,
+      "loss": 0.7974,
+      "step": 512
+    },
+    {
+      "epoch": 0.03152654867256637,
+      "grad_norm": 35.25,
+      "learning_rate": 6.826666666666667e-06,
+      "loss": 0.844,
+      "step": 513
+    },
+    {
+      "epoch": 0.031588003933136675,
+      "grad_norm": 25.75,
+      "learning_rate": 6.8400000000000014e-06,
+      "loss": 0.8082,
+      "step": 514
+    },
+    {
+      "epoch": 0.03164945919370698,
+      "grad_norm": 54.75,
+      "learning_rate": 6.853333333333334e-06,
+      "loss": 0.8529,
+      "step": 515
+    },
+    {
+      "epoch": 0.03171091445427729,
+      "grad_norm": 48.0,
+      "learning_rate": 6.866666666666667e-06,
+      "loss": 0.7857,
+      "step": 516
+    },
+    {
+      "epoch": 0.03177236971484759,
+      "grad_norm": 53.5,
+      "learning_rate": 6.88e-06,
+      "loss": 0.8327,
+      "step": 517
+    },
+    {
+      "epoch": 0.031833824975417896,
+      "grad_norm": 36.75,
+      "learning_rate": 6.893333333333334e-06,
+      "loss": 0.7622,
+      "step": 518
+    },
+    {
+      "epoch": 0.0318952802359882,
+      "grad_norm": 32.0,
+      "learning_rate": 6.906666666666667e-06,
+      "loss": 0.8164,
+      "step": 519
+    },
+    {
+      "epoch": 0.0319567354965585,
+      "grad_norm": 20.875,
+      "learning_rate": 6.92e-06,
+      "loss": 0.799,
+      "step": 520
+    },
+    {
+      "epoch": 0.03201819075712881,
+      "grad_norm": 36.5,
+      "learning_rate": 6.9333333333333344e-06,
+      "loss": 0.7903,
+      "step": 521
+    },
+    {
+      "epoch": 0.032079646017699116,
+      "grad_norm": 35.25,
+      "learning_rate": 6.946666666666667e-06,
+      "loss": 1.0056,
+      "step": 522
+    },
+    {
+      "epoch": 0.03214110127826942,
+      "grad_norm": 27.5,
+      "learning_rate": 6.96e-06,
+      "loss": 0.8031,
+      "step": 523
+    },
+    {
+      "epoch": 0.03220255653883972,
+      "grad_norm": 71.0,
+      "learning_rate": 6.973333333333334e-06,
+      "loss": 1.1609,
+      "step": 524
+    },
+    {
+      "epoch": 0.032264011799410026,
+      "grad_norm": 23.625,
+      "learning_rate": 6.986666666666667e-06,
+      "loss": 0.8413,
+      "step": 525
+    },
+    {
+      "epoch": 0.03232546705998034,
+      "grad_norm": 30.125,
+      "learning_rate": 7e-06,
+      "loss": 0.7228,
+      "step": 526
+    },
+    {
+      "epoch": 0.03238692232055064,
+      "grad_norm": 33.0,
+      "learning_rate": 7.0133333333333345e-06,
+      "loss": 0.8425,
+      "step": 527
+    },
+    {
+      "epoch": 0.032448377581120944,
+      "grad_norm": 46.75,
+      "learning_rate": 7.0266666666666674e-06,
+      "loss": 0.8788,
+      "step": 528
+    },
+    {
+      "epoch": 0.03250983284169125,
+      "grad_norm": 61.5,
+      "learning_rate": 7.04e-06,
+      "loss": 0.9771,
+      "step": 529
+    },
+    {
+      "epoch": 0.03257128810226155,
+      "grad_norm": 70.5,
+      "learning_rate": 7.053333333333334e-06,
+      "loss": 0.8288,
+      "step": 530
+    },
+    {
+      "epoch": 0.03263274336283186,
+      "grad_norm": 35.0,
+      "learning_rate": 7.066666666666667e-06,
+      "loss": 0.9278,
+      "step": 531
+    },
+    {
+      "epoch": 0.032694198623402164,
+      "grad_norm": 88.0,
+      "learning_rate": 7.08e-06,
+      "loss": 0.8835,
+      "step": 532
+    },
+    {
+      "epoch": 0.03275565388397247,
+      "grad_norm": 56.5,
+      "learning_rate": 7.093333333333335e-06,
+      "loss": 0.7722,
+      "step": 533
+    },
+    {
+      "epoch": 0.03281710914454277,
+      "grad_norm": 49.25,
+      "learning_rate": 7.1066666666666675e-06,
+      "loss": 0.7818,
+      "step": 534
+    },
+    {
+      "epoch": 0.03287856440511308,
+      "grad_norm": 69.5,
+      "learning_rate": 7.1200000000000004e-06,
+      "loss": 0.7787,
+      "step": 535
+    },
+    {
+      "epoch": 0.032940019665683384,
+      "grad_norm": 49.25,
+      "learning_rate": 7.133333333333334e-06,
+      "loss": 0.8444,
+      "step": 536
+    },
+    {
+      "epoch": 0.03300147492625369,
+      "grad_norm": 46.0,
+      "learning_rate": 7.146666666666667e-06,
+      "loss": 0.8784,
+      "step": 537
+    },
+    {
+      "epoch": 0.03306293018682399,
+      "grad_norm": 73.0,
+      "learning_rate": 7.16e-06,
+      "loss": 0.8475,
+      "step": 538
+    },
+    {
+      "epoch": 0.033124385447394294,
+      "grad_norm": 32.75,
+      "learning_rate": 7.173333333333335e-06,
+      "loss": 0.9105,
+      "step": 539
+    },
+    {
+      "epoch": 0.033185840707964605,
+      "grad_norm": 91.0,
+      "learning_rate": 7.186666666666668e-06,
+      "loss": 0.9562,
+      "step": 540
+    },
+    {
+      "epoch": 0.03324729596853491,
+      "grad_norm": 49.5,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 0.8389,
+      "step": 541
+    },
+    {
+      "epoch": 0.03330875122910521,
+      "grad_norm": 33.5,
+      "learning_rate": 7.213333333333334e-06,
+      "loss": 0.838,
+      "step": 542
+    },
+    {
+      "epoch": 0.033370206489675515,
+      "grad_norm": 43.25,
+      "learning_rate": 7.226666666666667e-06,
+      "loss": 0.9548,
+      "step": 543
+    },
+    {
+      "epoch": 0.03343166175024582,
+      "grad_norm": 32.0,
+      "learning_rate": 7.24e-06,
+      "loss": 0.8057,
+      "step": 544
+    },
+    {
+      "epoch": 0.03349311701081613,
+      "grad_norm": 33.0,
+      "learning_rate": 7.253333333333335e-06,
+      "loss": 0.8826,
+      "step": 545
+    },
+    {
+      "epoch": 0.03355457227138643,
+      "grad_norm": 40.0,
+      "learning_rate": 7.266666666666668e-06,
+      "loss": 0.802,
+      "step": 546
+    },
+    {
+      "epoch": 0.033616027531956735,
+      "grad_norm": 29.5,
+      "learning_rate": 7.280000000000001e-06,
+      "loss": 0.7812,
+      "step": 547
+    },
+    {
+      "epoch": 0.03367748279252704,
+      "grad_norm": 56.75,
+      "learning_rate": 7.2933333333333335e-06,
+      "loss": 0.8719,
+      "step": 548
+    },
+    {
+      "epoch": 0.03373893805309734,
+      "grad_norm": 33.75,
+      "learning_rate": 7.306666666666667e-06,
+      "loss": 0.9626,
+      "step": 549
+    },
+    {
+      "epoch": 0.03380039331366765,
+      "grad_norm": 51.5,
+      "learning_rate": 7.32e-06,
+      "loss": 1.0222,
+      "step": 550
+    },
+    {
+      "epoch": 0.033861848574237956,
+      "grad_norm": 38.75,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 0.7883,
+      "step": 551
+    },
+    {
+      "epoch": 0.03392330383480826,
+      "grad_norm": 77.5,
+      "learning_rate": 7.346666666666668e-06,
+      "loss": 0.8014,
+      "step": 552
+    },
+    {
+      "epoch": 0.03398475909537856,
+      "grad_norm": 70.5,
+      "learning_rate": 7.360000000000001e-06,
+      "loss": 0.862,
+      "step": 553
+    },
+    {
+      "epoch": 0.034046214355948866,
+      "grad_norm": 40.25,
+      "learning_rate": 7.373333333333334e-06,
+      "loss": 0.7617,
+      "step": 554
+    },
+    {
+      "epoch": 0.034107669616519176,
+      "grad_norm": 46.25,
+      "learning_rate": 7.386666666666667e-06,
+      "loss": 0.7726,
+      "step": 555
+    },
+    {
+      "epoch": 0.03416912487708948,
+      "grad_norm": 49.25,
+      "learning_rate": 7.4e-06,
+      "loss": 0.9869,
+      "step": 556
+    },
+    {
+      "epoch": 0.03423058013765978,
+      "grad_norm": 34.5,
+      "learning_rate": 7.413333333333333e-06,
+      "loss": 0.7539,
+      "step": 557
+    },
+    {
+      "epoch": 0.034292035398230086,
+      "grad_norm": 30.5,
+      "learning_rate": 7.426666666666668e-06,
+      "loss": 0.6686,
+      "step": 558
+    },
+    {
+      "epoch": 0.0343534906588004,
+      "grad_norm": 49.75,
+      "learning_rate": 7.440000000000001e-06,
+      "loss": 0.8961,
+      "step": 559
+    },
+    {
+      "epoch": 0.0344149459193707,
+      "grad_norm": 95.5,
+      "learning_rate": 7.453333333333334e-06,
+      "loss": 0.8031,
+      "step": 560
+    },
+    {
+      "epoch": 0.034476401179941,
+      "grad_norm": 54.25,
+      "learning_rate": 7.4666666666666675e-06,
+      "loss": 0.915,
+      "step": 561
+    },
+    {
+      "epoch": 0.03453785644051131,
+      "grad_norm": 102.0,
+      "learning_rate": 7.48e-06,
+      "loss": 1.0234,
+      "step": 562
+    },
+    {
+      "epoch": 0.03459931170108161,
+      "grad_norm": 70.5,
+      "learning_rate": 7.493333333333333e-06,
+      "loss": 0.9048,
+      "step": 563
+    },
+    {
+      "epoch": 0.03466076696165192,
+      "grad_norm": 101.0,
+      "learning_rate": 7.506666666666668e-06,
+      "loss": 0.989,
+      "step": 564
+    },
+    {
+      "epoch": 0.034722222222222224,
+      "grad_norm": 73.0,
+      "learning_rate": 7.520000000000001e-06,
+      "loss": 1.1432,
+      "step": 565
+    },
+    {
+      "epoch": 0.03478367748279253,
+      "grad_norm": 53.5,
+      "learning_rate": 7.533333333333334e-06,
+      "loss": 1.048,
+      "step": 566
+    },
+    {
+      "epoch": 0.03484513274336283,
+      "grad_norm": 34.75,
+      "learning_rate": 7.5466666666666675e-06,
+      "loss": 0.9056,
+      "step": 567
+    },
+    {
+      "epoch": 0.034906588003933134,
+      "grad_norm": 60.0,
+      "learning_rate": 7.5600000000000005e-06,
+      "loss": 0.9027,
+      "step": 568
+    },
+    {
+      "epoch": 0.034968043264503444,
+      "grad_norm": 32.25,
+      "learning_rate": 7.573333333333333e-06,
+      "loss": 0.7684,
+      "step": 569
+    },
+    {
+      "epoch": 0.03502949852507375,
+      "grad_norm": 57.0,
+      "learning_rate": 7.586666666666668e-06,
+      "loss": 0.7446,
+      "step": 570
+    },
+    {
+      "epoch": 0.03509095378564405,
+      "grad_norm": 52.75,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 0.8055,
+      "step": 571
+    },
+    {
+      "epoch": 0.035152409046214354,
+      "grad_norm": 36.75,
+      "learning_rate": 7.613333333333334e-06,
+      "loss": 0.8156,
+      "step": 572
+    },
+    {
+      "epoch": 0.03521386430678466,
+      "grad_norm": 27.0,
+      "learning_rate": 7.626666666666668e-06,
+      "loss": 0.7767,
+      "step": 573
+    },
+    {
+      "epoch": 0.03527531956735497,
+      "grad_norm": 32.5,
+      "learning_rate": 7.640000000000001e-06,
+      "loss": 0.812,
+      "step": 574
+    },
+    {
+      "epoch": 0.03533677482792527,
+      "grad_norm": 28.0,
+      "learning_rate": 7.653333333333333e-06,
+      "loss": 0.8056,
+      "step": 575
+    },
+    {
+      "epoch": 0.035398230088495575,
+      "grad_norm": 22.75,
+      "learning_rate": 7.666666666666667e-06,
+      "loss": 0.6915,
+      "step": 576
+    },
+    {
+      "epoch": 0.03545968534906588,
+      "grad_norm": 34.5,
+      "learning_rate": 7.680000000000001e-06,
+      "loss": 0.7822,
+      "step": 577
+    },
+    {
+      "epoch": 0.03552114060963618,
+      "grad_norm": 58.25,
+      "learning_rate": 7.693333333333333e-06,
+      "loss": 0.8356,
+      "step": 578
+    },
+    {
+      "epoch": 0.03558259587020649,
+      "grad_norm": 40.25,
+      "learning_rate": 7.706666666666669e-06,
+      "loss": 0.8046,
+      "step": 579
+    },
+    {
+      "epoch": 0.035644051130776795,
+      "grad_norm": 42.5,
+      "learning_rate": 7.72e-06,
+      "loss": 0.642,
+      "step": 580
+    },
+    {
+      "epoch": 0.0357055063913471,
+      "grad_norm": 36.75,
+      "learning_rate": 7.733333333333334e-06,
+      "loss": 0.7037,
+      "step": 581
+    },
+    {
+      "epoch": 0.0357669616519174,
+      "grad_norm": 53.25,
+      "learning_rate": 7.746666666666666e-06,
+      "loss": 0.9178,
+      "step": 582
+    },
+    {
+      "epoch": 0.03582841691248771,
+      "grad_norm": 61.5,
+      "learning_rate": 7.76e-06,
+      "loss": 0.8404,
+      "step": 583
+    },
+    {
+      "epoch": 0.035889872173058016,
+      "grad_norm": 34.5,
+      "learning_rate": 7.773333333333334e-06,
+      "loss": 0.7966,
+      "step": 584
+    },
+    {
+      "epoch": 0.03595132743362832,
+      "grad_norm": 40.25,
+      "learning_rate": 7.786666666666666e-06,
+      "loss": 0.8787,
+      "step": 585
+    },
+    {
+      "epoch": 0.03601278269419862,
+      "grad_norm": 53.0,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 1.002,
+      "step": 586
+    },
+    {
+      "epoch": 0.036074237954768926,
+      "grad_norm": 39.75,
+      "learning_rate": 7.813333333333334e-06,
+      "loss": 1.1473,
+      "step": 587
+    },
+    {
+      "epoch": 0.036135693215339236,
+      "grad_norm": 38.75,
+      "learning_rate": 7.826666666666667e-06,
+      "loss": 1.088,
+      "step": 588
+    },
+    {
+      "epoch": 0.03619714847590954,
+      "grad_norm": 26.375,
+      "learning_rate": 7.840000000000001e-06,
+      "loss": 0.9988,
+      "step": 589
+    },
+    {
+      "epoch": 0.03625860373647984,
+      "grad_norm": 45.25,
+      "learning_rate": 7.853333333333333e-06,
+      "loss": 0.9803,
+      "step": 590
+    },
+    {
+      "epoch": 0.036320058997050146,
+      "grad_norm": 36.75,
+      "learning_rate": 7.866666666666667e-06,
+      "loss": 1.039,
+      "step": 591
+    },
+    {
+      "epoch": 0.03638151425762045,
+      "grad_norm": 40.5,
+      "learning_rate": 7.88e-06,
+      "loss": 0.8799,
+      "step": 592
+    },
+    {
+      "epoch": 0.03644296951819076,
+      "grad_norm": 39.0,
+      "learning_rate": 7.893333333333335e-06,
+      "loss": 0.8865,
+      "step": 593
+    },
+    {
+      "epoch": 0.03650442477876106,
+      "grad_norm": 30.125,
+      "learning_rate": 7.906666666666667e-06,
+      "loss": 0.8302,
+      "step": 594
+    },
+    {
+      "epoch": 0.03656588003933137,
+      "grad_norm": 29.375,
+      "learning_rate": 7.92e-06,
+      "loss": 0.9001,
+      "step": 595
+    },
+    {
+      "epoch": 0.03662733529990167,
+      "grad_norm": 59.25,
+      "learning_rate": 7.933333333333334e-06,
+      "loss": 0.915,
+      "step": 596
+    },
+    {
+      "epoch": 0.03668879056047197,
+      "grad_norm": 36.75,
+      "learning_rate": 7.946666666666666e-06,
+      "loss": 0.7387,
+      "step": 597
+    },
+    {
+      "epoch": 0.036750245821042284,
+      "grad_norm": 48.75,
+      "learning_rate": 7.960000000000002e-06,
+      "loss": 0.8842,
+      "step": 598
+    },
+    {
+      "epoch": 0.03681170108161259,
+      "grad_norm": 16.875,
+      "learning_rate": 7.973333333333334e-06,
+      "loss": 0.6863,
+      "step": 599
+    },
+    {
+      "epoch": 0.03687315634218289,
+      "grad_norm": 35.75,
+      "learning_rate": 7.986666666666668e-06,
+      "loss": 0.796,
+      "step": 600
+    },
+    {
+      "epoch": 0.036934611602753194,
+      "grad_norm": 36.25,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.7564,
+      "step": 601
+    },
+    {
+      "epoch": 0.0369960668633235,
+      "grad_norm": 52.25,
+      "learning_rate": 8.013333333333333e-06,
+      "loss": 0.8301,
+      "step": 602
+    },
+    {
+      "epoch": 0.03705752212389381,
+      "grad_norm": 27.75,
+      "learning_rate": 8.026666666666667e-06,
+      "loss": 0.6874,
+      "step": 603
+    },
+    {
+      "epoch": 0.03711897738446411,
+      "grad_norm": 50.75,
+      "learning_rate": 8.040000000000001e-06,
+      "loss": 0.6375,
+      "step": 604
+    },
+    {
+      "epoch": 0.037180432645034414,
+      "grad_norm": 54.25,
+      "learning_rate": 8.053333333333335e-06,
+      "loss": 0.6989,
+      "step": 605
+    },
+    {
+      "epoch": 0.03724188790560472,
+      "grad_norm": 84.0,
+      "learning_rate": 8.066666666666667e-06,
+      "loss": 0.9357,
+      "step": 606
+    },
+    {
+      "epoch": 0.03730334316617503,
+      "grad_norm": 33.0,
+      "learning_rate": 8.08e-06,
+      "loss": 0.9372,
+      "step": 607
+    },
+    {
+      "epoch": 0.03736479842674533,
+      "grad_norm": 32.25,
+      "learning_rate": 8.093333333333334e-06,
+      "loss": 0.7379,
+      "step": 608
+    },
+    {
+      "epoch": 0.037426253687315635,
+      "grad_norm": 34.25,
+      "learning_rate": 8.106666666666666e-06,
+      "loss": 0.776,
+      "step": 609
+    },
+    {
+      "epoch": 0.03748770894788594,
+      "grad_norm": 25.125,
+      "learning_rate": 8.120000000000002e-06,
+      "loss": 0.7643,
+      "step": 610
+    },
+    {
+      "epoch": 0.03754916420845624,
+      "grad_norm": 55.75,
+      "learning_rate": 8.133333333333334e-06,
+      "loss": 0.8392,
+      "step": 611
+    },
+    {
+      "epoch": 0.03761061946902655,
+      "grad_norm": 51.25,
+      "learning_rate": 8.146666666666668e-06,
+      "loss": 0.726,
+      "step": 612
+    },
+    {
+      "epoch": 0.037672074729596855,
+      "grad_norm": 78.0,
+      "learning_rate": 8.16e-06,
+      "loss": 0.99,
+      "step": 613
+    },
+    {
+      "epoch": 0.03773352999016716,
+      "grad_norm": 53.5,
+      "learning_rate": 8.173333333333334e-06,
+      "loss": 0.8275,
+      "step": 614
+    },
+    {
+      "epoch": 0.03779498525073746,
+      "grad_norm": 48.0,
+      "learning_rate": 8.186666666666667e-06,
+      "loss": 0.827,
+      "step": 615
+    },
+    {
+      "epoch": 0.037856440511307765,
+      "grad_norm": 44.5,
+      "learning_rate": 8.2e-06,
+      "loss": 0.751,
+      "step": 616
+    },
+    {
+      "epoch": 0.037917895771878075,
+      "grad_norm": 58.25,
+      "learning_rate": 8.213333333333335e-06,
+      "loss": 0.8964,
+      "step": 617
+    },
+    {
+      "epoch": 0.03797935103244838,
+      "grad_norm": 37.5,
+      "learning_rate": 8.226666666666667e-06,
+      "loss": 1.0725,
+      "step": 618
+    },
+    {
+      "epoch": 0.03804080629301868,
+      "grad_norm": 73.0,
+      "learning_rate": 8.24e-06,
+      "loss": 0.8148,
+      "step": 619
+    },
+    {
+      "epoch": 0.038102261553588986,
+      "grad_norm": 45.25,
+      "learning_rate": 8.253333333333334e-06,
+      "loss": 0.7827,
+      "step": 620
+    },
+    {
+      "epoch": 0.03816371681415929,
+      "grad_norm": 69.0,
+      "learning_rate": 8.266666666666667e-06,
+      "loss": 0.9836,
+      "step": 621
+    },
+    {
+      "epoch": 0.0382251720747296,
+      "grad_norm": 56.0,
+      "learning_rate": 8.28e-06,
+      "loss": 0.9259,
+      "step": 622
+    },
+    {
+      "epoch": 0.0382866273352999,
+      "grad_norm": 42.5,
+      "learning_rate": 8.293333333333334e-06,
+      "loss": 0.8056,
+      "step": 623
+    },
+    {
+      "epoch": 0.038348082595870206,
+      "grad_norm": 40.25,
+      "learning_rate": 8.306666666666668e-06,
+      "loss": 0.8676,
+      "step": 624
+    },
+    {
+      "epoch": 0.03840953785644051,
+      "grad_norm": 61.5,
+      "learning_rate": 8.32e-06,
+      "loss": 1.0395,
+      "step": 625
+    },
+    {
+      "epoch": 0.03847099311701081,
+      "grad_norm": 42.75,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.8258,
+      "step": 626
+    },
+    {
+      "epoch": 0.03853244837758112,
+      "grad_norm": 65.5,
+      "learning_rate": 8.346666666666668e-06,
+      "loss": 0.7707,
+      "step": 627
+    },
+    {
+      "epoch": 0.038593903638151426,
+      "grad_norm": 44.75,
+      "learning_rate": 8.36e-06,
+      "loss": 0.8274,
+      "step": 628
+    },
+    {
+      "epoch": 0.03865535889872173,
+      "grad_norm": 66.5,
+      "learning_rate": 8.373333333333335e-06,
+      "loss": 0.8568,
+      "step": 629
+    },
+    {
+      "epoch": 0.03871681415929203,
+      "grad_norm": 68.5,
+      "learning_rate": 8.386666666666667e-06,
+      "loss": 0.7702,
+      "step": 630
+    },
+    {
+      "epoch": 0.03877826941986234,
+      "grad_norm": 46.5,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.7266,
+      "step": 631
+    },
+    {
+      "epoch": 0.03883972468043265,
+      "grad_norm": 24.75,
+      "learning_rate": 8.413333333333335e-06,
+      "loss": 0.6446,
+      "step": 632
+    },
+    {
+      "epoch": 0.03890117994100295,
+      "grad_norm": 74.0,
+      "learning_rate": 8.426666666666667e-06,
+      "loss": 0.9642,
+      "step": 633
+    },
+    {
+      "epoch": 0.038962635201573254,
+      "grad_norm": 66.0,
+      "learning_rate": 8.44e-06,
+      "loss": 1.119,
+      "step": 634
+    },
+    {
+      "epoch": 0.03902409046214356,
+      "grad_norm": 57.25,
+      "learning_rate": 8.453333333333334e-06,
+      "loss": 0.9793,
+      "step": 635
+    },
+    {
+      "epoch": 0.03908554572271387,
+      "grad_norm": 51.75,
+      "learning_rate": 8.466666666666668e-06,
+      "loss": 1.1962,
+      "step": 636
+    },
+    {
+      "epoch": 0.03914700098328417,
+      "grad_norm": 44.5,
+      "learning_rate": 8.48e-06,
+      "loss": 1.0263,
+      "step": 637
+    },
+    {
+      "epoch": 0.039208456243854474,
+      "grad_norm": 46.0,
+      "learning_rate": 8.493333333333334e-06,
+      "loss": 0.9972,
+      "step": 638
+    },
+    {
+      "epoch": 0.03926991150442478,
+      "grad_norm": 54.5,
+      "learning_rate": 8.506666666666668e-06,
+      "loss": 0.9303,
+      "step": 639
+    },
+    {
+      "epoch": 0.03933136676499508,
+      "grad_norm": 24.625,
+      "learning_rate": 8.52e-06,
+      "loss": 0.8192,
+      "step": 640
+    },
+    {
+      "epoch": 0.03939282202556539,
+      "grad_norm": 57.25,
+      "learning_rate": 8.533333333333335e-06,
+      "loss": 0.9214,
+      "step": 641
+    },
+    {
+      "epoch": 0.039454277286135694,
+      "grad_norm": 37.25,
+      "learning_rate": 8.546666666666667e-06,
+      "loss": 0.9645,
+      "step": 642
+    },
+    {
+      "epoch": 0.039515732546706,
+      "grad_norm": 26.125,
+      "learning_rate": 8.560000000000001e-06,
+      "loss": 0.971,
+      "step": 643
+    },
+    {
+      "epoch": 0.0395771878072763,
+      "grad_norm": 44.5,
+      "learning_rate": 8.573333333333333e-06,
+      "loss": 0.8452,
+      "step": 644
+    },
+    {
+      "epoch": 0.039638643067846605,
+      "grad_norm": 52.25,
+      "learning_rate": 8.586666666666667e-06,
+      "loss": 0.967,
+      "step": 645
+    },
+    {
+      "epoch": 0.039700098328416915,
+      "grad_norm": 79.0,
+      "learning_rate": 8.6e-06,
+      "loss": 1.0605,
+      "step": 646
+    },
+    {
+      "epoch": 0.03976155358898722,
+      "grad_norm": 19.125,
+      "learning_rate": 8.613333333333333e-06,
+      "loss": 0.8204,
+      "step": 647
+    },
+    {
+      "epoch": 0.03982300884955752,
+      "grad_norm": 28.25,
+      "learning_rate": 8.626666666666668e-06,
+      "loss": 0.8179,
+      "step": 648
+    },
+    {
+      "epoch": 0.039884464110127825,
+      "grad_norm": 30.375,
+      "learning_rate": 8.64e-06,
+      "loss": 0.8463,
+      "step": 649
+    },
+    {
+      "epoch": 0.03994591937069813,
+      "grad_norm": 28.375,
+      "learning_rate": 8.653333333333334e-06,
+      "loss": 0.7462,
+      "step": 650
+    },
+    {
+      "epoch": 0.04000737463126844,
+      "grad_norm": 64.5,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.7727,
+      "step": 651
+    },
+    {
+      "epoch": 0.04006882989183874,
+      "grad_norm": 80.5,
+      "learning_rate": 8.68e-06,
+      "loss": 0.9155,
+      "step": 652
+    },
+    {
+      "epoch": 0.040130285152409045,
+      "grad_norm": 47.25,
+      "learning_rate": 8.693333333333334e-06,
+      "loss": 0.728,
+      "step": 653
+    },
+    {
+      "epoch": 0.04019174041297935,
+      "grad_norm": 46.0,
+      "learning_rate": 8.706666666666667e-06,
+      "loss": 0.85,
+      "step": 654
+    },
+    {
+      "epoch": 0.04025319567354966,
+      "grad_norm": 30.0,
+      "learning_rate": 8.720000000000001e-06,
+      "loss": 0.8259,
+      "step": 655
+    },
+    {
+      "epoch": 0.04031465093411996,
+      "grad_norm": 51.25,
+      "learning_rate": 8.733333333333333e-06,
+      "loss": 0.729,
+      "step": 656
+    },
+    {
+      "epoch": 0.040376106194690266,
+      "grad_norm": 35.0,
+      "learning_rate": 8.746666666666667e-06,
+      "loss": 0.7594,
+      "step": 657
+    },
+    {
+      "epoch": 0.04043756145526057,
+      "grad_norm": 43.5,
+      "learning_rate": 8.76e-06,
+      "loss": 0.7039,
+      "step": 658
+    },
+    {
+      "epoch": 0.04049901671583087,
+      "grad_norm": 38.5,
+      "learning_rate": 8.773333333333333e-06,
+      "loss": 0.6428,
+      "step": 659
+    },
+    {
+      "epoch": 0.04056047197640118,
+      "grad_norm": 33.0,
+      "learning_rate": 8.786666666666668e-06,
+      "loss": 0.8518,
+      "step": 660
+    },
+    {
+      "epoch": 0.040621927236971486,
+      "grad_norm": 39.0,
+      "learning_rate": 8.8e-06,
+      "loss": 0.6758,
+      "step": 661
+    },
+    {
+      "epoch": 0.04068338249754179,
+      "grad_norm": 34.5,
+      "learning_rate": 8.813333333333334e-06,
+      "loss": 0.6863,
+      "step": 662
+    },
+    {
+      "epoch": 0.04074483775811209,
+      "grad_norm": 48.75,
+      "learning_rate": 8.826666666666668e-06,
+      "loss": 0.7819,
+      "step": 663
+    },
+    {
+      "epoch": 0.040806293018682396,
+      "grad_norm": 33.5,
+      "learning_rate": 8.84e-06,
+      "loss": 0.7992,
+      "step": 664
+    },
+    {
+      "epoch": 0.04086774827925271,
+      "grad_norm": 39.0,
+      "learning_rate": 8.853333333333334e-06,
+      "loss": 0.8031,
+      "step": 665
+    },
+    {
+      "epoch": 0.04092920353982301,
+      "grad_norm": 36.75,
+      "learning_rate": 8.866666666666668e-06,
+      "loss": 0.7268,
+      "step": 666
+    },
+    {
+      "epoch": 0.04099065880039331,
+      "grad_norm": 22.25,
+      "learning_rate": 8.880000000000001e-06,
+      "loss": 0.8118,
+      "step": 667
+    },
+    {
+      "epoch": 0.04105211406096362,
+      "grad_norm": 30.875,
+      "learning_rate": 8.893333333333333e-06,
+      "loss": 0.6847,
+      "step": 668
+    },
+    {
+      "epoch": 0.04111356932153392,
+      "grad_norm": 47.5,
+      "learning_rate": 8.906666666666667e-06,
+      "loss": 0.7691,
+      "step": 669
+    },
+    {
+      "epoch": 0.04117502458210423,
+      "grad_norm": 46.25,
+      "learning_rate": 8.920000000000001e-06,
+      "loss": 0.743,
+      "step": 670
+    },
+    {
+      "epoch": 0.041236479842674534,
+      "grad_norm": 34.25,
+      "learning_rate": 8.933333333333333e-06,
+      "loss": 0.6177,
+      "step": 671
+    },
+    {
+      "epoch": 0.04129793510324484,
+      "grad_norm": 37.75,
+      "learning_rate": 8.946666666666669e-06,
+      "loss": 0.7844,
+      "step": 672
+    },
+    {
+      "epoch": 0.04135939036381514,
+      "grad_norm": 49.0,
+      "learning_rate": 8.96e-06,
+      "loss": 0.7841,
+      "step": 673
+    },
+    {
+      "epoch": 0.041420845624385444,
+      "grad_norm": 50.5,
+      "learning_rate": 8.973333333333334e-06,
+      "loss": 0.7439,
+      "step": 674
+    },
+    {
+      "epoch": 0.041482300884955754,
+      "grad_norm": 37.75,
+      "learning_rate": 8.986666666666666e-06,
+      "loss": 0.7848,
+      "step": 675
+    },
+    {
+      "epoch": 0.04154375614552606,
+      "grad_norm": 59.0,
+      "learning_rate": 9e-06,
+      "loss": 0.7377,
+      "step": 676
+    },
+    {
+      "epoch": 0.04160521140609636,
+      "grad_norm": 27.125,
+      "learning_rate": 9.013333333333334e-06,
+      "loss": 0.7076,
+      "step": 677
+    },
+    {
+      "epoch": 0.041666666666666664,
+      "grad_norm": 42.0,
+      "learning_rate": 9.026666666666666e-06,
+      "loss": 0.7818,
+      "step": 678
+    },
+    {
+      "epoch": 0.041728121927236975,
+      "grad_norm": 35.0,
+      "learning_rate": 9.040000000000002e-06,
+      "loss": 0.6463,
+      "step": 679
+    },
+    {
+      "epoch": 0.04178957718780728,
+      "grad_norm": 51.25,
+      "learning_rate": 9.053333333333334e-06,
+      "loss": 0.7036,
+      "step": 680
+    },
+    {
+      "epoch": 0.04185103244837758,
+      "grad_norm": 39.25,
+      "learning_rate": 9.066666666666667e-06,
+      "loss": 0.7589,
+      "step": 681
+    },
+    {
+      "epoch": 0.041912487708947885,
+      "grad_norm": 44.0,
+      "learning_rate": 9.080000000000001e-06,
+      "loss": 0.7547,
+      "step": 682
+    },
+    {
+      "epoch": 0.04197394296951819,
+      "grad_norm": 48.0,
+      "learning_rate": 9.093333333333333e-06,
+      "loss": 0.7718,
+      "step": 683
+    },
+    {
+      "epoch": 0.0420353982300885,
+      "grad_norm": 31.625,
+      "learning_rate": 9.106666666666667e-06,
+      "loss": 0.816,
+      "step": 684
+    },
+    {
+      "epoch": 0.0420968534906588,
+      "grad_norm": 26.125,
+      "learning_rate": 9.12e-06,
+      "loss": 0.7412,
+      "step": 685
+    },
+    {
+      "epoch": 0.042158308751229105,
+      "grad_norm": 25.25,
+      "learning_rate": 9.133333333333335e-06,
+      "loss": 0.8087,
+      "step": 686
+    },
+    {
+      "epoch": 0.04221976401179941,
+      "grad_norm": 21.875,
+      "learning_rate": 9.146666666666667e-06,
+      "loss": 0.6321,
+      "step": 687
+    },
+    {
+      "epoch": 0.04228121927236971,
+      "grad_norm": 34.5,
+      "learning_rate": 9.16e-06,
+      "loss": 0.7168,
+      "step": 688
+    },
+    {
+      "epoch": 0.04234267453294002,
+      "grad_norm": 28.0,
+      "learning_rate": 9.173333333333334e-06,
+      "loss": 0.7317,
+      "step": 689
+    },
+    {
+      "epoch": 0.042404129793510326,
+      "grad_norm": 45.0,
+      "learning_rate": 9.186666666666666e-06,
+      "loss": 0.7807,
+      "step": 690
+    },
+    {
+      "epoch": 0.04246558505408063,
+      "grad_norm": 30.125,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 0.735,
+      "step": 691
+    },
+    {
+      "epoch": 0.04252704031465093,
+      "grad_norm": 35.75,
+      "learning_rate": 9.213333333333334e-06,
+      "loss": 0.8574,
+      "step": 692
+    },
+    {
+      "epoch": 0.042588495575221236,
+      "grad_norm": 41.25,
+      "learning_rate": 9.226666666666668e-06,
+      "loss": 0.6181,
+      "step": 693
+    },
+    {
+      "epoch": 0.042649950835791546,
+      "grad_norm": 39.0,
+      "learning_rate": 9.240000000000001e-06,
+      "loss": 0.8465,
+      "step": 694
+    },
+    {
+      "epoch": 0.04271140609636185,
+      "grad_norm": 26.25,
+      "learning_rate": 9.253333333333333e-06,
+      "loss": 0.7085,
+      "step": 695
+    },
+    {
+      "epoch": 0.04277286135693215,
+      "grad_norm": 35.25,
+      "learning_rate": 9.266666666666667e-06,
+      "loss": 0.7554,
+      "step": 696
+    },
+    {
+      "epoch": 0.042834316617502456,
+      "grad_norm": 82.5,
+      "learning_rate": 9.280000000000001e-06,
+      "loss": 1.2424,
+      "step": 697
+    },
+    {
+      "epoch": 0.042895771878072766,
+      "grad_norm": 33.75,
+      "learning_rate": 9.293333333333335e-06,
+      "loss": 0.9356,
+      "step": 698
+    },
+    {
+      "epoch": 0.04295722713864307,
+      "grad_norm": 30.25,
+      "learning_rate": 9.306666666666667e-06,
+      "loss": 0.9749,
+      "step": 699
+    },
+    {
+      "epoch": 0.04301868239921337,
+      "grad_norm": 52.0,
+      "learning_rate": 9.32e-06,
+      "loss": 1.1634,
+      "step": 700
+    },
+    {
+      "epoch": 0.04308013765978368,
+      "grad_norm": 32.25,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 1.1435,
+      "step": 701
+    },
+    {
+      "epoch": 0.04314159292035398,
+      "grad_norm": 44.0,
+      "learning_rate": 9.346666666666666e-06,
+      "loss": 1.0926,
+      "step": 702
+    },
+    {
+      "epoch": 0.04320304818092429,
+      "grad_norm": 32.75,
+      "learning_rate": 9.360000000000002e-06,
+      "loss": 0.8124,
+      "step": 703
+    },
+    {
+      "epoch": 0.043264503441494594,
+      "grad_norm": 25.125,
+      "learning_rate": 9.373333333333334e-06,
+      "loss": 0.978,
+      "step": 704
+    },
+    {
+      "epoch": 0.0433259587020649,
+      "grad_norm": 26.75,
+      "learning_rate": 9.386666666666668e-06,
+      "loss": 0.7538,
+      "step": 705
+    },
+    {
+      "epoch": 0.0433874139626352,
+      "grad_norm": 28.25,
+      "learning_rate": 9.4e-06,
+      "loss": 0.5974,
+      "step": 706
+    },
+    {
+      "epoch": 0.043448869223205504,
+      "grad_norm": 25.75,
+      "learning_rate": 9.413333333333334e-06,
+      "loss": 0.7366,
+      "step": 707
+    },
+    {
+      "epoch": 0.043510324483775814,
+      "grad_norm": 27.625,
+      "learning_rate": 9.426666666666667e-06,
+      "loss": 0.7622,
+      "step": 708
+    },
+    {
+      "epoch": 0.04357177974434612,
+      "grad_norm": 44.25,
+      "learning_rate": 9.440000000000001e-06,
+      "loss": 0.6432,
+      "step": 709
+    },
+    {
+      "epoch": 0.04363323500491642,
+      "grad_norm": 49.0,
+      "learning_rate": 9.453333333333335e-06,
+      "loss": 0.7007,
+      "step": 710
+    },
+    {
+      "epoch": 0.043694690265486724,
+      "grad_norm": 38.25,
+      "learning_rate": 9.466666666666667e-06,
+      "loss": 0.6815,
+      "step": 711
+    },
+    {
+      "epoch": 0.04375614552605703,
+      "grad_norm": 39.25,
+      "learning_rate": 9.48e-06,
+      "loss": 0.6481,
+      "step": 712
+    },
+    {
+      "epoch": 0.04381760078662734,
+      "grad_norm": 39.0,
+      "learning_rate": 9.493333333333334e-06,
+      "loss": 0.8907,
+      "step": 713
+    },
+    {
+      "epoch": 0.04387905604719764,
+      "grad_norm": 27.875,
+      "learning_rate": 9.506666666666667e-06,
+      "loss": 0.7847,
+      "step": 714
+    },
+    {
+      "epoch": 0.043940511307767945,
+      "grad_norm": 33.0,
+      "learning_rate": 9.52e-06,
+      "loss": 0.7244,
+      "step": 715
+    },
+    {
+      "epoch": 0.04400196656833825,
+      "grad_norm": 42.75,
+      "learning_rate": 9.533333333333334e-06,
+      "loss": 0.7727,
+      "step": 716
+    },
+    {
+      "epoch": 0.04406342182890855,
+      "grad_norm": 49.5,
+      "learning_rate": 9.546666666666668e-06,
+      "loss": 0.7163,
+      "step": 717
+    },
+    {
+      "epoch": 0.04412487708947886,
+      "grad_norm": 66.0,
+      "learning_rate": 9.56e-06,
+      "loss": 0.7222,
+      "step": 718
+    },
+    {
+      "epoch": 0.044186332350049165,
+      "grad_norm": 69.5,
+      "learning_rate": 9.573333333333334e-06,
+      "loss": 0.8451,
+      "step": 719
+    },
+    {
+      "epoch": 0.04424778761061947,
+      "grad_norm": 42.5,
+      "learning_rate": 9.586666666666667e-06,
+      "loss": 0.8206,
+      "step": 720
+    },
+    {
+      "epoch": 0.04430924287118977,
+      "grad_norm": 64.0,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.775,
+      "step": 721
+    },
+    {
+      "epoch": 0.04437069813176008,
+      "grad_norm": 30.25,
+      "learning_rate": 9.613333333333335e-06,
+      "loss": 0.6167,
+      "step": 722
+    },
+    {
+      "epoch": 0.044432153392330385,
+      "grad_norm": 49.0,
+      "learning_rate": 9.626666666666667e-06,
+      "loss": 0.5596,
+      "step": 723
+    },
+    {
+      "epoch": 0.04449360865290069,
+      "grad_norm": 39.5,
+      "learning_rate": 9.640000000000001e-06,
+      "loss": 0.7102,
+      "step": 724
+    },
+    {
+      "epoch": 0.04455506391347099,
+      "grad_norm": 41.25,
+      "learning_rate": 9.653333333333335e-06,
+      "loss": 0.79,
+      "step": 725
+    },
+    {
+      "epoch": 0.044616519174041296,
+      "grad_norm": 36.0,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 0.7565,
+      "step": 726
+    },
+    {
+      "epoch": 0.044677974434611606,
+      "grad_norm": 59.75,
+      "learning_rate": 9.68e-06,
+      "loss": 0.7336,
+      "step": 727
+    },
+    {
+      "epoch": 0.04473942969518191,
+      "grad_norm": 30.875,
+      "learning_rate": 9.693333333333334e-06,
+      "loss": 0.5786,
+      "step": 728
+    },
+    {
+      "epoch": 0.04480088495575221,
+      "grad_norm": 37.25,
+      "learning_rate": 9.706666666666668e-06,
+      "loss": 0.8139,
+      "step": 729
+    },
+    {
+      "epoch": 0.044862340216322516,
+      "grad_norm": 30.75,
+      "learning_rate": 9.72e-06,
+      "loss": 0.8694,
+      "step": 730
+    },
+    {
+      "epoch": 0.04492379547689282,
+      "grad_norm": 28.5,
+      "learning_rate": 9.733333333333334e-06,
+      "loss": 0.738,
+      "step": 731
+    },
+    {
+      "epoch": 0.04498525073746313,
+      "grad_norm": 26.5,
+      "learning_rate": 9.746666666666668e-06,
+      "loss": 0.6922,
+      "step": 732
+    },
+    {
+      "epoch": 0.04504670599803343,
+      "grad_norm": 50.75,
+      "learning_rate": 9.760000000000001e-06,
+      "loss": 0.7852,
+      "step": 733
+    },
+    {
+      "epoch": 0.045108161258603736,
+      "grad_norm": 39.5,
+      "learning_rate": 9.773333333333335e-06,
+      "loss": 0.8342,
+      "step": 734
+    },
+    {
+      "epoch": 0.04516961651917404,
+      "grad_norm": 27.5,
+      "learning_rate": 9.786666666666667e-06,
+      "loss": 0.6431,
+      "step": 735
+    },
+    {
+      "epoch": 0.04523107177974434,
+      "grad_norm": 27.375,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.757,
+      "step": 736
+    },
+    {
+      "epoch": 0.045292527040314653,
+      "grad_norm": 15.625,
+      "learning_rate": 9.813333333333333e-06,
+      "loss": 0.5835,
+      "step": 737
+    },
+    {
+      "epoch": 0.04535398230088496,
+      "grad_norm": 49.0,
+      "learning_rate": 9.826666666666667e-06,
+      "loss": 0.737,
+      "step": 738
+    },
+    {
+      "epoch": 0.04541543756145526,
+      "grad_norm": 54.0,
+      "learning_rate": 9.84e-06,
+      "loss": 0.661,
+      "step": 739
+    },
+    {
+      "epoch": 0.045476892822025564,
+      "grad_norm": 41.25,
+      "learning_rate": 9.853333333333334e-06,
+      "loss": 0.5831,
+      "step": 740
+    },
+    {
+      "epoch": 0.04553834808259587,
+      "grad_norm": 27.5,
+      "learning_rate": 9.866666666666668e-06,
+      "loss": 0.5602,
+      "step": 741
+    },
+    {
+      "epoch": 0.04559980334316618,
+      "grad_norm": 52.25,
+      "learning_rate": 9.88e-06,
+      "loss": 0.6011,
+      "step": 742
+    },
+    {
+      "epoch": 0.04566125860373648,
+      "grad_norm": 60.25,
+      "learning_rate": 9.893333333333334e-06,
+      "loss": 0.5949,
+      "step": 743
+    },
+    {
+      "epoch": 0.045722713864306784,
+      "grad_norm": 46.0,
+      "learning_rate": 9.906666666666668e-06,
+      "loss": 0.6577,
+      "step": 744
+    },
+    {
+      "epoch": 0.04578416912487709,
+      "grad_norm": 24.5,
+      "learning_rate": 9.920000000000002e-06,
+      "loss": 0.6808,
+      "step": 745
+    },
+    {
+      "epoch": 0.0458456243854474,
+      "grad_norm": 42.25,
+      "learning_rate": 9.933333333333334e-06,
+      "loss": 0.7542,
+      "step": 746
+    },
+    {
+      "epoch": 0.0459070796460177,
+      "grad_norm": 32.0,
+      "learning_rate": 9.946666666666667e-06,
+      "loss": 0.8687,
+      "step": 747
+    },
+    {
+      "epoch": 0.045968534906588004,
+      "grad_norm": 54.0,
+      "learning_rate": 9.960000000000001e-06,
+      "loss": 1.0703,
+      "step": 748
+    },
+    {
+      "epoch": 0.04602999016715831,
+      "grad_norm": 36.75,
+      "learning_rate": 9.973333333333333e-06,
+      "loss": 0.8712,
+      "step": 749
+    },
+    {
+      "epoch": 0.04609144542772861,
+      "grad_norm": 125.5,
+      "learning_rate": 9.986666666666667e-06,
+      "loss": 1.0925,
+      "step": 750
+    },
+    {
+      "epoch": 0.04615290068829892,
+      "grad_norm": 38.5,
+      "learning_rate": 1e-05,
+      "loss": 0.7888,
+      "step": 751
+    },
+    {
+      "epoch": 0.046214355948869225,
+      "grad_norm": 49.0,
+      "learning_rate": 1.0013333333333335e-05,
+      "loss": 0.6424,
+      "step": 752
+    },
+    {
+      "epoch": 0.04627581120943953,
+      "grad_norm": 48.5,
+      "learning_rate": 1.0026666666666667e-05,
+      "loss": 0.7062,
+      "step": 753
+    },
+    {
+      "epoch": 0.04633726647000983,
+      "grad_norm": 53.25,
+      "learning_rate": 1.004e-05,
+      "loss": 0.7203,
+      "step": 754
+    },
+    {
+      "epoch": 0.046398721730580135,
+      "grad_norm": 49.0,
+      "learning_rate": 1.0053333333333334e-05,
+      "loss": 0.8338,
+      "step": 755
+    },
+    {
+      "epoch": 0.046460176991150445,
+      "grad_norm": 77.0,
+      "learning_rate": 1.0066666666666666e-05,
+      "loss": 0.8558,
+      "step": 756
+    },
+    {
+      "epoch": 0.04652163225172075,
+      "grad_norm": 58.0,
+      "learning_rate": 1.008e-05,
+      "loss": 0.7071,
+      "step": 757
+    },
+    {
+      "epoch": 0.04658308751229105,
+      "grad_norm": 51.25,
+      "learning_rate": 1.0093333333333336e-05,
+      "loss": 0.6844,
+      "step": 758
+    },
+    {
+      "epoch": 0.046644542772861355,
+      "grad_norm": 36.0,
+      "learning_rate": 1.0106666666666668e-05,
+      "loss": 0.721,
+      "step": 759
+    },
+    {
+      "epoch": 0.04670599803343166,
+      "grad_norm": 46.75,
+      "learning_rate": 1.0120000000000001e-05,
+      "loss": 0.7316,
+      "step": 760
+    },
+    {
+      "epoch": 0.04676745329400197,
+      "grad_norm": 47.75,
+      "learning_rate": 1.0133333333333335e-05,
+      "loss": 0.7692,
+      "step": 761
+    },
+    {
+      "epoch": 0.04682890855457227,
+      "grad_norm": 37.5,
+      "learning_rate": 1.0146666666666667e-05,
+      "loss": 0.7745,
+      "step": 762
+    },
+    {
+      "epoch": 0.046890363815142576,
+      "grad_norm": 28.75,
+      "learning_rate": 1.0160000000000001e-05,
+      "loss": 0.7728,
+      "step": 763
+    },
+    {
+      "epoch": 0.04695181907571288,
+      "grad_norm": 36.25,
+      "learning_rate": 1.0173333333333335e-05,
+      "loss": 0.6003,
+      "step": 764
+    },
+    {
+      "epoch": 0.04701327433628318,
+      "grad_norm": 20.875,
+      "learning_rate": 1.0186666666666667e-05,
+      "loss": 0.7126,
+      "step": 765
+    },
+    {
+      "epoch": 0.04707472959685349,
+      "grad_norm": 22.75,
+      "learning_rate": 1.02e-05,
+      "loss": 0.5851,
+      "step": 766
+    },
+    {
+      "epoch": 0.047136184857423796,
+      "grad_norm": 32.75,
+      "learning_rate": 1.0213333333333334e-05,
+      "loss": 0.5361,
+      "step": 767
+    },
+    {
+      "epoch": 0.0471976401179941,
+      "grad_norm": 60.75,
+      "learning_rate": 1.0226666666666666e-05,
+      "loss": 0.6993,
+      "step": 768
+    },
+    {
+      "epoch": 0.0472590953785644,
+      "grad_norm": 40.5,
+      "learning_rate": 1.024e-05,
+      "loss": 0.7402,
+      "step": 769
+    },
+    {
+      "epoch": 0.04732055063913471,
+      "grad_norm": 32.75,
+      "learning_rate": 1.0253333333333336e-05,
+      "loss": 0.6583,
+      "step": 770
+    },
+    {
+      "epoch": 0.04738200589970502,
+      "grad_norm": 30.0,
+      "learning_rate": 1.0266666666666668e-05,
+      "loss": 0.6254,
+      "step": 771
+    },
+    {
+      "epoch": 0.04744346116027532,
+      "grad_norm": 51.0,
+      "learning_rate": 1.0280000000000002e-05,
+      "loss": 0.6938,
+      "step": 772
+    },
+    {
+      "epoch": 0.04750491642084562,
+      "grad_norm": 37.75,
+      "learning_rate": 1.0293333333333335e-05,
+      "loss": 0.697,
+      "step": 773
+    },
+    {
+      "epoch": 0.04756637168141593,
+      "grad_norm": 30.0,
+      "learning_rate": 1.0306666666666667e-05,
+      "loss": 0.7263,
+      "step": 774
+    },
+    {
+      "epoch": 0.04762782694198624,
+      "grad_norm": 49.0,
+      "learning_rate": 1.0320000000000001e-05,
+      "loss": 0.6839,
+      "step": 775
+    },
+    {
+      "epoch": 0.04768928220255654,
+      "grad_norm": 50.25,
+      "learning_rate": 1.0333333333333335e-05,
+      "loss": 0.6247,
+      "step": 776
+    },
+    {
+      "epoch": 0.047750737463126844,
+      "grad_norm": 53.0,
+      "learning_rate": 1.0346666666666667e-05,
+      "loss": 0.9384,
+      "step": 777
+    },
+    {
+      "epoch": 0.04781219272369715,
+      "grad_norm": 33.75,
+      "learning_rate": 1.036e-05,
+      "loss": 0.8235,
+      "step": 778
+    },
+    {
+      "epoch": 0.04787364798426745,
+      "grad_norm": 21.875,
+      "learning_rate": 1.0373333333333335e-05,
+      "loss": 0.6401,
+      "step": 779
+    },
+    {
+      "epoch": 0.04793510324483776,
+      "grad_norm": 31.875,
+      "learning_rate": 1.0386666666666667e-05,
+      "loss": 0.6479,
+      "step": 780
+    },
+    {
+      "epoch": 0.047996558505408064,
+      "grad_norm": 39.25,
+      "learning_rate": 1.04e-05,
+      "loss": 0.5009,
+      "step": 781
+    },
+    {
+      "epoch": 0.04805801376597837,
+      "grad_norm": 47.25,
+      "learning_rate": 1.0413333333333336e-05,
+      "loss": 0.7009,
+      "step": 782
+    },
+    {
+      "epoch": 0.04811946902654867,
+      "grad_norm": 28.375,
+      "learning_rate": 1.0426666666666668e-05,
+      "loss": 0.8483,
+      "step": 783
+    },
+    {
+      "epoch": 0.048180924287118974,
+      "grad_norm": 46.5,
+      "learning_rate": 1.0440000000000002e-05,
+      "loss": 0.6512,
+      "step": 784
+    },
+    {
+      "epoch": 0.048242379547689285,
+      "grad_norm": 40.0,
+      "learning_rate": 1.0453333333333334e-05,
+      "loss": 0.7431,
+      "step": 785
+    },
+    {
+      "epoch": 0.04830383480825959,
+      "grad_norm": 55.75,
+      "learning_rate": 1.0466666666666668e-05,
+      "loss": 0.8059,
+      "step": 786
+    },
+    {
+      "epoch": 0.04836529006882989,
+      "grad_norm": 22.875,
+      "learning_rate": 1.0480000000000001e-05,
+      "loss": 0.6876,
+      "step": 787
+    },
+    {
+      "epoch": 0.048426745329400195,
+      "grad_norm": 45.25,
+      "learning_rate": 1.0493333333333333e-05,
+      "loss": 0.6354,
+      "step": 788
+    },
+    {
+      "epoch": 0.0484882005899705,
+      "grad_norm": 36.25,
+      "learning_rate": 1.0506666666666667e-05,
+      "loss": 0.8947,
+      "step": 789
+    },
+    {
+      "epoch": 0.04854965585054081,
+      "grad_norm": 39.5,
+      "learning_rate": 1.0520000000000001e-05,
+      "loss": 0.593,
+      "step": 790
+    },
+    {
+      "epoch": 0.04861111111111111,
+      "grad_norm": 16.125,
+      "learning_rate": 1.0533333333333333e-05,
+      "loss": 0.6071,
+      "step": 791
+    },
+    {
+      "epoch": 0.048672566371681415,
+      "grad_norm": 60.75,
+      "learning_rate": 1.0546666666666667e-05,
+      "loss": 0.6698,
+      "step": 792
+    },
+    {
+      "epoch": 0.04873402163225172,
+      "grad_norm": 41.0,
+      "learning_rate": 1.056e-05,
+      "loss": 0.9218,
+      "step": 793
+    },
+    {
+      "epoch": 0.04879547689282203,
+      "grad_norm": 36.25,
+      "learning_rate": 1.0573333333333333e-05,
+      "loss": 0.8542,
+      "step": 794
+    },
+    {
+      "epoch": 0.04885693215339233,
+      "grad_norm": 35.25,
+      "learning_rate": 1.0586666666666668e-05,
+      "loss": 0.6295,
+      "step": 795
+    },
+    {
+      "epoch": 0.048918387413962636,
+      "grad_norm": 31.5,
+      "learning_rate": 1.0600000000000002e-05,
+      "loss": 0.7025,
+      "step": 796
+    },
+    {
+      "epoch": 0.04897984267453294,
+      "grad_norm": 32.75,
+      "learning_rate": 1.0613333333333334e-05,
+      "loss": 0.6968,
+      "step": 797
+    },
+    {
+      "epoch": 0.04904129793510324,
+      "grad_norm": 31.375,
+      "learning_rate": 1.0626666666666668e-05,
+      "loss": 0.7318,
+      "step": 798
+    },
+    {
+      "epoch": 0.04910275319567355,
+      "grad_norm": 44.5,
+      "learning_rate": 1.0640000000000001e-05,
+      "loss": 0.7562,
+      "step": 799
+    },
+    {
+      "epoch": 0.049164208456243856,
+      "grad_norm": 18.125,
+      "learning_rate": 1.0653333333333334e-05,
+      "loss": 0.6267,
+      "step": 800
+    },
+    {
+      "epoch": 0.04922566371681416,
+      "grad_norm": 36.75,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 0.665,
+      "step": 801
+    },
+    {
+      "epoch": 0.04928711897738446,
+      "grad_norm": 75.0,
+      "learning_rate": 1.0680000000000001e-05,
+      "loss": 0.6933,
+      "step": 802
+    },
+    {
+      "epoch": 0.049348574237954766,
+      "grad_norm": 57.75,
+      "learning_rate": 1.0693333333333333e-05,
+      "loss": 0.5902,
+      "step": 803
+    },
+    {
+      "epoch": 0.049410029498525077,
+      "grad_norm": 65.5,
+      "learning_rate": 1.0706666666666667e-05,
+      "loss": 0.6891,
+      "step": 804
+    },
+    {
+      "epoch": 0.04947148475909538,
+      "grad_norm": 51.75,
+      "learning_rate": 1.072e-05,
+      "loss": 0.7041,
+      "step": 805
+    },
+    {
+      "epoch": 0.04953294001966568,
+      "grad_norm": 35.0,
+      "learning_rate": 1.0733333333333333e-05,
+      "loss": 0.6583,
+      "step": 806
+    },
+    {
+      "epoch": 0.04959439528023599,
+      "grad_norm": 37.25,
+      "learning_rate": 1.0746666666666668e-05,
+      "loss": 0.7138,
+      "step": 807
+    },
+    {
+      "epoch": 0.04965585054080629,
+      "grad_norm": 52.25,
+      "learning_rate": 1.0760000000000002e-05,
+      "loss": 0.7472,
+      "step": 808
+    },
+    {
+      "epoch": 0.0497173058013766,
+      "grad_norm": 49.25,
+      "learning_rate": 1.0773333333333334e-05,
+      "loss": 0.6569,
+      "step": 809
+    },
+    {
+      "epoch": 0.049778761061946904,
+      "grad_norm": 50.5,
+      "learning_rate": 1.0786666666666668e-05,
+      "loss": 0.8802,
+      "step": 810
+    },
+    {
+      "epoch": 0.04984021632251721,
+      "grad_norm": 41.25,
+      "learning_rate": 1.0800000000000002e-05,
+      "loss": 0.7542,
+      "step": 811
+    },
+    {
+      "epoch": 0.04990167158308751,
+      "grad_norm": 27.25,
+      "learning_rate": 1.0813333333333334e-05,
+      "loss": 0.5113,
+      "step": 812
+    },
+    {
+      "epoch": 0.049963126843657814,
+      "grad_norm": 23.625,
+      "learning_rate": 1.0826666666666667e-05,
+      "loss": 0.677,
+      "step": 813
+    },
+    {
+      "epoch": 0.050024582104228124,
+      "grad_norm": 44.0,
+      "learning_rate": 1.0840000000000001e-05,
+      "loss": 0.6348,
+      "step": 814
+    },
+    {
+      "epoch": 0.05008603736479843,
+      "grad_norm": 22.375,
+      "learning_rate": 1.0853333333333333e-05,
+      "loss": 0.5962,
+      "step": 815
+    },
+    {
+      "epoch": 0.05014749262536873,
+      "grad_norm": 93.0,
+      "learning_rate": 1.0866666666666667e-05,
+      "loss": 0.6353,
+      "step": 816
+    },
+    {
+      "epoch": 0.050208947885939034,
+      "grad_norm": 65.0,
+      "learning_rate": 1.0880000000000001e-05,
+      "loss": 0.603,
+      "step": 817
+    },
+    {
+      "epoch": 0.050270403146509345,
+      "grad_norm": 44.0,
+      "learning_rate": 1.0893333333333333e-05,
+      "loss": 0.9362,
+      "step": 818
+    },
+    {
+      "epoch": 0.05033185840707965,
+      "grad_norm": 21.75,
+      "learning_rate": 1.0906666666666668e-05,
+      "loss": 0.7394,
+      "step": 819
+    },
+    {
+      "epoch": 0.05039331366764995,
+      "grad_norm": 43.0,
+      "learning_rate": 1.0920000000000002e-05,
+      "loss": 0.6651,
+      "step": 820
+    },
+    {
+      "epoch": 0.050454768928220255,
+      "grad_norm": 45.25,
+      "learning_rate": 1.0933333333333334e-05,
+      "loss": 0.6671,
+      "step": 821
+    },
+    {
+      "epoch": 0.05051622418879056,
+      "grad_norm": 69.0,
+      "learning_rate": 1.0946666666666668e-05,
+      "loss": 0.7588,
+      "step": 822
+    },
+    {
+      "epoch": 0.05057767944936087,
+      "grad_norm": 59.25,
+      "learning_rate": 1.0960000000000002e-05,
+      "loss": 0.7313,
+      "step": 823
+    },
+    {
+      "epoch": 0.05063913470993117,
+      "grad_norm": 44.25,
+      "learning_rate": 1.0973333333333334e-05,
+      "loss": 0.63,
+      "step": 824
+    },
+    {
+      "epoch": 0.050700589970501475,
+      "grad_norm": 45.25,
+      "learning_rate": 1.0986666666666668e-05,
+      "loss": 0.7048,
+      "step": 825
+    },
+    {
+      "epoch": 0.05076204523107178,
+      "grad_norm": 42.25,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.5822,
+      "step": 826
+    },
+    {
+      "epoch": 0.05082350049164208,
+      "grad_norm": 51.75,
+      "learning_rate": 1.1013333333333333e-05,
+      "loss": 0.7066,
+      "step": 827
+    },
+    {
+      "epoch": 0.05088495575221239,
+      "grad_norm": 55.0,
+      "learning_rate": 1.1026666666666667e-05,
+      "loss": 0.5021,
+      "step": 828
+    },
+    {
+      "epoch": 0.050946411012782696,
+      "grad_norm": 43.75,
+      "learning_rate": 1.1040000000000001e-05,
+      "loss": 0.7969,
+      "step": 829
+    },
+    {
+      "epoch": 0.051007866273353,
+      "grad_norm": 54.5,
+      "learning_rate": 1.1053333333333333e-05,
+      "loss": 0.6311,
+      "step": 830
+    },
+    {
+      "epoch": 0.0510693215339233,
+      "grad_norm": 25.875,
+      "learning_rate": 1.1066666666666669e-05,
+      "loss": 0.6224,
+      "step": 831
+    },
+    {
+      "epoch": 0.051130776794493606,
+      "grad_norm": 49.75,
+      "learning_rate": 1.1080000000000002e-05,
+      "loss": 0.9938,
+      "step": 832
+    },
+    {
+      "epoch": 0.051192232055063916,
+      "grad_norm": 35.0,
+      "learning_rate": 1.1093333333333334e-05,
+      "loss": 0.6986,
+      "step": 833
+    },
+    {
+      "epoch": 0.05125368731563422,
+      "grad_norm": 34.75,
+      "learning_rate": 1.1106666666666668e-05,
+      "loss": 0.7152,
+      "step": 834
+    },
+    {
+      "epoch": 0.05131514257620452,
+      "grad_norm": 20.25,
+      "learning_rate": 1.1120000000000002e-05,
+      "loss": 0.5822,
+      "step": 835
+    },
+    {
+      "epoch": 0.051376597836774826,
+      "grad_norm": 30.875,
+      "learning_rate": 1.1133333333333334e-05,
+      "loss": 0.6785,
+      "step": 836
+    },
+    {
+      "epoch": 0.05143805309734513,
+      "grad_norm": 22.25,
+      "learning_rate": 1.1146666666666668e-05,
+      "loss": 0.5322,
+      "step": 837
+    },
+    {
+      "epoch": 0.05149950835791544,
+      "grad_norm": 47.0,
+      "learning_rate": 1.1160000000000002e-05,
+      "loss": 0.7106,
+      "step": 838
+    },
+    {
+      "epoch": 0.05156096361848574,
+      "grad_norm": 32.75,
+      "learning_rate": 1.1173333333333334e-05,
+      "loss": 0.7764,
+      "step": 839
+    },
+    {
+      "epoch": 0.051622418879056046,
+      "grad_norm": 63.75,
+      "learning_rate": 1.1186666666666667e-05,
+      "loss": 0.9453,
+      "step": 840
+    },
+    {
+      "epoch": 0.05168387413962635,
+      "grad_norm": 27.875,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 0.7396,
+      "step": 841
+    },
+    {
+      "epoch": 0.05174532940019666,
+      "grad_norm": 25.75,
+      "learning_rate": 1.1213333333333333e-05,
+      "loss": 0.8482,
+      "step": 842
+    },
+    {
+      "epoch": 0.051806784660766964,
+      "grad_norm": 18.375,
+      "learning_rate": 1.1226666666666669e-05,
+      "loss": 0.6308,
+      "step": 843
+    },
+    {
+      "epoch": 0.05186823992133727,
+      "grad_norm": 22.25,
+      "learning_rate": 1.1240000000000002e-05,
+      "loss": 0.7905,
+      "step": 844
+    },
+    {
+      "epoch": 0.05192969518190757,
+      "grad_norm": 26.0,
+      "learning_rate": 1.1253333333333335e-05,
+      "loss": 0.6793,
+      "step": 845
+    },
+    {
+      "epoch": 0.051991150442477874,
+      "grad_norm": 24.375,
+      "learning_rate": 1.1266666666666668e-05,
+      "loss": 0.6244,
+      "step": 846
+    },
+    {
+      "epoch": 0.052052605703048184,
+      "grad_norm": 44.25,
+      "learning_rate": 1.128e-05,
+      "loss": 0.8591,
+      "step": 847
+    },
+    {
+      "epoch": 0.05211406096361849,
+      "grad_norm": 43.5,
+      "learning_rate": 1.1293333333333334e-05,
+      "loss": 0.7352,
+      "step": 848
+    },
+    {
+      "epoch": 0.05217551622418879,
+      "grad_norm": 41.5,
+      "learning_rate": 1.1306666666666668e-05,
+      "loss": 0.8791,
+      "step": 849
+    },
+    {
+      "epoch": 0.052236971484759094,
+      "grad_norm": 31.5,
+      "learning_rate": 1.132e-05,
+      "loss": 0.8298,
+      "step": 850
+    },
+    {
+      "epoch": 0.0522984267453294,
+      "grad_norm": 43.25,
+      "learning_rate": 1.1333333333333334e-05,
+      "loss": 0.7726,
+      "step": 851
+    },
+    {
+      "epoch": 0.05235988200589971,
+      "grad_norm": 44.75,
+      "learning_rate": 1.1346666666666668e-05,
+      "loss": 0.9036,
+      "step": 852
+    },
+    {
+      "epoch": 0.05242133726647001,
+      "grad_norm": 40.25,
+      "learning_rate": 1.136e-05,
+      "loss": 0.8183,
+      "step": 853
+    },
+    {
+      "epoch": 0.052482792527040315,
+      "grad_norm": 30.375,
+      "learning_rate": 1.1373333333333333e-05,
+      "loss": 0.7978,
+      "step": 854
+    },
+    {
+      "epoch": 0.05254424778761062,
+      "grad_norm": 28.0,
+      "learning_rate": 1.1386666666666669e-05,
+      "loss": 0.9153,
+      "step": 855
+    },
+    {
+      "epoch": 0.05260570304818092,
+      "grad_norm": 29.0,
+      "learning_rate": 1.14e-05,
+      "loss": 0.6611,
+      "step": 856
+    },
+    {
+      "epoch": 0.05266715830875123,
+      "grad_norm": 39.25,
+      "learning_rate": 1.1413333333333335e-05,
+      "loss": 0.7199,
+      "step": 857
+    },
+    {
+      "epoch": 0.052728613569321535,
+      "grad_norm": 40.25,
+      "learning_rate": 1.1426666666666669e-05,
+      "loss": 0.8384,
+      "step": 858
+    },
+    {
+      "epoch": 0.05279006882989184,
+      "grad_norm": 29.0,
+      "learning_rate": 1.144e-05,
+      "loss": 0.7427,
+      "step": 859
+    },
+    {
+      "epoch": 0.05285152409046214,
+      "grad_norm": 30.375,
+      "learning_rate": 1.1453333333333334e-05,
+      "loss": 0.7447,
+      "step": 860
+    },
+    {
+      "epoch": 0.052912979351032445,
+      "grad_norm": 55.0,
+      "learning_rate": 1.1466666666666668e-05,
+      "loss": 0.7126,
+      "step": 861
+    },
+    {
+      "epoch": 0.052974434611602755,
+      "grad_norm": 20.25,
+      "learning_rate": 1.148e-05,
+      "loss": 0.6529,
+      "step": 862
+    },
+    {
+      "epoch": 0.05303588987217306,
+      "grad_norm": 28.625,
+      "learning_rate": 1.1493333333333334e-05,
+      "loss": 0.5448,
+      "step": 863
+    },
+    {
+      "epoch": 0.05309734513274336,
+      "grad_norm": 34.25,
+      "learning_rate": 1.1506666666666668e-05,
+      "loss": 0.5725,
+      "step": 864
+    },
+    {
+      "epoch": 0.053158800393313665,
+      "grad_norm": 41.25,
+      "learning_rate": 1.152e-05,
+      "loss": 0.7562,
+      "step": 865
+    },
+    {
+      "epoch": 0.053220255653883976,
+      "grad_norm": 33.25,
+      "learning_rate": 1.1533333333333334e-05,
+      "loss": 0.9486,
+      "step": 866
+    },
+    {
+      "epoch": 0.05328171091445428,
+      "grad_norm": 26.25,
+      "learning_rate": 1.1546666666666669e-05,
+      "loss": 0.7683,
+      "step": 867
+    },
+    {
+      "epoch": 0.05334316617502458,
+      "grad_norm": 15.375,
+      "learning_rate": 1.156e-05,
+      "loss": 0.5429,
+      "step": 868
+    },
+    {
+      "epoch": 0.053404621435594886,
+      "grad_norm": 30.125,
+      "learning_rate": 1.1573333333333335e-05,
+      "loss": 0.5291,
+      "step": 869
+    },
+    {
+      "epoch": 0.05346607669616519,
+      "grad_norm": 29.5,
+      "learning_rate": 1.1586666666666669e-05,
+      "loss": 0.5687,
+      "step": 870
+    },
+    {
+      "epoch": 0.0535275319567355,
+      "grad_norm": 17.5,
+      "learning_rate": 1.16e-05,
+      "loss": 0.5489,
+      "step": 871
+    },
+    {
+      "epoch": 0.0535889872173058,
+      "grad_norm": 30.125,
+      "learning_rate": 1.1613333333333335e-05,
+      "loss": 0.6197,
+      "step": 872
+    },
+    {
+      "epoch": 0.053650442477876106,
+      "grad_norm": 24.5,
+      "learning_rate": 1.1626666666666668e-05,
+      "loss": 0.7239,
+      "step": 873
+    },
+    {
+      "epoch": 0.05371189773844641,
+      "grad_norm": 29.625,
+      "learning_rate": 1.164e-05,
+      "loss": 0.7267,
+      "step": 874
+    },
+    {
+      "epoch": 0.05377335299901671,
+      "grad_norm": 43.0,
+      "learning_rate": 1.1653333333333334e-05,
+      "loss": 0.6004,
+      "step": 875
+    },
+    {
+      "epoch": 0.05383480825958702,
+      "grad_norm": 34.0,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 0.7228,
+      "step": 876
+    },
+    {
+      "epoch": 0.05389626352015733,
+      "grad_norm": 29.125,
+      "learning_rate": 1.168e-05,
+      "loss": 0.6849,
+      "step": 877
+    },
+    {
+      "epoch": 0.05395771878072763,
+      "grad_norm": 29.75,
+      "learning_rate": 1.1693333333333334e-05,
+      "loss": 0.6881,
+      "step": 878
+    },
+    {
+      "epoch": 0.054019174041297933,
+      "grad_norm": 35.0,
+      "learning_rate": 1.170666666666667e-05,
+      "loss": 0.7448,
+      "step": 879
+    },
+    {
+      "epoch": 0.05408062930186824,
+      "grad_norm": 39.0,
+      "learning_rate": 1.172e-05,
+      "loss": 0.8501,
+      "step": 880
+    },
+    {
+      "epoch": 0.05414208456243855,
+      "grad_norm": 35.5,
+      "learning_rate": 1.1733333333333335e-05,
+      "loss": 0.7999,
+      "step": 881
+    },
+    {
+      "epoch": 0.05420353982300885,
+      "grad_norm": 38.5,
+      "learning_rate": 1.1746666666666669e-05,
+      "loss": 0.579,
+      "step": 882
+    },
+    {
+      "epoch": 0.054264995083579154,
+      "grad_norm": 16.25,
+      "learning_rate": 1.1760000000000001e-05,
+      "loss": 0.5202,
+      "step": 883
+    },
+    {
+      "epoch": 0.05432645034414946,
+      "grad_norm": 21.625,
+      "learning_rate": 1.1773333333333335e-05,
+      "loss": 0.5582,
+      "step": 884
+    },
+    {
+      "epoch": 0.05438790560471976,
+      "grad_norm": 32.25,
+      "learning_rate": 1.1786666666666668e-05,
+      "loss": 0.5072,
+      "step": 885
+    },
+    {
+      "epoch": 0.05444936086529007,
+      "grad_norm": 26.375,
+      "learning_rate": 1.18e-05,
+      "loss": 0.5798,
+      "step": 886
+    },
+    {
+      "epoch": 0.054510816125860374,
+      "grad_norm": 30.75,
+      "learning_rate": 1.1813333333333334e-05,
+      "loss": 0.4824,
+      "step": 887
+    },
+    {
+      "epoch": 0.05457227138643068,
+      "grad_norm": 27.375,
+      "learning_rate": 1.1826666666666668e-05,
+      "loss": 0.5179,
+      "step": 888
+    },
+    {
+      "epoch": 0.05463372664700098,
+      "grad_norm": 62.0,
+      "learning_rate": 1.184e-05,
+      "loss": 0.7642,
+      "step": 889
+    },
+    {
+      "epoch": 0.05469518190757129,
+      "grad_norm": 23.25,
+      "learning_rate": 1.1853333333333334e-05,
+      "loss": 0.692,
+      "step": 890
+    },
+    {
+      "epoch": 0.054756637168141595,
+      "grad_norm": 50.0,
+      "learning_rate": 1.186666666666667e-05,
+      "loss": 0.7653,
+      "step": 891
+    },
+    {
+      "epoch": 0.0548180924287119,
+      "grad_norm": 82.5,
+      "learning_rate": 1.188e-05,
+      "loss": 0.7653,
+      "step": 892
+    },
+    {
+      "epoch": 0.0548795476892822,
+      "grad_norm": 68.0,
+      "learning_rate": 1.1893333333333335e-05,
+      "loss": 0.742,
+      "step": 893
+    },
+    {
+      "epoch": 0.054941002949852505,
+      "grad_norm": 69.0,
+      "learning_rate": 1.1906666666666669e-05,
+      "loss": 0.7095,
+      "step": 894
+    },
+    {
+      "epoch": 0.055002458210422815,
+      "grad_norm": 45.25,
+      "learning_rate": 1.1920000000000001e-05,
+      "loss": 0.5097,
+      "step": 895
+    },
+    {
+      "epoch": 0.05506391347099312,
+      "grad_norm": 48.5,
+      "learning_rate": 1.1933333333333335e-05,
+      "loss": 0.5493,
+      "step": 896
+    },
+    {
+      "epoch": 0.05512536873156342,
+      "grad_norm": 40.75,
+      "learning_rate": 1.1946666666666669e-05,
+      "loss": 0.6108,
+      "step": 897
+    },
+    {
+      "epoch": 0.055186823992133725,
+      "grad_norm": 52.25,
+      "learning_rate": 1.196e-05,
+      "loss": 0.6981,
+      "step": 898
+    },
+    {
+      "epoch": 0.05524827925270403,
+      "grad_norm": 42.25,
+      "learning_rate": 1.1973333333333334e-05,
+      "loss": 0.7181,
+      "step": 899
+    },
+    {
+      "epoch": 0.05530973451327434,
+      "grad_norm": 22.625,
+      "learning_rate": 1.1986666666666668e-05,
+      "loss": 0.581,
+      "step": 900
+    },
+    {
+      "epoch": 0.05537118977384464,
+      "grad_norm": 34.75,
+      "learning_rate": 1.2e-05,
+      "loss": 0.6108,
+      "step": 901
+    },
+    {
+      "epoch": 0.055432645034414946,
+      "grad_norm": 29.875,
+      "learning_rate": 1.2013333333333334e-05,
+      "loss": 0.5962,
+      "step": 902
+    },
+    {
+      "epoch": 0.05549410029498525,
+      "grad_norm": 38.5,
+      "learning_rate": 1.202666666666667e-05,
+      "loss": 0.5995,
+      "step": 903
+    },
+    {
+      "epoch": 0.05555555555555555,
+      "grad_norm": 27.625,
+      "learning_rate": 1.204e-05,
+      "loss": 0.6027,
+      "step": 904
+    },
+    {
+      "epoch": 0.05561701081612586,
+      "grad_norm": 27.875,
+      "learning_rate": 1.2053333333333335e-05,
+      "loss": 0.771,
+      "step": 905
+    },
+    {
+      "epoch": 0.055678466076696166,
+      "grad_norm": 33.25,
+      "learning_rate": 1.206666666666667e-05,
+      "loss": 0.9046,
+      "step": 906
+    },
+    {
+      "epoch": 0.05573992133726647,
+      "grad_norm": 31.875,
+      "learning_rate": 1.2080000000000001e-05,
+      "loss": 0.8869,
+      "step": 907
+    },
+    {
+      "epoch": 0.05580137659783677,
+      "grad_norm": 38.25,
+      "learning_rate": 1.2093333333333335e-05,
+      "loss": 0.964,
+      "step": 908
+    },
+    {
+      "epoch": 0.055862831858407076,
+      "grad_norm": 30.5,
+      "learning_rate": 1.2106666666666667e-05,
+      "loss": 1.1805,
+      "step": 909
+    },
+    {
+      "epoch": 0.05592428711897739,
+      "grad_norm": 23.75,
+      "learning_rate": 1.2120000000000001e-05,
+      "loss": 0.74,
+      "step": 910
+    },
+    {
+      "epoch": 0.05598574237954769,
+      "grad_norm": 24.625,
+      "learning_rate": 1.2133333333333335e-05,
+      "loss": 0.8718,
+      "step": 911
+    },
+    {
+      "epoch": 0.05604719764011799,
+      "grad_norm": 25.625,
+      "learning_rate": 1.2146666666666667e-05,
+      "loss": 1.0713,
+      "step": 912
+    },
+    {
+      "epoch": 0.0561086529006883,
+      "grad_norm": 46.0,
+      "learning_rate": 1.216e-05,
+      "loss": 0.8758,
+      "step": 913
+    },
+    {
+      "epoch": 0.05617010816125861,
+      "grad_norm": 26.875,
+      "learning_rate": 1.2173333333333334e-05,
+      "loss": 0.771,
+      "step": 914
+    },
+    {
+      "epoch": 0.05623156342182891,
+      "grad_norm": 22.625,
+      "learning_rate": 1.2186666666666666e-05,
+      "loss": 0.8002,
+      "step": 915
+    },
+    {
+      "epoch": 0.056293018682399214,
+      "grad_norm": 32.25,
+      "learning_rate": 1.22e-05,
+      "loss": 0.6459,
+      "step": 916
+    },
+    {
+      "epoch": 0.05635447394296952,
+      "grad_norm": 25.125,
+      "learning_rate": 1.2213333333333336e-05,
+      "loss": 0.5845,
+      "step": 917
+    },
+    {
+      "epoch": 0.05641592920353982,
+      "grad_norm": 21.25,
+      "learning_rate": 1.2226666666666666e-05,
+      "loss": 0.6145,
+      "step": 918
+    },
+    {
+      "epoch": 0.05647738446411013,
+      "grad_norm": 23.875,
+      "learning_rate": 1.2240000000000001e-05,
+      "loss": 0.6707,
+      "step": 919
+    },
+    {
+      "epoch": 0.056538839724680434,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.2253333333333335e-05,
+      "loss": 0.5965,
+      "step": 920
+    },
+    {
+      "epoch": 0.05660029498525074,
+      "grad_norm": 33.5,
+      "learning_rate": 1.2266666666666667e-05,
+      "loss": 0.5909,
+      "step": 921
+    },
+    {
+      "epoch": 0.05666175024582104,
+      "grad_norm": 28.625,
+      "learning_rate": 1.2280000000000001e-05,
+      "loss": 0.7442,
+      "step": 922
+    },
+    {
+      "epoch": 0.056723205506391344,
+      "grad_norm": 41.25,
+      "learning_rate": 1.2293333333333335e-05,
+      "loss": 0.7375,
+      "step": 923
+    },
+    {
+      "epoch": 0.056784660766961655,
+      "grad_norm": 25.25,
+      "learning_rate": 1.2306666666666667e-05,
+      "loss": 0.5163,
+      "step": 924
+    },
+    {
+      "epoch": 0.05684611602753196,
+      "grad_norm": 37.75,
+      "learning_rate": 1.232e-05,
+      "loss": 0.6513,
+      "step": 925
+    },
+    {
+      "epoch": 0.05690757128810226,
+      "grad_norm": 34.5,
+      "learning_rate": 1.2333333333333334e-05,
+      "loss": 0.7151,
+      "step": 926
+    },
+    {
+      "epoch": 0.056969026548672565,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.2346666666666666e-05,
+      "loss": 0.6596,
+      "step": 927
+    },
+    {
+      "epoch": 0.05703048180924287,
+      "grad_norm": 34.0,
+      "learning_rate": 1.236e-05,
+      "loss": 0.6131,
+      "step": 928
+    },
+    {
+      "epoch": 0.05709193706981318,
+      "grad_norm": 39.25,
+      "learning_rate": 1.2373333333333336e-05,
+      "loss": 0.759,
+      "step": 929
+    },
+    {
+      "epoch": 0.05715339233038348,
+      "grad_norm": 29.25,
+      "learning_rate": 1.2386666666666666e-05,
+      "loss": 0.662,
+      "step": 930
+    },
+    {
+      "epoch": 0.057214847590953785,
+      "grad_norm": 72.0,
+      "learning_rate": 1.2400000000000002e-05,
+      "loss": 1.1421,
+      "step": 931
+    },
+    {
+      "epoch": 0.05727630285152409,
+      "grad_norm": 40.5,
+      "learning_rate": 1.2413333333333335e-05,
+      "loss": 0.7545,
+      "step": 932
+    },
+    {
+      "epoch": 0.05733775811209439,
+      "grad_norm": 50.5,
+      "learning_rate": 1.2426666666666667e-05,
+      "loss": 0.6465,
+      "step": 933
+    },
+    {
+      "epoch": 0.0573992133726647,
+      "grad_norm": 27.125,
+      "learning_rate": 1.2440000000000001e-05,
+      "loss": 0.811,
+      "step": 934
+    },
+    {
+      "epoch": 0.057460668633235006,
+      "grad_norm": 27.875,
+      "learning_rate": 1.2453333333333335e-05,
+      "loss": 0.6222,
+      "step": 935
+    },
+    {
+      "epoch": 0.05752212389380531,
+      "grad_norm": 31.375,
+      "learning_rate": 1.2466666666666667e-05,
+      "loss": 0.6505,
+      "step": 936
+    },
+    {
+      "epoch": 0.05758357915437561,
+      "grad_norm": 27.375,
+      "learning_rate": 1.248e-05,
+      "loss": 0.6831,
+      "step": 937
+    },
+    {
+      "epoch": 0.05764503441494592,
+      "grad_norm": 40.25,
+      "learning_rate": 1.2493333333333335e-05,
+      "loss": 0.7704,
+      "step": 938
+    },
+    {
+      "epoch": 0.057706489675516226,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.2506666666666667e-05,
+      "loss": 0.6004,
+      "step": 939
+    },
+    {
+      "epoch": 0.05776794493608653,
+      "grad_norm": 67.0,
+      "learning_rate": 1.252e-05,
+      "loss": 0.9434,
+      "step": 940
+    },
+    {
+      "epoch": 0.05782940019665683,
+      "grad_norm": 36.25,
+      "learning_rate": 1.2533333333333336e-05,
+      "loss": 0.8295,
+      "step": 941
+    },
+    {
+      "epoch": 0.057890855457227136,
+      "grad_norm": 47.75,
+      "learning_rate": 1.2546666666666666e-05,
+      "loss": 0.6951,
+      "step": 942
+    },
+    {
+      "epoch": 0.057952310717797446,
+      "grad_norm": 17.625,
+      "learning_rate": 1.2560000000000002e-05,
+      "loss": 0.7429,
+      "step": 943
+    },
+    {
+      "epoch": 0.05801376597836775,
+      "grad_norm": 22.25,
+      "learning_rate": 1.2573333333333335e-05,
+      "loss": 0.4897,
+      "step": 944
+    },
+    {
+      "epoch": 0.05807522123893805,
+      "grad_norm": 45.75,
+      "learning_rate": 1.2586666666666668e-05,
+      "loss": 0.7351,
+      "step": 945
+    },
+    {
+      "epoch": 0.058136676499508357,
+      "grad_norm": 34.5,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 0.8585,
+      "step": 946
+    },
+    {
+      "epoch": 0.05819813176007866,
+      "grad_norm": 22.875,
+      "learning_rate": 1.2613333333333335e-05,
+      "loss": 0.6328,
+      "step": 947
+    },
+    {
+      "epoch": 0.05825958702064897,
+      "grad_norm": 26.75,
+      "learning_rate": 1.2626666666666667e-05,
+      "loss": 0.7067,
+      "step": 948
+    },
+    {
+      "epoch": 0.058321042281219274,
+      "grad_norm": 23.0,
+      "learning_rate": 1.2640000000000001e-05,
+      "loss": 0.7098,
+      "step": 949
+    },
+    {
+      "epoch": 0.05838249754178958,
+      "grad_norm": 27.375,
+      "learning_rate": 1.2653333333333335e-05,
+      "loss": 0.7714,
+      "step": 950
+    },
+    {
+      "epoch": 0.05844395280235988,
+      "grad_norm": 40.75,
+      "learning_rate": 1.2666666666666667e-05,
+      "loss": 0.6829,
+      "step": 951
+    },
+    {
+      "epoch": 0.058505408062930184,
+      "grad_norm": 22.5,
+      "learning_rate": 1.268e-05,
+      "loss": 0.7709,
+      "step": 952
+    },
+    {
+      "epoch": 0.058566863323500494,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.2693333333333336e-05,
+      "loss": 0.7671,
+      "step": 953
+    },
+    {
+      "epoch": 0.0586283185840708,
+      "grad_norm": 28.75,
+      "learning_rate": 1.2706666666666666e-05,
+      "loss": 0.8198,
+      "step": 954
+    },
+    {
+      "epoch": 0.0586897738446411,
+      "grad_norm": 39.0,
+      "learning_rate": 1.2720000000000002e-05,
+      "loss": 0.754,
+      "step": 955
+    },
+    {
+      "epoch": 0.058751229105211404,
+      "grad_norm": 38.0,
+      "learning_rate": 1.2733333333333336e-05,
+      "loss": 0.8153,
+      "step": 956
+    },
+    {
+      "epoch": 0.05881268436578171,
+      "grad_norm": 19.75,
+      "learning_rate": 1.2746666666666668e-05,
+      "loss": 0.8053,
+      "step": 957
+    },
+    {
+      "epoch": 0.05887413962635202,
+      "grad_norm": 28.0,
+      "learning_rate": 1.2760000000000001e-05,
+      "loss": 0.8798,
+      "step": 958
+    },
+    {
+      "epoch": 0.05893559488692232,
+      "grad_norm": 29.375,
+      "learning_rate": 1.2773333333333335e-05,
+      "loss": 0.6067,
+      "step": 959
+    },
+    {
+      "epoch": 0.058997050147492625,
+      "grad_norm": 46.25,
+      "learning_rate": 1.2786666666666667e-05,
+      "loss": 0.7138,
+      "step": 960
+    },
+    {
+      "epoch": 0.05905850540806293,
+      "grad_norm": 21.375,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 0.8096,
+      "step": 961
+    },
+    {
+      "epoch": 0.05911996066863324,
+      "grad_norm": 35.5,
+      "learning_rate": 1.2813333333333335e-05,
+      "loss": 0.9131,
+      "step": 962
+    },
+    {
+      "epoch": 0.05918141592920354,
+      "grad_norm": 27.0,
+      "learning_rate": 1.2826666666666667e-05,
+      "loss": 0.7825,
+      "step": 963
+    },
+    {
+      "epoch": 0.059242871189773845,
+      "grad_norm": 28.5,
+      "learning_rate": 1.284e-05,
+      "loss": 0.7802,
+      "step": 964
+    },
+    {
+      "epoch": 0.05930432645034415,
+      "grad_norm": 29.25,
+      "learning_rate": 1.2853333333333336e-05,
+      "loss": 0.7461,
+      "step": 965
+    },
+    {
+      "epoch": 0.05936578171091445,
+      "grad_norm": 22.0,
+      "learning_rate": 1.2866666666666667e-05,
+      "loss": 0.5613,
+      "step": 966
+    },
+    {
+      "epoch": 0.05942723697148476,
+      "grad_norm": 21.25,
+      "learning_rate": 1.2880000000000002e-05,
+      "loss": 0.6754,
+      "step": 967
+    },
+    {
+      "epoch": 0.059488692232055065,
+      "grad_norm": 18.125,
+      "learning_rate": 1.2893333333333336e-05,
+      "loss": 0.6465,
+      "step": 968
+    },
+    {
+      "epoch": 0.05955014749262537,
+      "grad_norm": 21.125,
+      "learning_rate": 1.2906666666666668e-05,
+      "loss": 0.6107,
+      "step": 969
+    },
+    {
+      "epoch": 0.05961160275319567,
+      "grad_norm": 26.25,
+      "learning_rate": 1.2920000000000002e-05,
+      "loss": 0.7043,
+      "step": 970
+    },
+    {
+      "epoch": 0.059673058013765976,
+      "grad_norm": 20.25,
+      "learning_rate": 1.2933333333333334e-05,
+      "loss": 0.7324,
+      "step": 971
+    },
+    {
+      "epoch": 0.059734513274336286,
+      "grad_norm": 25.0,
+      "learning_rate": 1.2946666666666668e-05,
+      "loss": 0.5054,
+      "step": 972
+    },
+    {
+      "epoch": 0.05979596853490659,
+      "grad_norm": 45.75,
+      "learning_rate": 1.2960000000000001e-05,
+      "loss": 0.7858,
+      "step": 973
+    },
+    {
+      "epoch": 0.05985742379547689,
+      "grad_norm": 34.0,
+      "learning_rate": 1.2973333333333333e-05,
+      "loss": 0.654,
+      "step": 974
+    },
+    {
+      "epoch": 0.059918879056047196,
+      "grad_norm": 43.0,
+      "learning_rate": 1.2986666666666667e-05,
+      "loss": 0.6117,
+      "step": 975
+    },
+    {
+      "epoch": 0.0599803343166175,
+      "grad_norm": 17.75,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.5343,
+      "step": 976
+    },
+    {
+      "epoch": 0.06004178957718781,
+      "grad_norm": 62.5,
+      "learning_rate": 1.3013333333333333e-05,
+      "loss": 0.6651,
+      "step": 977
+    },
+    {
+      "epoch": 0.06010324483775811,
+      "grad_norm": 62.0,
+      "learning_rate": 1.3026666666666667e-05,
+      "loss": 0.7106,
+      "step": 978
+    },
+    {
+      "epoch": 0.060164700098328416,
+      "grad_norm": 21.5,
+      "learning_rate": 1.3040000000000002e-05,
+      "loss": 0.7053,
+      "step": 979
+    },
+    {
+      "epoch": 0.06022615535889872,
+      "grad_norm": 21.625,
+      "learning_rate": 1.3053333333333333e-05,
+      "loss": 0.7573,
+      "step": 980
+    },
+    {
+      "epoch": 0.06028761061946902,
+      "grad_norm": 17.0,
+      "learning_rate": 1.3066666666666668e-05,
+      "loss": 0.5481,
+      "step": 981
+    },
+    {
+      "epoch": 0.06034906588003933,
+      "grad_norm": 31.0,
+      "learning_rate": 1.3080000000000002e-05,
+      "loss": 0.4952,
+      "step": 982
+    },
+    {
+      "epoch": 0.06041052114060964,
+      "grad_norm": 28.25,
+      "learning_rate": 1.3093333333333334e-05,
+      "loss": 0.5908,
+      "step": 983
+    },
+    {
+      "epoch": 0.06047197640117994,
+      "grad_norm": 11.0,
+      "learning_rate": 1.3106666666666668e-05,
+      "loss": 0.7005,
+      "step": 984
+    },
+    {
+      "epoch": 0.060533431661750244,
+      "grad_norm": 66.5,
+      "learning_rate": 1.3120000000000001e-05,
+      "loss": 0.6408,
+      "step": 985
+    },
+    {
+      "epoch": 0.060594886922320554,
+      "grad_norm": 31.625,
+      "learning_rate": 1.3133333333333334e-05,
+      "loss": 0.7732,
+      "step": 986
+    },
+    {
+      "epoch": 0.06065634218289086,
+      "grad_norm": 38.5,
+      "learning_rate": 1.3146666666666667e-05,
+      "loss": 0.5178,
+      "step": 987
+    },
+    {
+      "epoch": 0.06071779744346116,
+      "grad_norm": 46.75,
+      "learning_rate": 1.3160000000000001e-05,
+      "loss": 0.679,
+      "step": 988
+    },
+    {
+      "epoch": 0.060779252704031464,
+      "grad_norm": 36.5,
+      "learning_rate": 1.3173333333333333e-05,
+      "loss": 0.7539,
+      "step": 989
+    },
+    {
+      "epoch": 0.06084070796460177,
+      "grad_norm": 43.5,
+      "learning_rate": 1.3186666666666667e-05,
+      "loss": 0.5826,
+      "step": 990
+    },
+    {
+      "epoch": 0.06090216322517208,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.3200000000000002e-05,
+      "loss": 0.6359,
+      "step": 991
+    },
+    {
+      "epoch": 0.06096361848574238,
+      "grad_norm": 34.75,
+      "learning_rate": 1.3213333333333333e-05,
+      "loss": 0.6015,
+      "step": 992
+    },
+    {
+      "epoch": 0.061025073746312684,
+      "grad_norm": 20.625,
+      "learning_rate": 1.3226666666666668e-05,
+      "loss": 0.5874,
+      "step": 993
+    },
+    {
+      "epoch": 0.06108652900688299,
+      "grad_norm": 27.25,
+      "learning_rate": 1.3240000000000002e-05,
+      "loss": 0.6563,
+      "step": 994
+    },
+    {
+      "epoch": 0.06114798426745329,
+      "grad_norm": 34.0,
+      "learning_rate": 1.3253333333333334e-05,
+      "loss": 0.6165,
+      "step": 995
+    },
+    {
+      "epoch": 0.0612094395280236,
+      "grad_norm": 22.5,
+      "learning_rate": 1.3266666666666668e-05,
+      "loss": 0.508,
+      "step": 996
+    },
+    {
+      "epoch": 0.061270894788593905,
+      "grad_norm": 32.75,
+      "learning_rate": 1.3280000000000002e-05,
+      "loss": 0.5479,
+      "step": 997
+    },
+    {
+      "epoch": 0.06133235004916421,
+      "grad_norm": 17.375,
+      "learning_rate": 1.3293333333333334e-05,
+      "loss": 0.5173,
+      "step": 998
+    },
+    {
+      "epoch": 0.06139380530973451,
+      "grad_norm": 29.625,
+      "learning_rate": 1.3306666666666667e-05,
+      "loss": 0.5615,
+      "step": 999
+    },
+    {
+      "epoch": 0.061455260570304815,
+      "grad_norm": 51.75,
+      "learning_rate": 1.3320000000000001e-05,
+      "loss": 0.6375,
+      "step": 1000
+    },
+    {
+      "epoch": 0.061516715830875125,
+      "grad_norm": 66.0,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.8289,
+      "step": 1001
+    },
+    {
+      "epoch": 0.06157817109144543,
+      "grad_norm": 55.75,
+      "learning_rate": 1.3346666666666667e-05,
+      "loss": 0.7942,
+      "step": 1002
+    },
+    {
+      "epoch": 0.06163962635201573,
+      "grad_norm": 31.625,
+      "learning_rate": 1.3360000000000003e-05,
+      "loss": 0.7145,
+      "step": 1003
+    },
+    {
+      "epoch": 0.061701081612586035,
+      "grad_norm": 21.875,
+      "learning_rate": 1.3373333333333333e-05,
+      "loss": 0.5583,
+      "step": 1004
+    },
+    {
+      "epoch": 0.06176253687315634,
+      "grad_norm": 30.75,
+      "learning_rate": 1.3386666666666668e-05,
+      "loss": 0.8524,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06182399213372665,
+      "grad_norm": 54.0,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 0.6351,
+      "step": 1006
+    },
+    {
+      "epoch": 0.06188544739429695,
+      "grad_norm": 23.25,
+      "learning_rate": 1.3413333333333334e-05,
+      "loss": 0.4771,
+      "step": 1007
+    },
+    {
+      "epoch": 0.061946902654867256,
+      "grad_norm": 40.25,
+      "learning_rate": 1.3426666666666668e-05,
+      "loss": 0.5642,
+      "step": 1008
+    },
+    {
+      "epoch": 0.06200835791543756,
+      "grad_norm": 29.625,
+      "learning_rate": 1.3440000000000002e-05,
+      "loss": 0.6781,
+      "step": 1009
+    },
+    {
+      "epoch": 0.06206981317600787,
+      "grad_norm": 41.5,
+      "learning_rate": 1.3453333333333334e-05,
+      "loss": 0.7648,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06213126843657817,
+      "grad_norm": 26.625,
+      "learning_rate": 1.3466666666666668e-05,
+      "loss": 0.8041,
+      "step": 1011
+    },
+    {
+      "epoch": 0.062192723697148476,
+      "grad_norm": 19.5,
+      "learning_rate": 1.3480000000000001e-05,
+      "loss": 0.5447,
+      "step": 1012
+    },
+    {
+      "epoch": 0.06225417895771878,
+      "grad_norm": 19.0,
+      "learning_rate": 1.3493333333333333e-05,
+      "loss": 0.5638,
+      "step": 1013
+    },
+    {
+      "epoch": 0.06231563421828908,
+      "grad_norm": 37.0,
+      "learning_rate": 1.3506666666666667e-05,
+      "loss": 0.5458,
+      "step": 1014
+    },
+    {
+      "epoch": 0.06237708947885939,
+      "grad_norm": 51.75,
+      "learning_rate": 1.3520000000000003e-05,
+      "loss": 0.6322,
+      "step": 1015
+    },
+    {
+      "epoch": 0.0624385447394297,
+      "grad_norm": 27.875,
+      "learning_rate": 1.3533333333333333e-05,
+      "loss": 0.6656,
+      "step": 1016
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 30.0,
+      "learning_rate": 1.3546666666666669e-05,
+      "loss": 0.8446,
+      "step": 1017
+    },
+    {
+      "epoch": 0.0625614552605703,
+      "grad_norm": 19.75,
+      "learning_rate": 1.3560000000000002e-05,
+      "loss": 0.777,
+      "step": 1018
+    },
+    {
+      "epoch": 0.0626229105211406,
+      "grad_norm": 46.0,
+      "learning_rate": 1.3573333333333334e-05,
+      "loss": 0.7143,
+      "step": 1019
+    },
+    {
+      "epoch": 0.06268436578171091,
+      "grad_norm": 41.5,
+      "learning_rate": 1.3586666666666668e-05,
+      "loss": 0.7387,
+      "step": 1020
+    },
+    {
+      "epoch": 0.06274582104228121,
+      "grad_norm": 31.875,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 0.7029,
+      "step": 1021
+    },
+    {
+      "epoch": 0.06280727630285153,
+      "grad_norm": 34.5,
+      "learning_rate": 1.3613333333333334e-05,
+      "loss": 0.7499,
+      "step": 1022
+    },
+    {
+      "epoch": 0.06286873156342183,
+      "grad_norm": 23.0,
+      "learning_rate": 1.3626666666666668e-05,
+      "loss": 0.6823,
+      "step": 1023
+    },
+    {
+      "epoch": 0.06293018682399214,
+      "grad_norm": 24.625,
+      "learning_rate": 1.3640000000000002e-05,
+      "loss": 0.6845,
+      "step": 1024
+    },
+    {
+      "epoch": 0.06299164208456244,
+      "grad_norm": 35.0,
+      "learning_rate": 1.3653333333333334e-05,
+      "loss": 0.7603,
+      "step": 1025
+    },
+    {
+      "epoch": 0.06305309734513274,
+      "grad_norm": 28.375,
+      "learning_rate": 1.3666666666666667e-05,
+      "loss": 0.7808,
+      "step": 1026
+    },
+    {
+      "epoch": 0.06311455260570305,
+      "grad_norm": 20.75,
+      "learning_rate": 1.3680000000000003e-05,
+      "loss": 0.6171,
+      "step": 1027
+    },
+    {
+      "epoch": 0.06317600786627335,
+      "grad_norm": 30.125,
+      "learning_rate": 1.3693333333333333e-05,
+      "loss": 0.549,
+      "step": 1028
+    },
+    {
+      "epoch": 0.06323746312684365,
+      "grad_norm": 23.5,
+      "learning_rate": 1.3706666666666669e-05,
+      "loss": 0.7087,
+      "step": 1029
+    },
+    {
+      "epoch": 0.06329891838741396,
+      "grad_norm": 18.875,
+      "learning_rate": 1.3720000000000002e-05,
+      "loss": 0.7751,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06336037364798426,
+      "grad_norm": 18.125,
+      "learning_rate": 1.3733333333333335e-05,
+      "loss": 0.6164,
+      "step": 1031
+    },
+    {
+      "epoch": 0.06342182890855458,
+      "grad_norm": 19.0,
+      "learning_rate": 1.3746666666666668e-05,
+      "loss": 0.5633,
+      "step": 1032
+    },
+    {
+      "epoch": 0.06348328416912488,
+      "grad_norm": 21.75,
+      "learning_rate": 1.376e-05,
+      "loss": 0.649,
+      "step": 1033
+    },
+    {
+      "epoch": 0.06354473942969519,
+      "grad_norm": 24.5,
+      "learning_rate": 1.3773333333333334e-05,
+      "loss": 0.764,
+      "step": 1034
+    },
+    {
+      "epoch": 0.06360619469026549,
+      "grad_norm": 46.5,
+      "learning_rate": 1.3786666666666668e-05,
+      "loss": 0.5924,
+      "step": 1035
+    },
+    {
+      "epoch": 0.06366764995083579,
+      "grad_norm": 32.5,
+      "learning_rate": 1.38e-05,
+      "loss": 0.6562,
+      "step": 1036
+    },
+    {
+      "epoch": 0.0637291052114061,
+      "grad_norm": 43.0,
+      "learning_rate": 1.3813333333333334e-05,
+      "loss": 0.731,
+      "step": 1037
+    },
+    {
+      "epoch": 0.0637905604719764,
+      "grad_norm": 25.375,
+      "learning_rate": 1.3826666666666668e-05,
+      "loss": 0.4474,
+      "step": 1038
+    },
+    {
+      "epoch": 0.0638520157325467,
+      "grad_norm": 29.5,
+      "learning_rate": 1.384e-05,
+      "loss": 0.5407,
+      "step": 1039
+    },
+    {
+      "epoch": 0.063913470993117,
+      "grad_norm": 17.75,
+      "learning_rate": 1.3853333333333333e-05,
+      "loss": 0.562,
+      "step": 1040
+    },
+    {
+      "epoch": 0.06397492625368732,
+      "grad_norm": 73.0,
+      "learning_rate": 1.3866666666666669e-05,
+      "loss": 0.6734,
+      "step": 1041
+    },
+    {
+      "epoch": 0.06403638151425763,
+      "grad_norm": 53.5,
+      "learning_rate": 1.3880000000000001e-05,
+      "loss": 0.9431,
+      "step": 1042
+    },
+    {
+      "epoch": 0.06409783677482793,
+      "grad_norm": 13.625,
+      "learning_rate": 1.3893333333333335e-05,
+      "loss": 0.5406,
+      "step": 1043
+    },
+    {
+      "epoch": 0.06415929203539823,
+      "grad_norm": 60.5,
+      "learning_rate": 1.3906666666666668e-05,
+      "loss": 0.7386,
+      "step": 1044
+    },
+    {
+      "epoch": 0.06422074729596854,
+      "grad_norm": 74.5,
+      "learning_rate": 1.392e-05,
+      "loss": 0.6873,
+      "step": 1045
+    },
+    {
+      "epoch": 0.06428220255653884,
+      "grad_norm": 45.5,
+      "learning_rate": 1.3933333333333334e-05,
+      "loss": 0.5061,
+      "step": 1046
+    },
+    {
+      "epoch": 0.06434365781710914,
+      "grad_norm": 41.5,
+      "learning_rate": 1.3946666666666668e-05,
+      "loss": 0.6824,
+      "step": 1047
+    },
+    {
+      "epoch": 0.06440511307767945,
+      "grad_norm": 34.0,
+      "learning_rate": 1.396e-05,
+      "loss": 0.6991,
+      "step": 1048
+    },
+    {
+      "epoch": 0.06446656833824975,
+      "grad_norm": 25.125,
+      "learning_rate": 1.3973333333333334e-05,
+      "loss": 0.7031,
+      "step": 1049
+    },
+    {
+      "epoch": 0.06452802359882005,
+      "grad_norm": 40.0,
+      "learning_rate": 1.3986666666666668e-05,
+      "loss": 0.7898,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06458947885939037,
+      "grad_norm": 17.25,
+      "learning_rate": 1.4e-05,
+      "loss": 0.6144,
+      "step": 1051
+    },
+    {
+      "epoch": 0.06465093411996067,
+      "grad_norm": 11.25,
+      "learning_rate": 1.4013333333333334e-05,
+      "loss": 0.5525,
+      "step": 1052
+    },
+    {
+      "epoch": 0.06471238938053098,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.4026666666666669e-05,
+      "loss": 0.4986,
+      "step": 1053
+    },
+    {
+      "epoch": 0.06477384464110128,
+      "grad_norm": 26.625,
+      "learning_rate": 1.4040000000000001e-05,
+      "loss": 0.5708,
+      "step": 1054
+    },
+    {
+      "epoch": 0.06483529990167158,
+      "grad_norm": 61.0,
+      "learning_rate": 1.4053333333333335e-05,
+      "loss": 0.5904,
+      "step": 1055
+    },
+    {
+      "epoch": 0.06489675516224189,
+      "grad_norm": 16.125,
+      "learning_rate": 1.4066666666666669e-05,
+      "loss": 0.5787,
+      "step": 1056
+    },
+    {
+      "epoch": 0.06495821042281219,
+      "grad_norm": 32.75,
+      "learning_rate": 1.408e-05,
+      "loss": 0.5654,
+      "step": 1057
+    },
+    {
+      "epoch": 0.0650196656833825,
+      "grad_norm": 40.5,
+      "learning_rate": 1.4093333333333334e-05,
+      "loss": 0.4767,
+      "step": 1058
+    },
+    {
+      "epoch": 0.0650811209439528,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4106666666666668e-05,
+      "loss": 0.5948,
+      "step": 1059
+    },
+    {
+      "epoch": 0.0651425762045231,
+      "grad_norm": 22.5,
+      "learning_rate": 1.412e-05,
+      "loss": 0.5722,
+      "step": 1060
+    },
+    {
+      "epoch": 0.06520403146509342,
+      "grad_norm": 40.0,
+      "learning_rate": 1.4133333333333334e-05,
+      "loss": 0.5792,
+      "step": 1061
+    },
+    {
+      "epoch": 0.06526548672566372,
+      "grad_norm": 21.625,
+      "learning_rate": 1.4146666666666668e-05,
+      "loss": 0.5644,
+      "step": 1062
+    },
+    {
+      "epoch": 0.06532694198623402,
+      "grad_norm": 61.75,
+      "learning_rate": 1.416e-05,
+      "loss": 0.5862,
+      "step": 1063
+    },
+    {
+      "epoch": 0.06538839724680433,
+      "grad_norm": 91.0,
+      "learning_rate": 1.4173333333333334e-05,
+      "loss": 0.6101,
+      "step": 1064
+    },
+    {
+      "epoch": 0.06544985250737463,
+      "grad_norm": 69.5,
+      "learning_rate": 1.418666666666667e-05,
+      "loss": 0.6225,
+      "step": 1065
+    },
+    {
+      "epoch": 0.06551130776794493,
+      "grad_norm": 38.25,
+      "learning_rate": 1.4200000000000001e-05,
+      "loss": 0.7773,
+      "step": 1066
+    },
+    {
+      "epoch": 0.06557276302851524,
+      "grad_norm": 48.0,
+      "learning_rate": 1.4213333333333335e-05,
+      "loss": 0.7213,
+      "step": 1067
+    },
+    {
+      "epoch": 0.06563421828908554,
+      "grad_norm": 47.25,
+      "learning_rate": 1.4226666666666669e-05,
+      "loss": 0.7991,
+      "step": 1068
+    },
+    {
+      "epoch": 0.06569567354965584,
+      "grad_norm": 89.0,
+      "learning_rate": 1.4240000000000001e-05,
+      "loss": 0.7871,
+      "step": 1069
+    },
+    {
+      "epoch": 0.06575712881022616,
+      "grad_norm": 73.5,
+      "learning_rate": 1.4253333333333335e-05,
+      "loss": 0.7182,
+      "step": 1070
+    },
+    {
+      "epoch": 0.06581858407079647,
+      "grad_norm": 41.75,
+      "learning_rate": 1.4266666666666668e-05,
+      "loss": 0.6139,
+      "step": 1071
+    },
+    {
+      "epoch": 0.06588003933136677,
+      "grad_norm": 66.0,
+      "learning_rate": 1.428e-05,
+      "loss": 0.8248,
+      "step": 1072
+    },
+    {
+      "epoch": 0.06594149459193707,
+      "grad_norm": 72.5,
+      "learning_rate": 1.4293333333333334e-05,
+      "loss": 0.8806,
+      "step": 1073
+    },
+    {
+      "epoch": 0.06600294985250738,
+      "grad_norm": 54.75,
+      "learning_rate": 1.4306666666666668e-05,
+      "loss": 1.0585,
+      "step": 1074
+    },
+    {
+      "epoch": 0.06606440511307768,
+      "grad_norm": 24.625,
+      "learning_rate": 1.432e-05,
+      "loss": 0.6862,
+      "step": 1075
+    },
+    {
+      "epoch": 0.06612586037364798,
+      "grad_norm": 35.25,
+      "learning_rate": 1.4333333333333334e-05,
+      "loss": 0.7819,
+      "step": 1076
+    },
+    {
+      "epoch": 0.06618731563421829,
+      "grad_norm": 32.0,
+      "learning_rate": 1.434666666666667e-05,
+      "loss": 0.7898,
+      "step": 1077
+    },
+    {
+      "epoch": 0.06624877089478859,
+      "grad_norm": 29.625,
+      "learning_rate": 1.4360000000000001e-05,
+      "loss": 0.601,
+      "step": 1078
+    },
+    {
+      "epoch": 0.06631022615535889,
+      "grad_norm": 28.5,
+      "learning_rate": 1.4373333333333335e-05,
+      "loss": 0.6504,
+      "step": 1079
+    },
+    {
+      "epoch": 0.06637168141592921,
+      "grad_norm": 30.25,
+      "learning_rate": 1.4386666666666669e-05,
+      "loss": 0.6407,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06643313667649951,
+      "grad_norm": 21.625,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 0.6019,
+      "step": 1081
+    },
+    {
+      "epoch": 0.06649459193706982,
+      "grad_norm": 18.75,
+      "learning_rate": 1.4413333333333335e-05,
+      "loss": 0.7393,
+      "step": 1082
+    },
+    {
+      "epoch": 0.06655604719764012,
+      "grad_norm": 21.75,
+      "learning_rate": 1.4426666666666669e-05,
+      "loss": 0.5901,
+      "step": 1083
+    },
+    {
+      "epoch": 0.06661750245821042,
+      "grad_norm": 26.25,
+      "learning_rate": 1.444e-05,
+      "loss": 0.5716,
+      "step": 1084
+    },
+    {
+      "epoch": 0.06667895771878073,
+      "grad_norm": 20.375,
+      "learning_rate": 1.4453333333333334e-05,
+      "loss": 0.5058,
+      "step": 1085
+    },
+    {
+      "epoch": 0.06674041297935103,
+      "grad_norm": 18.375,
+      "learning_rate": 1.4466666666666668e-05,
+      "loss": 0.6311,
+      "step": 1086
+    },
+    {
+      "epoch": 0.06680186823992133,
+      "grad_norm": 11.25,
+      "learning_rate": 1.448e-05,
+      "loss": 0.4169,
+      "step": 1087
+    },
+    {
+      "epoch": 0.06686332350049164,
+      "grad_norm": 10.625,
+      "learning_rate": 1.4493333333333334e-05,
+      "loss": 0.4954,
+      "step": 1088
+    },
+    {
+      "epoch": 0.06692477876106195,
+      "grad_norm": 32.0,
+      "learning_rate": 1.450666666666667e-05,
+      "loss": 0.7037,
+      "step": 1089
+    },
+    {
+      "epoch": 0.06698623402163226,
+      "grad_norm": 21.375,
+      "learning_rate": 1.4520000000000002e-05,
+      "loss": 0.7401,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06704768928220256,
+      "grad_norm": 7.625,
+      "learning_rate": 1.4533333333333335e-05,
+      "loss": 0.6295,
+      "step": 1091
+    },
+    {
+      "epoch": 0.06710914454277286,
+      "grad_norm": 56.0,
+      "learning_rate": 1.4546666666666669e-05,
+      "loss": 0.637,
+      "step": 1092
+    },
+    {
+      "epoch": 0.06717059980334317,
+      "grad_norm": 56.75,
+      "learning_rate": 1.4560000000000001e-05,
+      "loss": 0.6713,
+      "step": 1093
+    },
+    {
+      "epoch": 0.06723205506391347,
+      "grad_norm": 31.125,
+      "learning_rate": 1.4573333333333335e-05,
+      "loss": 0.4808,
+      "step": 1094
+    },
+    {
+      "epoch": 0.06729351032448377,
+      "grad_norm": 16.5,
+      "learning_rate": 1.4586666666666667e-05,
+      "loss": 0.3933,
+      "step": 1095
+    },
+    {
+      "epoch": 0.06735496558505408,
+      "grad_norm": 36.0,
+      "learning_rate": 1.46e-05,
+      "loss": 0.4986,
+      "step": 1096
+    },
+    {
+      "epoch": 0.06741642084562438,
+      "grad_norm": 50.75,
+      "learning_rate": 1.4613333333333335e-05,
+      "loss": 0.5888,
+      "step": 1097
+    },
+    {
+      "epoch": 0.06747787610619468,
+      "grad_norm": 91.0,
+      "learning_rate": 1.4626666666666667e-05,
+      "loss": 0.7183,
+      "step": 1098
+    },
+    {
+      "epoch": 0.067539331366765,
+      "grad_norm": 64.5,
+      "learning_rate": 1.464e-05,
+      "loss": 0.7257,
+      "step": 1099
+    },
+    {
+      "epoch": 0.0676007866273353,
+      "grad_norm": 48.75,
+      "learning_rate": 1.4653333333333334e-05,
+      "loss": 0.8161,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06766224188790561,
+      "grad_norm": 34.25,
+      "learning_rate": 1.4666666666666666e-05,
+      "loss": 0.7636,
+      "step": 1101
+    },
+    {
+      "epoch": 0.06772369714847591,
+      "grad_norm": 45.75,
+      "learning_rate": 1.4680000000000002e-05,
+      "loss": 0.6096,
+      "step": 1102
+    },
+    {
+      "epoch": 0.06778515240904621,
+      "grad_norm": 38.75,
+      "learning_rate": 1.4693333333333336e-05,
+      "loss": 0.7142,
+      "step": 1103
+    },
+    {
+      "epoch": 0.06784660766961652,
+      "grad_norm": 34.0,
+      "learning_rate": 1.4706666666666668e-05,
+      "loss": 0.6443,
+      "step": 1104
+    },
+    {
+      "epoch": 0.06790806293018682,
+      "grad_norm": 51.5,
+      "learning_rate": 1.4720000000000001e-05,
+      "loss": 0.7447,
+      "step": 1105
+    },
+    {
+      "epoch": 0.06796951819075712,
+      "grad_norm": 37.25,
+      "learning_rate": 1.4733333333333335e-05,
+      "loss": 0.8298,
+      "step": 1106
+    },
+    {
+      "epoch": 0.06803097345132743,
+      "grad_norm": 40.75,
+      "learning_rate": 1.4746666666666667e-05,
+      "loss": 0.7638,
+      "step": 1107
+    },
+    {
+      "epoch": 0.06809242871189773,
+      "grad_norm": 58.75,
+      "learning_rate": 1.4760000000000001e-05,
+      "loss": 0.7746,
+      "step": 1108
+    },
+    {
+      "epoch": 0.06815388397246805,
+      "grad_norm": 43.5,
+      "learning_rate": 1.4773333333333335e-05,
+      "loss": 0.9173,
+      "step": 1109
+    },
+    {
+      "epoch": 0.06821533923303835,
+      "grad_norm": 32.0,
+      "learning_rate": 1.4786666666666667e-05,
+      "loss": 0.7881,
+      "step": 1110
+    },
+    {
+      "epoch": 0.06827679449360866,
+      "grad_norm": 35.5,
+      "learning_rate": 1.48e-05,
+      "loss": 0.735,
+      "step": 1111
+    },
+    {
+      "epoch": 0.06833824975417896,
+      "grad_norm": 21.25,
+      "learning_rate": 1.4813333333333334e-05,
+      "loss": 0.6729,
+      "step": 1112
+    },
+    {
+      "epoch": 0.06839970501474926,
+      "grad_norm": 20.5,
+      "learning_rate": 1.4826666666666666e-05,
+      "loss": 1.0147,
+      "step": 1113
+    },
+    {
+      "epoch": 0.06846116027531957,
+      "grad_norm": 30.125,
+      "learning_rate": 1.4840000000000002e-05,
+      "loss": 0.9717,
+      "step": 1114
+    },
+    {
+      "epoch": 0.06852261553588987,
+      "grad_norm": 36.5,
+      "learning_rate": 1.4853333333333336e-05,
+      "loss": 0.7154,
+      "step": 1115
+    },
+    {
+      "epoch": 0.06858407079646017,
+      "grad_norm": 44.25,
+      "learning_rate": 1.4866666666666668e-05,
+      "loss": 1.0762,
+      "step": 1116
+    },
+    {
+      "epoch": 0.06864552605703048,
+      "grad_norm": 42.5,
+      "learning_rate": 1.4880000000000002e-05,
+      "loss": 0.8782,
+      "step": 1117
+    },
+    {
+      "epoch": 0.0687069813176008,
+      "grad_norm": 39.75,
+      "learning_rate": 1.4893333333333335e-05,
+      "loss": 0.9667,
+      "step": 1118
+    },
+    {
+      "epoch": 0.0687684365781711,
+      "grad_norm": 32.75,
+      "learning_rate": 1.4906666666666667e-05,
+      "loss": 0.8376,
+      "step": 1119
+    },
+    {
+      "epoch": 0.0688298918387414,
+      "grad_norm": 38.25,
+      "learning_rate": 1.4920000000000001e-05,
+      "loss": 0.8856,
+      "step": 1120
+    },
+    {
+      "epoch": 0.0688913470993117,
+      "grad_norm": 40.5,
+      "learning_rate": 1.4933333333333335e-05,
+      "loss": 0.7576,
+      "step": 1121
+    },
+    {
+      "epoch": 0.068952802359882,
+      "grad_norm": 31.5,
+      "learning_rate": 1.4946666666666667e-05,
+      "loss": 0.5704,
+      "step": 1122
+    },
+    {
+      "epoch": 0.06901425762045231,
+      "grad_norm": 29.625,
+      "learning_rate": 1.496e-05,
+      "loss": 0.6238,
+      "step": 1123
+    },
+    {
+      "epoch": 0.06907571288102261,
+      "grad_norm": 28.375,
+      "learning_rate": 1.4973333333333335e-05,
+      "loss": 0.6668,
+      "step": 1124
+    },
+    {
+      "epoch": 0.06913716814159292,
+      "grad_norm": 35.5,
+      "learning_rate": 1.4986666666666667e-05,
+      "loss": 0.8984,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06919862340216322,
+      "grad_norm": 27.125,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.5605,
+      "step": 1126
+    },
+    {
+      "epoch": 0.06926007866273352,
+      "grad_norm": 33.75,
+      "learning_rate": 1.5013333333333336e-05,
+      "loss": 0.6231,
+      "step": 1127
+    },
+    {
+      "epoch": 0.06932153392330384,
+      "grad_norm": 27.5,
+      "learning_rate": 1.5026666666666668e-05,
+      "loss": 0.5572,
+      "step": 1128
+    },
+    {
+      "epoch": 0.06938298918387414,
+      "grad_norm": 41.0,
+      "learning_rate": 1.5040000000000002e-05,
+      "loss": 0.938,
+      "step": 1129
+    },
+    {
+      "epoch": 0.06944444444444445,
+      "grad_norm": 30.125,
+      "learning_rate": 1.5053333333333335e-05,
+      "loss": 0.6742,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06950589970501475,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.5066666666666668e-05,
+      "loss": 0.657,
+      "step": 1131
+    },
+    {
+      "epoch": 0.06956735496558505,
+      "grad_norm": 46.0,
+      "learning_rate": 1.5080000000000001e-05,
+      "loss": 0.7191,
+      "step": 1132
+    },
+    {
+      "epoch": 0.06962881022615536,
+      "grad_norm": 25.625,
+      "learning_rate": 1.5093333333333335e-05,
+      "loss": 0.7605,
+      "step": 1133
+    },
+    {
+      "epoch": 0.06969026548672566,
+      "grad_norm": 38.0,
+      "learning_rate": 1.5106666666666667e-05,
+      "loss": 0.6979,
+      "step": 1134
+    },
+    {
+      "epoch": 0.06975172074729596,
+      "grad_norm": 34.25,
+      "learning_rate": 1.5120000000000001e-05,
+      "loss": 0.7091,
+      "step": 1135
+    },
+    {
+      "epoch": 0.06981317600786627,
+      "grad_norm": 32.25,
+      "learning_rate": 1.5133333333333335e-05,
+      "loss": 0.5476,
+      "step": 1136
+    },
+    {
+      "epoch": 0.06987463126843659,
+      "grad_norm": 26.375,
+      "learning_rate": 1.5146666666666667e-05,
+      "loss": 0.6273,
+      "step": 1137
+    },
+    {
+      "epoch": 0.06993608652900689,
+      "grad_norm": 30.875,
+      "learning_rate": 1.516e-05,
+      "loss": 0.8992,
+      "step": 1138
+    },
+    {
+      "epoch": 0.06999754178957719,
+      "grad_norm": 30.25,
+      "learning_rate": 1.5173333333333336e-05,
+      "loss": 0.87,
+      "step": 1139
+    },
+    {
+      "epoch": 0.0700589970501475,
+      "grad_norm": 53.0,
+      "learning_rate": 1.5186666666666668e-05,
+      "loss": 0.5385,
+      "step": 1140
+    },
+    {
+      "epoch": 0.0701204523107178,
+      "grad_norm": 55.25,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 0.6953,
+      "step": 1141
+    },
+    {
+      "epoch": 0.0701819075712881,
+      "grad_norm": 21.25,
+      "learning_rate": 1.5213333333333336e-05,
+      "loss": 0.4616,
+      "step": 1142
+    },
+    {
+      "epoch": 0.0702433628318584,
+      "grad_norm": 32.25,
+      "learning_rate": 1.5226666666666668e-05,
+      "loss": 0.6059,
+      "step": 1143
+    },
+    {
+      "epoch": 0.07030481809242871,
+      "grad_norm": 36.25,
+      "learning_rate": 1.5240000000000001e-05,
+      "loss": 0.6232,
+      "step": 1144
+    },
+    {
+      "epoch": 0.07036627335299901,
+      "grad_norm": 51.5,
+      "learning_rate": 1.5253333333333335e-05,
+      "loss": 0.6747,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07042772861356932,
+      "grad_norm": 23.625,
+      "learning_rate": 1.5266666666666667e-05,
+      "loss": 0.6227,
+      "step": 1146
+    },
+    {
+      "epoch": 0.07048918387413963,
+      "grad_norm": 19.375,
+      "learning_rate": 1.5280000000000003e-05,
+      "loss": 0.6565,
+      "step": 1147
+    },
+    {
+      "epoch": 0.07055063913470994,
+      "grad_norm": 48.5,
+      "learning_rate": 1.5293333333333335e-05,
+      "loss": 0.5863,
+      "step": 1148
+    },
+    {
+      "epoch": 0.07061209439528024,
+      "grad_norm": 37.25,
+      "learning_rate": 1.5306666666666667e-05,
+      "loss": 0.5731,
+      "step": 1149
+    },
+    {
+      "epoch": 0.07067354965585054,
+      "grad_norm": 48.0,
+      "learning_rate": 1.5320000000000002e-05,
+      "loss": 0.6599,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07073500491642085,
+      "grad_norm": 42.0,
+      "learning_rate": 1.5333333333333334e-05,
+      "loss": 0.6357,
+      "step": 1151
+    },
+    {
+      "epoch": 0.07079646017699115,
+      "grad_norm": 35.25,
+      "learning_rate": 1.5346666666666667e-05,
+      "loss": 0.7379,
+      "step": 1152
+    },
+    {
+      "epoch": 0.07085791543756145,
+      "grad_norm": 23.125,
+      "learning_rate": 1.5360000000000002e-05,
+      "loss": 0.7713,
+      "step": 1153
+    },
+    {
+      "epoch": 0.07091937069813176,
+      "grad_norm": 35.75,
+      "learning_rate": 1.5373333333333334e-05,
+      "loss": 0.7325,
+      "step": 1154
+    },
+    {
+      "epoch": 0.07098082595870206,
+      "grad_norm": 35.0,
+      "learning_rate": 1.5386666666666666e-05,
+      "loss": 0.7437,
+      "step": 1155
+    },
+    {
+      "epoch": 0.07104228121927236,
+      "grad_norm": 42.5,
+      "learning_rate": 1.54e-05,
+      "loss": 0.7071,
+      "step": 1156
+    },
+    {
+      "epoch": 0.07110373647984268,
+      "grad_norm": 45.0,
+      "learning_rate": 1.5413333333333337e-05,
+      "loss": 0.6112,
+      "step": 1157
+    },
+    {
+      "epoch": 0.07116519174041298,
+      "grad_norm": 54.0,
+      "learning_rate": 1.542666666666667e-05,
+      "loss": 0.685,
+      "step": 1158
+    },
+    {
+      "epoch": 0.07122664700098329,
+      "grad_norm": 43.0,
+      "learning_rate": 1.544e-05,
+      "loss": 0.6106,
+      "step": 1159
+    },
+    {
+      "epoch": 0.07128810226155359,
+      "grad_norm": 26.75,
+      "learning_rate": 1.5453333333333333e-05,
+      "loss": 0.534,
+      "step": 1160
+    },
+    {
+      "epoch": 0.0713495575221239,
+      "grad_norm": 46.25,
+      "learning_rate": 1.546666666666667e-05,
+      "loss": 0.6219,
+      "step": 1161
+    },
+    {
+      "epoch": 0.0714110127826942,
+      "grad_norm": 16.625,
+      "learning_rate": 1.548e-05,
+      "loss": 0.4795,
+      "step": 1162
+    },
+    {
+      "epoch": 0.0714724680432645,
+      "grad_norm": 58.5,
+      "learning_rate": 1.5493333333333333e-05,
+      "loss": 0.6535,
+      "step": 1163
+    },
+    {
+      "epoch": 0.0715339233038348,
+      "grad_norm": 21.0,
+      "learning_rate": 1.550666666666667e-05,
+      "loss": 0.6997,
+      "step": 1164
+    },
+    {
+      "epoch": 0.07159537856440511,
+      "grad_norm": 29.25,
+      "learning_rate": 1.552e-05,
+      "loss": 0.6972,
+      "step": 1165
+    },
+    {
+      "epoch": 0.07165683382497542,
+      "grad_norm": 32.0,
+      "learning_rate": 1.5533333333333333e-05,
+      "loss": 0.7933,
+      "step": 1166
+    },
+    {
+      "epoch": 0.07171828908554573,
+      "grad_norm": 12.875,
+      "learning_rate": 1.5546666666666668e-05,
+      "loss": 0.6228,
+      "step": 1167
+    },
+    {
+      "epoch": 0.07177974434611603,
+      "grad_norm": 36.0,
+      "learning_rate": 1.556e-05,
+      "loss": 0.5528,
+      "step": 1168
+    },
+    {
+      "epoch": 0.07184119960668633,
+      "grad_norm": 20.375,
+      "learning_rate": 1.5573333333333332e-05,
+      "loss": 0.6039,
+      "step": 1169
+    },
+    {
+      "epoch": 0.07190265486725664,
+      "grad_norm": 22.625,
+      "learning_rate": 1.5586666666666668e-05,
+      "loss": 0.5609,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07196411012782694,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.5600000000000003e-05,
+      "loss": 0.4003,
+      "step": 1171
+    },
+    {
+      "epoch": 0.07202556538839724,
+      "grad_norm": 25.0,
+      "learning_rate": 1.5613333333333335e-05,
+      "loss": 0.3726,
+      "step": 1172
+    },
+    {
+      "epoch": 0.07208702064896755,
+      "grad_norm": 46.25,
+      "learning_rate": 1.5626666666666667e-05,
+      "loss": 0.462,
+      "step": 1173
+    },
+    {
+      "epoch": 0.07214847590953785,
+      "grad_norm": 50.25,
+      "learning_rate": 1.5640000000000003e-05,
+      "loss": 0.4438,
+      "step": 1174
+    },
+    {
+      "epoch": 0.07220993117010815,
+      "grad_norm": 16.875,
+      "learning_rate": 1.5653333333333335e-05,
+      "loss": 0.4034,
+      "step": 1175
+    },
+    {
+      "epoch": 0.07227138643067847,
+      "grad_norm": 27.125,
+      "learning_rate": 1.5666666666666667e-05,
+      "loss": 0.4843,
+      "step": 1176
+    },
+    {
+      "epoch": 0.07233284169124878,
+      "grad_norm": 60.5,
+      "learning_rate": 1.5680000000000002e-05,
+      "loss": 0.6719,
+      "step": 1177
+    },
+    {
+      "epoch": 0.07239429695181908,
+      "grad_norm": 34.5,
+      "learning_rate": 1.5693333333333334e-05,
+      "loss": 0.7294,
+      "step": 1178
+    },
+    {
+      "epoch": 0.07245575221238938,
+      "grad_norm": 30.625,
+      "learning_rate": 1.5706666666666666e-05,
+      "loss": 0.5936,
+      "step": 1179
+    },
+    {
+      "epoch": 0.07251720747295969,
+      "grad_norm": 68.0,
+      "learning_rate": 1.5720000000000002e-05,
+      "loss": 0.67,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07257866273352999,
+      "grad_norm": 38.25,
+      "learning_rate": 1.5733333333333334e-05,
+      "loss": 0.5604,
+      "step": 1181
+    },
+    {
+      "epoch": 0.07264011799410029,
+      "grad_norm": 67.0,
+      "learning_rate": 1.574666666666667e-05,
+      "loss": 0.5524,
+      "step": 1182
+    },
+    {
+      "epoch": 0.0727015732546706,
+      "grad_norm": 67.0,
+      "learning_rate": 1.576e-05,
+      "loss": 0.4872,
+      "step": 1183
+    },
+    {
+      "epoch": 0.0727630285152409,
+      "grad_norm": 26.875,
+      "learning_rate": 1.5773333333333334e-05,
+      "loss": 0.495,
+      "step": 1184
+    },
+    {
+      "epoch": 0.07282448377581122,
+      "grad_norm": 72.0,
+      "learning_rate": 1.578666666666667e-05,
+      "loss": 0.8087,
+      "step": 1185
+    },
+    {
+      "epoch": 0.07288593903638152,
+      "grad_norm": 40.5,
+      "learning_rate": 1.58e-05,
+      "loss": 0.8484,
+      "step": 1186
+    },
+    {
+      "epoch": 0.07294739429695182,
+      "grad_norm": 30.75,
+      "learning_rate": 1.5813333333333333e-05,
+      "loss": 0.8589,
+      "step": 1187
+    },
+    {
+      "epoch": 0.07300884955752213,
+      "grad_norm": 39.75,
+      "learning_rate": 1.582666666666667e-05,
+      "loss": 0.7581,
+      "step": 1188
+    },
+    {
+      "epoch": 0.07307030481809243,
+      "grad_norm": 23.375,
+      "learning_rate": 1.584e-05,
+      "loss": 0.729,
+      "step": 1189
+    },
+    {
+      "epoch": 0.07313176007866273,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.5853333333333333e-05,
+      "loss": 0.6175,
+      "step": 1190
+    },
+    {
+      "epoch": 0.07319321533923304,
+      "grad_norm": 37.5,
+      "learning_rate": 1.586666666666667e-05,
+      "loss": 0.5987,
+      "step": 1191
+    },
+    {
+      "epoch": 0.07325467059980334,
+      "grad_norm": 49.0,
+      "learning_rate": 1.588e-05,
+      "loss": 0.7068,
+      "step": 1192
+    },
+    {
+      "epoch": 0.07331612586037364,
+      "grad_norm": 42.5,
+      "learning_rate": 1.5893333333333333e-05,
+      "loss": 0.8008,
+      "step": 1193
+    },
+    {
+      "epoch": 0.07337758112094395,
+      "grad_norm": 50.25,
+      "learning_rate": 1.5906666666666668e-05,
+      "loss": 0.7017,
+      "step": 1194
+    },
+    {
+      "epoch": 0.07343903638151426,
+      "grad_norm": 28.5,
+      "learning_rate": 1.5920000000000003e-05,
+      "loss": 0.6605,
+      "step": 1195
+    },
+    {
+      "epoch": 0.07350049164208457,
+      "grad_norm": 21.875,
+      "learning_rate": 1.5933333333333336e-05,
+      "loss": 0.527,
+      "step": 1196
+    },
+    {
+      "epoch": 0.07356194690265487,
+      "grad_norm": 24.25,
+      "learning_rate": 1.5946666666666668e-05,
+      "loss": 0.5601,
+      "step": 1197
+    },
+    {
+      "epoch": 0.07362340216322517,
+      "grad_norm": 42.0,
+      "learning_rate": 1.5960000000000003e-05,
+      "loss": 0.5686,
+      "step": 1198
+    },
+    {
+      "epoch": 0.07368485742379548,
+      "grad_norm": 26.125,
+      "learning_rate": 1.5973333333333335e-05,
+      "loss": 0.5352,
+      "step": 1199
+    },
+    {
+      "epoch": 0.07374631268436578,
+      "grad_norm": 12.25,
+      "learning_rate": 1.5986666666666667e-05,
+      "loss": 0.4315,
+      "step": 1200
+    },
+    {
+      "epoch": 0.07380776794493608,
+      "grad_norm": 26.0,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.7356,
+      "step": 1201
+    },
+    {
+      "epoch": 0.07386922320550639,
+      "grad_norm": 33.75,
+      "learning_rate": 1.6013333333333335e-05,
+      "loss": 0.6635,
+      "step": 1202
+    },
+    {
+      "epoch": 0.07393067846607669,
+      "grad_norm": 19.25,
+      "learning_rate": 1.6026666666666667e-05,
+      "loss": 0.6158,
+      "step": 1203
+    },
+    {
+      "epoch": 0.073992133726647,
+      "grad_norm": 28.0,
+      "learning_rate": 1.6040000000000002e-05,
+      "loss": 0.7113,
+      "step": 1204
+    },
+    {
+      "epoch": 0.07405358898721731,
+      "grad_norm": 26.75,
+      "learning_rate": 1.6053333333333334e-05,
+      "loss": 0.6945,
+      "step": 1205
+    },
+    {
+      "epoch": 0.07411504424778761,
+      "grad_norm": 17.75,
+      "learning_rate": 1.606666666666667e-05,
+      "loss": 0.5829,
+      "step": 1206
+    },
+    {
+      "epoch": 0.07417649950835792,
+      "grad_norm": 23.75,
+      "learning_rate": 1.6080000000000002e-05,
+      "loss": 0.6567,
+      "step": 1207
+    },
+    {
+      "epoch": 0.07423795476892822,
+      "grad_norm": 21.75,
+      "learning_rate": 1.6093333333333334e-05,
+      "loss": 0.5667,
+      "step": 1208
+    },
+    {
+      "epoch": 0.07429941002949852,
+      "grad_norm": 40.75,
+      "learning_rate": 1.610666666666667e-05,
+      "loss": 1.02,
+      "step": 1209
+    },
+    {
+      "epoch": 0.07436086529006883,
+      "grad_norm": 32.25,
+      "learning_rate": 1.612e-05,
+      "loss": 0.8284,
+      "step": 1210
+    },
+    {
+      "epoch": 0.07442232055063913,
+      "grad_norm": 16.375,
+      "learning_rate": 1.6133333333333334e-05,
+      "loss": 0.6401,
+      "step": 1211
+    },
+    {
+      "epoch": 0.07448377581120944,
+      "grad_norm": 21.125,
+      "learning_rate": 1.614666666666667e-05,
+      "loss": 0.6497,
+      "step": 1212
+    },
+    {
+      "epoch": 0.07454523107177974,
+      "grad_norm": 18.0,
+      "learning_rate": 1.616e-05,
+      "loss": 0.6779,
+      "step": 1213
+    },
+    {
+      "epoch": 0.07460668633235006,
+      "grad_norm": 38.0,
+      "learning_rate": 1.6173333333333333e-05,
+      "loss": 0.944,
+      "step": 1214
+    },
+    {
+      "epoch": 0.07466814159292036,
+      "grad_norm": 19.0,
+      "learning_rate": 1.618666666666667e-05,
+      "loss": 0.7312,
+      "step": 1215
+    },
+    {
+      "epoch": 0.07472959685349066,
+      "grad_norm": 13.25,
+      "learning_rate": 1.62e-05,
+      "loss": 0.4905,
+      "step": 1216
+    },
+    {
+      "epoch": 0.07479105211406097,
+      "grad_norm": 41.5,
+      "learning_rate": 1.6213333333333333e-05,
+      "loss": 0.8054,
+      "step": 1217
+    },
+    {
+      "epoch": 0.07485250737463127,
+      "grad_norm": 29.625,
+      "learning_rate": 1.6226666666666668e-05,
+      "loss": 0.6629,
+      "step": 1218
+    },
+    {
+      "epoch": 0.07491396263520157,
+      "grad_norm": 16.375,
+      "learning_rate": 1.6240000000000004e-05,
+      "loss": 0.661,
+      "step": 1219
+    },
+    {
+      "epoch": 0.07497541789577188,
+      "grad_norm": 33.25,
+      "learning_rate": 1.6253333333333336e-05,
+      "loss": 0.5849,
+      "step": 1220
+    },
+    {
+      "epoch": 0.07503687315634218,
+      "grad_norm": 54.75,
+      "learning_rate": 1.6266666666666668e-05,
+      "loss": 0.7616,
+      "step": 1221
+    },
+    {
+      "epoch": 0.07509832841691248,
+      "grad_norm": 38.25,
+      "learning_rate": 1.628e-05,
+      "loss": 0.6739,
+      "step": 1222
+    },
+    {
+      "epoch": 0.07515978367748279,
+      "grad_norm": 39.5,
+      "learning_rate": 1.6293333333333335e-05,
+      "loss": 0.7668,
+      "step": 1223
+    },
+    {
+      "epoch": 0.0752212389380531,
+      "grad_norm": 31.125,
+      "learning_rate": 1.6306666666666668e-05,
+      "loss": 0.4427,
+      "step": 1224
+    },
+    {
+      "epoch": 0.0752826941986234,
+      "grad_norm": 25.125,
+      "learning_rate": 1.632e-05,
+      "loss": 0.5461,
+      "step": 1225
+    },
+    {
+      "epoch": 0.07534414945919371,
+      "grad_norm": 29.25,
+      "learning_rate": 1.6333333333333335e-05,
+      "loss": 0.7374,
+      "step": 1226
+    },
+    {
+      "epoch": 0.07540560471976401,
+      "grad_norm": 17.75,
+      "learning_rate": 1.6346666666666667e-05,
+      "loss": 0.5886,
+      "step": 1227
+    },
+    {
+      "epoch": 0.07546705998033432,
+      "grad_norm": 23.25,
+      "learning_rate": 1.636e-05,
+      "loss": 0.54,
+      "step": 1228
+    },
+    {
+      "epoch": 0.07552851524090462,
+      "grad_norm": 18.625,
+      "learning_rate": 1.6373333333333335e-05,
+      "loss": 0.5391,
+      "step": 1229
+    },
+    {
+      "epoch": 0.07558997050147492,
+      "grad_norm": 18.5,
+      "learning_rate": 1.638666666666667e-05,
+      "loss": 0.569,
+      "step": 1230
+    },
+    {
+      "epoch": 0.07565142576204523,
+      "grad_norm": 25.375,
+      "learning_rate": 1.64e-05,
+      "loss": 0.5415,
+      "step": 1231
+    },
+    {
+      "epoch": 0.07571288102261553,
+      "grad_norm": 21.125,
+      "learning_rate": 1.6413333333333334e-05,
+      "loss": 0.5971,
+      "step": 1232
+    },
+    {
+      "epoch": 0.07577433628318585,
+      "grad_norm": 33.0,
+      "learning_rate": 1.642666666666667e-05,
+      "loss": 0.7882,
+      "step": 1233
+    },
+    {
+      "epoch": 0.07583579154375615,
+      "grad_norm": 35.25,
+      "learning_rate": 1.6440000000000002e-05,
+      "loss": 1.0296,
+      "step": 1234
+    },
+    {
+      "epoch": 0.07589724680432645,
+      "grad_norm": 29.0,
+      "learning_rate": 1.6453333333333334e-05,
+      "loss": 1.0611,
+      "step": 1235
+    },
+    {
+      "epoch": 0.07595870206489676,
+      "grad_norm": 35.0,
+      "learning_rate": 1.646666666666667e-05,
+      "loss": 0.8196,
+      "step": 1236
+    },
+    {
+      "epoch": 0.07602015732546706,
+      "grad_norm": 57.25,
+      "learning_rate": 1.648e-05,
+      "loss": 0.9906,
+      "step": 1237
+    },
+    {
+      "epoch": 0.07608161258603736,
+      "grad_norm": 17.125,
+      "learning_rate": 1.6493333333333334e-05,
+      "loss": 0.72,
+      "step": 1238
+    },
+    {
+      "epoch": 0.07614306784660767,
+      "grad_norm": 19.5,
+      "learning_rate": 1.650666666666667e-05,
+      "loss": 0.6491,
+      "step": 1239
+    },
+    {
+      "epoch": 0.07620452310717797,
+      "grad_norm": 18.875,
+      "learning_rate": 1.652e-05,
+      "loss": 0.6304,
+      "step": 1240
+    },
+    {
+      "epoch": 0.07626597836774827,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.6533333333333333e-05,
+      "loss": 0.7159,
+      "step": 1241
+    },
+    {
+      "epoch": 0.07632743362831858,
+      "grad_norm": 22.125,
+      "learning_rate": 1.654666666666667e-05,
+      "loss": 0.6239,
+      "step": 1242
+    },
+    {
+      "epoch": 0.0763888888888889,
+      "grad_norm": 24.375,
+      "learning_rate": 1.656e-05,
+      "loss": 0.7342,
+      "step": 1243
+    },
+    {
+      "epoch": 0.0764503441494592,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.6573333333333336e-05,
+      "loss": 0.5269,
+      "step": 1244
+    },
+    {
+      "epoch": 0.0765117994100295,
+      "grad_norm": 18.25,
+      "learning_rate": 1.6586666666666668e-05,
+      "loss": 0.4487,
+      "step": 1245
+    },
+    {
+      "epoch": 0.0765732546705998,
+      "grad_norm": 10.5,
+      "learning_rate": 1.66e-05,
+      "loss": 0.5805,
+      "step": 1246
+    },
+    {
+      "epoch": 0.07663470993117011,
+      "grad_norm": 27.625,
+      "learning_rate": 1.6613333333333336e-05,
+      "loss": 0.551,
+      "step": 1247
+    },
+    {
+      "epoch": 0.07669616519174041,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.6626666666666668e-05,
+      "loss": 0.3734,
+      "step": 1248
+    },
+    {
+      "epoch": 0.07675762045231072,
+      "grad_norm": 27.0,
+      "learning_rate": 1.664e-05,
+      "loss": 0.5799,
+      "step": 1249
+    },
+    {
+      "epoch": 0.07681907571288102,
+      "grad_norm": 25.0,
+      "learning_rate": 1.6653333333333335e-05,
+      "loss": 0.4928,
+      "step": 1250
+    },
+    {
+      "epoch": 0.07688053097345132,
+      "grad_norm": 39.0,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.5519,
+      "step": 1251
+    },
+    {
+      "epoch": 0.07694198623402163,
+      "grad_norm": 28.125,
+      "learning_rate": 1.668e-05,
+      "loss": 0.5392,
+      "step": 1252
+    },
+    {
+      "epoch": 0.07700344149459194,
+      "grad_norm": 20.375,
+      "learning_rate": 1.6693333333333335e-05,
+      "loss": 0.5647,
+      "step": 1253
+    },
+    {
+      "epoch": 0.07706489675516225,
+      "grad_norm": 40.25,
+      "learning_rate": 1.670666666666667e-05,
+      "loss": 0.6763,
+      "step": 1254
+    },
+    {
+      "epoch": 0.07712635201573255,
+      "grad_norm": 34.25,
+      "learning_rate": 1.672e-05,
+      "loss": 0.4502,
+      "step": 1255
+    },
+    {
+      "epoch": 0.07718780727630285,
+      "grad_norm": 8.75,
+      "learning_rate": 1.6733333333333335e-05,
+      "loss": 0.5004,
+      "step": 1256
+    },
+    {
+      "epoch": 0.07724926253687316,
+      "grad_norm": 94.5,
+      "learning_rate": 1.674666666666667e-05,
+      "loss": 0.9472,
+      "step": 1257
+    },
+    {
+      "epoch": 0.07731071779744346,
+      "grad_norm": 74.0,
+      "learning_rate": 1.6760000000000002e-05,
+      "loss": 0.8513,
+      "step": 1258
+    },
+    {
+      "epoch": 0.07737217305801376,
+      "grad_norm": 34.75,
+      "learning_rate": 1.6773333333333334e-05,
+      "loss": 0.8647,
+      "step": 1259
+    },
+    {
+      "epoch": 0.07743362831858407,
+      "grad_norm": 37.5,
+      "learning_rate": 1.678666666666667e-05,
+      "loss": 0.7916,
+      "step": 1260
+    },
+    {
+      "epoch": 0.07749508357915437,
+      "grad_norm": 35.0,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 0.7195,
+      "step": 1261
+    },
+    {
+      "epoch": 0.07755653883972469,
+      "grad_norm": 45.25,
+      "learning_rate": 1.6813333333333334e-05,
+      "loss": 0.8831,
+      "step": 1262
+    },
+    {
+      "epoch": 0.07761799410029499,
+      "grad_norm": 35.25,
+      "learning_rate": 1.682666666666667e-05,
+      "loss": 0.5772,
+      "step": 1263
+    },
+    {
+      "epoch": 0.0776794493608653,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.684e-05,
+      "loss": 0.5639,
+      "step": 1264
+    },
+    {
+      "epoch": 0.0777409046214356,
+      "grad_norm": 39.0,
+      "learning_rate": 1.6853333333333333e-05,
+      "loss": 0.6101,
+      "step": 1265
+    },
+    {
+      "epoch": 0.0778023598820059,
+      "grad_norm": 20.75,
+      "learning_rate": 1.686666666666667e-05,
+      "loss": 0.8422,
+      "step": 1266
+    },
+    {
+      "epoch": 0.0778638151425762,
+      "grad_norm": 36.25,
+      "learning_rate": 1.688e-05,
+      "loss": 0.7107,
+      "step": 1267
+    },
+    {
+      "epoch": 0.07792527040314651,
+      "grad_norm": 32.0,
+      "learning_rate": 1.6893333333333336e-05,
+      "loss": 0.6155,
+      "step": 1268
+    },
+    {
+      "epoch": 0.07798672566371681,
+      "grad_norm": 36.75,
+      "learning_rate": 1.690666666666667e-05,
+      "loss": 0.6383,
+      "step": 1269
+    },
+    {
+      "epoch": 0.07804818092428711,
+      "grad_norm": 29.125,
+      "learning_rate": 1.692e-05,
+      "loss": 0.6434,
+      "step": 1270
+    },
+    {
+      "epoch": 0.07810963618485742,
+      "grad_norm": 27.125,
+      "learning_rate": 1.6933333333333336e-05,
+      "loss": 0.4912,
+      "step": 1271
+    },
+    {
+      "epoch": 0.07817109144542773,
+      "grad_norm": 37.0,
+      "learning_rate": 1.6946666666666668e-05,
+      "loss": 0.5515,
+      "step": 1272
+    },
+    {
+      "epoch": 0.07823254670599804,
+      "grad_norm": 42.0,
+      "learning_rate": 1.696e-05,
+      "loss": 0.768,
+      "step": 1273
+    },
+    {
+      "epoch": 0.07829400196656834,
+      "grad_norm": 22.625,
+      "learning_rate": 1.6973333333333336e-05,
+      "loss": 0.8686,
+      "step": 1274
+    },
+    {
+      "epoch": 0.07835545722713864,
+      "grad_norm": 31.75,
+      "learning_rate": 1.6986666666666668e-05,
+      "loss": 0.8938,
+      "step": 1275
+    },
+    {
+      "epoch": 0.07841691248770895,
+      "grad_norm": 26.875,
+      "learning_rate": 1.7e-05,
+      "loss": 0.9088,
+      "step": 1276
+    },
+    {
+      "epoch": 0.07847836774827925,
+      "grad_norm": 34.5,
+      "learning_rate": 1.7013333333333335e-05,
+      "loss": 0.6139,
+      "step": 1277
+    },
+    {
+      "epoch": 0.07853982300884955,
+      "grad_norm": 17.375,
+      "learning_rate": 1.702666666666667e-05,
+      "loss": 0.4993,
+      "step": 1278
+    },
+    {
+      "epoch": 0.07860127826941986,
+      "grad_norm": 15.625,
+      "learning_rate": 1.704e-05,
+      "loss": 0.7102,
+      "step": 1279
+    },
+    {
+      "epoch": 0.07866273352999016,
+      "grad_norm": 11.5625,
+      "learning_rate": 1.7053333333333335e-05,
+      "loss": 0.6416,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07872418879056048,
+      "grad_norm": 25.625,
+      "learning_rate": 1.706666666666667e-05,
+      "loss": 0.707,
+      "step": 1281
+    },
+    {
+      "epoch": 0.07878564405113078,
+      "grad_norm": 29.0,
+      "learning_rate": 1.7080000000000002e-05,
+      "loss": 0.8199,
+      "step": 1282
+    },
+    {
+      "epoch": 0.07884709931170109,
+      "grad_norm": 22.5,
+      "learning_rate": 1.7093333333333335e-05,
+      "loss": 0.7795,
+      "step": 1283
+    },
+    {
+      "epoch": 0.07890855457227139,
+      "grad_norm": 40.25,
+      "learning_rate": 1.7106666666666667e-05,
+      "loss": 0.5938,
+      "step": 1284
+    },
+    {
+      "epoch": 0.07897000983284169,
+      "grad_norm": 18.375,
+      "learning_rate": 1.7120000000000002e-05,
+      "loss": 0.6192,
+      "step": 1285
+    },
+    {
+      "epoch": 0.079031465093412,
+      "grad_norm": 34.75,
+      "learning_rate": 1.7133333333333334e-05,
+      "loss": 0.7148,
+      "step": 1286
+    },
+    {
+      "epoch": 0.0790929203539823,
+      "grad_norm": 10.375,
+      "learning_rate": 1.7146666666666666e-05,
+      "loss": 0.6218,
+      "step": 1287
+    },
+    {
+      "epoch": 0.0791543756145526,
+      "grad_norm": 33.75,
+      "learning_rate": 1.7160000000000002e-05,
+      "loss": 0.5906,
+      "step": 1288
+    },
+    {
+      "epoch": 0.0792158308751229,
+      "grad_norm": 23.75,
+      "learning_rate": 1.7173333333333334e-05,
+      "loss": 0.803,
+      "step": 1289
+    },
+    {
+      "epoch": 0.07927728613569321,
+      "grad_norm": 27.75,
+      "learning_rate": 1.7186666666666666e-05,
+      "loss": 0.5048,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07933874139626353,
+      "grad_norm": 28.875,
+      "learning_rate": 1.72e-05,
+      "loss": 0.6709,
+      "step": 1291
+    },
+    {
+      "epoch": 0.07940019665683383,
+      "grad_norm": 24.5,
+      "learning_rate": 1.7213333333333337e-05,
+      "loss": 0.6672,
+      "step": 1292
+    },
+    {
+      "epoch": 0.07946165191740413,
+      "grad_norm": 23.5,
+      "learning_rate": 1.7226666666666665e-05,
+      "loss": 0.4858,
+      "step": 1293
+    },
+    {
+      "epoch": 0.07952310717797444,
+      "grad_norm": 18.875,
+      "learning_rate": 1.724e-05,
+      "loss": 0.5971,
+      "step": 1294
+    },
+    {
+      "epoch": 0.07958456243854474,
+      "grad_norm": 16.25,
+      "learning_rate": 1.7253333333333336e-05,
+      "loss": 0.5369,
+      "step": 1295
+    },
+    {
+      "epoch": 0.07964601769911504,
+      "grad_norm": 28.375,
+      "learning_rate": 1.726666666666667e-05,
+      "loss": 0.4849,
+      "step": 1296
+    },
+    {
+      "epoch": 0.07970747295968535,
+      "grad_norm": 29.25,
+      "learning_rate": 1.728e-05,
+      "loss": 0.5118,
+      "step": 1297
+    },
+    {
+      "epoch": 0.07976892822025565,
+      "grad_norm": 11.25,
+      "learning_rate": 1.7293333333333336e-05,
+      "loss": 0.4618,
+      "step": 1298
+    },
+    {
+      "epoch": 0.07983038348082595,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.7306666666666668e-05,
+      "loss": 0.5113,
+      "step": 1299
+    },
+    {
+      "epoch": 0.07989183874139626,
+      "grad_norm": 31.875,
+      "learning_rate": 1.732e-05,
+      "loss": 0.4626,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07995329400196657,
+      "grad_norm": 24.5,
+      "learning_rate": 1.7333333333333336e-05,
+      "loss": 0.5453,
+      "step": 1301
+    },
+    {
+      "epoch": 0.08001474926253688,
+      "grad_norm": 26.375,
+      "learning_rate": 1.7346666666666668e-05,
+      "loss": 0.4727,
+      "step": 1302
+    },
+    {
+      "epoch": 0.08007620452310718,
+      "grad_norm": 17.25,
+      "learning_rate": 1.736e-05,
+      "loss": 0.5201,
+      "step": 1303
+    },
+    {
+      "epoch": 0.08013765978367748,
+      "grad_norm": 10.25,
+      "learning_rate": 1.7373333333333335e-05,
+      "loss": 0.3878,
+      "step": 1304
+    },
+    {
+      "epoch": 0.08019911504424779,
+      "grad_norm": 54.5,
+      "learning_rate": 1.7386666666666667e-05,
+      "loss": 0.6364,
+      "step": 1305
+    },
+    {
+      "epoch": 0.08026057030481809,
+      "grad_norm": 18.5,
+      "learning_rate": 1.7400000000000003e-05,
+      "loss": 0.6689,
+      "step": 1306
+    },
+    {
+      "epoch": 0.0803220255653884,
+      "grad_norm": 10.6875,
+      "learning_rate": 1.7413333333333335e-05,
+      "loss": 0.5948,
+      "step": 1307
+    },
+    {
+      "epoch": 0.0803834808259587,
+      "grad_norm": 16.25,
+      "learning_rate": 1.7426666666666667e-05,
+      "loss": 0.7083,
+      "step": 1308
+    },
+    {
+      "epoch": 0.080444936086529,
+      "grad_norm": 36.5,
+      "learning_rate": 1.7440000000000002e-05,
+      "loss": 0.6281,
+      "step": 1309
+    },
+    {
+      "epoch": 0.08050639134709932,
+      "grad_norm": 26.25,
+      "learning_rate": 1.7453333333333335e-05,
+      "loss": 0.5585,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08056784660766962,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.7466666666666667e-05,
+      "loss": 0.386,
+      "step": 1311
+    },
+    {
+      "epoch": 0.08062930186823992,
+      "grad_norm": 24.125,
+      "learning_rate": 1.7480000000000002e-05,
+      "loss": 0.3992,
+      "step": 1312
+    },
+    {
+      "epoch": 0.08069075712881023,
+      "grad_norm": 45.0,
+      "learning_rate": 1.7493333333333334e-05,
+      "loss": 0.9013,
+      "step": 1313
+    },
+    {
+      "epoch": 0.08075221238938053,
+      "grad_norm": 29.875,
+      "learning_rate": 1.7506666666666666e-05,
+      "loss": 1.0279,
+      "step": 1314
+    },
+    {
+      "epoch": 0.08081366764995083,
+      "grad_norm": 21.125,
+      "learning_rate": 1.752e-05,
+      "loss": 0.7779,
+      "step": 1315
+    },
+    {
+      "epoch": 0.08087512291052114,
+      "grad_norm": 16.375,
+      "learning_rate": 1.7533333333333337e-05,
+      "loss": 0.6813,
+      "step": 1316
+    },
+    {
+      "epoch": 0.08093657817109144,
+      "grad_norm": 44.25,
+      "learning_rate": 1.7546666666666666e-05,
+      "loss": 0.9382,
+      "step": 1317
+    },
+    {
+      "epoch": 0.08099803343166175,
+      "grad_norm": 36.0,
+      "learning_rate": 1.756e-05,
+      "loss": 0.9156,
+      "step": 1318
+    },
+    {
+      "epoch": 0.08105948869223205,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.7573333333333337e-05,
+      "loss": 0.746,
+      "step": 1319
+    },
+    {
+      "epoch": 0.08112094395280237,
+      "grad_norm": 12.5,
+      "learning_rate": 1.758666666666667e-05,
+      "loss": 0.5663,
+      "step": 1320
+    },
+    {
+      "epoch": 0.08118239921337267,
+      "grad_norm": 18.625,
+      "learning_rate": 1.76e-05,
+      "loss": 0.6666,
+      "step": 1321
+    },
+    {
+      "epoch": 0.08124385447394297,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.7613333333333336e-05,
+      "loss": 0.5154,
+      "step": 1322
+    },
+    {
+      "epoch": 0.08130530973451328,
+      "grad_norm": 32.75,
+      "learning_rate": 1.762666666666667e-05,
+      "loss": 0.6638,
+      "step": 1323
+    },
+    {
+      "epoch": 0.08136676499508358,
+      "grad_norm": 15.625,
+      "learning_rate": 1.764e-05,
+      "loss": 0.5732,
+      "step": 1324
+    },
+    {
+      "epoch": 0.08142822025565388,
+      "grad_norm": 20.5,
+      "learning_rate": 1.7653333333333336e-05,
+      "loss": 0.5806,
+      "step": 1325
+    },
+    {
+      "epoch": 0.08148967551622419,
+      "grad_norm": 32.75,
+      "learning_rate": 1.7666666666666668e-05,
+      "loss": 0.6079,
+      "step": 1326
+    },
+    {
+      "epoch": 0.08155113077679449,
+      "grad_norm": 43.0,
+      "learning_rate": 1.768e-05,
+      "loss": 0.6406,
+      "step": 1327
+    },
+    {
+      "epoch": 0.08161258603736479,
+      "grad_norm": 25.5,
+      "learning_rate": 1.7693333333333336e-05,
+      "loss": 0.6069,
+      "step": 1328
+    },
+    {
+      "epoch": 0.08167404129793511,
+      "grad_norm": 26.25,
+      "learning_rate": 1.7706666666666668e-05,
+      "loss": 0.7652,
+      "step": 1329
+    },
+    {
+      "epoch": 0.08173549655850541,
+      "grad_norm": 21.125,
+      "learning_rate": 1.7720000000000003e-05,
+      "loss": 0.7136,
+      "step": 1330
+    },
+    {
+      "epoch": 0.08179695181907572,
+      "grad_norm": 46.0,
+      "learning_rate": 1.7733333333333335e-05,
+      "loss": 0.7269,
+      "step": 1331
+    },
+    {
+      "epoch": 0.08185840707964602,
+      "grad_norm": 58.5,
+      "learning_rate": 1.7746666666666667e-05,
+      "loss": 0.8524,
+      "step": 1332
+    },
+    {
+      "epoch": 0.08191986234021632,
+      "grad_norm": 19.375,
+      "learning_rate": 1.7760000000000003e-05,
+      "loss": 0.8192,
+      "step": 1333
+    },
+    {
+      "epoch": 0.08198131760078663,
+      "grad_norm": 35.25,
+      "learning_rate": 1.7773333333333335e-05,
+      "loss": 0.7183,
+      "step": 1334
+    },
+    {
+      "epoch": 0.08204277286135693,
+      "grad_norm": 24.5,
+      "learning_rate": 1.7786666666666667e-05,
+      "loss": 0.6622,
+      "step": 1335
+    },
+    {
+      "epoch": 0.08210422812192723,
+      "grad_norm": 27.375,
+      "learning_rate": 1.7800000000000002e-05,
+      "loss": 0.7104,
+      "step": 1336
+    },
+    {
+      "epoch": 0.08216568338249754,
+      "grad_norm": 39.0,
+      "learning_rate": 1.7813333333333334e-05,
+      "loss": 0.5864,
+      "step": 1337
+    },
+    {
+      "epoch": 0.08222713864306784,
+      "grad_norm": 21.875,
+      "learning_rate": 1.7826666666666667e-05,
+      "loss": 0.7185,
+      "step": 1338
+    },
+    {
+      "epoch": 0.08228859390363816,
+      "grad_norm": 22.5,
+      "learning_rate": 1.7840000000000002e-05,
+      "loss": 0.4773,
+      "step": 1339
+    },
+    {
+      "epoch": 0.08235004916420846,
+      "grad_norm": 31.375,
+      "learning_rate": 1.7853333333333337e-05,
+      "loss": 0.6331,
+      "step": 1340
+    },
+    {
+      "epoch": 0.08241150442477876,
+      "grad_norm": 25.75,
+      "learning_rate": 1.7866666666666666e-05,
+      "loss": 0.5024,
+      "step": 1341
+    },
+    {
+      "epoch": 0.08247295968534907,
+      "grad_norm": 24.125,
+      "learning_rate": 1.788e-05,
+      "loss": 0.5157,
+      "step": 1342
+    },
+    {
+      "epoch": 0.08253441494591937,
+      "grad_norm": 24.25,
+      "learning_rate": 1.7893333333333337e-05,
+      "loss": 0.6028,
+      "step": 1343
+    },
+    {
+      "epoch": 0.08259587020648967,
+      "grad_norm": 21.875,
+      "learning_rate": 1.790666666666667e-05,
+      "loss": 0.3299,
+      "step": 1344
+    },
+    {
+      "epoch": 0.08265732546705998,
+      "grad_norm": 25.5,
+      "learning_rate": 1.792e-05,
+      "loss": 0.4917,
+      "step": 1345
+    },
+    {
+      "epoch": 0.08271878072763028,
+      "grad_norm": 19.375,
+      "learning_rate": 1.7933333333333333e-05,
+      "loss": 0.6454,
+      "step": 1346
+    },
+    {
+      "epoch": 0.08278023598820058,
+      "grad_norm": 15.5,
+      "learning_rate": 1.794666666666667e-05,
+      "loss": 0.6313,
+      "step": 1347
+    },
+    {
+      "epoch": 0.08284169124877089,
+      "grad_norm": 27.25,
+      "learning_rate": 1.796e-05,
+      "loss": 0.7153,
+      "step": 1348
+    },
+    {
+      "epoch": 0.0829031465093412,
+      "grad_norm": 30.625,
+      "learning_rate": 1.7973333333333333e-05,
+      "loss": 0.5216,
+      "step": 1349
+    },
+    {
+      "epoch": 0.08296460176991151,
+      "grad_norm": 33.75,
+      "learning_rate": 1.798666666666667e-05,
+      "loss": 0.497,
+      "step": 1350
+    },
+    {
+      "epoch": 0.08302605703048181,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8e-05,
+      "loss": 0.4383,
+      "step": 1351
+    },
+    {
+      "epoch": 0.08308751229105212,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.8013333333333333e-05,
+      "loss": 0.5304,
+      "step": 1352
+    },
+    {
+      "epoch": 0.08314896755162242,
+      "grad_norm": 30.125,
+      "learning_rate": 1.8026666666666668e-05,
+      "loss": 0.542,
+      "step": 1353
+    },
+    {
+      "epoch": 0.08321042281219272,
+      "grad_norm": 24.5,
+      "learning_rate": 1.8040000000000003e-05,
+      "loss": 0.6674,
+      "step": 1354
+    },
+    {
+      "epoch": 0.08327187807276303,
+      "grad_norm": 10.0,
+      "learning_rate": 1.8053333333333332e-05,
+      "loss": 0.6858,
+      "step": 1355
+    },
+    {
+      "epoch": 0.08333333333333333,
+      "grad_norm": 16.125,
+      "learning_rate": 1.8066666666666668e-05,
+      "loss": 0.6673,
+      "step": 1356
+    },
+    {
+      "epoch": 0.08339478859390363,
+      "grad_norm": 9.75,
+      "learning_rate": 1.8080000000000003e-05,
+      "loss": 0.4523,
+      "step": 1357
+    },
+    {
+      "epoch": 0.08345624385447395,
+      "grad_norm": 52.25,
+      "learning_rate": 1.8093333333333335e-05,
+      "loss": 0.7677,
+      "step": 1358
+    },
+    {
+      "epoch": 0.08351769911504425,
+      "grad_norm": 33.0,
+      "learning_rate": 1.8106666666666667e-05,
+      "loss": 0.4808,
+      "step": 1359
+    },
+    {
+      "epoch": 0.08357915437561456,
+      "grad_norm": 51.75,
+      "learning_rate": 1.8120000000000003e-05,
+      "loss": 0.8363,
+      "step": 1360
+    },
+    {
+      "epoch": 0.08364060963618486,
+      "grad_norm": 33.0,
+      "learning_rate": 1.8133333333333335e-05,
+      "loss": 0.636,
+      "step": 1361
+    },
+    {
+      "epoch": 0.08370206489675516,
+      "grad_norm": 11.5625,
+      "learning_rate": 1.8146666666666667e-05,
+      "loss": 0.6303,
+      "step": 1362
+    },
+    {
+      "epoch": 0.08376352015732547,
+      "grad_norm": 21.25,
+      "learning_rate": 1.8160000000000002e-05,
+      "loss": 0.6055,
+      "step": 1363
+    },
+    {
+      "epoch": 0.08382497541789577,
+      "grad_norm": 43.0,
+      "learning_rate": 1.8173333333333334e-05,
+      "loss": 0.6292,
+      "step": 1364
+    },
+    {
+      "epoch": 0.08388643067846607,
+      "grad_norm": 29.375,
+      "learning_rate": 1.8186666666666666e-05,
+      "loss": 0.6603,
+      "step": 1365
+    },
+    {
+      "epoch": 0.08394788593903638,
+      "grad_norm": 29.75,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 0.5737,
+      "step": 1366
+    },
+    {
+      "epoch": 0.08400934119960668,
+      "grad_norm": 50.75,
+      "learning_rate": 1.8213333333333334e-05,
+      "loss": 0.7035,
+      "step": 1367
+    },
+    {
+      "epoch": 0.084070796460177,
+      "grad_norm": 76.5,
+      "learning_rate": 1.822666666666667e-05,
+      "loss": 0.5051,
+      "step": 1368
+    },
+    {
+      "epoch": 0.0841322517207473,
+      "grad_norm": 30.375,
+      "learning_rate": 1.824e-05,
+      "loss": 0.5556,
+      "step": 1369
+    },
+    {
+      "epoch": 0.0841937069813176,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.8253333333333334e-05,
+      "loss": 0.4911,
+      "step": 1370
+    },
+    {
+      "epoch": 0.08425516224188791,
+      "grad_norm": 20.375,
+      "learning_rate": 1.826666666666667e-05,
+      "loss": 0.6677,
+      "step": 1371
+    },
+    {
+      "epoch": 0.08431661750245821,
+      "grad_norm": 31.5,
+      "learning_rate": 1.828e-05,
+      "loss": 0.5995,
+      "step": 1372
+    },
+    {
+      "epoch": 0.08437807276302851,
+      "grad_norm": 28.625,
+      "learning_rate": 1.8293333333333333e-05,
+      "loss": 0.5916,
+      "step": 1373
+    },
+    {
+      "epoch": 0.08443952802359882,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.830666666666667e-05,
+      "loss": 0.3366,
+      "step": 1374
+    },
+    {
+      "epoch": 0.08450098328416912,
+      "grad_norm": 20.0,
+      "learning_rate": 1.832e-05,
+      "loss": 0.6325,
+      "step": 1375
+    },
+    {
+      "epoch": 0.08456243854473942,
+      "grad_norm": 54.75,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 0.4333,
+      "step": 1376
+    },
+    {
+      "epoch": 0.08462389380530974,
+      "grad_norm": 37.5,
+      "learning_rate": 1.834666666666667e-05,
+      "loss": 0.8487,
+      "step": 1377
+    },
+    {
+      "epoch": 0.08468534906588004,
+      "grad_norm": 39.5,
+      "learning_rate": 1.8360000000000004e-05,
+      "loss": 1.3229,
+      "step": 1378
+    },
+    {
+      "epoch": 0.08474680432645035,
+      "grad_norm": 26.25,
+      "learning_rate": 1.8373333333333332e-05,
+      "loss": 0.9511,
+      "step": 1379
+    },
+    {
+      "epoch": 0.08480825958702065,
+      "grad_norm": 27.125,
+      "learning_rate": 1.8386666666666668e-05,
+      "loss": 0.9488,
+      "step": 1380
+    },
+    {
+      "epoch": 0.08486971484759095,
+      "grad_norm": 20.25,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 0.89,
+      "step": 1381
+    },
+    {
+      "epoch": 0.08493117010816126,
+      "grad_norm": 15.4375,
+      "learning_rate": 1.8413333333333335e-05,
+      "loss": 0.7851,
+      "step": 1382
+    },
+    {
+      "epoch": 0.08499262536873156,
+      "grad_norm": 11.4375,
+      "learning_rate": 1.8426666666666668e-05,
+      "loss": 0.5351,
+      "step": 1383
+    },
+    {
+      "epoch": 0.08505408062930186,
+      "grad_norm": 16.0,
+      "learning_rate": 1.8440000000000003e-05,
+      "loss": 0.7284,
+      "step": 1384
+    },
+    {
+      "epoch": 0.08511553588987217,
+      "grad_norm": 20.875,
+      "learning_rate": 1.8453333333333335e-05,
+      "loss": 0.6495,
+      "step": 1385
+    },
+    {
+      "epoch": 0.08517699115044247,
+      "grad_norm": 16.25,
+      "learning_rate": 1.8466666666666667e-05,
+      "loss": 0.7219,
+      "step": 1386
+    },
+    {
+      "epoch": 0.08523844641101279,
+      "grad_norm": 16.375,
+      "learning_rate": 1.8480000000000003e-05,
+      "loss": 0.795,
+      "step": 1387
+    },
+    {
+      "epoch": 0.08529990167158309,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.8493333333333335e-05,
+      "loss": 0.5516,
+      "step": 1388
+    },
+    {
+      "epoch": 0.0853613569321534,
+      "grad_norm": 57.25,
+      "learning_rate": 1.8506666666666667e-05,
+      "loss": 0.6733,
+      "step": 1389
+    },
+    {
+      "epoch": 0.0854228121927237,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.8520000000000002e-05,
+      "loss": 0.5545,
+      "step": 1390
+    },
+    {
+      "epoch": 0.085484267453294,
+      "grad_norm": 24.875,
+      "learning_rate": 1.8533333333333334e-05,
+      "loss": 0.5134,
+      "step": 1391
+    },
+    {
+      "epoch": 0.0855457227138643,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.854666666666667e-05,
+      "loss": 0.4192,
+      "step": 1392
+    },
+    {
+      "epoch": 0.08560717797443461,
+      "grad_norm": 49.75,
+      "learning_rate": 1.8560000000000002e-05,
+      "loss": 0.6455,
+      "step": 1393
+    },
+    {
+      "epoch": 0.08566863323500491,
+      "grad_norm": 49.75,
+      "learning_rate": 1.8573333333333334e-05,
+      "loss": 0.5704,
+      "step": 1394
+    },
+    {
+      "epoch": 0.08573008849557522,
+      "grad_norm": 24.875,
+      "learning_rate": 1.858666666666667e-05,
+      "loss": 0.6552,
+      "step": 1395
+    },
+    {
+      "epoch": 0.08579154375614553,
+      "grad_norm": 28.125,
+      "learning_rate": 1.86e-05,
+      "loss": 0.5102,
+      "step": 1396
+    },
+    {
+      "epoch": 0.08585299901671584,
+      "grad_norm": 34.75,
+      "learning_rate": 1.8613333333333334e-05,
+      "loss": 0.8148,
+      "step": 1397
+    },
+    {
+      "epoch": 0.08591445427728614,
+      "grad_norm": 23.5,
+      "learning_rate": 1.862666666666667e-05,
+      "loss": 0.7509,
+      "step": 1398
+    },
+    {
+      "epoch": 0.08597590953785644,
+      "grad_norm": 36.0,
+      "learning_rate": 1.864e-05,
+      "loss": 0.5747,
+      "step": 1399
+    },
+    {
+      "epoch": 0.08603736479842675,
+      "grad_norm": 8.625,
+      "learning_rate": 1.8653333333333333e-05,
+      "loss": 0.4487,
+      "step": 1400
+    },
+    {
+      "epoch": 0.08609882005899705,
+      "grad_norm": 25.375,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 0.7071,
+      "step": 1401
+    },
+    {
+      "epoch": 0.08616027531956735,
+      "grad_norm": 83.0,
+      "learning_rate": 1.8680000000000004e-05,
+      "loss": 0.825,
+      "step": 1402
+    },
+    {
+      "epoch": 0.08622173058013766,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.8693333333333333e-05,
+      "loss": 0.5407,
+      "step": 1403
+    },
+    {
+      "epoch": 0.08628318584070796,
+      "grad_norm": 8.125,
+      "learning_rate": 1.8706666666666668e-05,
+      "loss": 0.5043,
+      "step": 1404
+    },
+    {
+      "epoch": 0.08634464110127826,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8720000000000004e-05,
+      "loss": 0.5296,
+      "step": 1405
+    },
+    {
+      "epoch": 0.08640609636184858,
+      "grad_norm": 19.75,
+      "learning_rate": 1.8733333333333336e-05,
+      "loss": 0.4216,
+      "step": 1406
+    },
+    {
+      "epoch": 0.08646755162241888,
+      "grad_norm": 5.875,
+      "learning_rate": 1.8746666666666668e-05,
+      "loss": 0.3793,
+      "step": 1407
+    },
+    {
+      "epoch": 0.08652900688298919,
+      "grad_norm": 24.625,
+      "learning_rate": 1.876e-05,
+      "loss": 0.5236,
+      "step": 1408
+    },
+    {
+      "epoch": 0.08659046214355949,
+      "grad_norm": 19.25,
+      "learning_rate": 1.8773333333333335e-05,
+      "loss": 0.5596,
+      "step": 1409
+    },
+    {
+      "epoch": 0.0866519174041298,
+      "grad_norm": 16.25,
+      "learning_rate": 1.8786666666666667e-05,
+      "loss": 0.4285,
+      "step": 1410
+    },
+    {
+      "epoch": 0.0867133726647001,
+      "grad_norm": 19.125,
+      "learning_rate": 1.88e-05,
+      "loss": 0.6017,
+      "step": 1411
+    },
+    {
+      "epoch": 0.0867748279252704,
+      "grad_norm": 15.125,
+      "learning_rate": 1.8813333333333335e-05,
+      "loss": 0.5549,
+      "step": 1412
+    },
+    {
+      "epoch": 0.0868362831858407,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.8826666666666667e-05,
+      "loss": 0.5747,
+      "step": 1413
+    },
+    {
+      "epoch": 0.08689773844641101,
+      "grad_norm": 27.875,
+      "learning_rate": 1.884e-05,
+      "loss": 0.503,
+      "step": 1414
+    },
+    {
+      "epoch": 0.08695919370698131,
+      "grad_norm": 34.5,
+      "learning_rate": 1.8853333333333335e-05,
+      "loss": 0.4358,
+      "step": 1415
+    },
+    {
+      "epoch": 0.08702064896755163,
+      "grad_norm": 40.75,
+      "learning_rate": 1.886666666666667e-05,
+      "loss": 0.5821,
+      "step": 1416
+    },
+    {
+      "epoch": 0.08708210422812193,
+      "grad_norm": 21.875,
+      "learning_rate": 1.8880000000000002e-05,
+      "loss": 0.3984,
+      "step": 1417
+    },
+    {
+      "epoch": 0.08714355948869223,
+      "grad_norm": 26.125,
+      "learning_rate": 1.8893333333333334e-05,
+      "loss": 0.4793,
+      "step": 1418
+    },
+    {
+      "epoch": 0.08720501474926254,
+      "grad_norm": 26.0,
+      "learning_rate": 1.890666666666667e-05,
+      "loss": 0.4882,
+      "step": 1419
+    },
+    {
+      "epoch": 0.08726647000983284,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8920000000000002e-05,
+      "loss": 0.7466,
+      "step": 1420
+    },
+    {
+      "epoch": 0.08732792527040315,
+      "grad_norm": 25.875,
+      "learning_rate": 1.8933333333333334e-05,
+      "loss": 0.4873,
+      "step": 1421
+    },
+    {
+      "epoch": 0.08738938053097345,
+      "grad_norm": 23.5,
+      "learning_rate": 1.894666666666667e-05,
+      "loss": 0.4732,
+      "step": 1422
+    },
+    {
+      "epoch": 0.08745083579154375,
+      "grad_norm": 49.5,
+      "learning_rate": 1.896e-05,
+      "loss": 0.5643,
+      "step": 1423
+    },
+    {
+      "epoch": 0.08751229105211406,
+      "grad_norm": 19.625,
+      "learning_rate": 1.8973333333333334e-05,
+      "loss": 0.4972,
+      "step": 1424
+    },
+    {
+      "epoch": 0.08757374631268437,
+      "grad_norm": 43.5,
+      "learning_rate": 1.898666666666667e-05,
+      "loss": 0.8297,
+      "step": 1425
+    },
+    {
+      "epoch": 0.08763520157325468,
+      "grad_norm": 44.0,
+      "learning_rate": 1.9e-05,
+      "loss": 0.8204,
+      "step": 1426
+    },
+    {
+      "epoch": 0.08769665683382498,
+      "grad_norm": 34.25,
+      "learning_rate": 1.9013333333333333e-05,
+      "loss": 0.9316,
+      "step": 1427
+    },
+    {
+      "epoch": 0.08775811209439528,
+      "grad_norm": 18.125,
+      "learning_rate": 1.902666666666667e-05,
+      "loss": 0.6756,
+      "step": 1428
+    },
+    {
+      "epoch": 0.08781956735496559,
+      "grad_norm": 37.5,
+      "learning_rate": 1.904e-05,
+      "loss": 0.7311,
+      "step": 1429
+    },
+    {
+      "epoch": 0.08788102261553589,
+      "grad_norm": 42.75,
+      "learning_rate": 1.9053333333333336e-05,
+      "loss": 0.6094,
+      "step": 1430
+    },
+    {
+      "epoch": 0.08794247787610619,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9066666666666668e-05,
+      "loss": 0.5781,
+      "step": 1431
+    },
+    {
+      "epoch": 0.0880039331366765,
+      "grad_norm": 12.875,
+      "learning_rate": 1.908e-05,
+      "loss": 0.6201,
+      "step": 1432
+    },
+    {
+      "epoch": 0.0880653883972468,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9093333333333336e-05,
+      "loss": 0.7821,
+      "step": 1433
+    },
+    {
+      "epoch": 0.0881268436578171,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.9106666666666668e-05,
+      "loss": 0.5977,
+      "step": 1434
+    },
+    {
+      "epoch": 0.08818829891838742,
+      "grad_norm": 36.75,
+      "learning_rate": 1.912e-05,
+      "loss": 0.6369,
+      "step": 1435
+    },
+    {
+      "epoch": 0.08824975417895772,
+      "grad_norm": 44.0,
+      "learning_rate": 1.9133333333333335e-05,
+      "loss": 0.7049,
+      "step": 1436
+    },
+    {
+      "epoch": 0.08831120943952803,
+      "grad_norm": 30.625,
+      "learning_rate": 1.9146666666666667e-05,
+      "loss": 0.5768,
+      "step": 1437
+    },
+    {
+      "epoch": 0.08837266470009833,
+      "grad_norm": 55.25,
+      "learning_rate": 1.916e-05,
+      "loss": 0.7788,
+      "step": 1438
+    },
+    {
+      "epoch": 0.08843411996066863,
+      "grad_norm": 27.25,
+      "learning_rate": 1.9173333333333335e-05,
+      "loss": 0.668,
+      "step": 1439
+    },
+    {
+      "epoch": 0.08849557522123894,
+      "grad_norm": 28.25,
+      "learning_rate": 1.918666666666667e-05,
+      "loss": 0.5766,
+      "step": 1440
+    },
+    {
+      "epoch": 0.08855703048180924,
+      "grad_norm": 63.0,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 0.6212,
+      "step": 1441
+    },
+    {
+      "epoch": 0.08861848574237954,
+      "grad_norm": 40.75,
+      "learning_rate": 1.9213333333333335e-05,
+      "loss": 0.6886,
+      "step": 1442
+    },
+    {
+      "epoch": 0.08867994100294985,
+      "grad_norm": 50.75,
+      "learning_rate": 1.922666666666667e-05,
+      "loss": 0.7232,
+      "step": 1443
+    },
+    {
+      "epoch": 0.08874139626352016,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.9240000000000002e-05,
+      "loss": 0.4958,
+      "step": 1444
+    },
+    {
+      "epoch": 0.08880285152409047,
+      "grad_norm": 31.5,
+      "learning_rate": 1.9253333333333334e-05,
+      "loss": 0.4745,
+      "step": 1445
+    },
+    {
+      "epoch": 0.08886430678466077,
+      "grad_norm": 18.25,
+      "learning_rate": 1.926666666666667e-05,
+      "loss": 0.4712,
+      "step": 1446
+    },
+    {
+      "epoch": 0.08892576204523107,
+      "grad_norm": 14.375,
+      "learning_rate": 1.9280000000000002e-05,
+      "loss": 0.3748,
+      "step": 1447
+    },
+    {
+      "epoch": 0.08898721730580138,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.9293333333333334e-05,
+      "loss": 0.3747,
+      "step": 1448
+    },
+    {
+      "epoch": 0.08904867256637168,
+      "grad_norm": 41.5,
+      "learning_rate": 1.930666666666667e-05,
+      "loss": 0.7149,
+      "step": 1449
+    },
+    {
+      "epoch": 0.08911012782694198,
+      "grad_norm": 17.375,
+      "learning_rate": 1.932e-05,
+      "loss": 0.4438,
+      "step": 1450
+    },
+    {
+      "epoch": 0.08917158308751229,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.9333333333333333e-05,
+      "loss": 0.4511,
+      "step": 1451
+    },
+    {
+      "epoch": 0.08923303834808259,
+      "grad_norm": 54.0,
+      "learning_rate": 1.934666666666667e-05,
+      "loss": 0.4975,
+      "step": 1452
+    },
+    {
+      "epoch": 0.0892944936086529,
+      "grad_norm": 34.75,
+      "learning_rate": 1.936e-05,
+      "loss": 0.5387,
+      "step": 1453
+    },
+    {
+      "epoch": 0.08935594886922321,
+      "grad_norm": 55.5,
+      "learning_rate": 1.9373333333333336e-05,
+      "loss": 0.5473,
+      "step": 1454
+    },
+    {
+      "epoch": 0.08941740412979352,
+      "grad_norm": 36.0,
+      "learning_rate": 1.938666666666667e-05,
+      "loss": 0.5324,
+      "step": 1455
+    },
+    {
+      "epoch": 0.08947885939036382,
+      "grad_norm": 41.5,
+      "learning_rate": 1.94e-05,
+      "loss": 0.4644,
+      "step": 1456
+    },
+    {
+      "epoch": 0.08954031465093412,
+      "grad_norm": 40.75,
+      "learning_rate": 1.9413333333333336e-05,
+      "loss": 0.9385,
+      "step": 1457
+    },
+    {
+      "epoch": 0.08960176991150443,
+      "grad_norm": 68.5,
+      "learning_rate": 1.9426666666666668e-05,
+      "loss": 0.5599,
+      "step": 1458
+    },
+    {
+      "epoch": 0.08966322517207473,
+      "grad_norm": 31.5,
+      "learning_rate": 1.944e-05,
+      "loss": 0.7873,
+      "step": 1459
+    },
+    {
+      "epoch": 0.08972468043264503,
+      "grad_norm": 40.25,
+      "learning_rate": 1.9453333333333336e-05,
+      "loss": 0.6027,
+      "step": 1460
+    },
+    {
+      "epoch": 0.08978613569321534,
+      "grad_norm": 37.25,
+      "learning_rate": 1.9466666666666668e-05,
+      "loss": 0.7739,
+      "step": 1461
+    },
+    {
+      "epoch": 0.08984759095378564,
+      "grad_norm": 8.375,
+      "learning_rate": 1.948e-05,
+      "loss": 0.4431,
+      "step": 1462
+    },
+    {
+      "epoch": 0.08990904621435594,
+      "grad_norm": 30.875,
+      "learning_rate": 1.9493333333333335e-05,
+      "loss": 0.4555,
+      "step": 1463
+    },
+    {
+      "epoch": 0.08997050147492626,
+      "grad_norm": 10.25,
+      "learning_rate": 1.950666666666667e-05,
+      "loss": 0.3987,
+      "step": 1464
+    },
+    {
+      "epoch": 0.09003195673549656,
+      "grad_norm": 19.5,
+      "learning_rate": 1.9520000000000003e-05,
+      "loss": 0.544,
+      "step": 1465
+    },
+    {
+      "epoch": 0.09009341199606687,
+      "grad_norm": 13.25,
+      "learning_rate": 1.9533333333333335e-05,
+      "loss": 0.6123,
+      "step": 1466
+    },
+    {
+      "epoch": 0.09015486725663717,
+      "grad_norm": 25.125,
+      "learning_rate": 1.954666666666667e-05,
+      "loss": 0.6146,
+      "step": 1467
+    },
+    {
+      "epoch": 0.09021632251720747,
+      "grad_norm": 53.5,
+      "learning_rate": 1.9560000000000002e-05,
+      "loss": 0.5794,
+      "step": 1468
+    },
+    {
+      "epoch": 0.09027777777777778,
+      "grad_norm": 30.5,
+      "learning_rate": 1.9573333333333335e-05,
+      "loss": 0.5069,
+      "step": 1469
+    },
+    {
+      "epoch": 0.09033923303834808,
+      "grad_norm": 23.125,
+      "learning_rate": 1.9586666666666667e-05,
+      "loss": 0.5018,
+      "step": 1470
+    },
+    {
+      "epoch": 0.09040068829891838,
+      "grad_norm": 16.5,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 0.4914,
+      "step": 1471
+    },
+    {
+      "epoch": 0.09046214355948869,
+      "grad_norm": 73.5,
+      "learning_rate": 1.9613333333333334e-05,
+      "loss": 0.4178,
+      "step": 1472
+    },
+    {
+      "epoch": 0.090523598820059,
+      "grad_norm": 28.5,
+      "learning_rate": 1.9626666666666666e-05,
+      "loss": 0.67,
+      "step": 1473
+    },
+    {
+      "epoch": 0.09058505408062931,
+      "grad_norm": 25.375,
+      "learning_rate": 1.9640000000000002e-05,
+      "loss": 0.6884,
+      "step": 1474
+    },
+    {
+      "epoch": 0.09064650934119961,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.9653333333333334e-05,
+      "loss": 0.3814,
+      "step": 1475
+    },
+    {
+      "epoch": 0.09070796460176991,
+      "grad_norm": 36.5,
+      "learning_rate": 1.9666666666666666e-05,
+      "loss": 0.5632,
+      "step": 1476
+    },
+    {
+      "epoch": 0.09076941986234022,
+      "grad_norm": 22.625,
+      "learning_rate": 1.968e-05,
+      "loss": 0.6045,
+      "step": 1477
+    },
+    {
+      "epoch": 0.09083087512291052,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.9693333333333337e-05,
+      "loss": 0.5875,
+      "step": 1478
+    },
+    {
+      "epoch": 0.09089233038348082,
+      "grad_norm": 10.375,
+      "learning_rate": 1.970666666666667e-05,
+      "loss": 0.4524,
+      "step": 1479
+    },
+    {
+      "epoch": 0.09095378564405113,
+      "grad_norm": 19.375,
+      "learning_rate": 1.972e-05,
+      "loss": 0.4269,
+      "step": 1480
+    },
+    {
+      "epoch": 0.09101524090462143,
+      "grad_norm": 20.125,
+      "learning_rate": 1.9733333333333336e-05,
+      "loss": 0.4363,
+      "step": 1481
+    },
+    {
+      "epoch": 0.09107669616519173,
+      "grad_norm": 26.625,
+      "learning_rate": 1.974666666666667e-05,
+      "loss": 0.6239,
+      "step": 1482
+    },
+    {
+      "epoch": 0.09113815142576205,
+      "grad_norm": 26.75,
+      "learning_rate": 1.976e-05,
+      "loss": 0.7371,
+      "step": 1483
+    },
+    {
+      "epoch": 0.09119960668633235,
+      "grad_norm": 15.4375,
+      "learning_rate": 1.9773333333333336e-05,
+      "loss": 0.5799,
+      "step": 1484
+    },
+    {
+      "epoch": 0.09126106194690266,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.9786666666666668e-05,
+      "loss": 0.3869,
+      "step": 1485
+    },
+    {
+      "epoch": 0.09132251720747296,
+      "grad_norm": 21.5,
+      "learning_rate": 1.98e-05,
+      "loss": 0.4029,
+      "step": 1486
+    },
+    {
+      "epoch": 0.09138397246804326,
+      "grad_norm": 7.625,
+      "learning_rate": 1.9813333333333336e-05,
+      "loss": 0.3257,
+      "step": 1487
+    },
+    {
+      "epoch": 0.09144542772861357,
+      "grad_norm": 31.25,
+      "learning_rate": 1.9826666666666668e-05,
+      "loss": 0.3925,
+      "step": 1488
+    },
+    {
+      "epoch": 0.09150688298918387,
+      "grad_norm": 25.375,
+      "learning_rate": 1.9840000000000003e-05,
+      "loss": 0.626,
+      "step": 1489
+    },
+    {
+      "epoch": 0.09156833824975417,
+      "grad_norm": 28.875,
+      "learning_rate": 1.9853333333333335e-05,
+      "loss": 0.6065,
+      "step": 1490
+    },
+    {
+      "epoch": 0.09162979351032448,
+      "grad_norm": 20.75,
+      "learning_rate": 1.9866666666666667e-05,
+      "loss": 0.4988,
+      "step": 1491
+    },
+    {
+      "epoch": 0.0916912487708948,
+      "grad_norm": 34.0,
+      "learning_rate": 1.9880000000000003e-05,
+      "loss": 0.7633,
+      "step": 1492
+    },
+    {
+      "epoch": 0.0917527040314651,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.9893333333333335e-05,
+      "loss": 0.5236,
+      "step": 1493
+    },
+    {
+      "epoch": 0.0918141592920354,
+      "grad_norm": 38.5,
+      "learning_rate": 1.9906666666666667e-05,
+      "loss": 0.744,
+      "step": 1494
+    },
+    {
+      "epoch": 0.0918756145526057,
+      "grad_norm": 60.25,
+      "learning_rate": 1.9920000000000002e-05,
+      "loss": 0.4128,
+      "step": 1495
+    },
+    {
+      "epoch": 0.09193706981317601,
+      "grad_norm": 11.0625,
+      "learning_rate": 1.9933333333333334e-05,
+      "loss": 0.4378,
+      "step": 1496
+    },
+    {
+      "epoch": 0.09199852507374631,
+      "grad_norm": 16.5,
+      "learning_rate": 1.9946666666666667e-05,
+      "loss": 0.5792,
+      "step": 1497
+    },
+    {
+      "epoch": 0.09205998033431662,
+      "grad_norm": 28.0,
+      "learning_rate": 1.9960000000000002e-05,
+      "loss": 0.7458,
+      "step": 1498
+    },
+    {
+      "epoch": 0.09212143559488692,
+      "grad_norm": 14.25,
+      "learning_rate": 1.9973333333333334e-05,
+      "loss": 0.8067,
+      "step": 1499
+    },
+    {
+      "epoch": 0.09218289085545722,
+      "grad_norm": 25.125,
+      "learning_rate": 1.9986666666666666e-05,
+      "loss": 0.6215,
+      "step": 1500
+    },
+    {
+      "epoch": 0.09224434611602753,
+      "grad_norm": 16.5,
+      "learning_rate": 2e-05,
+      "loss": 0.7529,
+      "step": 1501
+    },
+    {
+      "epoch": 0.09230580137659784,
+      "grad_norm": 15.4375,
+      "learning_rate": 1.9999999729228962e-05,
+      "loss": 0.718,
+      "step": 1502
+    },
+    {
+      "epoch": 0.09236725663716815,
+      "grad_norm": 9.375,
+      "learning_rate": 1.9999998916915862e-05,
+      "loss": 0.5488,
+      "step": 1503
+    },
+    {
+      "epoch": 0.09242871189773845,
+      "grad_norm": 20.75,
+      "learning_rate": 1.9999997563060744e-05,
+      "loss": 0.4644,
+      "step": 1504
+    },
+    {
+      "epoch": 0.09249016715830875,
+      "grad_norm": 14.75,
+      "learning_rate": 1.9999995667663676e-05,
+      "loss": 0.7382,
+      "step": 1505
+    },
+    {
+      "epoch": 0.09255162241887906,
+      "grad_norm": 22.125,
+      "learning_rate": 1.999999323072477e-05,
+      "loss": 0.6014,
+      "step": 1506
+    },
+    {
+      "epoch": 0.09261307767944936,
+      "grad_norm": 28.75,
+      "learning_rate": 1.9999990252244153e-05,
+      "loss": 0.6516,
+      "step": 1507
+    },
+    {
+      "epoch": 0.09267453294001966,
+      "grad_norm": 9.25,
+      "learning_rate": 1.999998673222198e-05,
+      "loss": 0.6359,
+      "step": 1508
+    },
+    {
+      "epoch": 0.09273598820058997,
+      "grad_norm": 49.0,
+      "learning_rate": 1.9999982670658457e-05,
+      "loss": 0.6436,
+      "step": 1509
+    },
+    {
+      "epoch": 0.09279744346116027,
+      "grad_norm": 11.625,
+      "learning_rate": 1.9999978067553796e-05,
+      "loss": 0.5704,
+      "step": 1510
+    },
+    {
+      "epoch": 0.09285889872173057,
+      "grad_norm": 25.0,
+      "learning_rate": 1.999997292290824e-05,
+      "loss": 0.6502,
+      "step": 1511
+    },
+    {
+      "epoch": 0.09292035398230089,
+      "grad_norm": 22.0,
+      "learning_rate": 1.9999967236722075e-05,
+      "loss": 0.5587,
+      "step": 1512
+    },
+    {
+      "epoch": 0.0929818092428712,
+      "grad_norm": 26.5,
+      "learning_rate": 1.9999961008995607e-05,
+      "loss": 0.4997,
+      "step": 1513
+    },
+    {
+      "epoch": 0.0930432645034415,
+      "grad_norm": 32.25,
+      "learning_rate": 1.9999954239729175e-05,
+      "loss": 0.8159,
+      "step": 1514
+    },
+    {
+      "epoch": 0.0931047197640118,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.999994692892314e-05,
+      "loss": 0.5844,
+      "step": 1515
+    },
+    {
+      "epoch": 0.0931661750245821,
+      "grad_norm": 51.5,
+      "learning_rate": 1.9999939076577906e-05,
+      "loss": 0.5972,
+      "step": 1516
+    },
+    {
+      "epoch": 0.09322763028515241,
+      "grad_norm": 25.625,
+      "learning_rate": 1.9999930682693892e-05,
+      "loss": 0.5903,
+      "step": 1517
+    },
+    {
+      "epoch": 0.09328908554572271,
+      "grad_norm": 38.5,
+      "learning_rate": 1.9999921747271554e-05,
+      "loss": 0.6251,
+      "step": 1518
+    },
+    {
+      "epoch": 0.09335054080629301,
+      "grad_norm": 41.5,
+      "learning_rate": 1.9999912270311376e-05,
+      "loss": 0.4034,
+      "step": 1519
+    },
+    {
+      "epoch": 0.09341199606686332,
+      "grad_norm": 63.0,
+      "learning_rate": 1.9999902251813873e-05,
+      "loss": 0.4594,
+      "step": 1520
+    },
+    {
+      "epoch": 0.09347345132743363,
+      "grad_norm": 51.75,
+      "learning_rate": 1.999989169177959e-05,
+      "loss": 0.6752,
+      "step": 1521
+    },
+    {
+      "epoch": 0.09353490658800394,
+      "grad_norm": 34.0,
+      "learning_rate": 1.999988059020909e-05,
+      "loss": 0.8115,
+      "step": 1522
+    },
+    {
+      "epoch": 0.09359636184857424,
+      "grad_norm": 4.75,
+      "learning_rate": 1.9999868947102982e-05,
+      "loss": 0.6671,
+      "step": 1523
+    },
+    {
+      "epoch": 0.09365781710914454,
+      "grad_norm": 36.25,
+      "learning_rate": 1.9999856762461892e-05,
+      "loss": 0.9567,
+      "step": 1524
+    },
+    {
+      "epoch": 0.09371927236971485,
+      "grad_norm": 33.0,
+      "learning_rate": 1.9999844036286483e-05,
+      "loss": 0.6517,
+      "step": 1525
+    },
+    {
+      "epoch": 0.09378072763028515,
+      "grad_norm": 34.25,
+      "learning_rate": 1.9999830768577445e-05,
+      "loss": 0.7101,
+      "step": 1526
+    },
+    {
+      "epoch": 0.09384218289085546,
+      "grad_norm": 29.25,
+      "learning_rate": 1.999981695933549e-05,
+      "loss": 0.5927,
+      "step": 1527
+    },
+    {
+      "epoch": 0.09390363815142576,
+      "grad_norm": 29.5,
+      "learning_rate": 1.999980260856137e-05,
+      "loss": 0.5613,
+      "step": 1528
+    },
+    {
+      "epoch": 0.09396509341199606,
+      "grad_norm": 21.0,
+      "learning_rate": 1.9999787716255867e-05,
+      "loss": 0.6787,
+      "step": 1529
+    },
+    {
+      "epoch": 0.09402654867256637,
+      "grad_norm": 10.0,
+      "learning_rate": 1.9999772282419783e-05,
+      "loss": 0.6078,
+      "step": 1530
+    },
+    {
+      "epoch": 0.09408800393313668,
+      "grad_norm": 23.625,
+      "learning_rate": 1.9999756307053947e-05,
+      "loss": 0.6432,
+      "step": 1531
+    },
+    {
+      "epoch": 0.09414945919370699,
+      "grad_norm": 30.5,
+      "learning_rate": 1.999973979015924e-05,
+      "loss": 0.7424,
+      "step": 1532
+    },
+    {
+      "epoch": 0.09421091445427729,
+      "grad_norm": 22.125,
+      "learning_rate": 1.999972273173654e-05,
+      "loss": 0.6113,
+      "step": 1533
+    },
+    {
+      "epoch": 0.09427236971484759,
+      "grad_norm": 48.25,
+      "learning_rate": 1.999970513178678e-05,
+      "loss": 0.6524,
+      "step": 1534
+    },
+    {
+      "epoch": 0.0943338249754179,
+      "grad_norm": 32.0,
+      "learning_rate": 1.9999686990310914e-05,
+      "loss": 0.8329,
+      "step": 1535
+    },
+    {
+      "epoch": 0.0943952802359882,
+      "grad_norm": 36.5,
+      "learning_rate": 1.999966830730992e-05,
+      "loss": 0.8255,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0944567354965585,
+      "grad_norm": 33.5,
+      "learning_rate": 1.9999649082784807e-05,
+      "loss": 0.7094,
+      "step": 1537
+    },
+    {
+      "epoch": 0.0945181907571288,
+      "grad_norm": 33.0,
+      "learning_rate": 1.9999629316736624e-05,
+      "loss": 0.7507,
+      "step": 1538
+    },
+    {
+      "epoch": 0.09457964601769911,
+      "grad_norm": 21.875,
+      "learning_rate": 1.999960900916644e-05,
+      "loss": 0.7474,
+      "step": 1539
+    },
+    {
+      "epoch": 0.09464110127826943,
+      "grad_norm": 17.875,
+      "learning_rate": 1.999958816007535e-05,
+      "loss": 0.5937,
+      "step": 1540
+    },
+    {
+      "epoch": 0.09470255653883973,
+      "grad_norm": 21.25,
+      "learning_rate": 1.9999566769464483e-05,
+      "loss": 0.709,
+      "step": 1541
+    },
+    {
+      "epoch": 0.09476401179941003,
+      "grad_norm": 16.0,
+      "learning_rate": 1.9999544837335007e-05,
+      "loss": 0.5635,
+      "step": 1542
+    },
+    {
+      "epoch": 0.09482546705998034,
+      "grad_norm": 18.625,
+      "learning_rate": 1.99995223636881e-05,
+      "loss": 0.53,
+      "step": 1543
+    },
+    {
+      "epoch": 0.09488692232055064,
+      "grad_norm": 10.125,
+      "learning_rate": 1.999949934852498e-05,
+      "loss": 0.5143,
+      "step": 1544
+    },
+    {
+      "epoch": 0.09494837758112094,
+      "grad_norm": 35.5,
+      "learning_rate": 1.9999475791846896e-05,
+      "loss": 0.7373,
+      "step": 1545
+    },
+    {
+      "epoch": 0.09500983284169125,
+      "grad_norm": 43.5,
+      "learning_rate": 1.9999451693655125e-05,
+      "loss": 0.7662,
+      "step": 1546
+    },
+    {
+      "epoch": 0.09507128810226155,
+      "grad_norm": 18.25,
+      "learning_rate": 1.9999427053950965e-05,
+      "loss": 0.6075,
+      "step": 1547
+    },
+    {
+      "epoch": 0.09513274336283185,
+      "grad_norm": 26.25,
+      "learning_rate": 1.999940187273576e-05,
+      "loss": 0.613,
+      "step": 1548
+    },
+    {
+      "epoch": 0.09519419862340216,
+      "grad_norm": 32.0,
+      "learning_rate": 1.9999376150010868e-05,
+      "loss": 0.5371,
+      "step": 1549
+    },
+    {
+      "epoch": 0.09525565388397247,
+      "grad_norm": 33.25,
+      "learning_rate": 1.999934988577768e-05,
+      "loss": 0.6484,
+      "step": 1550
+    },
+    {
+      "epoch": 0.09531710914454278,
+      "grad_norm": 35.75,
+      "learning_rate": 1.9999323080037623e-05,
+      "loss": 0.6118,
+      "step": 1551
+    },
+    {
+      "epoch": 0.09537856440511308,
+      "grad_norm": 42.0,
+      "learning_rate": 1.9999295732792146e-05,
+      "loss": 0.4874,
+      "step": 1552
+    },
+    {
+      "epoch": 0.09544001966568338,
+      "grad_norm": 50.25,
+      "learning_rate": 1.9999267844042734e-05,
+      "loss": 0.6037,
+      "step": 1553
+    },
+    {
+      "epoch": 0.09550147492625369,
+      "grad_norm": 16.75,
+      "learning_rate": 1.9999239413790895e-05,
+      "loss": 0.475,
+      "step": 1554
+    },
+    {
+      "epoch": 0.09556293018682399,
+      "grad_norm": 23.75,
+      "learning_rate": 1.9999210442038164e-05,
+      "loss": 0.7353,
+      "step": 1555
+    },
+    {
+      "epoch": 0.0956243854473943,
+      "grad_norm": 35.75,
+      "learning_rate": 1.9999180928786113e-05,
+      "loss": 0.6392,
+      "step": 1556
+    },
+    {
+      "epoch": 0.0956858407079646,
+      "grad_norm": 31.375,
+      "learning_rate": 1.9999150874036344e-05,
+      "loss": 0.5754,
+      "step": 1557
+    },
+    {
+      "epoch": 0.0957472959685349,
+      "grad_norm": 19.0,
+      "learning_rate": 1.9999120277790477e-05,
+      "loss": 0.5501,
+      "step": 1558
+    },
+    {
+      "epoch": 0.0958087512291052,
+      "grad_norm": 20.75,
+      "learning_rate": 1.9999089140050182e-05,
+      "loss": 0.4345,
+      "step": 1559
+    },
+    {
+      "epoch": 0.09587020648967552,
+      "grad_norm": 14.375,
+      "learning_rate": 1.999905746081713e-05,
+      "loss": 0.4183,
+      "step": 1560
+    },
+    {
+      "epoch": 0.09593166175024583,
+      "grad_norm": 23.5,
+      "learning_rate": 1.9999025240093045e-05,
+      "loss": 0.5495,
+      "step": 1561
+    },
+    {
+      "epoch": 0.09599311701081613,
+      "grad_norm": 16.5,
+      "learning_rate": 1.999899247787967e-05,
+      "loss": 0.7035,
+      "step": 1562
+    },
+    {
+      "epoch": 0.09605457227138643,
+      "grad_norm": 15.875,
+      "learning_rate": 1.9998959174178777e-05,
+      "loss": 0.539,
+      "step": 1563
+    },
+    {
+      "epoch": 0.09611602753195674,
+      "grad_norm": 21.5,
+      "learning_rate": 1.9998925328992175e-05,
+      "loss": 0.4575,
+      "step": 1564
+    },
+    {
+      "epoch": 0.09617748279252704,
+      "grad_norm": 18.75,
+      "learning_rate": 1.9998890942321693e-05,
+      "loss": 0.5762,
+      "step": 1565
+    },
+    {
+      "epoch": 0.09623893805309734,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.9998856014169193e-05,
+      "loss": 0.6934,
+      "step": 1566
+    },
+    {
+      "epoch": 0.09630039331366765,
+      "grad_norm": 11.75,
+      "learning_rate": 1.999882054453657e-05,
+      "loss": 0.629,
+      "step": 1567
+    },
+    {
+      "epoch": 0.09636184857423795,
+      "grad_norm": 23.125,
+      "learning_rate": 1.9998784533425743e-05,
+      "loss": 0.5115,
+      "step": 1568
+    },
+    {
+      "epoch": 0.09642330383480827,
+      "grad_norm": 23.5,
+      "learning_rate": 1.9998747980838655e-05,
+      "loss": 0.4956,
+      "step": 1569
+    },
+    {
+      "epoch": 0.09648475909537857,
+      "grad_norm": 63.5,
+      "learning_rate": 1.9998710886777298e-05,
+      "loss": 0.864,
+      "step": 1570
+    },
+    {
+      "epoch": 0.09654621435594887,
+      "grad_norm": 11.125,
+      "learning_rate": 1.9998673251243672e-05,
+      "loss": 0.5724,
+      "step": 1571
+    },
+    {
+      "epoch": 0.09660766961651918,
+      "grad_norm": 41.25,
+      "learning_rate": 1.999863507423982e-05,
+      "loss": 0.7549,
+      "step": 1572
+    },
+    {
+      "epoch": 0.09666912487708948,
+      "grad_norm": 18.75,
+      "learning_rate": 1.9998596355767805e-05,
+      "loss": 0.7428,
+      "step": 1573
+    },
+    {
+      "epoch": 0.09673058013765978,
+      "grad_norm": 47.5,
+      "learning_rate": 1.9998557095829727e-05,
+      "loss": 0.5957,
+      "step": 1574
+    },
+    {
+      "epoch": 0.09679203539823009,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.9998517294427712e-05,
+      "loss": 0.49,
+      "step": 1575
+    },
+    {
+      "epoch": 0.09685349065880039,
+      "grad_norm": 103.0,
+      "learning_rate": 1.9998476951563914e-05,
+      "loss": 0.5234,
+      "step": 1576
+    },
+    {
+      "epoch": 0.09691494591937069,
+      "grad_norm": 15.375,
+      "learning_rate": 1.9998436067240518e-05,
+      "loss": 0.5719,
+      "step": 1577
+    },
+    {
+      "epoch": 0.096976401179941,
+      "grad_norm": 58.25,
+      "learning_rate": 1.999839464145974e-05,
+      "loss": 0.6186,
+      "step": 1578
+    },
+    {
+      "epoch": 0.09703785644051131,
+      "grad_norm": 15.75,
+      "learning_rate": 1.9998352674223816e-05,
+      "loss": 0.5239,
+      "step": 1579
+    },
+    {
+      "epoch": 0.09709931170108162,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.999831016553503e-05,
+      "loss": 0.4837,
+      "step": 1580
+    },
+    {
+      "epoch": 0.09716076696165192,
+      "grad_norm": 35.75,
+      "learning_rate": 1.999826711539568e-05,
+      "loss": 0.3647,
+      "step": 1581
+    },
+    {
+      "epoch": 0.09722222222222222,
+      "grad_norm": 33.75,
+      "learning_rate": 1.9998223523808092e-05,
+      "loss": 0.426,
+      "step": 1582
+    },
+    {
+      "epoch": 0.09728367748279253,
+      "grad_norm": 70.5,
+      "learning_rate": 1.9998179390774632e-05,
+      "loss": 0.3515,
+      "step": 1583
+    },
+    {
+      "epoch": 0.09734513274336283,
+      "grad_norm": 28.0,
+      "learning_rate": 1.9998134716297688e-05,
+      "loss": 0.4279,
+      "step": 1584
+    },
+    {
+      "epoch": 0.09740658800393313,
+      "grad_norm": 20.75,
+      "learning_rate": 1.999808950037968e-05,
+      "loss": 0.4479,
+      "step": 1585
+    },
+    {
+      "epoch": 0.09746804326450344,
+      "grad_norm": 41.0,
+      "learning_rate": 1.9998043743023056e-05,
+      "loss": 0.6468,
+      "step": 1586
+    },
+    {
+      "epoch": 0.09752949852507374,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9997997444230294e-05,
+      "loss": 0.6227,
+      "step": 1587
+    },
+    {
+      "epoch": 0.09759095378564406,
+      "grad_norm": 44.0,
+      "learning_rate": 1.99979506040039e-05,
+      "loss": 0.5317,
+      "step": 1588
+    },
+    {
+      "epoch": 0.09765240904621436,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.9997903222346418e-05,
+      "loss": 0.3899,
+      "step": 1589
+    },
+    {
+      "epoch": 0.09771386430678466,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.9997855299260403e-05,
+      "loss": 0.3965,
+      "step": 1590
+    },
+    {
+      "epoch": 0.09777531956735497,
+      "grad_norm": 23.875,
+      "learning_rate": 1.9997806834748455e-05,
+      "loss": 0.449,
+      "step": 1591
+    },
+    {
+      "epoch": 0.09783677482792527,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9997757828813203e-05,
+      "loss": 0.3195,
+      "step": 1592
+    },
+    {
+      "epoch": 0.09789823008849557,
+      "grad_norm": 27.875,
+      "learning_rate": 1.9997708281457293e-05,
+      "loss": 0.628,
+      "step": 1593
+    },
+    {
+      "epoch": 0.09795968534906588,
+      "grad_norm": 53.75,
+      "learning_rate": 1.9997658192683412e-05,
+      "loss": 0.4898,
+      "step": 1594
+    },
+    {
+      "epoch": 0.09802114060963618,
+      "grad_norm": 21.75,
+      "learning_rate": 1.9997607562494277e-05,
+      "loss": 0.5189,
+      "step": 1595
+    },
+    {
+      "epoch": 0.09808259587020648,
+      "grad_norm": 20.25,
+      "learning_rate": 1.9997556390892623e-05,
+      "loss": 0.5038,
+      "step": 1596
+    },
+    {
+      "epoch": 0.09814405113077679,
+      "grad_norm": 31.5,
+      "learning_rate": 1.9997504677881224e-05,
+      "loss": 0.661,
+      "step": 1597
+    },
+    {
+      "epoch": 0.0982055063913471,
+      "grad_norm": 23.25,
+      "learning_rate": 1.999745242346288e-05,
+      "loss": 0.5747,
+      "step": 1598
+    },
+    {
+      "epoch": 0.09826696165191741,
+      "grad_norm": 42.5,
+      "learning_rate": 1.999739962764042e-05,
+      "loss": 0.4613,
+      "step": 1599
+    },
+    {
+      "epoch": 0.09832841691248771,
+      "grad_norm": 19.125,
+      "learning_rate": 1.9997346290416703e-05,
+      "loss": 0.4801,
+      "step": 1600
+    },
+    {
+      "epoch": 0.09838987217305802,
+      "grad_norm": 31.875,
+      "learning_rate": 1.999729241179462e-05,
+      "loss": 0.8283,
+      "step": 1601
+    },
+    {
+      "epoch": 0.09845132743362832,
+      "grad_norm": 39.25,
+      "learning_rate": 1.9997237991777085e-05,
+      "loss": 0.7647,
+      "step": 1602
+    },
+    {
+      "epoch": 0.09851278269419862,
+      "grad_norm": 25.625,
+      "learning_rate": 1.999718303036705e-05,
+      "loss": 0.7577,
+      "step": 1603
+    },
+    {
+      "epoch": 0.09857423795476893,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9997127527567487e-05,
+      "loss": 0.6192,
+      "step": 1604
+    },
+    {
+      "epoch": 0.09863569321533923,
+      "grad_norm": 58.25,
+      "learning_rate": 1.99970714833814e-05,
+      "loss": 0.6892,
+      "step": 1605
+    },
+    {
+      "epoch": 0.09869714847590953,
+      "grad_norm": 24.375,
+      "learning_rate": 1.9997014897811834e-05,
+      "loss": 0.5686,
+      "step": 1606
+    },
+    {
+      "epoch": 0.09875860373647984,
+      "grad_norm": 26.25,
+      "learning_rate": 1.9996957770861843e-05,
+      "loss": 0.5501,
+      "step": 1607
+    },
+    {
+      "epoch": 0.09882005899705015,
+      "grad_norm": 24.0,
+      "learning_rate": 1.9996900102534526e-05,
+      "loss": 0.4202,
+      "step": 1608
+    },
+    {
+      "epoch": 0.09888151425762046,
+      "grad_norm": 78.0,
+      "learning_rate": 1.9996841892833e-05,
+      "loss": 0.8203,
+      "step": 1609
+    },
+    {
+      "epoch": 0.09894296951819076,
+      "grad_norm": 38.0,
+      "learning_rate": 1.9996783141760425e-05,
+      "loss": 0.5394,
+      "step": 1610
+    },
+    {
+      "epoch": 0.09900442477876106,
+      "grad_norm": 45.25,
+      "learning_rate": 1.9996723849319978e-05,
+      "loss": 0.7046,
+      "step": 1611
+    },
+    {
+      "epoch": 0.09906588003933137,
+      "grad_norm": 27.125,
+      "learning_rate": 1.999666401551487e-05,
+      "loss": 0.6283,
+      "step": 1612
+    },
+    {
+      "epoch": 0.09912733529990167,
+      "grad_norm": 13.5,
+      "learning_rate": 1.999660364034835e-05,
+      "loss": 0.4878,
+      "step": 1613
+    },
+    {
+      "epoch": 0.09918879056047197,
+      "grad_norm": 24.25,
+      "learning_rate": 1.9996542723823673e-05,
+      "loss": 0.4808,
+      "step": 1614
+    },
+    {
+      "epoch": 0.09925024582104228,
+      "grad_norm": 22.125,
+      "learning_rate": 1.9996481265944146e-05,
+      "loss": 0.5646,
+      "step": 1615
+    },
+    {
+      "epoch": 0.09931170108161258,
+      "grad_norm": 34.0,
+      "learning_rate": 1.9996419266713097e-05,
+      "loss": 0.6158,
+      "step": 1616
+    },
+    {
+      "epoch": 0.0993731563421829,
+      "grad_norm": 24.0,
+      "learning_rate": 1.999635672613388e-05,
+      "loss": 0.3868,
+      "step": 1617
+    },
+    {
+      "epoch": 0.0994346116027532,
+      "grad_norm": 21.625,
+      "learning_rate": 1.9996293644209886e-05,
+      "loss": 0.4407,
+      "step": 1618
+    },
+    {
+      "epoch": 0.0994960668633235,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.9996230020944535e-05,
+      "loss": 0.4726,
+      "step": 1619
+    },
+    {
+      "epoch": 0.09955752212389381,
+      "grad_norm": 13.625,
+      "learning_rate": 1.999616585634126e-05,
+      "loss": 0.5784,
+      "step": 1620
+    },
+    {
+      "epoch": 0.09961897738446411,
+      "grad_norm": 23.0,
+      "learning_rate": 1.9996101150403543e-05,
+      "loss": 0.5214,
+      "step": 1621
+    },
+    {
+      "epoch": 0.09968043264503441,
+      "grad_norm": 36.25,
+      "learning_rate": 1.9996035903134893e-05,
+      "loss": 0.5937,
+      "step": 1622
+    },
+    {
+      "epoch": 0.09974188790560472,
+      "grad_norm": 6.0,
+      "learning_rate": 1.9995970114538835e-05,
+      "loss": 0.4794,
+      "step": 1623
+    },
+    {
+      "epoch": 0.09980334316617502,
+      "grad_norm": 15.125,
+      "learning_rate": 1.9995903784618936e-05,
+      "loss": 0.4414,
+      "step": 1624
+    },
+    {
+      "epoch": 0.09986479842674532,
+      "grad_norm": 36.0,
+      "learning_rate": 1.9995836913378786e-05,
+      "loss": 0.6542,
+      "step": 1625
+    },
+    {
+      "epoch": 0.09992625368731563,
+      "grad_norm": 16.875,
+      "learning_rate": 1.9995769500822007e-05,
+      "loss": 0.6602,
+      "step": 1626
+    },
+    {
+      "epoch": 0.09998770894788594,
+      "grad_norm": 21.5,
+      "learning_rate": 1.9995701546952252e-05,
+      "loss": 0.5246,
+      "step": 1627
+    },
+    {
+      "epoch": 0.10004916420845625,
+      "grad_norm": 14.125,
+      "learning_rate": 1.9995633051773197e-05,
+      "loss": 0.5604,
+      "step": 1628
+    },
+    {
+      "epoch": 0.10011061946902655,
+      "grad_norm": 18.0,
+      "learning_rate": 1.9995564015288555e-05,
+      "loss": 0.5439,
+      "step": 1629
+    },
+    {
+      "epoch": 0.10017207472959685,
+      "grad_norm": 29.0,
+      "learning_rate": 1.9995494437502064e-05,
+      "loss": 0.6075,
+      "step": 1630
+    },
+    {
+      "epoch": 0.10023352999016716,
+      "grad_norm": 23.875,
+      "learning_rate": 1.999542431841749e-05,
+      "loss": 0.6336,
+      "step": 1631
+    },
+    {
+      "epoch": 0.10029498525073746,
+      "grad_norm": 33.0,
+      "learning_rate": 1.999535365803863e-05,
+      "loss": 0.5192,
+      "step": 1632
+    },
+    {
+      "epoch": 0.10035644051130777,
+      "grad_norm": 13.625,
+      "learning_rate": 1.9995282456369313e-05,
+      "loss": 0.502,
+      "step": 1633
+    },
+    {
+      "epoch": 0.10041789577187807,
+      "grad_norm": 18.875,
+      "learning_rate": 1.9995210713413393e-05,
+      "loss": 0.5259,
+      "step": 1634
+    },
+    {
+      "epoch": 0.10047935103244837,
+      "grad_norm": 10.875,
+      "learning_rate": 1.9995138429174757e-05,
+      "loss": 0.4521,
+      "step": 1635
+    },
+    {
+      "epoch": 0.10054080629301869,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9995065603657317e-05,
+      "loss": 0.6616,
+      "step": 1636
+    },
+    {
+      "epoch": 0.10060226155358899,
+      "grad_norm": 22.875,
+      "learning_rate": 1.9994992236865017e-05,
+      "loss": 0.6891,
+      "step": 1637
+    },
+    {
+      "epoch": 0.1006637168141593,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.9994918328801835e-05,
+      "loss": 0.5476,
+      "step": 1638
+    },
+    {
+      "epoch": 0.1007251720747296,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.999484387947177e-05,
+      "loss": 0.4832,
+      "step": 1639
+    },
+    {
+      "epoch": 0.1007866273352999,
+      "grad_norm": 32.25,
+      "learning_rate": 1.9994768888878848e-05,
+      "loss": 0.3903,
+      "step": 1640
+    },
+    {
+      "epoch": 0.1008480825958702,
+      "grad_norm": 31.125,
+      "learning_rate": 1.999469335702714e-05,
+      "loss": 0.7796,
+      "step": 1641
+    },
+    {
+      "epoch": 0.10090953785644051,
+      "grad_norm": 27.125,
+      "learning_rate": 1.999461728392073e-05,
+      "loss": 0.8646,
+      "step": 1642
+    },
+    {
+      "epoch": 0.10097099311701081,
+      "grad_norm": 38.5,
+      "learning_rate": 1.9994540669563742e-05,
+      "loss": 0.6055,
+      "step": 1643
+    },
+    {
+      "epoch": 0.10103244837758112,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.999446351396032e-05,
+      "loss": 0.5912,
+      "step": 1644
+    },
+    {
+      "epoch": 0.10109390363815142,
+      "grad_norm": 18.75,
+      "learning_rate": 1.9994385817114644e-05,
+      "loss": 0.5026,
+      "step": 1645
+    },
+    {
+      "epoch": 0.10115535889872174,
+      "grad_norm": 9.625,
+      "learning_rate": 1.9994307579030925e-05,
+      "loss": 0.5101,
+      "step": 1646
+    },
+    {
+      "epoch": 0.10121681415929204,
+      "grad_norm": 21.75,
+      "learning_rate": 1.9994228799713396e-05,
+      "loss": 0.5674,
+      "step": 1647
+    },
+    {
+      "epoch": 0.10127826941986234,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.9994149479166324e-05,
+      "loss": 0.4756,
+      "step": 1648
+    },
+    {
+      "epoch": 0.10133972468043265,
+      "grad_norm": 30.375,
+      "learning_rate": 1.999406961739401e-05,
+      "loss": 0.7109,
+      "step": 1649
+    },
+    {
+      "epoch": 0.10140117994100295,
+      "grad_norm": 24.5,
+      "learning_rate": 1.9993989214400765e-05,
+      "loss": 0.943,
+      "step": 1650
+    },
+    {
+      "epoch": 0.10146263520157325,
+      "grad_norm": 22.625,
+      "learning_rate": 1.999390827019096e-05,
+      "loss": 0.6687,
+      "step": 1651
+    },
+    {
+      "epoch": 0.10152409046214356,
+      "grad_norm": 34.75,
+      "learning_rate": 1.9993826784768967e-05,
+      "loss": 0.6459,
+      "step": 1652
+    },
+    {
+      "epoch": 0.10158554572271386,
+      "grad_norm": 28.75,
+      "learning_rate": 1.9993744758139203e-05,
+      "loss": 0.7094,
+      "step": 1653
+    },
+    {
+      "epoch": 0.10164700098328416,
+      "grad_norm": 19.375,
+      "learning_rate": 1.999366219030611e-05,
+      "loss": 0.5492,
+      "step": 1654
+    },
+    {
+      "epoch": 0.10170845624385447,
+      "grad_norm": 30.25,
+      "learning_rate": 1.999357908127416e-05,
+      "loss": 0.5481,
+      "step": 1655
+    },
+    {
+      "epoch": 0.10176991150442478,
+      "grad_norm": 26.25,
+      "learning_rate": 1.999349543104785e-05,
+      "loss": 0.764,
+      "step": 1656
+    },
+    {
+      "epoch": 0.10183136676499509,
+      "grad_norm": 24.625,
+      "learning_rate": 1.9993411239631713e-05,
+      "loss": 0.6592,
+      "step": 1657
+    },
+    {
+      "epoch": 0.10189282202556539,
+      "grad_norm": 32.0,
+      "learning_rate": 1.9993326507030312e-05,
+      "loss": 0.9689,
+      "step": 1658
+    },
+    {
+      "epoch": 0.1019542772861357,
+      "grad_norm": 24.5,
+      "learning_rate": 1.9993241233248227e-05,
+      "loss": 0.6126,
+      "step": 1659
+    },
+    {
+      "epoch": 0.102015732546706,
+      "grad_norm": 17.125,
+      "learning_rate": 1.999315541829008e-05,
+      "loss": 0.5927,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1020771878072763,
+      "grad_norm": 17.875,
+      "learning_rate": 1.999306906216052e-05,
+      "loss": 0.6562,
+      "step": 1661
+    },
+    {
+      "epoch": 0.1021386430678466,
+      "grad_norm": 30.375,
+      "learning_rate": 1.9992982164864223e-05,
+      "loss": 0.4678,
+      "step": 1662
+    },
+    {
+      "epoch": 0.10220009832841691,
+      "grad_norm": 17.5,
+      "learning_rate": 1.9992894726405894e-05,
+      "loss": 0.4465,
+      "step": 1663
+    },
+    {
+      "epoch": 0.10226155358898721,
+      "grad_norm": 17.0,
+      "learning_rate": 1.9992806746790266e-05,
+      "loss": 0.4717,
+      "step": 1664
+    },
+    {
+      "epoch": 0.10232300884955753,
+      "grad_norm": 56.5,
+      "learning_rate": 1.999271822602211e-05,
+      "loss": 0.8307,
+      "step": 1665
+    },
+    {
+      "epoch": 0.10238446411012783,
+      "grad_norm": 41.25,
+      "learning_rate": 1.999262916410621e-05,
+      "loss": 0.9471,
+      "step": 1666
+    },
+    {
+      "epoch": 0.10244591937069814,
+      "grad_norm": 16.25,
+      "learning_rate": 1.9992539561047403e-05,
+      "loss": 0.9567,
+      "step": 1667
+    },
+    {
+      "epoch": 0.10250737463126844,
+      "grad_norm": 31.625,
+      "learning_rate": 1.9992449416850527e-05,
+      "loss": 0.834,
+      "step": 1668
+    },
+    {
+      "epoch": 0.10256882989183874,
+      "grad_norm": 29.375,
+      "learning_rate": 1.999235873152047e-05,
+      "loss": 0.7224,
+      "step": 1669
+    },
+    {
+      "epoch": 0.10263028515240905,
+      "grad_norm": 30.5,
+      "learning_rate": 1.9992267505062143e-05,
+      "loss": 0.6981,
+      "step": 1670
+    },
+    {
+      "epoch": 0.10269174041297935,
+      "grad_norm": 19.125,
+      "learning_rate": 1.9992175737480487e-05,
+      "loss": 0.5298,
+      "step": 1671
+    },
+    {
+      "epoch": 0.10275319567354965,
+      "grad_norm": 17.25,
+      "learning_rate": 1.999208342878047e-05,
+      "loss": 0.5783,
+      "step": 1672
+    },
+    {
+      "epoch": 0.10281465093411996,
+      "grad_norm": 35.75,
+      "learning_rate": 1.999199057896709e-05,
+      "loss": 0.7088,
+      "step": 1673
+    },
+    {
+      "epoch": 0.10287610619469026,
+      "grad_norm": 30.875,
+      "learning_rate": 1.999189718804538e-05,
+      "loss": 0.6098,
+      "step": 1674
+    },
+    {
+      "epoch": 0.10293756145526058,
+      "grad_norm": 27.25,
+      "learning_rate": 1.9991803256020393e-05,
+      "loss": 0.886,
+      "step": 1675
+    },
+    {
+      "epoch": 0.10299901671583088,
+      "grad_norm": 29.25,
+      "learning_rate": 1.9991708782897214e-05,
+      "loss": 0.504,
+      "step": 1676
+    },
+    {
+      "epoch": 0.10306047197640118,
+      "grad_norm": 24.375,
+      "learning_rate": 1.9991613768680965e-05,
+      "loss": 0.7169,
+      "step": 1677
+    },
+    {
+      "epoch": 0.10312192723697149,
+      "grad_norm": 26.75,
+      "learning_rate": 1.9991518213376787e-05,
+      "loss": 0.6876,
+      "step": 1678
+    },
+    {
+      "epoch": 0.10318338249754179,
+      "grad_norm": 43.0,
+      "learning_rate": 1.999142211698986e-05,
+      "loss": 0.4543,
+      "step": 1679
+    },
+    {
+      "epoch": 0.10324483775811209,
+      "grad_norm": 41.5,
+      "learning_rate": 1.999132547952538e-05,
+      "loss": 0.7736,
+      "step": 1680
+    },
+    {
+      "epoch": 0.1033062930186824,
+      "grad_norm": 15.0,
+      "learning_rate": 1.9991228300988586e-05,
+      "loss": 0.5768,
+      "step": 1681
+    },
+    {
+      "epoch": 0.1033677482792527,
+      "grad_norm": 46.75,
+      "learning_rate": 1.9991130581384737e-05,
+      "loss": 0.6743,
+      "step": 1682
+    },
+    {
+      "epoch": 0.103429203539823,
+      "grad_norm": 20.625,
+      "learning_rate": 1.9991032320719132e-05,
+      "loss": 0.5048,
+      "step": 1683
+    },
+    {
+      "epoch": 0.10349065880039332,
+      "grad_norm": 27.75,
+      "learning_rate": 1.9990933518997086e-05,
+      "loss": 0.539,
+      "step": 1684
+    },
+    {
+      "epoch": 0.10355211406096362,
+      "grad_norm": 30.25,
+      "learning_rate": 1.999083417622395e-05,
+      "loss": 0.4855,
+      "step": 1685
+    },
+    {
+      "epoch": 0.10361356932153393,
+      "grad_norm": 15.75,
+      "learning_rate": 1.9990734292405102e-05,
+      "loss": 0.4706,
+      "step": 1686
+    },
+    {
+      "epoch": 0.10367502458210423,
+      "grad_norm": 11.4375,
+      "learning_rate": 1.9990633867545956e-05,
+      "loss": 0.4691,
+      "step": 1687
+    },
+    {
+      "epoch": 0.10373647984267453,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.9990532901651948e-05,
+      "loss": 0.4734,
+      "step": 1688
+    },
+    {
+      "epoch": 0.10379793510324484,
+      "grad_norm": 10.5,
+      "learning_rate": 1.9990431394728547e-05,
+      "loss": 0.4649,
+      "step": 1689
+    },
+    {
+      "epoch": 0.10385939036381514,
+      "grad_norm": 10.75,
+      "learning_rate": 1.999032934678125e-05,
+      "loss": 0.5539,
+      "step": 1690
+    },
+    {
+      "epoch": 0.10392084562438544,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.9990226757815582e-05,
+      "loss": 0.39,
+      "step": 1691
+    },
+    {
+      "epoch": 0.10398230088495575,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.9990123627837094e-05,
+      "loss": 0.3595,
+      "step": 1692
+    },
+    {
+      "epoch": 0.10404375614552605,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.9990019956851384e-05,
+      "loss": 0.5008,
+      "step": 1693
+    },
+    {
+      "epoch": 0.10410521140609637,
+      "grad_norm": 19.75,
+      "learning_rate": 1.998991574486405e-05,
+      "loss": 0.5157,
+      "step": 1694
+    },
+    {
+      "epoch": 0.10416666666666667,
+      "grad_norm": 18.875,
+      "learning_rate": 1.998981099188075e-05,
+      "loss": 0.3759,
+      "step": 1695
+    },
+    {
+      "epoch": 0.10422812192723697,
+      "grad_norm": 28.875,
+      "learning_rate": 1.998970569790715e-05,
+      "loss": 0.3985,
+      "step": 1696
+    },
+    {
+      "epoch": 0.10428957718780728,
+      "grad_norm": 50.5,
+      "learning_rate": 1.998959986294895e-05,
+      "loss": 0.6884,
+      "step": 1697
+    },
+    {
+      "epoch": 0.10435103244837758,
+      "grad_norm": 38.5,
+      "learning_rate": 1.9989493487011883e-05,
+      "loss": 0.713,
+      "step": 1698
+    },
+    {
+      "epoch": 0.10441248770894788,
+      "grad_norm": 54.25,
+      "learning_rate": 1.9989386570101716e-05,
+      "loss": 0.7491,
+      "step": 1699
+    },
+    {
+      "epoch": 0.10447394296951819,
+      "grad_norm": 39.0,
+      "learning_rate": 1.9989279112224226e-05,
+      "loss": 0.6194,
+      "step": 1700
+    },
+    {
+      "epoch": 0.10453539823008849,
+      "grad_norm": 26.125,
+      "learning_rate": 1.998917111338525e-05,
+      "loss": 0.4844,
+      "step": 1701
+    },
+    {
+      "epoch": 0.1045968534906588,
+      "grad_norm": 24.875,
+      "learning_rate": 1.9989062573590618e-05,
+      "loss": 0.6842,
+      "step": 1702
+    },
+    {
+      "epoch": 0.10465830875122911,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.9988953492846218e-05,
+      "loss": 0.3785,
+      "step": 1703
+    },
+    {
+      "epoch": 0.10471976401179942,
+      "grad_norm": 29.375,
+      "learning_rate": 1.998884387115796e-05,
+      "loss": 0.4311,
+      "step": 1704
+    },
+    {
+      "epoch": 0.10478121927236972,
+      "grad_norm": 38.5,
+      "learning_rate": 1.9988733708531772e-05,
+      "loss": 0.7294,
+      "step": 1705
+    },
+    {
+      "epoch": 0.10484267453294002,
+      "grad_norm": 28.375,
+      "learning_rate": 1.9988623004973625e-05,
+      "loss": 0.7229,
+      "step": 1706
+    },
+    {
+      "epoch": 0.10490412979351033,
+      "grad_norm": 6.28125,
+      "learning_rate": 1.9988511760489514e-05,
+      "loss": 0.6447,
+      "step": 1707
+    },
+    {
+      "epoch": 0.10496558505408063,
+      "grad_norm": 20.25,
+      "learning_rate": 1.998839997508546e-05,
+      "loss": 0.6963,
+      "step": 1708
+    },
+    {
+      "epoch": 0.10502704031465093,
+      "grad_norm": 63.0,
+      "learning_rate": 1.9988287648767523e-05,
+      "loss": 0.6341,
+      "step": 1709
+    },
+    {
+      "epoch": 0.10508849557522124,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.9988174781541774e-05,
+      "loss": 0.3937,
+      "step": 1710
+    },
+    {
+      "epoch": 0.10514995083579154,
+      "grad_norm": 22.5,
+      "learning_rate": 1.9988061373414342e-05,
+      "loss": 0.4364,
+      "step": 1711
+    },
+    {
+      "epoch": 0.10521140609636184,
+      "grad_norm": 33.5,
+      "learning_rate": 1.9987947424391358e-05,
+      "loss": 0.4652,
+      "step": 1712
+    },
+    {
+      "epoch": 0.10527286135693216,
+      "grad_norm": 28.625,
+      "learning_rate": 1.9987832934478992e-05,
+      "loss": 0.6301,
+      "step": 1713
+    },
+    {
+      "epoch": 0.10533431661750246,
+      "grad_norm": 22.625,
+      "learning_rate": 1.9987717903683447e-05,
+      "loss": 0.6495,
+      "step": 1714
+    },
+    {
+      "epoch": 0.10539577187807277,
+      "grad_norm": 35.5,
+      "learning_rate": 1.9987602332010952e-05,
+      "loss": 0.668,
+      "step": 1715
+    },
+    {
+      "epoch": 0.10545722713864307,
+      "grad_norm": 16.25,
+      "learning_rate": 1.9987486219467764e-05,
+      "loss": 0.5514,
+      "step": 1716
+    },
+    {
+      "epoch": 0.10551868239921337,
+      "grad_norm": 17.0,
+      "learning_rate": 1.998736956606018e-05,
+      "loss": 0.5,
+      "step": 1717
+    },
+    {
+      "epoch": 0.10558013765978368,
+      "grad_norm": 35.25,
+      "learning_rate": 1.99872523717945e-05,
+      "loss": 0.5595,
+      "step": 1718
+    },
+    {
+      "epoch": 0.10564159292035398,
+      "grad_norm": 11.25,
+      "learning_rate": 1.998713463667709e-05,
+      "loss": 0.3511,
+      "step": 1719
+    },
+    {
+      "epoch": 0.10570304818092428,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.9987016360714307e-05,
+      "loss": 0.3298,
+      "step": 1720
+    },
+    {
+      "epoch": 0.10576450344149459,
+      "grad_norm": 42.5,
+      "learning_rate": 1.998689754391257e-05,
+      "loss": 0.6871,
+      "step": 1721
+    },
+    {
+      "epoch": 0.10582595870206489,
+      "grad_norm": 33.0,
+      "learning_rate": 1.998677818627831e-05,
+      "loss": 0.7382,
+      "step": 1722
+    },
+    {
+      "epoch": 0.10588741396263521,
+      "grad_norm": 24.625,
+      "learning_rate": 1.998665828781799e-05,
+      "loss": 0.7832,
+      "step": 1723
+    },
+    {
+      "epoch": 0.10594886922320551,
+      "grad_norm": 37.5,
+      "learning_rate": 1.99865378485381e-05,
+      "loss": 0.6788,
+      "step": 1724
+    },
+    {
+      "epoch": 0.10601032448377581,
+      "grad_norm": 34.5,
+      "learning_rate": 1.9986416868445167e-05,
+      "loss": 0.7302,
+      "step": 1725
+    },
+    {
+      "epoch": 0.10607177974434612,
+      "grad_norm": 50.0,
+      "learning_rate": 1.9986295347545738e-05,
+      "loss": 0.6242,
+      "step": 1726
+    },
+    {
+      "epoch": 0.10613323500491642,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.99861732858464e-05,
+      "loss": 0.3836,
+      "step": 1727
+    },
+    {
+      "epoch": 0.10619469026548672,
+      "grad_norm": 26.0,
+      "learning_rate": 1.998605068335376e-05,
+      "loss": 0.6436,
+      "step": 1728
+    },
+    {
+      "epoch": 0.10625614552605703,
+      "grad_norm": 46.0,
+      "learning_rate": 1.9985927540074453e-05,
+      "loss": 0.5314,
+      "step": 1729
+    },
+    {
+      "epoch": 0.10631760078662733,
+      "grad_norm": 40.0,
+      "learning_rate": 1.9985803856015155e-05,
+      "loss": 0.623,
+      "step": 1730
+    },
+    {
+      "epoch": 0.10637905604719763,
+      "grad_norm": 10.25,
+      "learning_rate": 1.998567963118256e-05,
+      "loss": 0.606,
+      "step": 1731
+    },
+    {
+      "epoch": 0.10644051130776795,
+      "grad_norm": 29.375,
+      "learning_rate": 1.9985554865583394e-05,
+      "loss": 0.5354,
+      "step": 1732
+    },
+    {
+      "epoch": 0.10650196656833825,
+      "grad_norm": 29.125,
+      "learning_rate": 1.998542955922442e-05,
+      "loss": 0.4986,
+      "step": 1733
+    },
+    {
+      "epoch": 0.10656342182890856,
+      "grad_norm": 10.875,
+      "learning_rate": 1.9985303712112415e-05,
+      "loss": 0.4722,
+      "step": 1734
+    },
+    {
+      "epoch": 0.10662487708947886,
+      "grad_norm": 48.75,
+      "learning_rate": 1.99851773242542e-05,
+      "loss": 0.4711,
+      "step": 1735
+    },
+    {
+      "epoch": 0.10668633235004917,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.9985050395656617e-05,
+      "loss": 0.4718,
+      "step": 1736
+    },
+    {
+      "epoch": 0.10674778761061947,
+      "grad_norm": 21.375,
+      "learning_rate": 1.9984922926326543e-05,
+      "loss": 0.4948,
+      "step": 1737
+    },
+    {
+      "epoch": 0.10680924287118977,
+      "grad_norm": 53.25,
+      "learning_rate": 1.9984794916270876e-05,
+      "loss": 0.6574,
+      "step": 1738
+    },
+    {
+      "epoch": 0.10687069813176008,
+      "grad_norm": 33.5,
+      "learning_rate": 1.9984666365496555e-05,
+      "loss": 0.5539,
+      "step": 1739
+    },
+    {
+      "epoch": 0.10693215339233038,
+      "grad_norm": 15.25,
+      "learning_rate": 1.9984537274010536e-05,
+      "loss": 0.6014,
+      "step": 1740
+    },
+    {
+      "epoch": 0.10699360865290068,
+      "grad_norm": 11.75,
+      "learning_rate": 1.9984407641819812e-05,
+      "loss": 0.4793,
+      "step": 1741
+    },
+    {
+      "epoch": 0.107055063913471,
+      "grad_norm": 11.25,
+      "learning_rate": 1.99842774689314e-05,
+      "loss": 0.4216,
+      "step": 1742
+    },
+    {
+      "epoch": 0.1071165191740413,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.9984146755352354e-05,
+      "loss": 0.4465,
+      "step": 1743
+    },
+    {
+      "epoch": 0.1071779744346116,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.998401550108975e-05,
+      "loss": 0.4818,
+      "step": 1744
+    },
+    {
+      "epoch": 0.10723942969518191,
+      "grad_norm": 28.625,
+      "learning_rate": 1.99838837061507e-05,
+      "loss": 0.6439,
+      "step": 1745
+    },
+    {
+      "epoch": 0.10730088495575221,
+      "grad_norm": 106.0,
+      "learning_rate": 1.9983751370542334e-05,
+      "loss": 0.4896,
+      "step": 1746
+    },
+    {
+      "epoch": 0.10736234021632252,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.9983618494271825e-05,
+      "loss": 0.5001,
+      "step": 1747
+    },
+    {
+      "epoch": 0.10742379547689282,
+      "grad_norm": 45.75,
+      "learning_rate": 1.998348507734636e-05,
+      "loss": 0.4929,
+      "step": 1748
+    },
+    {
+      "epoch": 0.10748525073746312,
+      "grad_norm": 31.25,
+      "learning_rate": 1.998335111977318e-05,
+      "loss": 0.3756,
+      "step": 1749
+    },
+    {
+      "epoch": 0.10754670599803343,
+      "grad_norm": 35.25,
+      "learning_rate": 1.9983216621559525e-05,
+      "loss": 0.4998,
+      "step": 1750
+    },
+    {
+      "epoch": 0.10760816125860374,
+      "grad_norm": 30.75,
+      "learning_rate": 1.9983081582712684e-05,
+      "loss": 0.4718,
+      "step": 1751
+    },
+    {
+      "epoch": 0.10766961651917405,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.998294600323997e-05,
+      "loss": 0.3708,
+      "step": 1752
+    },
+    {
+      "epoch": 0.10773107177974435,
+      "grad_norm": 37.0,
+      "learning_rate": 1.998280988314872e-05,
+      "loss": 0.5807,
+      "step": 1753
+    },
+    {
+      "epoch": 0.10779252704031465,
+      "grad_norm": 58.5,
+      "learning_rate": 1.9982673222446312e-05,
+      "loss": 0.5386,
+      "step": 1754
+    },
+    {
+      "epoch": 0.10785398230088496,
+      "grad_norm": 5.34375,
+      "learning_rate": 1.998253602114015e-05,
+      "loss": 0.5628,
+      "step": 1755
+    },
+    {
+      "epoch": 0.10791543756145526,
+      "grad_norm": 65.5,
+      "learning_rate": 1.9982398279237657e-05,
+      "loss": 0.6032,
+      "step": 1756
+    },
+    {
+      "epoch": 0.10797689282202556,
+      "grad_norm": 60.5,
+      "learning_rate": 1.998225999674629e-05,
+      "loss": 0.5538,
+      "step": 1757
+    },
+    {
+      "epoch": 0.10803834808259587,
+      "grad_norm": 34.5,
+      "learning_rate": 1.9982121173673542e-05,
+      "loss": 0.5654,
+      "step": 1758
+    },
+    {
+      "epoch": 0.10809980334316617,
+      "grad_norm": 35.5,
+      "learning_rate": 1.9981981810026932e-05,
+      "loss": 0.3955,
+      "step": 1759
+    },
+    {
+      "epoch": 0.10816125860373647,
+      "grad_norm": 26.0,
+      "learning_rate": 1.998184190581401e-05,
+      "loss": 0.4848,
+      "step": 1760
+    },
+    {
+      "epoch": 0.10822271386430679,
+      "grad_norm": 19.0,
+      "learning_rate": 1.998170146104234e-05,
+      "loss": 0.5413,
+      "step": 1761
+    },
+    {
+      "epoch": 0.1082841691248771,
+      "grad_norm": 10.875,
+      "learning_rate": 1.998156047571954e-05,
+      "loss": 0.5592,
+      "step": 1762
+    },
+    {
+      "epoch": 0.1083456243854474,
+      "grad_norm": 11.875,
+      "learning_rate": 1.998141894985324e-05,
+      "loss": 0.4617,
+      "step": 1763
+    },
+    {
+      "epoch": 0.1084070796460177,
+      "grad_norm": 38.5,
+      "learning_rate": 1.9981276883451107e-05,
+      "loss": 0.486,
+      "step": 1764
+    },
+    {
+      "epoch": 0.108468534906588,
+      "grad_norm": 6.1875,
+      "learning_rate": 1.9981134276520828e-05,
+      "loss": 0.3595,
+      "step": 1765
+    },
+    {
+      "epoch": 0.10852999016715831,
+      "grad_norm": 47.75,
+      "learning_rate": 1.998099112907013e-05,
+      "loss": 0.5574,
+      "step": 1766
+    },
+    {
+      "epoch": 0.10859144542772861,
+      "grad_norm": 27.625,
+      "learning_rate": 1.998084744110677e-05,
+      "loss": 0.4474,
+      "step": 1767
+    },
+    {
+      "epoch": 0.10865290068829891,
+      "grad_norm": 39.25,
+      "learning_rate": 1.9980703212638522e-05,
+      "loss": 0.4586,
+      "step": 1768
+    },
+    {
+      "epoch": 0.10871435594886922,
+      "grad_norm": 22.375,
+      "learning_rate": 1.99805584436732e-05,
+      "loss": 0.6056,
+      "step": 1769
+    },
+    {
+      "epoch": 0.10877581120943952,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.998041313421864e-05,
+      "loss": 0.7046,
+      "step": 1770
+    },
+    {
+      "epoch": 0.10883726647000984,
+      "grad_norm": 23.25,
+      "learning_rate": 1.9980267284282718e-05,
+      "loss": 0.7499,
+      "step": 1771
+    },
+    {
+      "epoch": 0.10889872173058014,
+      "grad_norm": 41.75,
+      "learning_rate": 1.9980120893873324e-05,
+      "loss": 0.7948,
+      "step": 1772
+    },
+    {
+      "epoch": 0.10896017699115045,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.9979973962998396e-05,
+      "loss": 0.5272,
+      "step": 1773
+    },
+    {
+      "epoch": 0.10902163225172075,
+      "grad_norm": 39.75,
+      "learning_rate": 1.997982649166588e-05,
+      "loss": 0.7371,
+      "step": 1774
+    },
+    {
+      "epoch": 0.10908308751229105,
+      "grad_norm": 5.84375,
+      "learning_rate": 1.997967847988377e-05,
+      "loss": 0.4769,
+      "step": 1775
+    },
+    {
+      "epoch": 0.10914454277286136,
+      "grad_norm": 20.75,
+      "learning_rate": 1.9979529927660076e-05,
+      "loss": 0.3689,
+      "step": 1776
+    },
+    {
+      "epoch": 0.10920599803343166,
+      "grad_norm": 28.25,
+      "learning_rate": 1.9979380835002846e-05,
+      "loss": 0.5754,
+      "step": 1777
+    },
+    {
+      "epoch": 0.10926745329400196,
+      "grad_norm": 72.0,
+      "learning_rate": 1.9979231201920154e-05,
+      "loss": 0.5654,
+      "step": 1778
+    },
+    {
+      "epoch": 0.10932890855457227,
+      "grad_norm": 26.625,
+      "learning_rate": 1.9979081028420105e-05,
+      "loss": 0.5503,
+      "step": 1779
+    },
+    {
+      "epoch": 0.10939036381514258,
+      "grad_norm": 20.625,
+      "learning_rate": 1.9978930314510826e-05,
+      "loss": 0.5304,
+      "step": 1780
+    },
+    {
+      "epoch": 0.10945181907571289,
+      "grad_norm": 28.0,
+      "learning_rate": 1.9978779060200483e-05,
+      "loss": 0.5729,
+      "step": 1781
+    },
+    {
+      "epoch": 0.10951327433628319,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.9978627265497262e-05,
+      "loss": 0.4955,
+      "step": 1782
+    },
+    {
+      "epoch": 0.10957472959685349,
+      "grad_norm": 29.0,
+      "learning_rate": 1.9978474930409396e-05,
+      "loss": 0.598,
+      "step": 1783
+    },
+    {
+      "epoch": 0.1096361848574238,
+      "grad_norm": 17.0,
+      "learning_rate": 1.997832205494512e-05,
+      "loss": 0.6184,
+      "step": 1784
+    },
+    {
+      "epoch": 0.1096976401179941,
+      "grad_norm": 16.625,
+      "learning_rate": 1.997816863911272e-05,
+      "loss": 0.5125,
+      "step": 1785
+    },
+    {
+      "epoch": 0.1097590953785644,
+      "grad_norm": 38.75,
+      "learning_rate": 1.9978014682920503e-05,
+      "loss": 0.6683,
+      "step": 1786
+    },
+    {
+      "epoch": 0.1098205506391347,
+      "grad_norm": 50.75,
+      "learning_rate": 1.9977860186376804e-05,
+      "loss": 0.6064,
+      "step": 1787
+    },
+    {
+      "epoch": 0.10988200589970501,
+      "grad_norm": 9.0,
+      "learning_rate": 1.9977705149489996e-05,
+      "loss": 0.3939,
+      "step": 1788
+    },
+    {
+      "epoch": 0.10994346116027531,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.997754957226847e-05,
+      "loss": 0.3928,
+      "step": 1789
+    },
+    {
+      "epoch": 0.11000491642084563,
+      "grad_norm": 23.375,
+      "learning_rate": 1.9977393454720646e-05,
+      "loss": 0.4789,
+      "step": 1790
+    },
+    {
+      "epoch": 0.11006637168141593,
+      "grad_norm": 12.5,
+      "learning_rate": 1.997723679685499e-05,
+      "loss": 0.4058,
+      "step": 1791
+    },
+    {
+      "epoch": 0.11012782694198624,
+      "grad_norm": 9.0,
+      "learning_rate": 1.9977079598679978e-05,
+      "loss": 0.4885,
+      "step": 1792
+    },
+    {
+      "epoch": 0.11018928220255654,
+      "grad_norm": 44.75,
+      "learning_rate": 1.997692186020413e-05,
+      "loss": 0.4216,
+      "step": 1793
+    },
+    {
+      "epoch": 0.11025073746312684,
+      "grad_norm": 39.0,
+      "learning_rate": 1.997676358143598e-05,
+      "loss": 0.4548,
+      "step": 1794
+    },
+    {
+      "epoch": 0.11031219272369715,
+      "grad_norm": 20.125,
+      "learning_rate": 1.99766047623841e-05,
+      "loss": 0.4144,
+      "step": 1795
+    },
+    {
+      "epoch": 0.11037364798426745,
+      "grad_norm": 41.0,
+      "learning_rate": 1.9976445403057095e-05,
+      "loss": 0.283,
+      "step": 1796
+    },
+    {
+      "epoch": 0.11043510324483775,
+      "grad_norm": 9.0,
+      "learning_rate": 1.9976285503463593e-05,
+      "loss": 0.3986,
+      "step": 1797
+    },
+    {
+      "epoch": 0.11049655850540806,
+      "grad_norm": 10.75,
+      "learning_rate": 1.9976125063612254e-05,
+      "loss": 0.3914,
+      "step": 1798
+    },
+    {
+      "epoch": 0.11055801376597837,
+      "grad_norm": 37.5,
+      "learning_rate": 1.9975964083511768e-05,
+      "loss": 0.4547,
+      "step": 1799
+    },
+    {
+      "epoch": 0.11061946902654868,
+      "grad_norm": 14.75,
+      "learning_rate": 1.9975802563170846e-05,
+      "loss": 0.403,
+      "step": 1800
+    },
+    {
+      "epoch": 0.11068092428711898,
+      "grad_norm": 15.0,
+      "learning_rate": 1.9975640502598243e-05,
+      "loss": 0.4122,
+      "step": 1801
+    },
+    {
+      "epoch": 0.11074237954768928,
+      "grad_norm": 24.125,
+      "learning_rate": 1.9975477901802734e-05,
+      "loss": 0.5104,
+      "step": 1802
+    },
+    {
+      "epoch": 0.11080383480825959,
+      "grad_norm": 17.125,
+      "learning_rate": 1.997531476079312e-05,
+      "loss": 0.4515,
+      "step": 1803
+    },
+    {
+      "epoch": 0.11086529006882989,
+      "grad_norm": 78.0,
+      "learning_rate": 1.9975151079578238e-05,
+      "loss": 0.4663,
+      "step": 1804
+    },
+    {
+      "epoch": 0.1109267453294002,
+      "grad_norm": 33.5,
+      "learning_rate": 1.9974986858166954e-05,
+      "loss": 0.4599,
+      "step": 1805
+    },
+    {
+      "epoch": 0.1109882005899705,
+      "grad_norm": 10.4375,
+      "learning_rate": 1.9974822096568157e-05,
+      "loss": 0.5938,
+      "step": 1806
+    },
+    {
+      "epoch": 0.1110496558505408,
+      "grad_norm": 58.5,
+      "learning_rate": 1.9974656794790777e-05,
+      "loss": 0.6857,
+      "step": 1807
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 25.25,
+      "learning_rate": 1.997449095284376e-05,
+      "loss": 0.5636,
+      "step": 1808
+    },
+    {
+      "epoch": 0.11117256637168142,
+      "grad_norm": 43.5,
+      "learning_rate": 1.9974324570736087e-05,
+      "loss": 0.7875,
+      "step": 1809
+    },
+    {
+      "epoch": 0.11123402163225173,
+      "grad_norm": 36.0,
+      "learning_rate": 1.9974157648476768e-05,
+      "loss": 0.6296,
+      "step": 1810
+    },
+    {
+      "epoch": 0.11129547689282203,
+      "grad_norm": 73.5,
+      "learning_rate": 1.9973990186074844e-05,
+      "loss": 0.7205,
+      "step": 1811
+    },
+    {
+      "epoch": 0.11135693215339233,
+      "grad_norm": 43.25,
+      "learning_rate": 1.9973822183539384e-05,
+      "loss": 0.6143,
+      "step": 1812
+    },
+    {
+      "epoch": 0.11141838741396264,
+      "grad_norm": 36.25,
+      "learning_rate": 1.9973653640879486e-05,
+      "loss": 0.7,
+      "step": 1813
+    },
+    {
+      "epoch": 0.11147984267453294,
+      "grad_norm": 44.25,
+      "learning_rate": 1.997348455810428e-05,
+      "loss": 0.5676,
+      "step": 1814
+    },
+    {
+      "epoch": 0.11154129793510324,
+      "grad_norm": 28.5,
+      "learning_rate": 1.9973314935222917e-05,
+      "loss": 0.6433,
+      "step": 1815
+    },
+    {
+      "epoch": 0.11160275319567355,
+      "grad_norm": 45.0,
+      "learning_rate": 1.997314477224458e-05,
+      "loss": 0.5414,
+      "step": 1816
+    },
+    {
+      "epoch": 0.11166420845624385,
+      "grad_norm": 45.5,
+      "learning_rate": 1.9972974069178498e-05,
+      "loss": 0.4317,
+      "step": 1817
+    },
+    {
+      "epoch": 0.11172566371681415,
+      "grad_norm": 25.125,
+      "learning_rate": 1.99728028260339e-05,
+      "loss": 0.4421,
+      "step": 1818
+    },
+    {
+      "epoch": 0.11178711897738447,
+      "grad_norm": 24.625,
+      "learning_rate": 1.997263104282007e-05,
+      "loss": 0.4677,
+      "step": 1819
+    },
+    {
+      "epoch": 0.11184857423795477,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.9972458719546307e-05,
+      "loss": 0.3855,
+      "step": 1820
+    },
+    {
+      "epoch": 0.11191002949852508,
+      "grad_norm": 27.25,
+      "learning_rate": 1.9972285856221944e-05,
+      "loss": 0.4087,
+      "step": 1821
+    },
+    {
+      "epoch": 0.11197148475909538,
+      "grad_norm": 6.40625,
+      "learning_rate": 1.997211245285634e-05,
+      "loss": 0.3288,
+      "step": 1822
+    },
+    {
+      "epoch": 0.11203294001966568,
+      "grad_norm": 21.125,
+      "learning_rate": 1.9971938509458888e-05,
+      "loss": 0.3195,
+      "step": 1823
+    },
+    {
+      "epoch": 0.11209439528023599,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9971764026039005e-05,
+      "loss": 0.5184,
+      "step": 1824
+    },
+    {
+      "epoch": 0.11215585054080629,
+      "grad_norm": 17.125,
+      "learning_rate": 1.997158900260614e-05,
+      "loss": 0.5094,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1122173058013766,
+      "grad_norm": 22.75,
+      "learning_rate": 1.9971413439169777e-05,
+      "loss": 0.5112,
+      "step": 1826
+    },
+    {
+      "epoch": 0.1122787610619469,
+      "grad_norm": 26.5,
+      "learning_rate": 1.9971237335739416e-05,
+      "loss": 0.5052,
+      "step": 1827
+    },
+    {
+      "epoch": 0.11234021632251721,
+      "grad_norm": 18.25,
+      "learning_rate": 1.99710606923246e-05,
+      "loss": 0.5202,
+      "step": 1828
+    },
+    {
+      "epoch": 0.11240167158308752,
+      "grad_norm": 17.625,
+      "learning_rate": 1.997088350893489e-05,
+      "loss": 0.5362,
+      "step": 1829
+    },
+    {
+      "epoch": 0.11246312684365782,
+      "grad_norm": 26.75,
+      "learning_rate": 1.9970705785579882e-05,
+      "loss": 0.6664,
+      "step": 1830
+    },
+    {
+      "epoch": 0.11252458210422812,
+      "grad_norm": 24.875,
+      "learning_rate": 1.9970527522269204e-05,
+      "loss": 0.5387,
+      "step": 1831
+    },
+    {
+      "epoch": 0.11258603736479843,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.997034871901251e-05,
+      "loss": 0.3272,
+      "step": 1832
+    },
+    {
+      "epoch": 0.11264749262536873,
+      "grad_norm": 18.625,
+      "learning_rate": 1.9970169375819475e-05,
+      "loss": 0.6959,
+      "step": 1833
+    },
+    {
+      "epoch": 0.11270894788593903,
+      "grad_norm": 17.625,
+      "learning_rate": 1.996998949269982e-05,
+      "loss": 0.6524,
+      "step": 1834
+    },
+    {
+      "epoch": 0.11277040314650934,
+      "grad_norm": 25.0,
+      "learning_rate": 1.996980906966328e-05,
+      "loss": 0.8149,
+      "step": 1835
+    },
+    {
+      "epoch": 0.11283185840707964,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9969628106719632e-05,
+      "loss": 0.6222,
+      "step": 1836
+    },
+    {
+      "epoch": 0.11289331366764994,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.9969446603878673e-05,
+      "loss": 0.4338,
+      "step": 1837
+    },
+    {
+      "epoch": 0.11295476892822026,
+      "grad_norm": 25.75,
+      "learning_rate": 1.996926456115023e-05,
+      "loss": 0.6691,
+      "step": 1838
+    },
+    {
+      "epoch": 0.11301622418879056,
+      "grad_norm": 13.0,
+      "learning_rate": 1.9969081978544163e-05,
+      "loss": 0.5539,
+      "step": 1839
+    },
+    {
+      "epoch": 0.11307767944936087,
+      "grad_norm": 30.5,
+      "learning_rate": 1.996889885607036e-05,
+      "loss": 0.7009,
+      "step": 1840
+    },
+    {
+      "epoch": 0.11313913470993117,
+      "grad_norm": 43.5,
+      "learning_rate": 1.9968715193738738e-05,
+      "loss": 0.74,
+      "step": 1841
+    },
+    {
+      "epoch": 0.11320058997050148,
+      "grad_norm": 41.25,
+      "learning_rate": 1.9968530991559244e-05,
+      "loss": 0.592,
+      "step": 1842
+    },
+    {
+      "epoch": 0.11326204523107178,
+      "grad_norm": 39.75,
+      "learning_rate": 1.9968346249541848e-05,
+      "loss": 0.5829,
+      "step": 1843
+    },
+    {
+      "epoch": 0.11332350049164208,
+      "grad_norm": 30.875,
+      "learning_rate": 1.9968160967696565e-05,
+      "loss": 0.6194,
+      "step": 1844
+    },
+    {
+      "epoch": 0.11338495575221239,
+      "grad_norm": 35.75,
+      "learning_rate": 1.9967975146033417e-05,
+      "loss": 0.7347,
+      "step": 1845
+    },
+    {
+      "epoch": 0.11344641101278269,
+      "grad_norm": 10.5625,
+      "learning_rate": 1.9967788784562474e-05,
+      "loss": 0.5696,
+      "step": 1846
+    },
+    {
+      "epoch": 0.113507866273353,
+      "grad_norm": 41.75,
+      "learning_rate": 1.9967601883293826e-05,
+      "loss": 0.5031,
+      "step": 1847
+    },
+    {
+      "epoch": 0.11356932153392331,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.9967414442237595e-05,
+      "loss": 0.4634,
+      "step": 1848
+    },
+    {
+      "epoch": 0.11363077679449361,
+      "grad_norm": 27.375,
+      "learning_rate": 1.9967226461403934e-05,
+      "loss": 0.5535,
+      "step": 1849
+    },
+    {
+      "epoch": 0.11369223205506392,
+      "grad_norm": 32.25,
+      "learning_rate": 1.9967037940803016e-05,
+      "loss": 0.6167,
+      "step": 1850
+    },
+    {
+      "epoch": 0.11375368731563422,
+      "grad_norm": 18.75,
+      "learning_rate": 1.996684888044506e-05,
+      "loss": 0.5572,
+      "step": 1851
+    },
+    {
+      "epoch": 0.11381514257620452,
+      "grad_norm": 12.75,
+      "learning_rate": 1.99666592803403e-05,
+      "loss": 0.4397,
+      "step": 1852
+    },
+    {
+      "epoch": 0.11387659783677483,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.9966469140499e-05,
+      "loss": 0.609,
+      "step": 1853
+    },
+    {
+      "epoch": 0.11393805309734513,
+      "grad_norm": 13.625,
+      "learning_rate": 1.9966278460931463e-05,
+      "loss": 0.3777,
+      "step": 1854
+    },
+    {
+      "epoch": 0.11399950835791543,
+      "grad_norm": 20.125,
+      "learning_rate": 1.996608724164801e-05,
+      "loss": 0.4651,
+      "step": 1855
+    },
+    {
+      "epoch": 0.11406096361848574,
+      "grad_norm": 12.5,
+      "learning_rate": 1.9965895482659e-05,
+      "loss": 0.5072,
+      "step": 1856
+    },
+    {
+      "epoch": 0.11412241887905605,
+      "grad_norm": 20.0,
+      "learning_rate": 1.9965703183974813e-05,
+      "loss": 0.6233,
+      "step": 1857
+    },
+    {
+      "epoch": 0.11418387413962636,
+      "grad_norm": 20.875,
+      "learning_rate": 1.9965510345605866e-05,
+      "loss": 0.6148,
+      "step": 1858
+    },
+    {
+      "epoch": 0.11424532940019666,
+      "grad_norm": 24.125,
+      "learning_rate": 1.9965316967562606e-05,
+      "loss": 0.7652,
+      "step": 1859
+    },
+    {
+      "epoch": 0.11430678466076696,
+      "grad_norm": 42.75,
+      "learning_rate": 1.9965123049855498e-05,
+      "loss": 0.7911,
+      "step": 1860
+    },
+    {
+      "epoch": 0.11436823992133727,
+      "grad_norm": 22.375,
+      "learning_rate": 1.9964928592495046e-05,
+      "loss": 0.7689,
+      "step": 1861
+    },
+    {
+      "epoch": 0.11442969518190757,
+      "grad_norm": 18.0,
+      "learning_rate": 1.996473359549178e-05,
+      "loss": 0.5049,
+      "step": 1862
+    },
+    {
+      "epoch": 0.11449115044247787,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9964538058856263e-05,
+      "loss": 0.6148,
+      "step": 1863
+    },
+    {
+      "epoch": 0.11455260570304818,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.996434198259908e-05,
+      "loss": 0.435,
+      "step": 1864
+    },
+    {
+      "epoch": 0.11461406096361848,
+      "grad_norm": 34.25,
+      "learning_rate": 1.9964145366730854e-05,
+      "loss": 0.4826,
+      "step": 1865
+    },
+    {
+      "epoch": 0.11467551622418878,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.9963948211262233e-05,
+      "loss": 0.5486,
+      "step": 1866
+    },
+    {
+      "epoch": 0.1147369714847591,
+      "grad_norm": 13.875,
+      "learning_rate": 1.9963750516203887e-05,
+      "loss": 0.522,
+      "step": 1867
+    },
+    {
+      "epoch": 0.1147984267453294,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.9963552281566523e-05,
+      "loss": 0.5836,
+      "step": 1868
+    },
+    {
+      "epoch": 0.11485988200589971,
+      "grad_norm": 36.5,
+      "learning_rate": 1.9963353507360884e-05,
+      "loss": 0.4916,
+      "step": 1869
+    },
+    {
+      "epoch": 0.11492133726647001,
+      "grad_norm": 29.0,
+      "learning_rate": 1.9963154193597728e-05,
+      "loss": 0.5142,
+      "step": 1870
+    },
+    {
+      "epoch": 0.11498279252704031,
+      "grad_norm": 14.5,
+      "learning_rate": 1.996295434028785e-05,
+      "loss": 0.3301,
+      "step": 1871
+    },
+    {
+      "epoch": 0.11504424778761062,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.9962753947442072e-05,
+      "loss": 0.3774,
+      "step": 1872
+    },
+    {
+      "epoch": 0.11510570304818092,
+      "grad_norm": 29.875,
+      "learning_rate": 1.996255301507125e-05,
+      "loss": 0.6848,
+      "step": 1873
+    },
+    {
+      "epoch": 0.11516715830875122,
+      "grad_norm": 5.9375,
+      "learning_rate": 1.996235154318626e-05,
+      "loss": 0.717,
+      "step": 1874
+    },
+    {
+      "epoch": 0.11522861356932153,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.9962149531798015e-05,
+      "loss": 0.6016,
+      "step": 1875
+    },
+    {
+      "epoch": 0.11529006882989185,
+      "grad_norm": 11.5625,
+      "learning_rate": 1.9961946980917457e-05,
+      "loss": 0.5468,
+      "step": 1876
+    },
+    {
+      "epoch": 0.11535152409046215,
+      "grad_norm": 33.75,
+      "learning_rate": 1.9961743890555553e-05,
+      "loss": 0.4646,
+      "step": 1877
+    },
+    {
+      "epoch": 0.11541297935103245,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.99615402607233e-05,
+      "loss": 0.4641,
+      "step": 1878
+    },
+    {
+      "epoch": 0.11547443461160276,
+      "grad_norm": 13.625,
+      "learning_rate": 1.9961336091431728e-05,
+      "loss": 0.5797,
+      "step": 1879
+    },
+    {
+      "epoch": 0.11553588987217306,
+      "grad_norm": 11.125,
+      "learning_rate": 1.996113138269189e-05,
+      "loss": 0.3914,
+      "step": 1880
+    },
+    {
+      "epoch": 0.11559734513274336,
+      "grad_norm": 40.5,
+      "learning_rate": 1.9960926134514875e-05,
+      "loss": 0.6931,
+      "step": 1881
+    },
+    {
+      "epoch": 0.11565880039331367,
+      "grad_norm": 53.0,
+      "learning_rate": 1.9960720346911798e-05,
+      "loss": 0.7989,
+      "step": 1882
+    },
+    {
+      "epoch": 0.11572025565388397,
+      "grad_norm": 48.25,
+      "learning_rate": 1.99605140198938e-05,
+      "loss": 0.8516,
+      "step": 1883
+    },
+    {
+      "epoch": 0.11578171091445427,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.996030715347206e-05,
+      "loss": 0.5719,
+      "step": 1884
+    },
+    {
+      "epoch": 0.11584316617502458,
+      "grad_norm": 12.375,
+      "learning_rate": 1.9960099747657774e-05,
+      "loss": 0.6435,
+      "step": 1885
+    },
+    {
+      "epoch": 0.11590462143559489,
+      "grad_norm": 12.75,
+      "learning_rate": 1.995989180246218e-05,
+      "loss": 0.509,
+      "step": 1886
+    },
+    {
+      "epoch": 0.1159660766961652,
+      "grad_norm": 22.25,
+      "learning_rate": 1.9959683317896535e-05,
+      "loss": 0.5274,
+      "step": 1887
+    },
+    {
+      "epoch": 0.1160275319567355,
+      "grad_norm": 47.75,
+      "learning_rate": 1.995947429397213e-05,
+      "loss": 0.7349,
+      "step": 1888
+    },
+    {
+      "epoch": 0.1160889872173058,
+      "grad_norm": 33.0,
+      "learning_rate": 1.995926473070029e-05,
+      "loss": 0.4721,
+      "step": 1889
+    },
+    {
+      "epoch": 0.1161504424778761,
+      "grad_norm": 18.75,
+      "learning_rate": 1.9959054628092353e-05,
+      "loss": 0.5948,
+      "step": 1890
+    },
+    {
+      "epoch": 0.11621189773844641,
+      "grad_norm": 46.5,
+      "learning_rate": 1.9958843986159705e-05,
+      "loss": 0.5243,
+      "step": 1891
+    },
+    {
+      "epoch": 0.11627335299901671,
+      "grad_norm": 65.5,
+      "learning_rate": 1.9958632804913753e-05,
+      "loss": 0.5539,
+      "step": 1892
+    },
+    {
+      "epoch": 0.11633480825958702,
+      "grad_norm": 33.5,
+      "learning_rate": 1.9958421084365928e-05,
+      "loss": 0.4856,
+      "step": 1893
+    },
+    {
+      "epoch": 0.11639626352015732,
+      "grad_norm": 30.125,
+      "learning_rate": 1.9958208824527702e-05,
+      "loss": 0.5976,
+      "step": 1894
+    },
+    {
+      "epoch": 0.11645771878072764,
+      "grad_norm": 63.25,
+      "learning_rate": 1.9957996025410565e-05,
+      "loss": 0.4858,
+      "step": 1895
+    },
+    {
+      "epoch": 0.11651917404129794,
+      "grad_norm": 27.25,
+      "learning_rate": 1.9957782687026046e-05,
+      "loss": 0.5822,
+      "step": 1896
+    },
+    {
+      "epoch": 0.11658062930186824,
+      "grad_norm": 46.25,
+      "learning_rate": 1.9957568809385693e-05,
+      "loss": 0.4152,
+      "step": 1897
+    },
+    {
+      "epoch": 0.11664208456243855,
+      "grad_norm": 45.25,
+      "learning_rate": 1.9957354392501092e-05,
+      "loss": 0.4972,
+      "step": 1898
+    },
+    {
+      "epoch": 0.11670353982300885,
+      "grad_norm": 44.75,
+      "learning_rate": 1.995713943638385e-05,
+      "loss": 0.5448,
+      "step": 1899
+    },
+    {
+      "epoch": 0.11676499508357915,
+      "grad_norm": 33.25,
+      "learning_rate": 1.9956923941045613e-05,
+      "loss": 0.4045,
+      "step": 1900
+    },
+    {
+      "epoch": 0.11682645034414946,
+      "grad_norm": 38.75,
+      "learning_rate": 1.9956707906498046e-05,
+      "loss": 0.4992,
+      "step": 1901
+    },
+    {
+      "epoch": 0.11688790560471976,
+      "grad_norm": 20.75,
+      "learning_rate": 1.995649133275285e-05,
+      "loss": 0.4209,
+      "step": 1902
+    },
+    {
+      "epoch": 0.11694936086529006,
+      "grad_norm": 35.0,
+      "learning_rate": 1.995627421982176e-05,
+      "loss": 0.4744,
+      "step": 1903
+    },
+    {
+      "epoch": 0.11701081612586037,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.995605656771652e-05,
+      "loss": 0.4039,
+      "step": 1904
+    },
+    {
+      "epoch": 0.11707227138643068,
+      "grad_norm": 23.0,
+      "learning_rate": 1.9955838376448933e-05,
+      "loss": 0.4875,
+      "step": 1905
+    },
+    {
+      "epoch": 0.11713372664700099,
+      "grad_norm": 23.375,
+      "learning_rate": 1.99556196460308e-05,
+      "loss": 0.6234,
+      "step": 1906
+    },
+    {
+      "epoch": 0.11719518190757129,
+      "grad_norm": 26.625,
+      "learning_rate": 1.9955400376473977e-05,
+      "loss": 0.5063,
+      "step": 1907
+    },
+    {
+      "epoch": 0.1172566371681416,
+      "grad_norm": 20.75,
+      "learning_rate": 1.9955180567790332e-05,
+      "loss": 0.6176,
+      "step": 1908
+    },
+    {
+      "epoch": 0.1173180924287119,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.995496021999177e-05,
+      "loss": 0.4577,
+      "step": 1909
+    },
+    {
+      "epoch": 0.1173795476892822,
+      "grad_norm": 24.625,
+      "learning_rate": 1.995473933309023e-05,
+      "loss": 0.4337,
+      "step": 1910
+    },
+    {
+      "epoch": 0.1174410029498525,
+      "grad_norm": 7.25,
+      "learning_rate": 1.9954517907097663e-05,
+      "loss": 0.4849,
+      "step": 1911
+    },
+    {
+      "epoch": 0.11750245821042281,
+      "grad_norm": 16.125,
+      "learning_rate": 1.9954295942026065e-05,
+      "loss": 0.585,
+      "step": 1912
+    },
+    {
+      "epoch": 0.11756391347099311,
+      "grad_norm": 45.25,
+      "learning_rate": 1.9954073437887457e-05,
+      "loss": 0.5748,
+      "step": 1913
+    },
+    {
+      "epoch": 0.11762536873156341,
+      "grad_norm": 38.25,
+      "learning_rate": 1.995385039469389e-05,
+      "loss": 0.4938,
+      "step": 1914
+    },
+    {
+      "epoch": 0.11768682399213373,
+      "grad_norm": 29.875,
+      "learning_rate": 1.995362681245744e-05,
+      "loss": 0.7121,
+      "step": 1915
+    },
+    {
+      "epoch": 0.11774827925270404,
+      "grad_norm": 34.5,
+      "learning_rate": 1.9953402691190218e-05,
+      "loss": 0.5834,
+      "step": 1916
+    },
+    {
+      "epoch": 0.11780973451327434,
+      "grad_norm": 9.125,
+      "learning_rate": 1.9953178030904357e-05,
+      "loss": 0.5756,
+      "step": 1917
+    },
+    {
+      "epoch": 0.11787118977384464,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9952952831612027e-05,
+      "loss": 0.531,
+      "step": 1918
+    },
+    {
+      "epoch": 0.11793264503441495,
+      "grad_norm": 20.25,
+      "learning_rate": 1.9952727093325423e-05,
+      "loss": 0.3872,
+      "step": 1919
+    },
+    {
+      "epoch": 0.11799410029498525,
+      "grad_norm": 21.25,
+      "learning_rate": 1.9952500816056765e-05,
+      "loss": 0.3882,
+      "step": 1920
+    },
+    {
+      "epoch": 0.11805555555555555,
+      "grad_norm": 5.21875,
+      "learning_rate": 1.9952273999818312e-05,
+      "loss": 0.527,
+      "step": 1921
+    },
+    {
+      "epoch": 0.11811701081612586,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.9952046644622346e-05,
+      "loss": 0.4709,
+      "step": 1922
+    },
+    {
+      "epoch": 0.11817846607669616,
+      "grad_norm": 27.5,
+      "learning_rate": 1.995181875048118e-05,
+      "loss": 0.3579,
+      "step": 1923
+    },
+    {
+      "epoch": 0.11823992133726648,
+      "grad_norm": 30.875,
+      "learning_rate": 1.9951590317407152e-05,
+      "loss": 0.4172,
+      "step": 1924
+    },
+    {
+      "epoch": 0.11830137659783678,
+      "grad_norm": 33.0,
+      "learning_rate": 1.9951361345412636e-05,
+      "loss": 0.6696,
+      "step": 1925
+    },
+    {
+      "epoch": 0.11836283185840708,
+      "grad_norm": 31.125,
+      "learning_rate": 1.9951131834510034e-05,
+      "loss": 0.3767,
+      "step": 1926
+    },
+    {
+      "epoch": 0.11842428711897739,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.9950901784711765e-05,
+      "loss": 0.4291,
+      "step": 1927
+    },
+    {
+      "epoch": 0.11848574237954769,
+      "grad_norm": 69.0,
+      "learning_rate": 1.9950671196030298e-05,
+      "loss": 0.4201,
+      "step": 1928
+    },
+    {
+      "epoch": 0.118547197640118,
+      "grad_norm": 28.75,
+      "learning_rate": 1.9950440068478115e-05,
+      "loss": 0.7555,
+      "step": 1929
+    },
+    {
+      "epoch": 0.1186086529006883,
+      "grad_norm": 58.75,
+      "learning_rate": 1.9950208402067735e-05,
+      "loss": 0.7644,
+      "step": 1930
+    },
+    {
+      "epoch": 0.1186701081612586,
+      "grad_norm": 47.25,
+      "learning_rate": 1.99499761968117e-05,
+      "loss": 0.7416,
+      "step": 1931
+    },
+    {
+      "epoch": 0.1187315634218289,
+      "grad_norm": 38.25,
+      "learning_rate": 1.994974345272259e-05,
+      "loss": 0.498,
+      "step": 1932
+    },
+    {
+      "epoch": 0.1187930186823992,
+      "grad_norm": 36.5,
+      "learning_rate": 1.9949510169813006e-05,
+      "loss": 0.6546,
+      "step": 1933
+    },
+    {
+      "epoch": 0.11885447394296952,
+      "grad_norm": 38.75,
+      "learning_rate": 1.9949276348095576e-05,
+      "loss": 0.6238,
+      "step": 1934
+    },
+    {
+      "epoch": 0.11891592920353983,
+      "grad_norm": 21.625,
+      "learning_rate": 1.9949041987582975e-05,
+      "loss": 0.5189,
+      "step": 1935
+    },
+    {
+      "epoch": 0.11897738446411013,
+      "grad_norm": 23.75,
+      "learning_rate": 1.9948807088287884e-05,
+      "loss": 0.4802,
+      "step": 1936
+    },
+    {
+      "epoch": 0.11903883972468043,
+      "grad_norm": 41.25,
+      "learning_rate": 1.9948571650223027e-05,
+      "loss": 0.6451,
+      "step": 1937
+    },
+    {
+      "epoch": 0.11910029498525074,
+      "grad_norm": 27.25,
+      "learning_rate": 1.9948335673401157e-05,
+      "loss": 0.6864,
+      "step": 1938
+    },
+    {
+      "epoch": 0.11916175024582104,
+      "grad_norm": 38.0,
+      "learning_rate": 1.994809915783505e-05,
+      "loss": 0.5372,
+      "step": 1939
+    },
+    {
+      "epoch": 0.11922320550639134,
+      "grad_norm": 60.0,
+      "learning_rate": 1.994786210353751e-05,
+      "loss": 0.5996,
+      "step": 1940
+    },
+    {
+      "epoch": 0.11928466076696165,
+      "grad_norm": 35.75,
+      "learning_rate": 1.9947624510521385e-05,
+      "loss": 0.4957,
+      "step": 1941
+    },
+    {
+      "epoch": 0.11934611602753195,
+      "grad_norm": 25.625,
+      "learning_rate": 1.9947386378799534e-05,
+      "loss": 0.4601,
+      "step": 1942
+    },
+    {
+      "epoch": 0.11940757128810227,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.9947147708384858e-05,
+      "loss": 0.3769,
+      "step": 1943
+    },
+    {
+      "epoch": 0.11946902654867257,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.9946908499290275e-05,
+      "loss": 0.4424,
+      "step": 1944
+    },
+    {
+      "epoch": 0.11953048180924288,
+      "grad_norm": 21.0,
+      "learning_rate": 1.9946668751528745e-05,
+      "loss": 0.501,
+      "step": 1945
+    },
+    {
+      "epoch": 0.11959193706981318,
+      "grad_norm": 25.75,
+      "learning_rate": 1.9946428465113244e-05,
+      "loss": 0.4752,
+      "step": 1946
+    },
+    {
+      "epoch": 0.11965339233038348,
+      "grad_norm": 21.625,
+      "learning_rate": 1.9946187640056794e-05,
+      "loss": 0.5517,
+      "step": 1947
+    },
+    {
+      "epoch": 0.11971484759095379,
+      "grad_norm": 23.625,
+      "learning_rate": 1.9945946276372435e-05,
+      "loss": 0.4254,
+      "step": 1948
+    },
+    {
+      "epoch": 0.11977630285152409,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.994570437407323e-05,
+      "loss": 0.463,
+      "step": 1949
+    },
+    {
+      "epoch": 0.11983775811209439,
+      "grad_norm": 5.1875,
+      "learning_rate": 1.994546193317229e-05,
+      "loss": 0.4918,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1198992133726647,
+      "grad_norm": 12.625,
+      "learning_rate": 1.9945218953682736e-05,
+      "loss": 0.3255,
+      "step": 1951
+    },
+    {
+      "epoch": 0.119960668633235,
+      "grad_norm": 20.0,
+      "learning_rate": 1.994497543561773e-05,
+      "loss": 0.2954,
+      "step": 1952
+    },
+    {
+      "epoch": 0.12002212389380532,
+      "grad_norm": 28.25,
+      "learning_rate": 1.9944731378990452e-05,
+      "loss": 0.6378,
+      "step": 1953
+    },
+    {
+      "epoch": 0.12008357915437562,
+      "grad_norm": 25.625,
+      "learning_rate": 1.9944486783814135e-05,
+      "loss": 0.526,
+      "step": 1954
+    },
+    {
+      "epoch": 0.12014503441494592,
+      "grad_norm": 15.25,
+      "learning_rate": 1.994424165010201e-05,
+      "loss": 0.3548,
+      "step": 1955
+    },
+    {
+      "epoch": 0.12020648967551623,
+      "grad_norm": 10.8125,
+      "learning_rate": 1.9943995977867358e-05,
+      "loss": 0.5702,
+      "step": 1956
+    },
+    {
+      "epoch": 0.12026794493608653,
+      "grad_norm": 9.375,
+      "learning_rate": 1.994374976712348e-05,
+      "loss": 0.4634,
+      "step": 1957
+    },
+    {
+      "epoch": 0.12032940019665683,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.9943503017883716e-05,
+      "loss": 0.5298,
+      "step": 1958
+    },
+    {
+      "epoch": 0.12039085545722714,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.994325573016142e-05,
+      "loss": 0.3792,
+      "step": 1959
+    },
+    {
+      "epoch": 0.12045231071779744,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.994300790396999e-05,
+      "loss": 0.5573,
+      "step": 1960
+    },
+    {
+      "epoch": 0.12051376597836774,
+      "grad_norm": 36.0,
+      "learning_rate": 1.9942759539322845e-05,
+      "loss": 0.5288,
+      "step": 1961
+    },
+    {
+      "epoch": 0.12057522123893805,
+      "grad_norm": 13.375,
+      "learning_rate": 1.9942510636233435e-05,
+      "loss": 0.4077,
+      "step": 1962
+    },
+    {
+      "epoch": 0.12063667649950836,
+      "grad_norm": 48.0,
+      "learning_rate": 1.9942261194715236e-05,
+      "loss": 0.6043,
+      "step": 1963
+    },
+    {
+      "epoch": 0.12069813176007867,
+      "grad_norm": 15.75,
+      "learning_rate": 1.994201121478176e-05,
+      "loss": 0.4062,
+      "step": 1964
+    },
+    {
+      "epoch": 0.12075958702064897,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.9941760696446546e-05,
+      "loss": 0.3833,
+      "step": 1965
+    },
+    {
+      "epoch": 0.12082104228121927,
+      "grad_norm": 40.0,
+      "learning_rate": 1.9941509639723155e-05,
+      "loss": 0.65,
+      "step": 1966
+    },
+    {
+      "epoch": 0.12088249754178958,
+      "grad_norm": 51.25,
+      "learning_rate": 1.994125804462519e-05,
+      "loss": 0.4981,
+      "step": 1967
+    },
+    {
+      "epoch": 0.12094395280235988,
+      "grad_norm": 24.625,
+      "learning_rate": 1.994100591116627e-05,
+      "loss": 0.476,
+      "step": 1968
+    },
+    {
+      "epoch": 0.12100540806293018,
+      "grad_norm": 60.5,
+      "learning_rate": 1.9940753239360047e-05,
+      "loss": 0.6537,
+      "step": 1969
+    },
+    {
+      "epoch": 0.12106686332350049,
+      "grad_norm": 33.5,
+      "learning_rate": 1.994050002922021e-05,
+      "loss": 0.4306,
+      "step": 1970
+    },
+    {
+      "epoch": 0.12112831858407079,
+      "grad_norm": 48.0,
+      "learning_rate": 1.9940246280760473e-05,
+      "loss": 0.5095,
+      "step": 1971
+    },
+    {
+      "epoch": 0.12118977384464111,
+      "grad_norm": 58.25,
+      "learning_rate": 1.993999199399457e-05,
+      "loss": 0.506,
+      "step": 1972
+    },
+    {
+      "epoch": 0.12125122910521141,
+      "grad_norm": 36.5,
+      "learning_rate": 1.993973716893628e-05,
+      "loss": 0.4068,
+      "step": 1973
+    },
+    {
+      "epoch": 0.12131268436578171,
+      "grad_norm": 22.0,
+      "learning_rate": 1.9939481805599398e-05,
+      "loss": 0.4259,
+      "step": 1974
+    },
+    {
+      "epoch": 0.12137413962635202,
+      "grad_norm": 18.25,
+      "learning_rate": 1.9939225903997748e-05,
+      "loss": 0.4216,
+      "step": 1975
+    },
+    {
+      "epoch": 0.12143559488692232,
+      "grad_norm": 11.875,
+      "learning_rate": 1.99389694641452e-05,
+      "loss": 0.4085,
+      "step": 1976
+    },
+    {
+      "epoch": 0.12149705014749262,
+      "grad_norm": 48.25,
+      "learning_rate": 1.9938712486055632e-05,
+      "loss": 0.5013,
+      "step": 1977
+    },
+    {
+      "epoch": 0.12155850540806293,
+      "grad_norm": 22.5,
+      "learning_rate": 1.993845496974297e-05,
+      "loss": 0.418,
+      "step": 1978
+    },
+    {
+      "epoch": 0.12161996066863323,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.9938196915221147e-05,
+      "loss": 0.3448,
+      "step": 1979
+    },
+    {
+      "epoch": 0.12168141592920353,
+      "grad_norm": 13.25,
+      "learning_rate": 1.9937938322504144e-05,
+      "loss": 0.4219,
+      "step": 1980
+    },
+    {
+      "epoch": 0.12174287118977384,
+      "grad_norm": 11.5625,
+      "learning_rate": 1.9937679191605964e-05,
+      "loss": 0.3498,
+      "step": 1981
+    },
+    {
+      "epoch": 0.12180432645034416,
+      "grad_norm": 5.3125,
+      "learning_rate": 1.9937419522540644e-05,
+      "loss": 0.3105,
+      "step": 1982
+    },
+    {
+      "epoch": 0.12186578171091446,
+      "grad_norm": 43.5,
+      "learning_rate": 1.9937159315322242e-05,
+      "loss": 0.4721,
+      "step": 1983
+    },
+    {
+      "epoch": 0.12192723697148476,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.993689856996485e-05,
+      "loss": 0.3457,
+      "step": 1984
+    },
+    {
+      "epoch": 0.12198869223205507,
+      "grad_norm": 46.0,
+      "learning_rate": 1.993663728648259e-05,
+      "loss": 0.6441,
+      "step": 1985
+    },
+    {
+      "epoch": 0.12205014749262537,
+      "grad_norm": 51.75,
+      "learning_rate": 1.9936375464889608e-05,
+      "loss": 0.6362,
+      "step": 1986
+    },
+    {
+      "epoch": 0.12211160275319567,
+      "grad_norm": 27.5,
+      "learning_rate": 1.9936113105200085e-05,
+      "loss": 0.594,
+      "step": 1987
+    },
+    {
+      "epoch": 0.12217305801376598,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.993585020742823e-05,
+      "loss": 0.767,
+      "step": 1988
+    },
+    {
+      "epoch": 0.12223451327433628,
+      "grad_norm": 64.5,
+      "learning_rate": 1.993558677158828e-05,
+      "loss": 0.4731,
+      "step": 1989
+    },
+    {
+      "epoch": 0.12229596853490658,
+      "grad_norm": 37.25,
+      "learning_rate": 1.99353227976945e-05,
+      "loss": 0.6903,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1223574237954769,
+      "grad_norm": 23.0,
+      "learning_rate": 1.9935058285761185e-05,
+      "loss": 0.3168,
+      "step": 1991
+    },
+    {
+      "epoch": 0.1224188790560472,
+      "grad_norm": 25.625,
+      "learning_rate": 1.993479323580266e-05,
+      "loss": 0.527,
+      "step": 1992
+    },
+    {
+      "epoch": 0.1224803343166175,
+      "grad_norm": 24.125,
+      "learning_rate": 1.9934527647833276e-05,
+      "loss": 0.8246,
+      "step": 1993
+    },
+    {
+      "epoch": 0.12254178957718781,
+      "grad_norm": 30.5,
+      "learning_rate": 1.993426152186742e-05,
+      "loss": 0.7521,
+      "step": 1994
+    },
+    {
+      "epoch": 0.12260324483775811,
+      "grad_norm": 17.75,
+      "learning_rate": 1.9933994857919505e-05,
+      "loss": 0.5836,
+      "step": 1995
+    },
+    {
+      "epoch": 0.12266470009832842,
+      "grad_norm": 22.0,
+      "learning_rate": 1.9933727656003964e-05,
+      "loss": 0.5834,
+      "step": 1996
+    },
+    {
+      "epoch": 0.12272615535889872,
+      "grad_norm": 26.5,
+      "learning_rate": 1.9933459916135277e-05,
+      "loss": 0.6181,
+      "step": 1997
+    },
+    {
+      "epoch": 0.12278761061946902,
+      "grad_norm": 16.375,
+      "learning_rate": 1.9933191638327935e-05,
+      "loss": 0.5982,
+      "step": 1998
+    },
+    {
+      "epoch": 0.12284906588003933,
+      "grad_norm": 22.625,
+      "learning_rate": 1.993292282259647e-05,
+      "loss": 0.5736,
+      "step": 1999
+    },
+    {
+      "epoch": 0.12291052114060963,
+      "grad_norm": 16.875,
+      "learning_rate": 1.9932653468955442e-05,
+      "loss": 0.4823,
+      "step": 2000
+    },
+    {
+      "epoch": 0.12297197640117995,
+      "grad_norm": 21.625,
+      "learning_rate": 1.9932383577419432e-05,
+      "loss": 0.6774,
+      "step": 2001
+    },
+    {
+      "epoch": 0.12303343166175025,
+      "grad_norm": 20.75,
+      "learning_rate": 1.9932113148003057e-05,
+      "loss": 0.5491,
+      "step": 2002
+    },
+    {
+      "epoch": 0.12309488692232055,
+      "grad_norm": 22.125,
+      "learning_rate": 1.9931842180720967e-05,
+      "loss": 0.5375,
+      "step": 2003
+    },
+    {
+      "epoch": 0.12315634218289086,
+      "grad_norm": 13.625,
+      "learning_rate": 1.9931570675587833e-05,
+      "loss": 0.4592,
+      "step": 2004
+    },
+    {
+      "epoch": 0.12321779744346116,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9931298632618355e-05,
+      "loss": 0.4537,
+      "step": 2005
+    },
+    {
+      "epoch": 0.12327925270403146,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.993102605182727e-05,
+      "loss": 0.4392,
+      "step": 2006
+    },
+    {
+      "epoch": 0.12334070796460177,
+      "grad_norm": 19.625,
+      "learning_rate": 1.9930752933229335e-05,
+      "loss": 0.5462,
+      "step": 2007
+    },
+    {
+      "epoch": 0.12340216322517207,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.9930479276839347e-05,
+      "loss": 0.4,
+      "step": 2008
+    },
+    {
+      "epoch": 0.12346361848574237,
+      "grad_norm": 24.0,
+      "learning_rate": 1.9930205082672115e-05,
+      "loss": 0.513,
+      "step": 2009
+    },
+    {
+      "epoch": 0.12352507374631268,
+      "grad_norm": 17.0,
+      "learning_rate": 1.99299303507425e-05,
+      "loss": 0.4483,
+      "step": 2010
+    },
+    {
+      "epoch": 0.123586529006883,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.992965508106537e-05,
+      "loss": 0.4814,
+      "step": 2011
+    },
+    {
+      "epoch": 0.1236479842674533,
+      "grad_norm": 31.5,
+      "learning_rate": 1.992937927365564e-05,
+      "loss": 0.507,
+      "step": 2012
+    },
+    {
+      "epoch": 0.1237094395280236,
+      "grad_norm": 22.75,
+      "learning_rate": 1.992910292852824e-05,
+      "loss": 0.4852,
+      "step": 2013
+    },
+    {
+      "epoch": 0.1237708947885939,
+      "grad_norm": 11.5,
+      "learning_rate": 1.9928826045698138e-05,
+      "loss": 0.4664,
+      "step": 2014
+    },
+    {
+      "epoch": 0.12383235004916421,
+      "grad_norm": 10.0,
+      "learning_rate": 1.992854862518033e-05,
+      "loss": 0.4074,
+      "step": 2015
+    },
+    {
+      "epoch": 0.12389380530973451,
+      "grad_norm": 15.75,
+      "learning_rate": 1.9928270666989835e-05,
+      "loss": 0.4631,
+      "step": 2016
+    },
+    {
+      "epoch": 0.12395526057030481,
+      "grad_norm": 24.125,
+      "learning_rate": 1.9927992171141707e-05,
+      "loss": 0.7821,
+      "step": 2017
+    },
+    {
+      "epoch": 0.12401671583087512,
+      "grad_norm": 32.5,
+      "learning_rate": 1.992771313765103e-05,
+      "loss": 0.5627,
+      "step": 2018
+    },
+    {
+      "epoch": 0.12407817109144542,
+      "grad_norm": 50.25,
+      "learning_rate": 1.9927433566532913e-05,
+      "loss": 0.4954,
+      "step": 2019
+    },
+    {
+      "epoch": 0.12413962635201574,
+      "grad_norm": 17.25,
+      "learning_rate": 1.99271534578025e-05,
+      "loss": 0.7139,
+      "step": 2020
+    },
+    {
+      "epoch": 0.12420108161258604,
+      "grad_norm": 27.375,
+      "learning_rate": 1.9926872811474952e-05,
+      "loss": 0.7874,
+      "step": 2021
+    },
+    {
+      "epoch": 0.12426253687315635,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.9926591627565476e-05,
+      "loss": 0.4781,
+      "step": 2022
+    },
+    {
+      "epoch": 0.12432399213372665,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.992630990608929e-05,
+      "loss": 0.3339,
+      "step": 2023
+    },
+    {
+      "epoch": 0.12438544739429695,
+      "grad_norm": 56.0,
+      "learning_rate": 1.9926027647061658e-05,
+      "loss": 0.668,
+      "step": 2024
+    },
+    {
+      "epoch": 0.12444690265486726,
+      "grad_norm": 28.375,
+      "learning_rate": 1.9925744850497866e-05,
+      "loss": 0.7087,
+      "step": 2025
+    },
+    {
+      "epoch": 0.12450835791543756,
+      "grad_norm": 29.0,
+      "learning_rate": 1.9925461516413224e-05,
+      "loss": 0.5799,
+      "step": 2026
+    },
+    {
+      "epoch": 0.12456981317600786,
+      "grad_norm": 20.0,
+      "learning_rate": 1.9925177644823073e-05,
+      "loss": 0.5245,
+      "step": 2027
+    },
+    {
+      "epoch": 0.12463126843657817,
+      "grad_norm": 21.125,
+      "learning_rate": 1.9924893235742795e-05,
+      "loss": 0.4386,
+      "step": 2028
+    },
+    {
+      "epoch": 0.12469272369714847,
+      "grad_norm": 24.25,
+      "learning_rate": 1.9924608289187786e-05,
+      "loss": 0.535,
+      "step": 2029
+    },
+    {
+      "epoch": 0.12475417895771879,
+      "grad_norm": 10.125,
+      "learning_rate": 1.992432280517348e-05,
+      "loss": 0.6802,
+      "step": 2030
+    },
+    {
+      "epoch": 0.12481563421828909,
+      "grad_norm": 16.5,
+      "learning_rate": 1.992403678371533e-05,
+      "loss": 0.3988,
+      "step": 2031
+    },
+    {
+      "epoch": 0.1248770894788594,
+      "grad_norm": 11.625,
+      "learning_rate": 1.9923750224828833e-05,
+      "loss": 0.4574,
+      "step": 2032
+    },
+    {
+      "epoch": 0.1249385447394297,
+      "grad_norm": 11.1875,
+      "learning_rate": 1.9923463128529507e-05,
+      "loss": 0.5725,
+      "step": 2033
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 19.875,
+      "learning_rate": 1.9923175494832895e-05,
+      "loss": 0.6356,
+      "step": 2034
+    },
+    {
+      "epoch": 0.1250614552605703,
+      "grad_norm": 23.0,
+      "learning_rate": 1.992288732375458e-05,
+      "loss": 0.4285,
+      "step": 2035
+    },
+    {
+      "epoch": 0.1251229105211406,
+      "grad_norm": 18.875,
+      "learning_rate": 1.9922598615310157e-05,
+      "loss": 0.5139,
+      "step": 2036
+    },
+    {
+      "epoch": 0.1251843657817109,
+      "grad_norm": 32.5,
+      "learning_rate": 1.9922309369515272e-05,
+      "loss": 0.528,
+      "step": 2037
+    },
+    {
+      "epoch": 0.1252458210422812,
+      "grad_norm": 30.125,
+      "learning_rate": 1.9922019586385587e-05,
+      "loss": 0.4491,
+      "step": 2038
+    },
+    {
+      "epoch": 0.12530727630285152,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.992172926593679e-05,
+      "loss": 0.4352,
+      "step": 2039
+    },
+    {
+      "epoch": 0.12536873156342182,
+      "grad_norm": 5.28125,
+      "learning_rate": 1.99214384081846e-05,
+      "loss": 0.3763,
+      "step": 2040
+    },
+    {
+      "epoch": 0.12543018682399212,
+      "grad_norm": 28.25,
+      "learning_rate": 1.9921147013144782e-05,
+      "loss": 0.6106,
+      "step": 2041
+    },
+    {
+      "epoch": 0.12549164208456243,
+      "grad_norm": 38.0,
+      "learning_rate": 1.99208550808331e-05,
+      "loss": 0.6626,
+      "step": 2042
+    },
+    {
+      "epoch": 0.12555309734513273,
+      "grad_norm": 27.5,
+      "learning_rate": 1.9920562611265378e-05,
+      "loss": 0.5614,
+      "step": 2043
+    },
+    {
+      "epoch": 0.12561455260570306,
+      "grad_norm": 32.5,
+      "learning_rate": 1.9920269604457444e-05,
+      "loss": 0.7316,
+      "step": 2044
+    },
+    {
+      "epoch": 0.12567600786627336,
+      "grad_norm": 36.25,
+      "learning_rate": 1.991997606042517e-05,
+      "loss": 0.5965,
+      "step": 2045
+    },
+    {
+      "epoch": 0.12573746312684367,
+      "grad_norm": 17.375,
+      "learning_rate": 1.9919681979184452e-05,
+      "loss": 0.484,
+      "step": 2046
+    },
+    {
+      "epoch": 0.12579891838741397,
+      "grad_norm": 62.5,
+      "learning_rate": 1.9919387360751216e-05,
+      "loss": 0.729,
+      "step": 2047
+    },
+    {
+      "epoch": 0.12586037364798427,
+      "grad_norm": 21.0,
+      "learning_rate": 1.9919092205141414e-05,
+      "loss": 0.4672,
+      "step": 2048
+    },
+    {
+      "epoch": 0.12592182890855458,
+      "grad_norm": 53.25,
+      "learning_rate": 1.9918796512371035e-05,
+      "loss": 0.8433,
+      "step": 2049
+    },
+    {
+      "epoch": 0.12598328416912488,
+      "grad_norm": 24.0,
+      "learning_rate": 1.991850028245609e-05,
+      "loss": 0.5589,
+      "step": 2050
+    },
+    {
+      "epoch": 0.12604473942969519,
+      "grad_norm": 27.625,
+      "learning_rate": 1.9918203515412616e-05,
+      "loss": 0.5541,
+      "step": 2051
+    },
+    {
+      "epoch": 0.1261061946902655,
+      "grad_norm": 34.25,
+      "learning_rate": 1.9917906211256696e-05,
+      "loss": 0.6042,
+      "step": 2052
+    },
+    {
+      "epoch": 0.1261676499508358,
+      "grad_norm": 24.0,
+      "learning_rate": 1.9917608370004417e-05,
+      "loss": 0.5591,
+      "step": 2053
+    },
+    {
+      "epoch": 0.1262291052114061,
+      "grad_norm": 35.5,
+      "learning_rate": 1.9917309991671915e-05,
+      "loss": 0.4642,
+      "step": 2054
+    },
+    {
+      "epoch": 0.1262905604719764,
+      "grad_norm": 53.25,
+      "learning_rate": 1.991701107627535e-05,
+      "loss": 0.6261,
+      "step": 2055
+    },
+    {
+      "epoch": 0.1263520157325467,
+      "grad_norm": 12.875,
+      "learning_rate": 1.9916711623830904e-05,
+      "loss": 0.5418,
+      "step": 2056
+    },
+    {
+      "epoch": 0.126413470993117,
+      "grad_norm": 10.3125,
+      "learning_rate": 1.99164116343548e-05,
+      "loss": 0.4516,
+      "step": 2057
+    },
+    {
+      "epoch": 0.1264749262536873,
+      "grad_norm": 16.5,
+      "learning_rate": 1.991611110786328e-05,
+      "loss": 0.5993,
+      "step": 2058
+    },
+    {
+      "epoch": 0.1265363815142576,
+      "grad_norm": 13.25,
+      "learning_rate": 1.9915810044372618e-05,
+      "loss": 0.4228,
+      "step": 2059
+    },
+    {
+      "epoch": 0.12659783677482792,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.991550844389912e-05,
+      "loss": 0.7607,
+      "step": 2060
+    },
+    {
+      "epoch": 0.12665929203539822,
+      "grad_norm": 13.875,
+      "learning_rate": 1.9915206306459117e-05,
+      "loss": 0.4728,
+      "step": 2061
+    },
+    {
+      "epoch": 0.12672074729596852,
+      "grad_norm": 28.25,
+      "learning_rate": 1.9914903632068975e-05,
+      "loss": 0.3913,
+      "step": 2062
+    },
+    {
+      "epoch": 0.12678220255653885,
+      "grad_norm": 11.0625,
+      "learning_rate": 1.991460042074508e-05,
+      "loss": 0.6293,
+      "step": 2063
+    },
+    {
+      "epoch": 0.12684365781710916,
+      "grad_norm": 29.125,
+      "learning_rate": 1.9914296672503854e-05,
+      "loss": 0.4878,
+      "step": 2064
+    },
+    {
+      "epoch": 0.12690511307767946,
+      "grad_norm": 22.0,
+      "learning_rate": 1.9913992387361747e-05,
+      "loss": 0.4785,
+      "step": 2065
+    },
+    {
+      "epoch": 0.12696656833824976,
+      "grad_norm": 11.5625,
+      "learning_rate": 1.9913687565335237e-05,
+      "loss": 0.5729,
+      "step": 2066
+    },
+    {
+      "epoch": 0.12702802359882007,
+      "grad_norm": 26.625,
+      "learning_rate": 1.9913382206440834e-05,
+      "loss": 0.5256,
+      "step": 2067
+    },
+    {
+      "epoch": 0.12708947885939037,
+      "grad_norm": 28.5,
+      "learning_rate": 1.9913076310695068e-05,
+      "loss": 0.6305,
+      "step": 2068
+    },
+    {
+      "epoch": 0.12715093411996067,
+      "grad_norm": 24.5,
+      "learning_rate": 1.9912769878114508e-05,
+      "loss": 0.5512,
+      "step": 2069
+    },
+    {
+      "epoch": 0.12721238938053098,
+      "grad_norm": 28.5,
+      "learning_rate": 1.9912462908715752e-05,
+      "loss": 0.5863,
+      "step": 2070
+    },
+    {
+      "epoch": 0.12727384464110128,
+      "grad_norm": 21.0,
+      "learning_rate": 1.991215540251542e-05,
+      "loss": 0.4194,
+      "step": 2071
+    },
+    {
+      "epoch": 0.12733529990167158,
+      "grad_norm": 16.25,
+      "learning_rate": 1.9911847359530165e-05,
+      "loss": 0.5458,
+      "step": 2072
+    },
+    {
+      "epoch": 0.1273967551622419,
+      "grad_norm": 36.75,
+      "learning_rate": 1.991153877977667e-05,
+      "loss": 0.4815,
+      "step": 2073
+    },
+    {
+      "epoch": 0.1274582104228122,
+      "grad_norm": 34.25,
+      "learning_rate": 1.991122966327164e-05,
+      "loss": 0.5823,
+      "step": 2074
+    },
+    {
+      "epoch": 0.1275196656833825,
+      "grad_norm": 26.5,
+      "learning_rate": 1.9910920010031825e-05,
+      "loss": 0.4819,
+      "step": 2075
+    },
+    {
+      "epoch": 0.1275811209439528,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.9910609820073986e-05,
+      "loss": 0.3158,
+      "step": 2076
+    },
+    {
+      "epoch": 0.1276425762045231,
+      "grad_norm": 23.875,
+      "learning_rate": 1.991029909341493e-05,
+      "loss": 0.4087,
+      "step": 2077
+    },
+    {
+      "epoch": 0.1277040314650934,
+      "grad_norm": 26.75,
+      "learning_rate": 1.9909987830071477e-05,
+      "loss": 0.4744,
+      "step": 2078
+    },
+    {
+      "epoch": 0.1277654867256637,
+      "grad_norm": 42.5,
+      "learning_rate": 1.990967603006048e-05,
+      "loss": 0.3962,
+      "step": 2079
+    },
+    {
+      "epoch": 0.127826941986234,
+      "grad_norm": 21.875,
+      "learning_rate": 1.9909363693398828e-05,
+      "loss": 0.3299,
+      "step": 2080
+    },
+    {
+      "epoch": 0.12788839724680431,
+      "grad_norm": 60.0,
+      "learning_rate": 1.990905082010344e-05,
+      "loss": 0.5674,
+      "step": 2081
+    },
+    {
+      "epoch": 0.12794985250737465,
+      "grad_norm": 62.75,
+      "learning_rate": 1.9908737410191255e-05,
+      "loss": 0.6488,
+      "step": 2082
+    },
+    {
+      "epoch": 0.12801130776794495,
+      "grad_norm": 28.5,
+      "learning_rate": 1.9908423463679246e-05,
+      "loss": 0.4973,
+      "step": 2083
+    },
+    {
+      "epoch": 0.12807276302851525,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.9908108980584417e-05,
+      "loss": 0.3893,
+      "step": 2084
+    },
+    {
+      "epoch": 0.12813421828908556,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.9907793960923795e-05,
+      "loss": 0.4961,
+      "step": 2085
+    },
+    {
+      "epoch": 0.12819567354965586,
+      "grad_norm": 9.0,
+      "learning_rate": 1.9907478404714438e-05,
+      "loss": 0.3962,
+      "step": 2086
+    },
+    {
+      "epoch": 0.12825712881022616,
+      "grad_norm": 48.0,
+      "learning_rate": 1.990716231197344e-05,
+      "loss": 0.3513,
+      "step": 2087
+    },
+    {
+      "epoch": 0.12831858407079647,
+      "grad_norm": 5.53125,
+      "learning_rate": 1.9906845682717914e-05,
+      "loss": 0.4186,
+      "step": 2088
+    },
+    {
+      "epoch": 0.12838003933136677,
+      "grad_norm": 23.0,
+      "learning_rate": 1.990652851696501e-05,
+      "loss": 0.3368,
+      "step": 2089
+    },
+    {
+      "epoch": 0.12844149459193707,
+      "grad_norm": 31.125,
+      "learning_rate": 1.9906210814731905e-05,
+      "loss": 0.6681,
+      "step": 2090
+    },
+    {
+      "epoch": 0.12850294985250738,
+      "grad_norm": 17.875,
+      "learning_rate": 1.9905892576035798e-05,
+      "loss": 0.4653,
+      "step": 2091
+    },
+    {
+      "epoch": 0.12856440511307768,
+      "grad_norm": 25.25,
+      "learning_rate": 1.990557380089393e-05,
+      "loss": 0.4199,
+      "step": 2092
+    },
+    {
+      "epoch": 0.12862586037364798,
+      "grad_norm": 21.375,
+      "learning_rate": 1.990525448932356e-05,
+      "loss": 0.4106,
+      "step": 2093
+    },
+    {
+      "epoch": 0.12868731563421829,
+      "grad_norm": 20.75,
+      "learning_rate": 1.9904934641341983e-05,
+      "loss": 0.3241,
+      "step": 2094
+    },
+    {
+      "epoch": 0.1287487708947886,
+      "grad_norm": 30.875,
+      "learning_rate": 1.9904614256966514e-05,
+      "loss": 0.3988,
+      "step": 2095
+    },
+    {
+      "epoch": 0.1288102261553589,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.9904293336214508e-05,
+      "loss": 0.3013,
+      "step": 2096
+    },
+    {
+      "epoch": 0.1288716814159292,
+      "grad_norm": 44.75,
+      "learning_rate": 1.9903971879103344e-05,
+      "loss": 0.5463,
+      "step": 2097
+    },
+    {
+      "epoch": 0.1289331366764995,
+      "grad_norm": 63.25,
+      "learning_rate": 1.990364988565043e-05,
+      "loss": 0.5981,
+      "step": 2098
+    },
+    {
+      "epoch": 0.1289945919370698,
+      "grad_norm": 49.25,
+      "learning_rate": 1.9903327355873205e-05,
+      "loss": 0.5066,
+      "step": 2099
+    },
+    {
+      "epoch": 0.1290560471976401,
+      "grad_norm": 26.5,
+      "learning_rate": 1.990300428978913e-05,
+      "loss": 0.7064,
+      "step": 2100
+    },
+    {
+      "epoch": 0.1291175024582104,
+      "grad_norm": 39.75,
+      "learning_rate": 1.9902680687415704e-05,
+      "loss": 0.594,
+      "step": 2101
+    },
+    {
+      "epoch": 0.12917895771878074,
+      "grad_norm": 75.5,
+      "learning_rate": 1.9902356548770455e-05,
+      "loss": 0.5776,
+      "step": 2102
+    },
+    {
+      "epoch": 0.12924041297935104,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.9902031873870927e-05,
+      "loss": 0.3013,
+      "step": 2103
+    },
+    {
+      "epoch": 0.12930186823992135,
+      "grad_norm": 31.5,
+      "learning_rate": 1.990170666273471e-05,
+      "loss": 0.3609,
+      "step": 2104
+    },
+    {
+      "epoch": 0.12936332350049165,
+      "grad_norm": 18.25,
+      "learning_rate": 1.9901380915379413e-05,
+      "loss": 0.4397,
+      "step": 2105
+    },
+    {
+      "epoch": 0.12942477876106195,
+      "grad_norm": 21.625,
+      "learning_rate": 1.990105463182268e-05,
+      "loss": 0.6121,
+      "step": 2106
+    },
+    {
+      "epoch": 0.12948623402163226,
+      "grad_norm": 23.375,
+      "learning_rate": 1.9900727812082177e-05,
+      "loss": 0.5462,
+      "step": 2107
+    },
+    {
+      "epoch": 0.12954768928220256,
+      "grad_norm": 32.25,
+      "learning_rate": 1.99004004561756e-05,
+      "loss": 0.5745,
+      "step": 2108
+    },
+    {
+      "epoch": 0.12960914454277286,
+      "grad_norm": 34.0,
+      "learning_rate": 1.9900072564120686e-05,
+      "loss": 0.7375,
+      "step": 2109
+    },
+    {
+      "epoch": 0.12967059980334317,
+      "grad_norm": 22.75,
+      "learning_rate": 1.989974413593518e-05,
+      "loss": 0.5015,
+      "step": 2110
+    },
+    {
+      "epoch": 0.12973205506391347,
+      "grad_norm": 8.375,
+      "learning_rate": 1.989941517163688e-05,
+      "loss": 0.4905,
+      "step": 2111
+    },
+    {
+      "epoch": 0.12979351032448377,
+      "grad_norm": 37.0,
+      "learning_rate": 1.989908567124359e-05,
+      "loss": 0.5282,
+      "step": 2112
+    },
+    {
+      "epoch": 0.12985496558505408,
+      "grad_norm": 41.25,
+      "learning_rate": 1.989875563477316e-05,
+      "loss": 0.719,
+      "step": 2113
+    },
+    {
+      "epoch": 0.12991642084562438,
+      "grad_norm": 31.625,
+      "learning_rate": 1.989842506224346e-05,
+      "loss": 0.5717,
+      "step": 2114
+    },
+    {
+      "epoch": 0.12997787610619468,
+      "grad_norm": 31.125,
+      "learning_rate": 1.9898093953672392e-05,
+      "loss": 0.6358,
+      "step": 2115
+    },
+    {
+      "epoch": 0.130039331366765,
+      "grad_norm": 30.5,
+      "learning_rate": 1.989776230907789e-05,
+      "loss": 0.5398,
+      "step": 2116
+    },
+    {
+      "epoch": 0.1301007866273353,
+      "grad_norm": 17.0,
+      "learning_rate": 1.9897430128477912e-05,
+      "loss": 0.4648,
+      "step": 2117
+    },
+    {
+      "epoch": 0.1301622418879056,
+      "grad_norm": 10.8125,
+      "learning_rate": 1.989709741189045e-05,
+      "loss": 0.4408,
+      "step": 2118
+    },
+    {
+      "epoch": 0.1302236971484759,
+      "grad_norm": 31.25,
+      "learning_rate": 1.989676415933351e-05,
+      "loss": 0.5942,
+      "step": 2119
+    },
+    {
+      "epoch": 0.1302851524090462,
+      "grad_norm": 22.25,
+      "learning_rate": 1.9896430370825156e-05,
+      "loss": 0.5058,
+      "step": 2120
+    },
+    {
+      "epoch": 0.13034660766961653,
+      "grad_norm": 22.0,
+      "learning_rate": 1.9896096046383456e-05,
+      "loss": 0.6241,
+      "step": 2121
+    },
+    {
+      "epoch": 0.13040806293018684,
+      "grad_norm": 20.25,
+      "learning_rate": 1.989576118602651e-05,
+      "loss": 0.4868,
+      "step": 2122
+    },
+    {
+      "epoch": 0.13046951819075714,
+      "grad_norm": 21.75,
+      "learning_rate": 1.989542578977246e-05,
+      "loss": 0.5702,
+      "step": 2123
+    },
+    {
+      "epoch": 0.13053097345132744,
+      "grad_norm": 33.75,
+      "learning_rate": 1.989508985763947e-05,
+      "loss": 0.4937,
+      "step": 2124
+    },
+    {
+      "epoch": 0.13059242871189775,
+      "grad_norm": 13.0,
+      "learning_rate": 1.9894753389645723e-05,
+      "loss": 0.4441,
+      "step": 2125
+    },
+    {
+      "epoch": 0.13065388397246805,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9894416385809444e-05,
+      "loss": 0.4078,
+      "step": 2126
+    },
+    {
+      "epoch": 0.13071533923303835,
+      "grad_norm": 19.875,
+      "learning_rate": 1.989407884614889e-05,
+      "loss": 0.3699,
+      "step": 2127
+    },
+    {
+      "epoch": 0.13077679449360866,
+      "grad_norm": 16.0,
+      "learning_rate": 1.9893740770682334e-05,
+      "loss": 0.4088,
+      "step": 2128
+    },
+    {
+      "epoch": 0.13083824975417896,
+      "grad_norm": 24.875,
+      "learning_rate": 1.9893402159428084e-05,
+      "loss": 0.5049,
+      "step": 2129
+    },
+    {
+      "epoch": 0.13089970501474926,
+      "grad_norm": 9.875,
+      "learning_rate": 1.989306301240448e-05,
+      "loss": 0.4817,
+      "step": 2130
+    },
+    {
+      "epoch": 0.13096116027531957,
+      "grad_norm": 9.625,
+      "learning_rate": 1.9892723329629885e-05,
+      "loss": 0.5094,
+      "step": 2131
+    },
+    {
+      "epoch": 0.13102261553588987,
+      "grad_norm": 13.875,
+      "learning_rate": 1.9892383111122696e-05,
+      "loss": 0.4608,
+      "step": 2132
+    },
+    {
+      "epoch": 0.13108407079646017,
+      "grad_norm": 23.25,
+      "learning_rate": 1.9892042356901337e-05,
+      "loss": 0.4225,
+      "step": 2133
+    },
+    {
+      "epoch": 0.13114552605703048,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9891701066984264e-05,
+      "loss": 0.3494,
+      "step": 2134
+    },
+    {
+      "epoch": 0.13120698131760078,
+      "grad_norm": 27.375,
+      "learning_rate": 1.9891359241389955e-05,
+      "loss": 0.5765,
+      "step": 2135
+    },
+    {
+      "epoch": 0.13126843657817108,
+      "grad_norm": 14.625,
+      "learning_rate": 1.9891016880136923e-05,
+      "loss": 0.5409,
+      "step": 2136
+    },
+    {
+      "epoch": 0.13132989183874139,
+      "grad_norm": 18.0,
+      "learning_rate": 1.9890673983243708e-05,
+      "loss": 0.4229,
+      "step": 2137
+    },
+    {
+      "epoch": 0.1313913470993117,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.989033055072888e-05,
+      "loss": 0.4937,
+      "step": 2138
+    },
+    {
+      "epoch": 0.131452802359882,
+      "grad_norm": 16.25,
+      "learning_rate": 1.9889986582611036e-05,
+      "loss": 0.6144,
+      "step": 2139
+    },
+    {
+      "epoch": 0.13151425762045232,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.9889642078908805e-05,
+      "loss": 0.466,
+      "step": 2140
+    },
+    {
+      "epoch": 0.13157571288102263,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.988929703964084e-05,
+      "loss": 0.4866,
+      "step": 2141
+    },
+    {
+      "epoch": 0.13163716814159293,
+      "grad_norm": 27.75,
+      "learning_rate": 1.9888951464825835e-05,
+      "loss": 0.5341,
+      "step": 2142
+    },
+    {
+      "epoch": 0.13169862340216323,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.9888605354482494e-05,
+      "loss": 0.3986,
+      "step": 2143
+    },
+    {
+      "epoch": 0.13176007866273354,
+      "grad_norm": 11.75,
+      "learning_rate": 1.9888258708629565e-05,
+      "loss": 0.3388,
+      "step": 2144
+    },
+    {
+      "epoch": 0.13182153392330384,
+      "grad_norm": 32.75,
+      "learning_rate": 1.988791152728582e-05,
+      "loss": 0.704,
+      "step": 2145
+    },
+    {
+      "epoch": 0.13188298918387414,
+      "grad_norm": 17.625,
+      "learning_rate": 1.988756381047006e-05,
+      "loss": 0.6299,
+      "step": 2146
+    },
+    {
+      "epoch": 0.13194444444444445,
+      "grad_norm": 30.0,
+      "learning_rate": 1.9887215558201115e-05,
+      "loss": 0.6765,
+      "step": 2147
+    },
+    {
+      "epoch": 0.13200589970501475,
+      "grad_norm": 16.0,
+      "learning_rate": 1.9886866770497847e-05,
+      "loss": 0.669,
+      "step": 2148
+    },
+    {
+      "epoch": 0.13206735496558505,
+      "grad_norm": 13.5,
+      "learning_rate": 1.988651744737914e-05,
+      "loss": 0.8379,
+      "step": 2149
+    },
+    {
+      "epoch": 0.13212881022615536,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.9886167588863918e-05,
+      "loss": 0.5354,
+      "step": 2150
+    },
+    {
+      "epoch": 0.13219026548672566,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9885817194971116e-05,
+      "loss": 0.3848,
+      "step": 2151
+    },
+    {
+      "epoch": 0.13225172074729596,
+      "grad_norm": 46.25,
+      "learning_rate": 1.9885466265719723e-05,
+      "loss": 0.5101,
+      "step": 2152
+    },
+    {
+      "epoch": 0.13231317600786627,
+      "grad_norm": 13.875,
+      "learning_rate": 1.988511480112873e-05,
+      "loss": 0.4344,
+      "step": 2153
+    },
+    {
+      "epoch": 0.13237463126843657,
+      "grad_norm": 12.5,
+      "learning_rate": 1.9884762801217185e-05,
+      "loss": 0.5549,
+      "step": 2154
+    },
+    {
+      "epoch": 0.13243608652900687,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.9884410266004134e-05,
+      "loss": 0.3857,
+      "step": 2155
+    },
+    {
+      "epoch": 0.13249754178957718,
+      "grad_norm": 21.75,
+      "learning_rate": 1.9884057195508683e-05,
+      "loss": 0.4965,
+      "step": 2156
+    },
+    {
+      "epoch": 0.13255899705014748,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.9883703589749938e-05,
+      "loss": 0.3709,
+      "step": 2157
+    },
+    {
+      "epoch": 0.13262045231071778,
+      "grad_norm": 25.875,
+      "learning_rate": 1.988334944874706e-05,
+      "loss": 0.3618,
+      "step": 2158
+    },
+    {
+      "epoch": 0.13268190757128812,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.9882994772519225e-05,
+      "loss": 0.3453,
+      "step": 2159
+    },
+    {
+      "epoch": 0.13274336283185842,
+      "grad_norm": 48.0,
+      "learning_rate": 1.9882639561085634e-05,
+      "loss": 0.449,
+      "step": 2160
+    },
+    {
+      "epoch": 0.13280481809242872,
+      "grad_norm": 30.625,
+      "learning_rate": 1.988228381446553e-05,
+      "loss": 0.5761,
+      "step": 2161
+    },
+    {
+      "epoch": 0.13286627335299903,
+      "grad_norm": 28.75,
+      "learning_rate": 1.9881927532678174e-05,
+      "loss": 0.5464,
+      "step": 2162
+    },
+    {
+      "epoch": 0.13292772861356933,
+      "grad_norm": 52.0,
+      "learning_rate": 1.9881570715742865e-05,
+      "loss": 0.5055,
+      "step": 2163
+    },
+    {
+      "epoch": 0.13298918387413963,
+      "grad_norm": 9.625,
+      "learning_rate": 1.988121336367892e-05,
+      "loss": 0.5087,
+      "step": 2164
+    },
+    {
+      "epoch": 0.13305063913470994,
+      "grad_norm": 47.25,
+      "learning_rate": 1.9880855476505694e-05,
+      "loss": 0.5997,
+      "step": 2165
+    },
+    {
+      "epoch": 0.13311209439528024,
+      "grad_norm": 10.625,
+      "learning_rate": 1.9880497054242566e-05,
+      "loss": 0.5849,
+      "step": 2166
+    },
+    {
+      "epoch": 0.13317354965585054,
+      "grad_norm": 5.3125,
+      "learning_rate": 1.9880138096908955e-05,
+      "loss": 0.4537,
+      "step": 2167
+    },
+    {
+      "epoch": 0.13323500491642085,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.9879778604524286e-05,
+      "loss": 0.5885,
+      "step": 2168
+    },
+    {
+      "epoch": 0.13329646017699115,
+      "grad_norm": 37.25,
+      "learning_rate": 1.9879418577108037e-05,
+      "loss": 0.5533,
+      "step": 2169
+    },
+    {
+      "epoch": 0.13335791543756145,
+      "grad_norm": 26.375,
+      "learning_rate": 1.9879058014679704e-05,
+      "loss": 0.4738,
+      "step": 2170
+    },
+    {
+      "epoch": 0.13341937069813176,
+      "grad_norm": 46.25,
+      "learning_rate": 1.987869691725881e-05,
+      "loss": 0.4987,
+      "step": 2171
+    },
+    {
+      "epoch": 0.13348082595870206,
+      "grad_norm": 73.5,
+      "learning_rate": 1.987833528486491e-05,
+      "loss": 0.5424,
+      "step": 2172
+    },
+    {
+      "epoch": 0.13354228121927236,
+      "grad_norm": 43.0,
+      "learning_rate": 1.987797311751759e-05,
+      "loss": 0.427,
+      "step": 2173
+    },
+    {
+      "epoch": 0.13360373647984267,
+      "grad_norm": 18.75,
+      "learning_rate": 1.9877610415236464e-05,
+      "loss": 0.4269,
+      "step": 2174
+    },
+    {
+      "epoch": 0.13366519174041297,
+      "grad_norm": 53.0,
+      "learning_rate": 1.9877247178041167e-05,
+      "loss": 0.5603,
+      "step": 2175
+    },
+    {
+      "epoch": 0.13372664700098327,
+      "grad_norm": 5.875,
+      "learning_rate": 1.9876883405951378e-05,
+      "loss": 0.4074,
+      "step": 2176
+    },
+    {
+      "epoch": 0.13378810226155358,
+      "grad_norm": 21.875,
+      "learning_rate": 1.987651909898679e-05,
+      "loss": 0.4472,
+      "step": 2177
+    },
+    {
+      "epoch": 0.1338495575221239,
+      "grad_norm": 37.25,
+      "learning_rate": 1.9876154257167143e-05,
+      "loss": 0.639,
+      "step": 2178
+    },
+    {
+      "epoch": 0.1339110127826942,
+      "grad_norm": 32.75,
+      "learning_rate": 1.9875788880512183e-05,
+      "loss": 0.6329,
+      "step": 2179
+    },
+    {
+      "epoch": 0.13397246804326451,
+      "grad_norm": 10.875,
+      "learning_rate": 1.9875422969041697e-05,
+      "loss": 0.5516,
+      "step": 2180
+    },
+    {
+      "epoch": 0.13403392330383482,
+      "grad_norm": 12.25,
+      "learning_rate": 1.9875056522775506e-05,
+      "loss": 0.5832,
+      "step": 2181
+    },
+    {
+      "epoch": 0.13409537856440512,
+      "grad_norm": 19.25,
+      "learning_rate": 1.9874689541733455e-05,
+      "loss": 0.4347,
+      "step": 2182
+    },
+    {
+      "epoch": 0.13415683382497542,
+      "grad_norm": 37.5,
+      "learning_rate": 1.9874322025935414e-05,
+      "loss": 0.3835,
+      "step": 2183
+    },
+    {
+      "epoch": 0.13421828908554573,
+      "grad_norm": 28.125,
+      "learning_rate": 1.9873953975401286e-05,
+      "loss": 0.4577,
+      "step": 2184
+    },
+    {
+      "epoch": 0.13427974434611603,
+      "grad_norm": 42.75,
+      "learning_rate": 1.9873585390151003e-05,
+      "loss": 0.4348,
+      "step": 2185
+    },
+    {
+      "epoch": 0.13434119960668633,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.987321627020453e-05,
+      "loss": 0.5868,
+      "step": 2186
+    },
+    {
+      "epoch": 0.13440265486725664,
+      "grad_norm": 26.875,
+      "learning_rate": 1.9872846615581848e-05,
+      "loss": 0.5404,
+      "step": 2187
+    },
+    {
+      "epoch": 0.13446411012782694,
+      "grad_norm": 16.875,
+      "learning_rate": 1.9872476426302983e-05,
+      "loss": 0.4522,
+      "step": 2188
+    },
+    {
+      "epoch": 0.13452556538839724,
+      "grad_norm": 34.0,
+      "learning_rate": 1.9872105702387976e-05,
+      "loss": 0.5009,
+      "step": 2189
+    },
+    {
+      "epoch": 0.13458702064896755,
+      "grad_norm": 19.75,
+      "learning_rate": 1.9871734443856907e-05,
+      "loss": 0.489,
+      "step": 2190
+    },
+    {
+      "epoch": 0.13464847590953785,
+      "grad_norm": 7.0,
+      "learning_rate": 1.987136265072988e-05,
+      "loss": 0.3451,
+      "step": 2191
+    },
+    {
+      "epoch": 0.13470993117010815,
+      "grad_norm": 27.75,
+      "learning_rate": 1.9870990323027032e-05,
+      "loss": 0.4976,
+      "step": 2192
+    },
+    {
+      "epoch": 0.13477138643067846,
+      "grad_norm": 23.75,
+      "learning_rate": 1.987061746076852e-05,
+      "loss": 0.4327,
+      "step": 2193
+    },
+    {
+      "epoch": 0.13483284169124876,
+      "grad_norm": 31.125,
+      "learning_rate": 1.987024406397454e-05,
+      "loss": 0.5763,
+      "step": 2194
+    },
+    {
+      "epoch": 0.13489429695181906,
+      "grad_norm": 25.5,
+      "learning_rate": 1.9869870132665318e-05,
+      "loss": 0.5103,
+      "step": 2195
+    },
+    {
+      "epoch": 0.13495575221238937,
+      "grad_norm": 19.875,
+      "learning_rate": 1.9869495666861094e-05,
+      "loss": 0.5635,
+      "step": 2196
+    },
+    {
+      "epoch": 0.13501720747295967,
+      "grad_norm": 16.125,
+      "learning_rate": 1.9869120666582153e-05,
+      "loss": 0.3279,
+      "step": 2197
+    },
+    {
+      "epoch": 0.13507866273353,
+      "grad_norm": 20.25,
+      "learning_rate": 1.9868745131848798e-05,
+      "loss": 0.4138,
+      "step": 2198
+    },
+    {
+      "epoch": 0.1351401179941003,
+      "grad_norm": 28.125,
+      "learning_rate": 1.9868369062681373e-05,
+      "loss": 0.4851,
+      "step": 2199
+    },
+    {
+      "epoch": 0.1352015732546706,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.986799245910024e-05,
+      "loss": 0.3865,
+      "step": 2200
+    },
+    {
+      "epoch": 0.1352630285152409,
+      "grad_norm": 25.875,
+      "learning_rate": 1.9867615321125796e-05,
+      "loss": 0.5552,
+      "step": 2201
+    },
+    {
+      "epoch": 0.13532448377581122,
+      "grad_norm": 27.875,
+      "learning_rate": 1.986723764877846e-05,
+      "loss": 0.6568,
+      "step": 2202
+    },
+    {
+      "epoch": 0.13538593903638152,
+      "grad_norm": 21.375,
+      "learning_rate": 1.986685944207868e-05,
+      "loss": 0.574,
+      "step": 2203
+    },
+    {
+      "epoch": 0.13544739429695182,
+      "grad_norm": 41.5,
+      "learning_rate": 1.9866480701046953e-05,
+      "loss": 0.6094,
+      "step": 2204
+    },
+    {
+      "epoch": 0.13550884955752213,
+      "grad_norm": 41.0,
+      "learning_rate": 1.9866101425703776e-05,
+      "loss": 0.6465,
+      "step": 2205
+    },
+    {
+      "epoch": 0.13557030481809243,
+      "grad_norm": 36.25,
+      "learning_rate": 1.9865721616069695e-05,
+      "loss": 0.6278,
+      "step": 2206
+    },
+    {
+      "epoch": 0.13563176007866273,
+      "grad_norm": 16.25,
+      "learning_rate": 1.9865341272165276e-05,
+      "loss": 0.393,
+      "step": 2207
+    },
+    {
+      "epoch": 0.13569321533923304,
+      "grad_norm": 16.625,
+      "learning_rate": 1.9864960394011115e-05,
+      "loss": 0.4922,
+      "step": 2208
+    },
+    {
+      "epoch": 0.13575467059980334,
+      "grad_norm": 28.625,
+      "learning_rate": 1.9864578981627844e-05,
+      "loss": 0.5771,
+      "step": 2209
+    },
+    {
+      "epoch": 0.13581612586037364,
+      "grad_norm": 28.5,
+      "learning_rate": 1.9864197035036112e-05,
+      "loss": 0.5807,
+      "step": 2210
+    },
+    {
+      "epoch": 0.13587758112094395,
+      "grad_norm": 16.5,
+      "learning_rate": 1.98638145542566e-05,
+      "loss": 0.4254,
+      "step": 2211
+    },
+    {
+      "epoch": 0.13593903638151425,
+      "grad_norm": 13.375,
+      "learning_rate": 1.9863431539310033e-05,
+      "loss": 0.6729,
+      "step": 2212
+    },
+    {
+      "epoch": 0.13600049164208455,
+      "grad_norm": 16.5,
+      "learning_rate": 1.9863047990217143e-05,
+      "loss": 0.527,
+      "step": 2213
+    },
+    {
+      "epoch": 0.13606194690265486,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.98626639069987e-05,
+      "loss": 0.3867,
+      "step": 2214
+    },
+    {
+      "epoch": 0.13612340216322516,
+      "grad_norm": 12.75,
+      "learning_rate": 1.986227928967551e-05,
+      "loss": 0.4007,
+      "step": 2215
+    },
+    {
+      "epoch": 0.13618485742379546,
+      "grad_norm": 16.25,
+      "learning_rate": 1.9861894138268402e-05,
+      "loss": 0.4178,
+      "step": 2216
+    },
+    {
+      "epoch": 0.1362463126843658,
+      "grad_norm": 27.875,
+      "learning_rate": 1.9861508452798224e-05,
+      "loss": 0.5855,
+      "step": 2217
+    },
+    {
+      "epoch": 0.1363077679449361,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.9861122233285873e-05,
+      "loss": 0.4761,
+      "step": 2218
+    },
+    {
+      "epoch": 0.1363692232055064,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.986073547975226e-05,
+      "loss": 0.372,
+      "step": 2219
+    },
+    {
+      "epoch": 0.1364306784660767,
+      "grad_norm": 38.75,
+      "learning_rate": 1.9860348192218327e-05,
+      "loss": 0.5724,
+      "step": 2220
+    },
+    {
+      "epoch": 0.136492133726647,
+      "grad_norm": 29.875,
+      "learning_rate": 1.985996037070505e-05,
+      "loss": 0.4056,
+      "step": 2221
+    },
+    {
+      "epoch": 0.1365535889872173,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.9859572015233436e-05,
+      "loss": 0.5368,
+      "step": 2222
+    },
+    {
+      "epoch": 0.13661504424778761,
+      "grad_norm": 22.125,
+      "learning_rate": 1.9859183125824506e-05,
+      "loss": 0.3944,
+      "step": 2223
+    },
+    {
+      "epoch": 0.13667649950835792,
+      "grad_norm": 33.0,
+      "learning_rate": 1.9858793702499322e-05,
+      "loss": 0.5275,
+      "step": 2224
+    },
+    {
+      "epoch": 0.13673795476892822,
+      "grad_norm": 27.0,
+      "learning_rate": 1.9858403745278982e-05,
+      "loss": 0.5205,
+      "step": 2225
+    },
+    {
+      "epoch": 0.13679941002949852,
+      "grad_norm": 23.875,
+      "learning_rate": 1.9858013254184597e-05,
+      "loss": 0.6841,
+      "step": 2226
+    },
+    {
+      "epoch": 0.13686086529006883,
+      "grad_norm": 24.875,
+      "learning_rate": 1.9857622229237315e-05,
+      "loss": 0.7128,
+      "step": 2227
+    },
+    {
+      "epoch": 0.13692232055063913,
+      "grad_norm": 33.75,
+      "learning_rate": 1.985723067045831e-05,
+      "loss": 0.6001,
+      "step": 2228
+    },
+    {
+      "epoch": 0.13698377581120944,
+      "grad_norm": 33.5,
+      "learning_rate": 1.985683857786879e-05,
+      "loss": 0.6496,
+      "step": 2229
+    },
+    {
+      "epoch": 0.13704523107177974,
+      "grad_norm": 26.5,
+      "learning_rate": 1.9856445951489984e-05,
+      "loss": 0.601,
+      "step": 2230
+    },
+    {
+      "epoch": 0.13710668633235004,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.9856052791343153e-05,
+      "loss": 0.3879,
+      "step": 2231
+    },
+    {
+      "epoch": 0.13716814159292035,
+      "grad_norm": 23.875,
+      "learning_rate": 1.9855659097449597e-05,
+      "loss": 0.4743,
+      "step": 2232
+    },
+    {
+      "epoch": 0.13722959685349065,
+      "grad_norm": 17.5,
+      "learning_rate": 1.985526486983063e-05,
+      "loss": 0.7091,
+      "step": 2233
+    },
+    {
+      "epoch": 0.13729105211406095,
+      "grad_norm": 18.625,
+      "learning_rate": 1.98548701085076e-05,
+      "loss": 0.6984,
+      "step": 2234
+    },
+    {
+      "epoch": 0.13735250737463126,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.985447481350189e-05,
+      "loss": 0.5653,
+      "step": 2235
+    },
+    {
+      "epoch": 0.1374139626352016,
+      "grad_norm": 19.75,
+      "learning_rate": 1.9854078984834904e-05,
+      "loss": 0.4881,
+      "step": 2236
+    },
+    {
+      "epoch": 0.1374754178957719,
+      "grad_norm": 28.375,
+      "learning_rate": 1.9853682622528074e-05,
+      "loss": 0.6677,
+      "step": 2237
+    },
+    {
+      "epoch": 0.1375368731563422,
+      "grad_norm": 12.125,
+      "learning_rate": 1.9853285726602874e-05,
+      "loss": 0.4534,
+      "step": 2238
+    },
+    {
+      "epoch": 0.1375983284169125,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.985288829708079e-05,
+      "loss": 0.4135,
+      "step": 2239
+    },
+    {
+      "epoch": 0.1376597836774828,
+      "grad_norm": 13.25,
+      "learning_rate": 1.9852490333983342e-05,
+      "loss": 0.5622,
+      "step": 2240
+    },
+    {
+      "epoch": 0.1377212389380531,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.985209183733209e-05,
+      "loss": 0.3986,
+      "step": 2241
+    },
+    {
+      "epoch": 0.1377826941986234,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.9851692807148612e-05,
+      "loss": 0.3636,
+      "step": 2242
+    },
+    {
+      "epoch": 0.1378441494591937,
+      "grad_norm": 23.375,
+      "learning_rate": 1.9851293243454513e-05,
+      "loss": 0.3933,
+      "step": 2243
+    },
+    {
+      "epoch": 0.137905604719764,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.985089314627143e-05,
+      "loss": 0.42,
+      "step": 2244
+    },
+    {
+      "epoch": 0.13796705998033432,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.9850492515621038e-05,
+      "loss": 0.4015,
+      "step": 2245
+    },
+    {
+      "epoch": 0.13802851524090462,
+      "grad_norm": 14.75,
+      "learning_rate": 1.985009135152503e-05,
+      "loss": 0.3528,
+      "step": 2246
+    },
+    {
+      "epoch": 0.13808997050147492,
+      "grad_norm": 5.125,
+      "learning_rate": 1.9849689654005124e-05,
+      "loss": 0.2579,
+      "step": 2247
+    },
+    {
+      "epoch": 0.13815142576204523,
+      "grad_norm": 25.75,
+      "learning_rate": 1.984928742308308e-05,
+      "loss": 0.3416,
+      "step": 2248
+    },
+    {
+      "epoch": 0.13821288102261553,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.984888465878068e-05,
+      "loss": 0.3824,
+      "step": 2249
+    },
+    {
+      "epoch": 0.13827433628318583,
+      "grad_norm": 25.875,
+      "learning_rate": 1.9848481361119733e-05,
+      "loss": 0.5539,
+      "step": 2250
+    },
+    {
+      "epoch": 0.13833579154375614,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.9848077530122083e-05,
+      "loss": 0.5354,
+      "step": 2251
+    },
+    {
+      "epoch": 0.13839724680432644,
+      "grad_norm": 26.25,
+      "learning_rate": 1.9847673165809595e-05,
+      "loss": 0.5823,
+      "step": 2252
+    },
+    {
+      "epoch": 0.13845870206489674,
+      "grad_norm": 63.0,
+      "learning_rate": 1.9847268268204168e-05,
+      "loss": 0.716,
+      "step": 2253
+    },
+    {
+      "epoch": 0.13852015732546705,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9846862837327733e-05,
+      "loss": 0.5441,
+      "step": 2254
+    },
+    {
+      "epoch": 0.13858161258603738,
+      "grad_norm": 27.125,
+      "learning_rate": 1.984645687320224e-05,
+      "loss": 0.5445,
+      "step": 2255
+    },
+    {
+      "epoch": 0.13864306784660768,
+      "grad_norm": 51.0,
+      "learning_rate": 1.9846050375849674e-05,
+      "loss": 0.5248,
+      "step": 2256
+    },
+    {
+      "epoch": 0.13870452310717798,
+      "grad_norm": 12.0,
+      "learning_rate": 1.9845643345292055e-05,
+      "loss": 0.6126,
+      "step": 2257
+    },
+    {
+      "epoch": 0.1387659783677483,
+      "grad_norm": 26.625,
+      "learning_rate": 1.984523578155142e-05,
+      "loss": 0.5107,
+      "step": 2258
+    },
+    {
+      "epoch": 0.1388274336283186,
+      "grad_norm": 29.25,
+      "learning_rate": 1.984482768464984e-05,
+      "loss": 0.506,
+      "step": 2259
+    },
+    {
+      "epoch": 0.1388888888888889,
+      "grad_norm": 20.625,
+      "learning_rate": 1.9844419054609418e-05,
+      "loss": 0.595,
+      "step": 2260
+    },
+    {
+      "epoch": 0.1389503441494592,
+      "grad_norm": 28.375,
+      "learning_rate": 1.984400989145228e-05,
+      "loss": 0.4633,
+      "step": 2261
+    },
+    {
+      "epoch": 0.1390117994100295,
+      "grad_norm": 20.625,
+      "learning_rate": 1.984360019520059e-05,
+      "loss": 0.5401,
+      "step": 2262
+    },
+    {
+      "epoch": 0.1390732546705998,
+      "grad_norm": 9.125,
+      "learning_rate": 1.9843189965876525e-05,
+      "loss": 0.3445,
+      "step": 2263
+    },
+    {
+      "epoch": 0.1391347099311701,
+      "grad_norm": 29.625,
+      "learning_rate": 1.984277920350231e-05,
+      "loss": 0.3929,
+      "step": 2264
+    },
+    {
+      "epoch": 0.1391961651917404,
+      "grad_norm": 45.0,
+      "learning_rate": 1.9842367908100182e-05,
+      "loss": 0.5897,
+      "step": 2265
+    },
+    {
+      "epoch": 0.13925762045231072,
+      "grad_norm": 17.75,
+      "learning_rate": 1.984195607969242e-05,
+      "loss": 0.439,
+      "step": 2266
+    },
+    {
+      "epoch": 0.13931907571288102,
+      "grad_norm": 65.0,
+      "learning_rate": 1.9841543718301324e-05,
+      "loss": 0.5191,
+      "step": 2267
+    },
+    {
+      "epoch": 0.13938053097345132,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.9841130823949223e-05,
+      "loss": 0.5323,
+      "step": 2268
+    },
+    {
+      "epoch": 0.13944198623402163,
+      "grad_norm": 35.25,
+      "learning_rate": 1.9840717396658483e-05,
+      "loss": 0.5804,
+      "step": 2269
+    },
+    {
+      "epoch": 0.13950344149459193,
+      "grad_norm": 34.5,
+      "learning_rate": 1.9840303436451485e-05,
+      "loss": 0.5179,
+      "step": 2270
+    },
+    {
+      "epoch": 0.13956489675516223,
+      "grad_norm": 13.75,
+      "learning_rate": 1.9839888943350656e-05,
+      "loss": 0.3176,
+      "step": 2271
+    },
+    {
+      "epoch": 0.13962635201573254,
+      "grad_norm": 11.4375,
+      "learning_rate": 1.9839473917378432e-05,
+      "loss": 0.5403,
+      "step": 2272
+    },
+    {
+      "epoch": 0.13968780727630284,
+      "grad_norm": 28.625,
+      "learning_rate": 1.9839058358557296e-05,
+      "loss": 0.8265,
+      "step": 2273
+    },
+    {
+      "epoch": 0.13974926253687317,
+      "grad_norm": 42.25,
+      "learning_rate": 1.983864226690975e-05,
+      "loss": 0.5885,
+      "step": 2274
+    },
+    {
+      "epoch": 0.13981071779744347,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.983822564245833e-05,
+      "loss": 0.4583,
+      "step": 2275
+    },
+    {
+      "epoch": 0.13987217305801378,
+      "grad_norm": 10.75,
+      "learning_rate": 1.983780848522559e-05,
+      "loss": 0.4722,
+      "step": 2276
+    },
+    {
+      "epoch": 0.13993362831858408,
+      "grad_norm": 21.375,
+      "learning_rate": 1.9837390795234128e-05,
+      "loss": 0.5182,
+      "step": 2277
+    },
+    {
+      "epoch": 0.13999508357915438,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.9836972572506557e-05,
+      "loss": 0.4182,
+      "step": 2278
+    },
+    {
+      "epoch": 0.1400565388397247,
+      "grad_norm": 28.375,
+      "learning_rate": 1.983655381706554e-05,
+      "loss": 0.4289,
+      "step": 2279
+    },
+    {
+      "epoch": 0.140117994100295,
+      "grad_norm": 9.0,
+      "learning_rate": 1.9836134528933733e-05,
+      "loss": 0.4892,
+      "step": 2280
+    },
+    {
+      "epoch": 0.1401794493608653,
+      "grad_norm": 21.375,
+      "learning_rate": 1.983571470813386e-05,
+      "loss": 0.5442,
+      "step": 2281
+    },
+    {
+      "epoch": 0.1402409046214356,
+      "grad_norm": 24.875,
+      "learning_rate": 1.983529435468865e-05,
+      "loss": 0.7,
+      "step": 2282
+    },
+    {
+      "epoch": 0.1403023598820059,
+      "grad_norm": 9.5,
+      "learning_rate": 1.9834873468620864e-05,
+      "loss": 0.3771,
+      "step": 2283
+    },
+    {
+      "epoch": 0.1403638151425762,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.98344520499533e-05,
+      "loss": 0.4826,
+      "step": 2284
+    },
+    {
+      "epoch": 0.1404252704031465,
+      "grad_norm": 36.5,
+      "learning_rate": 1.9834030098708773e-05,
+      "loss": 0.766,
+      "step": 2285
+    },
+    {
+      "epoch": 0.1404867256637168,
+      "grad_norm": 36.5,
+      "learning_rate": 1.983360761491014e-05,
+      "loss": 0.5972,
+      "step": 2286
+    },
+    {
+      "epoch": 0.1405481809242871,
+      "grad_norm": 20.5,
+      "learning_rate": 1.983318459858028e-05,
+      "loss": 0.5578,
+      "step": 2287
+    },
+    {
+      "epoch": 0.14060963618485742,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.9832761049742092e-05,
+      "loss": 0.4374,
+      "step": 2288
+    },
+    {
+      "epoch": 0.14067109144542772,
+      "grad_norm": 21.375,
+      "learning_rate": 1.9832336968418527e-05,
+      "loss": 0.5643,
+      "step": 2289
+    },
+    {
+      "epoch": 0.14073254670599802,
+      "grad_norm": 19.25,
+      "learning_rate": 1.9831912354632537e-05,
+      "loss": 0.4419,
+      "step": 2290
+    },
+    {
+      "epoch": 0.14079400196656833,
+      "grad_norm": 10.25,
+      "learning_rate": 1.9831487208407126e-05,
+      "loss": 0.6119,
+      "step": 2291
+    },
+    {
+      "epoch": 0.14085545722713863,
+      "grad_norm": 32.25,
+      "learning_rate": 1.9831061529765316e-05,
+      "loss": 0.6325,
+      "step": 2292
+    },
+    {
+      "epoch": 0.14091691248770893,
+      "grad_norm": 51.5,
+      "learning_rate": 1.9830635318730155e-05,
+      "loss": 0.5302,
+      "step": 2293
+    },
+    {
+      "epoch": 0.14097836774827927,
+      "grad_norm": 19.5,
+      "learning_rate": 1.983020857532473e-05,
+      "loss": 0.5718,
+      "step": 2294
+    },
+    {
+      "epoch": 0.14103982300884957,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.9829781299572147e-05,
+      "loss": 0.4449,
+      "step": 2295
+    },
+    {
+      "epoch": 0.14110127826941987,
+      "grad_norm": 30.375,
+      "learning_rate": 1.9829353491495545e-05,
+      "loss": 0.4569,
+      "step": 2296
+    },
+    {
+      "epoch": 0.14116273352999018,
+      "grad_norm": 30.125,
+      "learning_rate": 1.9828925151118092e-05,
+      "loss": 0.7089,
+      "step": 2297
+    },
+    {
+      "epoch": 0.14122418879056048,
+      "grad_norm": 24.25,
+      "learning_rate": 1.9828496278462987e-05,
+      "loss": 0.7536,
+      "step": 2298
+    },
+    {
+      "epoch": 0.14128564405113078,
+      "grad_norm": 25.875,
+      "learning_rate": 1.982806687355345e-05,
+      "loss": 0.5499,
+      "step": 2299
+    },
+    {
+      "epoch": 0.14134709931170109,
+      "grad_norm": 17.75,
+      "learning_rate": 1.9827636936412736e-05,
+      "loss": 0.6142,
+      "step": 2300
+    },
+    {
+      "epoch": 0.1414085545722714,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.9827206467064133e-05,
+      "loss": 0.5247,
+      "step": 2301
+    },
+    {
+      "epoch": 0.1414700098328417,
+      "grad_norm": 17.125,
+      "learning_rate": 1.982677546553095e-05,
+      "loss": 0.5337,
+      "step": 2302
+    },
+    {
+      "epoch": 0.141531465093412,
+      "grad_norm": 25.75,
+      "learning_rate": 1.982634393183653e-05,
+      "loss": 0.6573,
+      "step": 2303
+    },
+    {
+      "epoch": 0.1415929203539823,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.9825911866004235e-05,
+      "loss": 0.5288,
+      "step": 2304
+    },
+    {
+      "epoch": 0.1416543756145526,
+      "grad_norm": 18.25,
+      "learning_rate": 1.982547926805747e-05,
+      "loss": 0.4985,
+      "step": 2305
+    },
+    {
+      "epoch": 0.1417158308751229,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.9825046138019658e-05,
+      "loss": 0.607,
+      "step": 2306
+    },
+    {
+      "epoch": 0.1417772861356932,
+      "grad_norm": 18.0,
+      "learning_rate": 1.982461247591426e-05,
+      "loss": 0.4951,
+      "step": 2307
+    },
+    {
+      "epoch": 0.1418387413962635,
+      "grad_norm": 32.0,
+      "learning_rate": 1.9824178281764753e-05,
+      "loss": 0.4146,
+      "step": 2308
+    },
+    {
+      "epoch": 0.14190019665683382,
+      "grad_norm": 27.125,
+      "learning_rate": 1.9823743555594655e-05,
+      "loss": 0.4983,
+      "step": 2309
+    },
+    {
+      "epoch": 0.14196165191740412,
+      "grad_norm": 26.875,
+      "learning_rate": 1.9823308297427514e-05,
+      "loss": 0.3965,
+      "step": 2310
+    },
+    {
+      "epoch": 0.14202310717797442,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.982287250728689e-05,
+      "loss": 0.3937,
+      "step": 2311
+    },
+    {
+      "epoch": 0.14208456243854473,
+      "grad_norm": 16.125,
+      "learning_rate": 1.982243618519639e-05,
+      "loss": 0.4282,
+      "step": 2312
+    },
+    {
+      "epoch": 0.14214601769911506,
+      "grad_norm": 39.25,
+      "learning_rate": 1.9821999331179635e-05,
+      "loss": 0.5139,
+      "step": 2313
+    },
+    {
+      "epoch": 0.14220747295968536,
+      "grad_norm": 26.125,
+      "learning_rate": 1.9821561945260292e-05,
+      "loss": 0.4733,
+      "step": 2314
+    },
+    {
+      "epoch": 0.14226892822025566,
+      "grad_norm": 15.75,
+      "learning_rate": 1.9821124027462043e-05,
+      "loss": 0.5178,
+      "step": 2315
+    },
+    {
+      "epoch": 0.14233038348082597,
+      "grad_norm": 28.875,
+      "learning_rate": 1.9820685577808604e-05,
+      "loss": 0.6117,
+      "step": 2316
+    },
+    {
+      "epoch": 0.14239183874139627,
+      "grad_norm": 25.75,
+      "learning_rate": 1.982024659632372e-05,
+      "loss": 0.5936,
+      "step": 2317
+    },
+    {
+      "epoch": 0.14245329400196657,
+      "grad_norm": 27.625,
+      "learning_rate": 1.981980708303116e-05,
+      "loss": 0.4173,
+      "step": 2318
+    },
+    {
+      "epoch": 0.14251474926253688,
+      "grad_norm": 14.0,
+      "learning_rate": 1.9819367037954725e-05,
+      "loss": 0.4138,
+      "step": 2319
+    },
+    {
+      "epoch": 0.14257620452310718,
+      "grad_norm": 19.25,
+      "learning_rate": 1.9818926461118254e-05,
+      "loss": 0.4707,
+      "step": 2320
+    },
+    {
+      "epoch": 0.14263765978367748,
+      "grad_norm": 23.875,
+      "learning_rate": 1.9818485352545595e-05,
+      "loss": 0.7985,
+      "step": 2321
+    },
+    {
+      "epoch": 0.1426991150442478,
+      "grad_norm": 20.125,
+      "learning_rate": 1.981804371226064e-05,
+      "loss": 0.6754,
+      "step": 2322
+    },
+    {
+      "epoch": 0.1427605703048181,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.981760154028731e-05,
+      "loss": 0.6029,
+      "step": 2323
+    },
+    {
+      "epoch": 0.1428220255653884,
+      "grad_norm": 24.125,
+      "learning_rate": 1.9817158836649542e-05,
+      "loss": 0.5088,
+      "step": 2324
+    },
+    {
+      "epoch": 0.1428834808259587,
+      "grad_norm": 22.25,
+      "learning_rate": 1.981671560137132e-05,
+      "loss": 0.4542,
+      "step": 2325
+    },
+    {
+      "epoch": 0.142944936086529,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.9816271834476642e-05,
+      "loss": 0.7558,
+      "step": 2326
+    },
+    {
+      "epoch": 0.1430063913470993,
+      "grad_norm": 22.5,
+      "learning_rate": 1.981582753598954e-05,
+      "loss": 0.5192,
+      "step": 2327
+    },
+    {
+      "epoch": 0.1430678466076696,
+      "grad_norm": 31.875,
+      "learning_rate": 1.981538270593407e-05,
+      "loss": 0.4768,
+      "step": 2328
+    },
+    {
+      "epoch": 0.1431293018682399,
+      "grad_norm": 37.5,
+      "learning_rate": 1.981493734433433e-05,
+      "loss": 0.4595,
+      "step": 2329
+    },
+    {
+      "epoch": 0.14319075712881021,
+      "grad_norm": 32.5,
+      "learning_rate": 1.9814491451214434e-05,
+      "loss": 0.6074,
+      "step": 2330
+    },
+    {
+      "epoch": 0.14325221238938052,
+      "grad_norm": 27.75,
+      "learning_rate": 1.981404502659853e-05,
+      "loss": 0.5599,
+      "step": 2331
+    },
+    {
+      "epoch": 0.14331366764995085,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.981359807051079e-05,
+      "loss": 0.5609,
+      "step": 2332
+    },
+    {
+      "epoch": 0.14337512291052115,
+      "grad_norm": 11.0625,
+      "learning_rate": 1.9813150582975425e-05,
+      "loss": 0.4313,
+      "step": 2333
+    },
+    {
+      "epoch": 0.14343657817109146,
+      "grad_norm": 18.25,
+      "learning_rate": 1.9812702564016664e-05,
+      "loss": 0.4355,
+      "step": 2334
+    },
+    {
+      "epoch": 0.14349803343166176,
+      "grad_norm": 8.375,
+      "learning_rate": 1.981225401365877e-05,
+      "loss": 0.352,
+      "step": 2335
+    },
+    {
+      "epoch": 0.14355948869223206,
+      "grad_norm": 13.125,
+      "learning_rate": 1.981180493192603e-05,
+      "loss": 0.4398,
+      "step": 2336
+    },
+    {
+      "epoch": 0.14362094395280237,
+      "grad_norm": 25.375,
+      "learning_rate": 1.9811355318842776e-05,
+      "loss": 0.5923,
+      "step": 2337
+    },
+    {
+      "epoch": 0.14368239921337267,
+      "grad_norm": 16.5,
+      "learning_rate": 1.981090517443334e-05,
+      "loss": 0.4834,
+      "step": 2338
+    },
+    {
+      "epoch": 0.14374385447394297,
+      "grad_norm": 21.125,
+      "learning_rate": 1.9810454498722114e-05,
+      "loss": 0.4312,
+      "step": 2339
+    },
+    {
+      "epoch": 0.14380530973451328,
+      "grad_norm": 6.375,
+      "learning_rate": 1.9810003291733492e-05,
+      "loss": 0.3149,
+      "step": 2340
+    },
+    {
+      "epoch": 0.14386676499508358,
+      "grad_norm": 49.25,
+      "learning_rate": 1.9809551553491918e-05,
+      "loss": 0.4829,
+      "step": 2341
+    },
+    {
+      "epoch": 0.14392822025565388,
+      "grad_norm": 5.90625,
+      "learning_rate": 1.980909928402185e-05,
+      "loss": 0.378,
+      "step": 2342
+    },
+    {
+      "epoch": 0.14398967551622419,
+      "grad_norm": 13.0,
+      "learning_rate": 1.980864648334778e-05,
+      "loss": 0.3847,
+      "step": 2343
+    },
+    {
+      "epoch": 0.1440511307767945,
+      "grad_norm": 5.03125,
+      "learning_rate": 1.9808193151494233e-05,
+      "loss": 0.3351,
+      "step": 2344
+    },
+    {
+      "epoch": 0.1441125860373648,
+      "grad_norm": 10.8125,
+      "learning_rate": 1.9807739288485758e-05,
+      "loss": 0.3433,
+      "step": 2345
+    },
+    {
+      "epoch": 0.1441740412979351,
+      "grad_norm": 49.75,
+      "learning_rate": 1.980728489434693e-05,
+      "loss": 0.5013,
+      "step": 2346
+    },
+    {
+      "epoch": 0.1442354965585054,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.9806829969102356e-05,
+      "loss": 0.373,
+      "step": 2347
+    },
+    {
+      "epoch": 0.1442969518190757,
+      "grad_norm": 17.5,
+      "learning_rate": 1.980637451277668e-05,
+      "loss": 0.4202,
+      "step": 2348
+    },
+    {
+      "epoch": 0.144358407079646,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.980591852539456e-05,
+      "loss": 0.4677,
+      "step": 2349
+    },
+    {
+      "epoch": 0.1444198623402163,
+      "grad_norm": 20.125,
+      "learning_rate": 1.9805462006980688e-05,
+      "loss": 0.4328,
+      "step": 2350
+    },
+    {
+      "epoch": 0.14448131760078664,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9805004957559795e-05,
+      "loss": 0.4094,
+      "step": 2351
+    },
+    {
+      "epoch": 0.14454277286135694,
+      "grad_norm": 29.375,
+      "learning_rate": 1.9804547377156624e-05,
+      "loss": 0.3067,
+      "step": 2352
+    },
+    {
+      "epoch": 0.14460422812192725,
+      "grad_norm": 28.625,
+      "learning_rate": 1.980408926579596e-05,
+      "loss": 0.7151,
+      "step": 2353
+    },
+    {
+      "epoch": 0.14466568338249755,
+      "grad_norm": 20.75,
+      "learning_rate": 1.9803630623502606e-05,
+      "loss": 0.5994,
+      "step": 2354
+    },
+    {
+      "epoch": 0.14472713864306785,
+      "grad_norm": 25.75,
+      "learning_rate": 1.9803171450301405e-05,
+      "loss": 0.6675,
+      "step": 2355
+    },
+    {
+      "epoch": 0.14478859390363816,
+      "grad_norm": 24.125,
+      "learning_rate": 1.9802711746217222e-05,
+      "loss": 0.6266,
+      "step": 2356
+    },
+    {
+      "epoch": 0.14485004916420846,
+      "grad_norm": 26.5,
+      "learning_rate": 1.9802251511274946e-05,
+      "loss": 0.7621,
+      "step": 2357
+    },
+    {
+      "epoch": 0.14491150442477876,
+      "grad_norm": 19.5,
+      "learning_rate": 1.980179074549951e-05,
+      "loss": 0.5386,
+      "step": 2358
+    },
+    {
+      "epoch": 0.14497295968534907,
+      "grad_norm": 13.875,
+      "learning_rate": 1.9801329448915863e-05,
+      "loss": 0.4819,
+      "step": 2359
+    },
+    {
+      "epoch": 0.14503441494591937,
+      "grad_norm": 10.6875,
+      "learning_rate": 1.980086762154898e-05,
+      "loss": 0.4122,
+      "step": 2360
+    },
+    {
+      "epoch": 0.14509587020648967,
+      "grad_norm": 16.625,
+      "learning_rate": 1.980040526342388e-05,
+      "loss": 0.608,
+      "step": 2361
+    },
+    {
+      "epoch": 0.14515732546705998,
+      "grad_norm": 21.5,
+      "learning_rate": 1.9799942374565597e-05,
+      "loss": 0.5776,
+      "step": 2362
+    },
+    {
+      "epoch": 0.14521878072763028,
+      "grad_norm": 21.0,
+      "learning_rate": 1.9799478954999198e-05,
+      "loss": 0.7758,
+      "step": 2363
+    },
+    {
+      "epoch": 0.14528023598820058,
+      "grad_norm": 20.0,
+      "learning_rate": 1.979901500474978e-05,
+      "loss": 0.7533,
+      "step": 2364
+    },
+    {
+      "epoch": 0.1453416912487709,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.979855052384247e-05,
+      "loss": 0.679,
+      "step": 2365
+    },
+    {
+      "epoch": 0.1454031465093412,
+      "grad_norm": 11.4375,
+      "learning_rate": 1.9798085512302418e-05,
+      "loss": 0.3526,
+      "step": 2366
+    },
+    {
+      "epoch": 0.1454646017699115,
+      "grad_norm": 17.375,
+      "learning_rate": 1.9797619970154807e-05,
+      "loss": 0.4703,
+      "step": 2367
+    },
+    {
+      "epoch": 0.1455260570304818,
+      "grad_norm": 15.625,
+      "learning_rate": 1.9797153897424854e-05,
+      "loss": 0.3675,
+      "step": 2368
+    },
+    {
+      "epoch": 0.1455875122910521,
+      "grad_norm": 19.0,
+      "learning_rate": 1.979668729413779e-05,
+      "loss": 0.4406,
+      "step": 2369
+    },
+    {
+      "epoch": 0.14564896755162243,
+      "grad_norm": 14.875,
+      "learning_rate": 1.9796220160318885e-05,
+      "loss": 0.5295,
+      "step": 2370
+    },
+    {
+      "epoch": 0.14571042281219274,
+      "grad_norm": 25.875,
+      "learning_rate": 1.979575249599344e-05,
+      "loss": 0.5895,
+      "step": 2371
+    },
+    {
+      "epoch": 0.14577187807276304,
+      "grad_norm": 14.625,
+      "learning_rate": 1.9795284301186782e-05,
+      "loss": 0.4618,
+      "step": 2372
+    },
+    {
+      "epoch": 0.14583333333333334,
+      "grad_norm": 28.625,
+      "learning_rate": 1.9794815575924264e-05,
+      "loss": 0.5992,
+      "step": 2373
+    },
+    {
+      "epoch": 0.14589478859390365,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.9794346320231265e-05,
+      "loss": 0.3861,
+      "step": 2374
+    },
+    {
+      "epoch": 0.14595624385447395,
+      "grad_norm": 8.625,
+      "learning_rate": 1.9793876534133205e-05,
+      "loss": 0.3257,
+      "step": 2375
+    },
+    {
+      "epoch": 0.14601769911504425,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.9793406217655516e-05,
+      "loss": 0.3326,
+      "step": 2376
+    },
+    {
+      "epoch": 0.14607915437561456,
+      "grad_norm": 18.125,
+      "learning_rate": 1.9792935370823676e-05,
+      "loss": 0.4156,
+      "step": 2377
+    },
+    {
+      "epoch": 0.14614060963618486,
+      "grad_norm": 18.75,
+      "learning_rate": 1.979246399366318e-05,
+      "loss": 0.4597,
+      "step": 2378
+    },
+    {
+      "epoch": 0.14620206489675516,
+      "grad_norm": 20.625,
+      "learning_rate": 1.9791992086199552e-05,
+      "loss": 0.4928,
+      "step": 2379
+    },
+    {
+      "epoch": 0.14626352015732547,
+      "grad_norm": 19.5,
+      "learning_rate": 1.9791519648458352e-05,
+      "loss": 0.3948,
+      "step": 2380
+    },
+    {
+      "epoch": 0.14632497541789577,
+      "grad_norm": 26.625,
+      "learning_rate": 1.979104668046516e-05,
+      "loss": 0.4023,
+      "step": 2381
+    },
+    {
+      "epoch": 0.14638643067846607,
+      "grad_norm": 15.75,
+      "learning_rate": 1.97905731822456e-05,
+      "loss": 0.4679,
+      "step": 2382
+    },
+    {
+      "epoch": 0.14644788593903638,
+      "grad_norm": 12.125,
+      "learning_rate": 1.97900991538253e-05,
+      "loss": 0.4376,
+      "step": 2383
+    },
+    {
+      "epoch": 0.14650934119960668,
+      "grad_norm": 23.25,
+      "learning_rate": 1.9789624595229937e-05,
+      "loss": 0.4788,
+      "step": 2384
+    },
+    {
+      "epoch": 0.14657079646017698,
+      "grad_norm": 23.875,
+      "learning_rate": 1.978914950648521e-05,
+      "loss": 0.5495,
+      "step": 2385
+    },
+    {
+      "epoch": 0.1466322517207473,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.9788673887616852e-05,
+      "loss": 0.4129,
+      "step": 2386
+    },
+    {
+      "epoch": 0.1466937069813176,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.9788197738650613e-05,
+      "loss": 0.6316,
+      "step": 2387
+    },
+    {
+      "epoch": 0.1467551622418879,
+      "grad_norm": 12.125,
+      "learning_rate": 1.9787721059612276e-05,
+      "loss": 0.4781,
+      "step": 2388
+    },
+    {
+      "epoch": 0.14681661750245822,
+      "grad_norm": 24.25,
+      "learning_rate": 1.9787243850527663e-05,
+      "loss": 0.3069,
+      "step": 2389
+    },
+    {
+      "epoch": 0.14687807276302853,
+      "grad_norm": 23.5,
+      "learning_rate": 1.9786766111422616e-05,
+      "loss": 0.3726,
+      "step": 2390
+    },
+    {
+      "epoch": 0.14693952802359883,
+      "grad_norm": 31.375,
+      "learning_rate": 1.9786287842323002e-05,
+      "loss": 0.3191,
+      "step": 2391
+    },
+    {
+      "epoch": 0.14700098328416913,
+      "grad_norm": 22.375,
+      "learning_rate": 1.978580904325472e-05,
+      "loss": 0.4062,
+      "step": 2392
+    },
+    {
+      "epoch": 0.14706243854473944,
+      "grad_norm": 49.25,
+      "learning_rate": 1.978532971424371e-05,
+      "loss": 0.4739,
+      "step": 2393
+    },
+    {
+      "epoch": 0.14712389380530974,
+      "grad_norm": 5.875,
+      "learning_rate": 1.9784849855315914e-05,
+      "loss": 0.3387,
+      "step": 2394
+    },
+    {
+      "epoch": 0.14718534906588004,
+      "grad_norm": 11.1875,
+      "learning_rate": 1.9784369466497333e-05,
+      "loss": 0.6701,
+      "step": 2395
+    },
+    {
+      "epoch": 0.14724680432645035,
+      "grad_norm": 23.375,
+      "learning_rate": 1.978388854781397e-05,
+      "loss": 0.5333,
+      "step": 2396
+    },
+    {
+      "epoch": 0.14730825958702065,
+      "grad_norm": 18.625,
+      "learning_rate": 1.978340709929188e-05,
+      "loss": 0.5456,
+      "step": 2397
+    },
+    {
+      "epoch": 0.14736971484759095,
+      "grad_norm": 19.375,
+      "learning_rate": 1.9782925120957123e-05,
+      "loss": 0.3883,
+      "step": 2398
+    },
+    {
+      "epoch": 0.14743117010816126,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.9782442612835812e-05,
+      "loss": 0.3444,
+      "step": 2399
+    },
+    {
+      "epoch": 0.14749262536873156,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.978195957495407e-05,
+      "loss": 0.3027,
+      "step": 2400
+    },
+    {
+      "epoch": 0.14755408062930186,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.9781476007338058e-05,
+      "loss": 0.4413,
+      "step": 2401
+    },
+    {
+      "epoch": 0.14761553588987217,
+      "grad_norm": 30.25,
+      "learning_rate": 1.978099191001396e-05,
+      "loss": 0.5716,
+      "step": 2402
+    },
+    {
+      "epoch": 0.14767699115044247,
+      "grad_norm": 5.21875,
+      "learning_rate": 1.9780507283008e-05,
+      "loss": 0.4114,
+      "step": 2403
+    },
+    {
+      "epoch": 0.14773844641101277,
+      "grad_norm": 25.875,
+      "learning_rate": 1.9780022126346413e-05,
+      "loss": 0.6259,
+      "step": 2404
+    },
+    {
+      "epoch": 0.14779990167158308,
+      "grad_norm": 30.125,
+      "learning_rate": 1.9779536440055474e-05,
+      "loss": 0.5516,
+      "step": 2405
+    },
+    {
+      "epoch": 0.14786135693215338,
+      "grad_norm": 7.625,
+      "learning_rate": 1.9779050224161494e-05,
+      "loss": 0.2358,
+      "step": 2406
+    },
+    {
+      "epoch": 0.14792281219272368,
+      "grad_norm": 6.0,
+      "learning_rate": 1.977856347869079e-05,
+      "loss": 0.4169,
+      "step": 2407
+    },
+    {
+      "epoch": 0.147984267453294,
+      "grad_norm": 47.25,
+      "learning_rate": 1.977807620366973e-05,
+      "loss": 0.3913,
+      "step": 2408
+    },
+    {
+      "epoch": 0.14804572271386432,
+      "grad_norm": 19.5,
+      "learning_rate": 1.9777588399124702e-05,
+      "loss": 0.7274,
+      "step": 2409
+    },
+    {
+      "epoch": 0.14810717797443462,
+      "grad_norm": 32.75,
+      "learning_rate": 1.977710006508212e-05,
+      "loss": 0.5983,
+      "step": 2410
+    },
+    {
+      "epoch": 0.14816863323500493,
+      "grad_norm": 48.0,
+      "learning_rate": 1.9776611201568434e-05,
+      "loss": 0.7134,
+      "step": 2411
+    },
+    {
+      "epoch": 0.14823008849557523,
+      "grad_norm": 43.5,
+      "learning_rate": 1.977612180861011e-05,
+      "loss": 0.5775,
+      "step": 2412
+    },
+    {
+      "epoch": 0.14829154375614553,
+      "grad_norm": 23.625,
+      "learning_rate": 1.9775631886233655e-05,
+      "loss": 0.6386,
+      "step": 2413
+    },
+    {
+      "epoch": 0.14835299901671584,
+      "grad_norm": 18.625,
+      "learning_rate": 1.97751414344656e-05,
+      "loss": 0.4715,
+      "step": 2414
+    },
+    {
+      "epoch": 0.14841445427728614,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.9774650453332506e-05,
+      "loss": 0.6077,
+      "step": 2415
+    },
+    {
+      "epoch": 0.14847590953785644,
+      "grad_norm": 10.25,
+      "learning_rate": 1.9774158942860962e-05,
+      "loss": 0.6176,
+      "step": 2416
+    },
+    {
+      "epoch": 0.14853736479842675,
+      "grad_norm": 14.625,
+      "learning_rate": 1.977366690307758e-05,
+      "loss": 0.4794,
+      "step": 2417
+    },
+    {
+      "epoch": 0.14859882005899705,
+      "grad_norm": 8.75,
+      "learning_rate": 1.9773174334009016e-05,
+      "loss": 0.3786,
+      "step": 2418
+    },
+    {
+      "epoch": 0.14866027531956735,
+      "grad_norm": 37.25,
+      "learning_rate": 1.9772681235681936e-05,
+      "loss": 0.5023,
+      "step": 2419
+    },
+    {
+      "epoch": 0.14872173058013766,
+      "grad_norm": 32.25,
+      "learning_rate": 1.9772187608123047e-05,
+      "loss": 0.499,
+      "step": 2420
+    },
+    {
+      "epoch": 0.14878318584070796,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.977169345135908e-05,
+      "loss": 0.4382,
+      "step": 2421
+    },
+    {
+      "epoch": 0.14884464110127826,
+      "grad_norm": 30.75,
+      "learning_rate": 1.97711987654168e-05,
+      "loss": 0.4144,
+      "step": 2422
+    },
+    {
+      "epoch": 0.14890609636184857,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.9770703550322986e-05,
+      "loss": 0.3927,
+      "step": 2423
+    },
+    {
+      "epoch": 0.14896755162241887,
+      "grad_norm": 6.875,
+      "learning_rate": 1.9770207806104466e-05,
+      "loss": 0.2667,
+      "step": 2424
+    },
+    {
+      "epoch": 0.14902900688298917,
+      "grad_norm": 17.0,
+      "learning_rate": 1.9769711532788083e-05,
+      "loss": 0.5883,
+      "step": 2425
+    },
+    {
+      "epoch": 0.14909046214355948,
+      "grad_norm": 4.71875,
+      "learning_rate": 1.976921473040071e-05,
+      "loss": 0.4883,
+      "step": 2426
+    },
+    {
+      "epoch": 0.14915191740412978,
+      "grad_norm": 40.5,
+      "learning_rate": 1.9768717398969258e-05,
+      "loss": 0.6213,
+      "step": 2427
+    },
+    {
+      "epoch": 0.1492133726647001,
+      "grad_norm": 60.0,
+      "learning_rate": 1.976821953852065e-05,
+      "loss": 0.5008,
+      "step": 2428
+    },
+    {
+      "epoch": 0.14927482792527041,
+      "grad_norm": 36.25,
+      "learning_rate": 1.9767721149081852e-05,
+      "loss": 0.5369,
+      "step": 2429
+    },
+    {
+      "epoch": 0.14933628318584072,
+      "grad_norm": 36.0,
+      "learning_rate": 1.9767222230679858e-05,
+      "loss": 0.473,
+      "step": 2430
+    },
+    {
+      "epoch": 0.14939773844641102,
+      "grad_norm": 47.25,
+      "learning_rate": 1.9766722783341682e-05,
+      "loss": 0.5133,
+      "step": 2431
+    },
+    {
+      "epoch": 0.14945919370698132,
+      "grad_norm": 34.25,
+      "learning_rate": 1.976622280709437e-05,
+      "loss": 0.5707,
+      "step": 2432
+    },
+    {
+      "epoch": 0.14952064896755163,
+      "grad_norm": 12.5,
+      "learning_rate": 1.9765722301965e-05,
+      "loss": 0.5615,
+      "step": 2433
+    },
+    {
+      "epoch": 0.14958210422812193,
+      "grad_norm": 22.625,
+      "learning_rate": 1.9765221267980675e-05,
+      "loss": 0.4909,
+      "step": 2434
+    },
+    {
+      "epoch": 0.14964355948869223,
+      "grad_norm": 41.0,
+      "learning_rate": 1.9764719705168528e-05,
+      "loss": 0.3712,
+      "step": 2435
+    },
+    {
+      "epoch": 0.14970501474926254,
+      "grad_norm": 12.0,
+      "learning_rate": 1.976421761355572e-05,
+      "loss": 0.3455,
+      "step": 2436
+    },
+    {
+      "epoch": 0.14976647000983284,
+      "grad_norm": 29.625,
+      "learning_rate": 1.976371499316945e-05,
+      "loss": 0.4265,
+      "step": 2437
+    },
+    {
+      "epoch": 0.14982792527040315,
+      "grad_norm": 13.875,
+      "learning_rate": 1.9763211844036928e-05,
+      "loss": 0.5158,
+      "step": 2438
+    },
+    {
+      "epoch": 0.14988938053097345,
+      "grad_norm": 22.875,
+      "learning_rate": 1.97627081661854e-05,
+      "loss": 0.5486,
+      "step": 2439
+    },
+    {
+      "epoch": 0.14995083579154375,
+      "grad_norm": 38.25,
+      "learning_rate": 1.976220395964215e-05,
+      "loss": 0.4846,
+      "step": 2440
+    },
+    {
+      "epoch": 0.15001229105211406,
+      "grad_norm": 40.75,
+      "learning_rate": 1.9761699224434476e-05,
+      "loss": 0.6749,
+      "step": 2441
+    },
+    {
+      "epoch": 0.15007374631268436,
+      "grad_norm": 27.375,
+      "learning_rate": 1.976119396058972e-05,
+      "loss": 0.47,
+      "step": 2442
+    },
+    {
+      "epoch": 0.15013520157325466,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.9760688168135233e-05,
+      "loss": 0.4736,
+      "step": 2443
+    },
+    {
+      "epoch": 0.15019665683382497,
+      "grad_norm": 43.75,
+      "learning_rate": 1.9760181847098412e-05,
+      "loss": 0.4598,
+      "step": 2444
+    },
+    {
+      "epoch": 0.15025811209439527,
+      "grad_norm": 21.0,
+      "learning_rate": 1.975967499750668e-05,
+      "loss": 0.3586,
+      "step": 2445
+    },
+    {
+      "epoch": 0.15031956735496557,
+      "grad_norm": 37.75,
+      "learning_rate": 1.9759167619387474e-05,
+      "loss": 0.503,
+      "step": 2446
+    },
+    {
+      "epoch": 0.1503810226155359,
+      "grad_norm": 13.375,
+      "learning_rate": 1.9758659712768283e-05,
+      "loss": 0.4314,
+      "step": 2447
+    },
+    {
+      "epoch": 0.1504424778761062,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.9758151277676603e-05,
+      "loss": 0.5163,
+      "step": 2448
+    },
+    {
+      "epoch": 0.1505039331366765,
+      "grad_norm": 25.75,
+      "learning_rate": 1.9757642314139977e-05,
+      "loss": 0.458,
+      "step": 2449
+    },
+    {
+      "epoch": 0.1505653883972468,
+      "grad_norm": 27.375,
+      "learning_rate": 1.975713282218596e-05,
+      "loss": 0.5305,
+      "step": 2450
+    },
+    {
+      "epoch": 0.15062684365781712,
+      "grad_norm": 14.25,
+      "learning_rate": 1.9756622801842144e-05,
+      "loss": 0.5893,
+      "step": 2451
+    },
+    {
+      "epoch": 0.15068829891838742,
+      "grad_norm": 31.125,
+      "learning_rate": 1.9756112253136154e-05,
+      "loss": 0.421,
+      "step": 2452
+    },
+    {
+      "epoch": 0.15074975417895772,
+      "grad_norm": 16.0,
+      "learning_rate": 1.975560117609563e-05,
+      "loss": 0.3991,
+      "step": 2453
+    },
+    {
+      "epoch": 0.15081120943952803,
+      "grad_norm": 16.875,
+      "learning_rate": 1.9755089570748255e-05,
+      "loss": 0.4059,
+      "step": 2454
+    },
+    {
+      "epoch": 0.15087266470009833,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.9754577437121733e-05,
+      "loss": 0.3474,
+      "step": 2455
+    },
+    {
+      "epoch": 0.15093411996066863,
+      "grad_norm": 20.0,
+      "learning_rate": 1.9754064775243797e-05,
+      "loss": 0.5048,
+      "step": 2456
+    },
+    {
+      "epoch": 0.15099557522123894,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.9753551585142214e-05,
+      "loss": 0.4424,
+      "step": 2457
+    },
+    {
+      "epoch": 0.15105703048180924,
+      "grad_norm": 31.0,
+      "learning_rate": 1.975303786684477e-05,
+      "loss": 0.5277,
+      "step": 2458
+    },
+    {
+      "epoch": 0.15111848574237954,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9752523620379288e-05,
+      "loss": 0.4678,
+      "step": 2459
+    },
+    {
+      "epoch": 0.15117994100294985,
+      "grad_norm": 14.75,
+      "learning_rate": 1.9752008845773618e-05,
+      "loss": 0.4845,
+      "step": 2460
+    },
+    {
+      "epoch": 0.15124139626352015,
+      "grad_norm": 34.25,
+      "learning_rate": 1.9751493543055634e-05,
+      "loss": 0.6685,
+      "step": 2461
+    },
+    {
+      "epoch": 0.15130285152409045,
+      "grad_norm": 15.5,
+      "learning_rate": 1.9750977712253242e-05,
+      "loss": 0.4018,
+      "step": 2462
+    },
+    {
+      "epoch": 0.15136430678466076,
+      "grad_norm": 8.625,
+      "learning_rate": 1.975046135339438e-05,
+      "loss": 0.4305,
+      "step": 2463
+    },
+    {
+      "epoch": 0.15142576204523106,
+      "grad_norm": 9.875,
+      "learning_rate": 1.9749944466507007e-05,
+      "loss": 0.5154,
+      "step": 2464
+    },
+    {
+      "epoch": 0.15148721730580136,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.9749427051619118e-05,
+      "loss": 0.3798,
+      "step": 2465
+    },
+    {
+      "epoch": 0.1515486725663717,
+      "grad_norm": 9.875,
+      "learning_rate": 1.9748909108758727e-05,
+      "loss": 0.5277,
+      "step": 2466
+    },
+    {
+      "epoch": 0.151610127826942,
+      "grad_norm": 17.375,
+      "learning_rate": 1.974839063795389e-05,
+      "loss": 0.5633,
+      "step": 2467
+    },
+    {
+      "epoch": 0.1516715830875123,
+      "grad_norm": 14.75,
+      "learning_rate": 1.974787163923268e-05,
+      "loss": 0.588,
+      "step": 2468
+    },
+    {
+      "epoch": 0.1517330383480826,
+      "grad_norm": 16.125,
+      "learning_rate": 1.9747352112623207e-05,
+      "loss": 0.4127,
+      "step": 2469
+    },
+    {
+      "epoch": 0.1517944936086529,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.9746832058153602e-05,
+      "loss": 0.3546,
+      "step": 2470
+    },
+    {
+      "epoch": 0.1518559488692232,
+      "grad_norm": 7.25,
+      "learning_rate": 1.9746311475852028e-05,
+      "loss": 0.3774,
+      "step": 2471
+    },
+    {
+      "epoch": 0.15191740412979352,
+      "grad_norm": 36.25,
+      "learning_rate": 1.9745790365746678e-05,
+      "loss": 0.3119,
+      "step": 2472
+    },
+    {
+      "epoch": 0.15197885939036382,
+      "grad_norm": 28.0,
+      "learning_rate": 1.9745268727865774e-05,
+      "loss": 0.6003,
+      "step": 2473
+    },
+    {
+      "epoch": 0.15204031465093412,
+      "grad_norm": 32.75,
+      "learning_rate": 1.9744746562237563e-05,
+      "loss": 0.4319,
+      "step": 2474
+    },
+    {
+      "epoch": 0.15210176991150443,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9744223868890317e-05,
+      "loss": 0.4008,
+      "step": 2475
+    },
+    {
+      "epoch": 0.15216322517207473,
+      "grad_norm": 5.90625,
+      "learning_rate": 1.9743700647852356e-05,
+      "loss": 0.329,
+      "step": 2476
+    },
+    {
+      "epoch": 0.15222468043264503,
+      "grad_norm": 23.875,
+      "learning_rate": 1.9743176899151998e-05,
+      "loss": 0.4819,
+      "step": 2477
+    },
+    {
+      "epoch": 0.15228613569321534,
+      "grad_norm": 26.125,
+      "learning_rate": 1.9742652622817622e-05,
+      "loss": 0.4721,
+      "step": 2478
+    },
+    {
+      "epoch": 0.15234759095378564,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.9742127818877605e-05,
+      "loss": 0.5077,
+      "step": 2479
+    },
+    {
+      "epoch": 0.15240904621435594,
+      "grad_norm": 9.5,
+      "learning_rate": 1.9741602487360382e-05,
+      "loss": 0.3709,
+      "step": 2480
+    },
+    {
+      "epoch": 0.15247050147492625,
+      "grad_norm": 28.875,
+      "learning_rate": 1.9741076628294387e-05,
+      "loss": 0.586,
+      "step": 2481
+    },
+    {
+      "epoch": 0.15253195673549655,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.974055024170811e-05,
+      "loss": 0.4819,
+      "step": 2482
+    },
+    {
+      "epoch": 0.15259341199606685,
+      "grad_norm": 35.5,
+      "learning_rate": 1.974002332763005e-05,
+      "loss": 0.4552,
+      "step": 2483
+    },
+    {
+      "epoch": 0.15265486725663716,
+      "grad_norm": 19.125,
+      "learning_rate": 1.9739495886088745e-05,
+      "loss": 0.4949,
+      "step": 2484
+    },
+    {
+      "epoch": 0.1527163225172075,
+      "grad_norm": 49.25,
+      "learning_rate": 1.9738967917112752e-05,
+      "loss": 0.5519,
+      "step": 2485
+    },
+    {
+      "epoch": 0.1527777777777778,
+      "grad_norm": 28.5,
+      "learning_rate": 1.9738439420730674e-05,
+      "loss": 0.4282,
+      "step": 2486
+    },
+    {
+      "epoch": 0.1528392330383481,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9737910396971122e-05,
+      "loss": 0.4504,
+      "step": 2487
+    },
+    {
+      "epoch": 0.1529006882989184,
+      "grad_norm": 39.5,
+      "learning_rate": 1.9737380845862745e-05,
+      "loss": 0.4778,
+      "step": 2488
+    },
+    {
+      "epoch": 0.1529621435594887,
+      "grad_norm": 25.625,
+      "learning_rate": 1.9736850767434226e-05,
+      "loss": 0.4263,
+      "step": 2489
+    },
+    {
+      "epoch": 0.153023598820059,
+      "grad_norm": 25.625,
+      "learning_rate": 1.9736320161714267e-05,
+      "loss": 0.4195,
+      "step": 2490
+    },
+    {
+      "epoch": 0.1530850540806293,
+      "grad_norm": 32.25,
+      "learning_rate": 1.9735789028731603e-05,
+      "loss": 0.3562,
+      "step": 2491
+    },
+    {
+      "epoch": 0.1531465093411996,
+      "grad_norm": 31.375,
+      "learning_rate": 1.9735257368515e-05,
+      "loss": 0.5781,
+      "step": 2492
+    },
+    {
+      "epoch": 0.1532079646017699,
+      "grad_norm": 16.625,
+      "learning_rate": 1.9734725181093246e-05,
+      "loss": 0.4943,
+      "step": 2493
+    },
+    {
+      "epoch": 0.15326941986234022,
+      "grad_norm": 10.375,
+      "learning_rate": 1.9734192466495162e-05,
+      "loss": 0.4052,
+      "step": 2494
+    },
+    {
+      "epoch": 0.15333087512291052,
+      "grad_norm": 16.375,
+      "learning_rate": 1.9733659224749596e-05,
+      "loss": 0.365,
+      "step": 2495
+    },
+    {
+      "epoch": 0.15339233038348082,
+      "grad_norm": 18.625,
+      "learning_rate": 1.973312545588543e-05,
+      "loss": 0.4682,
+      "step": 2496
+    },
+    {
+      "epoch": 0.15345378564405113,
+      "grad_norm": 6.25,
+      "learning_rate": 1.9732591159931564e-05,
+      "loss": 0.3856,
+      "step": 2497
+    },
+    {
+      "epoch": 0.15351524090462143,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.9732056336916936e-05,
+      "loss": 0.6198,
+      "step": 2498
+    },
+    {
+      "epoch": 0.15357669616519173,
+      "grad_norm": 25.75,
+      "learning_rate": 1.9731520986870505e-05,
+      "loss": 0.4293,
+      "step": 2499
+    },
+    {
+      "epoch": 0.15363815142576204,
+      "grad_norm": 22.125,
+      "learning_rate": 1.9730985109821268e-05,
+      "loss": 0.6935,
+      "step": 2500
+    },
+    {
+      "epoch": 0.15369960668633234,
+      "grad_norm": 22.625,
+      "learning_rate": 1.973044870579824e-05,
+      "loss": 0.2953,
+      "step": 2501
+    },
+    {
+      "epoch": 0.15376106194690264,
+      "grad_norm": 40.25,
+      "learning_rate": 1.9729911774830472e-05,
+      "loss": 0.4655,
+      "step": 2502
+    },
+    {
+      "epoch": 0.15382251720747295,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.972937431694704e-05,
+      "loss": 0.3424,
+      "step": 2503
+    },
+    {
+      "epoch": 0.15388397246804325,
+      "grad_norm": 31.625,
+      "learning_rate": 1.9728836332177054e-05,
+      "loss": 0.5138,
+      "step": 2504
+    },
+    {
+      "epoch": 0.15394542772861358,
+      "grad_norm": 28.25,
+      "learning_rate": 1.972829782054964e-05,
+      "loss": 0.6226,
+      "step": 2505
+    },
+    {
+      "epoch": 0.15400688298918389,
+      "grad_norm": 29.5,
+      "learning_rate": 1.972775878209397e-05,
+      "loss": 0.7932,
+      "step": 2506
+    },
+    {
+      "epoch": 0.1540683382497542,
+      "grad_norm": 19.375,
+      "learning_rate": 1.9727219216839223e-05,
+      "loss": 0.5903,
+      "step": 2507
+    },
+    {
+      "epoch": 0.1541297935103245,
+      "grad_norm": 46.25,
+      "learning_rate": 1.9726679124814633e-05,
+      "loss": 0.7314,
+      "step": 2508
+    },
+    {
+      "epoch": 0.1541912487708948,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.9726138506049438e-05,
+      "loss": 0.6516,
+      "step": 2509
+    },
+    {
+      "epoch": 0.1542527040314651,
+      "grad_norm": 36.25,
+      "learning_rate": 1.9725597360572918e-05,
+      "loss": 0.8025,
+      "step": 2510
+    },
+    {
+      "epoch": 0.1543141592920354,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.9725055688414378e-05,
+      "loss": 0.6405,
+      "step": 2511
+    },
+    {
+      "epoch": 0.1543756145526057,
+      "grad_norm": 18.25,
+      "learning_rate": 1.9724513489603153e-05,
+      "loss": 0.6724,
+      "step": 2512
+    },
+    {
+      "epoch": 0.154437069813176,
+      "grad_norm": 35.0,
+      "learning_rate": 1.9723970764168606e-05,
+      "loss": 0.4985,
+      "step": 2513
+    },
+    {
+      "epoch": 0.1544985250737463,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.9723427512140124e-05,
+      "loss": 0.4238,
+      "step": 2514
+    },
+    {
+      "epoch": 0.15455998033431662,
+      "grad_norm": 20.625,
+      "learning_rate": 1.9722883733547128e-05,
+      "loss": 0.4777,
+      "step": 2515
+    },
+    {
+      "epoch": 0.15462143559488692,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.972233942841907e-05,
+      "loss": 0.3481,
+      "step": 2516
+    },
+    {
+      "epoch": 0.15468289085545722,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.9721794596785423e-05,
+      "loss": 0.4643,
+      "step": 2517
+    },
+    {
+      "epoch": 0.15474434611602753,
+      "grad_norm": 8.625,
+      "learning_rate": 1.9721249238675688e-05,
+      "loss": 0.3452,
+      "step": 2518
+    },
+    {
+      "epoch": 0.15480580137659783,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9720703354119407e-05,
+      "loss": 0.3795,
+      "step": 2519
+    },
+    {
+      "epoch": 0.15486725663716813,
+      "grad_norm": 24.0,
+      "learning_rate": 1.9720156943146133e-05,
+      "loss": 0.4835,
+      "step": 2520
+    },
+    {
+      "epoch": 0.15492871189773844,
+      "grad_norm": 17.75,
+      "learning_rate": 1.9719610005785466e-05,
+      "loss": 0.5871,
+      "step": 2521
+    },
+    {
+      "epoch": 0.15499016715830874,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.9719062542067014e-05,
+      "loss": 0.5449,
+      "step": 2522
+    },
+    {
+      "epoch": 0.15505162241887904,
+      "grad_norm": 20.375,
+      "learning_rate": 1.9718514552020434e-05,
+      "loss": 0.4589,
+      "step": 2523
+    },
+    {
+      "epoch": 0.15511307767944937,
+      "grad_norm": 13.75,
+      "learning_rate": 1.97179660356754e-05,
+      "loss": 0.5105,
+      "step": 2524
+    },
+    {
+      "epoch": 0.15517453294001968,
+      "grad_norm": 18.5,
+      "learning_rate": 1.971741699306161e-05,
+      "loss": 0.4686,
+      "step": 2525
+    },
+    {
+      "epoch": 0.15523598820058998,
+      "grad_norm": 17.75,
+      "learning_rate": 1.9716867424208805e-05,
+      "loss": 0.539,
+      "step": 2526
+    },
+    {
+      "epoch": 0.15529744346116028,
+      "grad_norm": 20.75,
+      "learning_rate": 1.971631732914674e-05,
+      "loss": 0.4935,
+      "step": 2527
+    },
+    {
+      "epoch": 0.1553588987217306,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.971576670790521e-05,
+      "loss": 0.3302,
+      "step": 2528
+    },
+    {
+      "epoch": 0.1554203539823009,
+      "grad_norm": 16.75,
+      "learning_rate": 1.9715215560514032e-05,
+      "loss": 0.5374,
+      "step": 2529
+    },
+    {
+      "epoch": 0.1554818092428712,
+      "grad_norm": 19.375,
+      "learning_rate": 1.9714663887003055e-05,
+      "loss": 0.6986,
+      "step": 2530
+    },
+    {
+      "epoch": 0.1555432645034415,
+      "grad_norm": 21.625,
+      "learning_rate": 1.9714111687402146e-05,
+      "loss": 0.7489,
+      "step": 2531
+    },
+    {
+      "epoch": 0.1556047197640118,
+      "grad_norm": 22.375,
+      "learning_rate": 1.971355896174122e-05,
+      "loss": 0.646,
+      "step": 2532
+    },
+    {
+      "epoch": 0.1556661750245821,
+      "grad_norm": 29.875,
+      "learning_rate": 1.9713005710050203e-05,
+      "loss": 0.4847,
+      "step": 2533
+    },
+    {
+      "epoch": 0.1557276302851524,
+      "grad_norm": 16.75,
+      "learning_rate": 1.9712451932359056e-05,
+      "loss": 0.4794,
+      "step": 2534
+    },
+    {
+      "epoch": 0.1557890855457227,
+      "grad_norm": 21.75,
+      "learning_rate": 1.971189762869777e-05,
+      "loss": 0.5296,
+      "step": 2535
+    },
+    {
+      "epoch": 0.15585054080629301,
+      "grad_norm": 13.5,
+      "learning_rate": 1.971134279909636e-05,
+      "loss": 0.4521,
+      "step": 2536
+    },
+    {
+      "epoch": 0.15591199606686332,
+      "grad_norm": 24.375,
+      "learning_rate": 1.971078744358488e-05,
+      "loss": 0.5604,
+      "step": 2537
+    },
+    {
+      "epoch": 0.15597345132743362,
+      "grad_norm": 22.375,
+      "learning_rate": 1.9710231562193397e-05,
+      "loss": 0.6101,
+      "step": 2538
+    },
+    {
+      "epoch": 0.15603490658800392,
+      "grad_norm": 31.125,
+      "learning_rate": 1.9709675154952017e-05,
+      "loss": 0.6847,
+      "step": 2539
+    },
+    {
+      "epoch": 0.15609636184857423,
+      "grad_norm": 14.75,
+      "learning_rate": 1.9709118221890872e-05,
+      "loss": 0.5018,
+      "step": 2540
+    },
+    {
+      "epoch": 0.15615781710914453,
+      "grad_norm": 36.25,
+      "learning_rate": 1.970856076304012e-05,
+      "loss": 0.7679,
+      "step": 2541
+    },
+    {
+      "epoch": 0.15621927236971483,
+      "grad_norm": 27.25,
+      "learning_rate": 1.9708002778429957e-05,
+      "loss": 0.4504,
+      "step": 2542
+    },
+    {
+      "epoch": 0.15628072763028517,
+      "grad_norm": 12.0,
+      "learning_rate": 1.970744426809059e-05,
+      "loss": 0.4568,
+      "step": 2543
+    },
+    {
+      "epoch": 0.15634218289085547,
+      "grad_norm": 15.125,
+      "learning_rate": 1.9706885232052274e-05,
+      "loss": 0.4079,
+      "step": 2544
+    },
+    {
+      "epoch": 0.15640363815142577,
+      "grad_norm": 16.0,
+      "learning_rate": 1.9706325670345276e-05,
+      "loss": 0.6281,
+      "step": 2545
+    },
+    {
+      "epoch": 0.15646509341199608,
+      "grad_norm": 17.375,
+      "learning_rate": 1.97057655829999e-05,
+      "loss": 0.7279,
+      "step": 2546
+    },
+    {
+      "epoch": 0.15652654867256638,
+      "grad_norm": 13.25,
+      "learning_rate": 1.970520497004648e-05,
+      "loss": 0.624,
+      "step": 2547
+    },
+    {
+      "epoch": 0.15658800393313668,
+      "grad_norm": 17.875,
+      "learning_rate": 1.9704643831515377e-05,
+      "loss": 0.5684,
+      "step": 2548
+    },
+    {
+      "epoch": 0.15664945919370699,
+      "grad_norm": 15.0,
+      "learning_rate": 1.9704082167436974e-05,
+      "loss": 0.5858,
+      "step": 2549
+    },
+    {
+      "epoch": 0.1567109144542773,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9703519977841686e-05,
+      "loss": 0.4229,
+      "step": 2550
+    },
+    {
+      "epoch": 0.1567723697148476,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.9702957262759964e-05,
+      "loss": 0.3657,
+      "step": 2551
+    },
+    {
+      "epoch": 0.1568338249754179,
+      "grad_norm": 22.125,
+      "learning_rate": 1.9702394022222282e-05,
+      "loss": 0.3621,
+      "step": 2552
+    },
+    {
+      "epoch": 0.1568952802359882,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.9701830256259136e-05,
+      "loss": 0.4467,
+      "step": 2553
+    },
+    {
+      "epoch": 0.1569567354965585,
+      "grad_norm": 20.125,
+      "learning_rate": 1.970126596490106e-05,
+      "loss": 0.4679,
+      "step": 2554
+    },
+    {
+      "epoch": 0.1570181907571288,
+      "grad_norm": 21.625,
+      "learning_rate": 1.970070114817861e-05,
+      "loss": 0.5566,
+      "step": 2555
+    },
+    {
+      "epoch": 0.1570796460176991,
+      "grad_norm": 17.125,
+      "learning_rate": 1.9700135806122378e-05,
+      "loss": 0.4786,
+      "step": 2556
+    },
+    {
+      "epoch": 0.1571411012782694,
+      "grad_norm": 10.5625,
+      "learning_rate": 1.9699569938762975e-05,
+      "loss": 0.3824,
+      "step": 2557
+    },
+    {
+      "epoch": 0.15720255653883972,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.969900354613105e-05,
+      "loss": 0.2935,
+      "step": 2558
+    },
+    {
+      "epoch": 0.15726401179941002,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.9698436628257266e-05,
+      "loss": 0.5448,
+      "step": 2559
+    },
+    {
+      "epoch": 0.15732546705998032,
+      "grad_norm": 13.875,
+      "learning_rate": 1.969786918517233e-05,
+      "loss": 0.3638,
+      "step": 2560
+    },
+    {
+      "epoch": 0.15738692232055063,
+      "grad_norm": 20.25,
+      "learning_rate": 1.969730121690698e-05,
+      "loss": 0.3758,
+      "step": 2561
+    },
+    {
+      "epoch": 0.15744837758112096,
+      "grad_norm": 19.125,
+      "learning_rate": 1.969673272349196e-05,
+      "loss": 0.4196,
+      "step": 2562
+    },
+    {
+      "epoch": 0.15750983284169126,
+      "grad_norm": 20.125,
+      "learning_rate": 1.969616370495806e-05,
+      "loss": 0.6692,
+      "step": 2563
+    },
+    {
+      "epoch": 0.15757128810226156,
+      "grad_norm": 20.625,
+      "learning_rate": 1.9695594161336102e-05,
+      "loss": 0.4147,
+      "step": 2564
+    },
+    {
+      "epoch": 0.15763274336283187,
+      "grad_norm": 10.5625,
+      "learning_rate": 1.969502409265692e-05,
+      "loss": 0.3686,
+      "step": 2565
+    },
+    {
+      "epoch": 0.15769419862340217,
+      "grad_norm": 5.6875,
+      "learning_rate": 1.9694453498951392e-05,
+      "loss": 0.4473,
+      "step": 2566
+    },
+    {
+      "epoch": 0.15775565388397247,
+      "grad_norm": 6.375,
+      "learning_rate": 1.9693882380250412e-05,
+      "loss": 0.2839,
+      "step": 2567
+    },
+    {
+      "epoch": 0.15781710914454278,
+      "grad_norm": 34.75,
+      "learning_rate": 1.9693310736584913e-05,
+      "loss": 0.3555,
+      "step": 2568
+    },
+    {
+      "epoch": 0.15787856440511308,
+      "grad_norm": 39.5,
+      "learning_rate": 1.9692738567985853e-05,
+      "loss": 0.383,
+      "step": 2569
+    },
+    {
+      "epoch": 0.15794001966568338,
+      "grad_norm": 11.125,
+      "learning_rate": 1.9692165874484216e-05,
+      "loss": 0.5434,
+      "step": 2570
+    },
+    {
+      "epoch": 0.1580014749262537,
+      "grad_norm": 20.75,
+      "learning_rate": 1.969159265611101e-05,
+      "loss": 0.3975,
+      "step": 2571
+    },
+    {
+      "epoch": 0.158062930186824,
+      "grad_norm": 33.5,
+      "learning_rate": 1.9691018912897285e-05,
+      "loss": 0.4194,
+      "step": 2572
+    },
+    {
+      "epoch": 0.1581243854473943,
+      "grad_norm": 20.0,
+      "learning_rate": 1.9690444644874112e-05,
+      "loss": 0.3679,
+      "step": 2573
+    },
+    {
+      "epoch": 0.1581858407079646,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.9689869852072584e-05,
+      "loss": 0.4648,
+      "step": 2574
+    },
+    {
+      "epoch": 0.1582472959685349,
+      "grad_norm": 16.75,
+      "learning_rate": 1.968929453452383e-05,
+      "loss": 0.3546,
+      "step": 2575
+    },
+    {
+      "epoch": 0.1583087512291052,
+      "grad_norm": 12.625,
+      "learning_rate": 1.9688718692259007e-05,
+      "loss": 0.6424,
+      "step": 2576
+    },
+    {
+      "epoch": 0.1583702064896755,
+      "grad_norm": 22.375,
+      "learning_rate": 1.9688142325309302e-05,
+      "loss": 0.6393,
+      "step": 2577
+    },
+    {
+      "epoch": 0.1584316617502458,
+      "grad_norm": 44.0,
+      "learning_rate": 1.9687565433705926e-05,
+      "loss": 0.7496,
+      "step": 2578
+    },
+    {
+      "epoch": 0.15849311701081611,
+      "grad_norm": 36.75,
+      "learning_rate": 1.9686988017480114e-05,
+      "loss": 0.5012,
+      "step": 2579
+    },
+    {
+      "epoch": 0.15855457227138642,
+      "grad_norm": 28.0,
+      "learning_rate": 1.968641007666315e-05,
+      "loss": 0.4225,
+      "step": 2580
+    },
+    {
+      "epoch": 0.15861602753195675,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.9685831611286312e-05,
+      "loss": 0.507,
+      "step": 2581
+    },
+    {
+      "epoch": 0.15867748279252705,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9685252621380942e-05,
+      "loss": 0.5538,
+      "step": 2582
+    },
+    {
+      "epoch": 0.15873893805309736,
+      "grad_norm": 31.625,
+      "learning_rate": 1.9684673106978392e-05,
+      "loss": 0.4698,
+      "step": 2583
+    },
+    {
+      "epoch": 0.15880039331366766,
+      "grad_norm": 23.75,
+      "learning_rate": 1.968409306811004e-05,
+      "loss": 0.4338,
+      "step": 2584
+    },
+    {
+      "epoch": 0.15886184857423796,
+      "grad_norm": 40.25,
+      "learning_rate": 1.9683512504807298e-05,
+      "loss": 0.5239,
+      "step": 2585
+    },
+    {
+      "epoch": 0.15892330383480827,
+      "grad_norm": 37.5,
+      "learning_rate": 1.968293141710161e-05,
+      "loss": 0.491,
+      "step": 2586
+    },
+    {
+      "epoch": 0.15898475909537857,
+      "grad_norm": 30.625,
+      "learning_rate": 1.9682349805024447e-05,
+      "loss": 0.4252,
+      "step": 2587
+    },
+    {
+      "epoch": 0.15904621435594887,
+      "grad_norm": 19.375,
+      "learning_rate": 1.9681767668607298e-05,
+      "loss": 0.4828,
+      "step": 2588
+    },
+    {
+      "epoch": 0.15910766961651918,
+      "grad_norm": 17.75,
+      "learning_rate": 1.968118500788169e-05,
+      "loss": 0.4736,
+      "step": 2589
+    },
+    {
+      "epoch": 0.15916912487708948,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.968060182287918e-05,
+      "loss": 0.4417,
+      "step": 2590
+    },
+    {
+      "epoch": 0.15923058013765978,
+      "grad_norm": 10.0,
+      "learning_rate": 1.9680018113631347e-05,
+      "loss": 0.4989,
+      "step": 2591
+    },
+    {
+      "epoch": 0.1592920353982301,
+      "grad_norm": 33.0,
+      "learning_rate": 1.9679433880169805e-05,
+      "loss": 0.3964,
+      "step": 2592
+    },
+    {
+      "epoch": 0.1593534906588004,
+      "grad_norm": 29.75,
+      "learning_rate": 1.967884912252619e-05,
+      "loss": 0.4144,
+      "step": 2593
+    },
+    {
+      "epoch": 0.1594149459193707,
+      "grad_norm": 5.53125,
+      "learning_rate": 1.9678263840732167e-05,
+      "loss": 0.6681,
+      "step": 2594
+    },
+    {
+      "epoch": 0.159476401179941,
+      "grad_norm": 16.25,
+      "learning_rate": 1.9677678034819433e-05,
+      "loss": 0.6497,
+      "step": 2595
+    },
+    {
+      "epoch": 0.1595378564405113,
+      "grad_norm": 19.375,
+      "learning_rate": 1.9677091704819714e-05,
+      "loss": 0.4177,
+      "step": 2596
+    },
+    {
+      "epoch": 0.1595993117010816,
+      "grad_norm": 13.5,
+      "learning_rate": 1.967650485076476e-05,
+      "loss": 0.3584,
+      "step": 2597
+    },
+    {
+      "epoch": 0.1596607669616519,
+      "grad_norm": 10.0,
+      "learning_rate": 1.9675917472686352e-05,
+      "loss": 0.3692,
+      "step": 2598
+    },
+    {
+      "epoch": 0.1597222222222222,
+      "grad_norm": 11.375,
+      "learning_rate": 1.96753295706163e-05,
+      "loss": 0.3856,
+      "step": 2599
+    },
+    {
+      "epoch": 0.1597836774827925,
+      "grad_norm": 22.0,
+      "learning_rate": 1.9674741144586442e-05,
+      "loss": 0.4771,
+      "step": 2600
+    },
+    {
+      "epoch": 0.15984513274336284,
+      "grad_norm": 25.375,
+      "learning_rate": 1.967415219462864e-05,
+      "loss": 0.5236,
+      "step": 2601
+    },
+    {
+      "epoch": 0.15990658800393315,
+      "grad_norm": 32.0,
+      "learning_rate": 1.9673562720774792e-05,
+      "loss": 0.402,
+      "step": 2602
+    },
+    {
+      "epoch": 0.15996804326450345,
+      "grad_norm": 32.75,
+      "learning_rate": 1.967297272305682e-05,
+      "loss": 0.4073,
+      "step": 2603
+    },
+    {
+      "epoch": 0.16002949852507375,
+      "grad_norm": 16.25,
+      "learning_rate": 1.9672382201506674e-05,
+      "loss": 0.4683,
+      "step": 2604
+    },
+    {
+      "epoch": 0.16009095378564406,
+      "grad_norm": 20.625,
+      "learning_rate": 1.967179115615633e-05,
+      "loss": 0.5189,
+      "step": 2605
+    },
+    {
+      "epoch": 0.16015240904621436,
+      "grad_norm": 16.25,
+      "learning_rate": 1.96711995870378e-05,
+      "loss": 0.3884,
+      "step": 2606
+    },
+    {
+      "epoch": 0.16021386430678466,
+      "grad_norm": 5.1875,
+      "learning_rate": 1.967060749418312e-05,
+      "loss": 0.4963,
+      "step": 2607
+    },
+    {
+      "epoch": 0.16027531956735497,
+      "grad_norm": 5.78125,
+      "learning_rate": 1.9670014877624353e-05,
+      "loss": 0.4457,
+      "step": 2608
+    },
+    {
+      "epoch": 0.16033677482792527,
+      "grad_norm": 23.625,
+      "learning_rate": 1.966942173739359e-05,
+      "loss": 0.4183,
+      "step": 2609
+    },
+    {
+      "epoch": 0.16039823008849557,
+      "grad_norm": 27.125,
+      "learning_rate": 1.9668828073522955e-05,
+      "loss": 0.4612,
+      "step": 2610
+    },
+    {
+      "epoch": 0.16045968534906588,
+      "grad_norm": 30.125,
+      "learning_rate": 1.9668233886044597e-05,
+      "loss": 0.5442,
+      "step": 2611
+    },
+    {
+      "epoch": 0.16052114060963618,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.9667639174990693e-05,
+      "loss": 0.3133,
+      "step": 2612
+    },
+    {
+      "epoch": 0.16058259587020648,
+      "grad_norm": 29.75,
+      "learning_rate": 1.9667043940393445e-05,
+      "loss": 0.5385,
+      "step": 2613
+    },
+    {
+      "epoch": 0.1606440511307768,
+      "grad_norm": 26.75,
+      "learning_rate": 1.9666448182285095e-05,
+      "loss": 0.4815,
+      "step": 2614
+    },
+    {
+      "epoch": 0.1607055063913471,
+      "grad_norm": 21.75,
+      "learning_rate": 1.9665851900697902e-05,
+      "loss": 0.4334,
+      "step": 2615
+    },
+    {
+      "epoch": 0.1607669616519174,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.9665255095664155e-05,
+      "loss": 0.3862,
+      "step": 2616
+    },
+    {
+      "epoch": 0.1608284169124877,
+      "grad_norm": 51.75,
+      "learning_rate": 1.9664657767216176e-05,
+      "loss": 0.6288,
+      "step": 2617
+    },
+    {
+      "epoch": 0.160889872173058,
+      "grad_norm": 25.5,
+      "learning_rate": 1.9664059915386316e-05,
+      "loss": 0.4779,
+      "step": 2618
+    },
+    {
+      "epoch": 0.1609513274336283,
+      "grad_norm": 25.125,
+      "learning_rate": 1.9663461540206947e-05,
+      "loss": 0.6359,
+      "step": 2619
+    },
+    {
+      "epoch": 0.16101278269419864,
+      "grad_norm": 20.0,
+      "learning_rate": 1.966286264171047e-05,
+      "loss": 0.4382,
+      "step": 2620
+    },
+    {
+      "epoch": 0.16107423795476894,
+      "grad_norm": 33.25,
+      "learning_rate": 1.966226321992933e-05,
+      "loss": 0.5504,
+      "step": 2621
+    },
+    {
+      "epoch": 0.16113569321533924,
+      "grad_norm": 22.75,
+      "learning_rate": 1.966166327489598e-05,
+      "loss": 0.5014,
+      "step": 2622
+    },
+    {
+      "epoch": 0.16119714847590955,
+      "grad_norm": 38.5,
+      "learning_rate": 1.9661062806642903e-05,
+      "loss": 0.5305,
+      "step": 2623
+    },
+    {
+      "epoch": 0.16125860373647985,
+      "grad_norm": 24.875,
+      "learning_rate": 1.9660461815202627e-05,
+      "loss": 0.582,
+      "step": 2624
+    },
+    {
+      "epoch": 0.16132005899705015,
+      "grad_norm": 26.375,
+      "learning_rate": 1.9659860300607695e-05,
+      "loss": 0.6754,
+      "step": 2625
+    },
+    {
+      "epoch": 0.16138151425762046,
+      "grad_norm": 37.25,
+      "learning_rate": 1.9659258262890683e-05,
+      "loss": 0.7695,
+      "step": 2626
+    },
+    {
+      "epoch": 0.16144296951819076,
+      "grad_norm": 19.25,
+      "learning_rate": 1.9658655702084192e-05,
+      "loss": 0.6172,
+      "step": 2627
+    },
+    {
+      "epoch": 0.16150442477876106,
+      "grad_norm": 25.75,
+      "learning_rate": 1.9658052618220857e-05,
+      "loss": 0.6578,
+      "step": 2628
+    },
+    {
+      "epoch": 0.16156588003933137,
+      "grad_norm": 21.875,
+      "learning_rate": 1.9657449011333328e-05,
+      "loss": 0.6463,
+      "step": 2629
+    },
+    {
+      "epoch": 0.16162733529990167,
+      "grad_norm": 29.125,
+      "learning_rate": 1.9656844881454304e-05,
+      "loss": 0.6031,
+      "step": 2630
+    },
+    {
+      "epoch": 0.16168879056047197,
+      "grad_norm": 9.375,
+      "learning_rate": 1.9656240228616496e-05,
+      "loss": 0.5417,
+      "step": 2631
+    },
+    {
+      "epoch": 0.16175024582104228,
+      "grad_norm": 18.125,
+      "learning_rate": 1.9655635052852648e-05,
+      "loss": 0.4604,
+      "step": 2632
+    },
+    {
+      "epoch": 0.16181170108161258,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.9655029354195534e-05,
+      "loss": 0.4965,
+      "step": 2633
+    },
+    {
+      "epoch": 0.16187315634218288,
+      "grad_norm": 20.625,
+      "learning_rate": 1.9654423132677956e-05,
+      "loss": 0.4449,
+      "step": 2634
+    },
+    {
+      "epoch": 0.1619346116027532,
+      "grad_norm": 20.875,
+      "learning_rate": 1.965381638833274e-05,
+      "loss": 0.4853,
+      "step": 2635
+    },
+    {
+      "epoch": 0.1619960668633235,
+      "grad_norm": 13.0,
+      "learning_rate": 1.9653209121192747e-05,
+      "loss": 0.4964,
+      "step": 2636
+    },
+    {
+      "epoch": 0.1620575221238938,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.9652601331290864e-05,
+      "loss": 0.4388,
+      "step": 2637
+    },
+    {
+      "epoch": 0.1621189773844641,
+      "grad_norm": 17.75,
+      "learning_rate": 1.9651993018660002e-05,
+      "loss": 0.3994,
+      "step": 2638
+    },
+    {
+      "epoch": 0.16218043264503443,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.9651384183333105e-05,
+      "loss": 0.392,
+      "step": 2639
+    },
+    {
+      "epoch": 0.16224188790560473,
+      "grad_norm": 15.875,
+      "learning_rate": 1.965077482534314e-05,
+      "loss": 0.3243,
+      "step": 2640
+    },
+    {
+      "epoch": 0.16230334316617503,
+      "grad_norm": 11.125,
+      "learning_rate": 1.9650164944723116e-05,
+      "loss": 0.2711,
+      "step": 2641
+    },
+    {
+      "epoch": 0.16236479842674534,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.9649554541506056e-05,
+      "loss": 0.3926,
+      "step": 2642
+    },
+    {
+      "epoch": 0.16242625368731564,
+      "grad_norm": 6.125,
+      "learning_rate": 1.9648943615725012e-05,
+      "loss": 0.407,
+      "step": 2643
+    },
+    {
+      "epoch": 0.16248770894788594,
+      "grad_norm": 21.75,
+      "learning_rate": 1.9648332167413067e-05,
+      "loss": 0.3339,
+      "step": 2644
+    },
+    {
+      "epoch": 0.16254916420845625,
+      "grad_norm": 37.25,
+      "learning_rate": 1.9647720196603343e-05,
+      "loss": 0.3392,
+      "step": 2645
+    },
+    {
+      "epoch": 0.16261061946902655,
+      "grad_norm": 17.5,
+      "learning_rate": 1.964710770332898e-05,
+      "loss": 0.1862,
+      "step": 2646
+    },
+    {
+      "epoch": 0.16267207472959685,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9646494687623135e-05,
+      "loss": 0.2309,
+      "step": 2647
+    },
+    {
+      "epoch": 0.16273352999016716,
+      "grad_norm": 15.0,
+      "learning_rate": 1.9645881149519014e-05,
+      "loss": 0.2427,
+      "step": 2648
+    },
+    {
+      "epoch": 0.16279498525073746,
+      "grad_norm": 17.5,
+      "learning_rate": 1.9645267089049844e-05,
+      "loss": 0.7125,
+      "step": 2649
+    },
+    {
+      "epoch": 0.16285644051130777,
+      "grad_norm": 24.75,
+      "learning_rate": 1.9644652506248872e-05,
+      "loss": 0.4653,
+      "step": 2650
+    },
+    {
+      "epoch": 0.16291789577187807,
+      "grad_norm": 15.875,
+      "learning_rate": 1.964403740114939e-05,
+      "loss": 0.4796,
+      "step": 2651
+    },
+    {
+      "epoch": 0.16297935103244837,
+      "grad_norm": 15.25,
+      "learning_rate": 1.9643421773784702e-05,
+      "loss": 0.4809,
+      "step": 2652
+    },
+    {
+      "epoch": 0.16304080629301868,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.964280562418815e-05,
+      "loss": 0.6492,
+      "step": 2653
+    },
+    {
+      "epoch": 0.16310226155358898,
+      "grad_norm": 13.625,
+      "learning_rate": 1.9642188952393096e-05,
+      "loss": 0.4483,
+      "step": 2654
+    },
+    {
+      "epoch": 0.16316371681415928,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.964157175843294e-05,
+      "loss": 0.4374,
+      "step": 2655
+    },
+    {
+      "epoch": 0.16322517207472959,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.96409540423411e-05,
+      "loss": 0.3426,
+      "step": 2656
+    },
+    {
+      "epoch": 0.1632866273352999,
+      "grad_norm": 48.0,
+      "learning_rate": 1.964033580415104e-05,
+      "loss": 0.4748,
+      "step": 2657
+    },
+    {
+      "epoch": 0.16334808259587022,
+      "grad_norm": 8.625,
+      "learning_rate": 1.9639717043896226e-05,
+      "loss": 0.4228,
+      "step": 2658
+    },
+    {
+      "epoch": 0.16340953785644052,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.9639097761610174e-05,
+      "loss": 0.6938,
+      "step": 2659
+    },
+    {
+      "epoch": 0.16347099311701083,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.9638477957326424e-05,
+      "loss": 0.5007,
+      "step": 2660
+    },
+    {
+      "epoch": 0.16353244837758113,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.9637857631078532e-05,
+      "loss": 0.3132,
+      "step": 2661
+    },
+    {
+      "epoch": 0.16359390363815143,
+      "grad_norm": 39.5,
+      "learning_rate": 1.96372367829001e-05,
+      "loss": 0.5564,
+      "step": 2662
+    },
+    {
+      "epoch": 0.16365535889872174,
+      "grad_norm": 24.75,
+      "learning_rate": 1.963661541282474e-05,
+      "loss": 0.5625,
+      "step": 2663
+    },
+    {
+      "epoch": 0.16371681415929204,
+      "grad_norm": 18.0,
+      "learning_rate": 1.9635993520886113e-05,
+      "loss": 0.4031,
+      "step": 2664
+    },
+    {
+      "epoch": 0.16377826941986234,
+      "grad_norm": 38.0,
+      "learning_rate": 1.963537110711789e-05,
+      "loss": 0.505,
+      "step": 2665
+    },
+    {
+      "epoch": 0.16383972468043265,
+      "grad_norm": 50.5,
+      "learning_rate": 1.9634748171553775e-05,
+      "loss": 0.6045,
+      "step": 2666
+    },
+    {
+      "epoch": 0.16390117994100295,
+      "grad_norm": 53.0,
+      "learning_rate": 1.9634124714227507e-05,
+      "loss": 0.7988,
+      "step": 2667
+    },
+    {
+      "epoch": 0.16396263520157325,
+      "grad_norm": 27.375,
+      "learning_rate": 1.963350073517285e-05,
+      "loss": 0.6065,
+      "step": 2668
+    },
+    {
+      "epoch": 0.16402409046214356,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.9632876234423592e-05,
+      "loss": 0.3484,
+      "step": 2669
+    },
+    {
+      "epoch": 0.16408554572271386,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.9632251212013554e-05,
+      "loss": 0.3596,
+      "step": 2670
+    },
+    {
+      "epoch": 0.16414700098328416,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.9631625667976584e-05,
+      "loss": 0.3507,
+      "step": 2671
+    },
+    {
+      "epoch": 0.16420845624385447,
+      "grad_norm": 9.25,
+      "learning_rate": 1.9630999602346558e-05,
+      "loss": 0.3919,
+      "step": 2672
+    },
+    {
+      "epoch": 0.16426991150442477,
+      "grad_norm": 34.25,
+      "learning_rate": 1.9630373015157374e-05,
+      "loss": 0.6056,
+      "step": 2673
+    },
+    {
+      "epoch": 0.16433136676499507,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.9629745906442973e-05,
+      "loss": 0.4807,
+      "step": 2674
+    },
+    {
+      "epoch": 0.16439282202556538,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.962911827623731e-05,
+      "loss": 0.4916,
+      "step": 2675
+    },
+    {
+      "epoch": 0.16445427728613568,
+      "grad_norm": 26.5,
+      "learning_rate": 1.962849012457438e-05,
+      "loss": 0.5823,
+      "step": 2676
+    },
+    {
+      "epoch": 0.164515732546706,
+      "grad_norm": 40.0,
+      "learning_rate": 1.962786145148819e-05,
+      "loss": 0.5755,
+      "step": 2677
+    },
+    {
+      "epoch": 0.16457718780727632,
+      "grad_norm": 25.5,
+      "learning_rate": 1.9627232257012793e-05,
+      "loss": 0.7444,
+      "step": 2678
+    },
+    {
+      "epoch": 0.16463864306784662,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9626602541182264e-05,
+      "loss": 0.4651,
+      "step": 2679
+    },
+    {
+      "epoch": 0.16470009832841692,
+      "grad_norm": 41.75,
+      "learning_rate": 1.9625972304030697e-05,
+      "loss": 0.6109,
+      "step": 2680
+    },
+    {
+      "epoch": 0.16476155358898723,
+      "grad_norm": 11.75,
+      "learning_rate": 1.9625341545592226e-05,
+      "loss": 0.5222,
+      "step": 2681
+    },
+    {
+      "epoch": 0.16482300884955753,
+      "grad_norm": 26.625,
+      "learning_rate": 1.9624710265901012e-05,
+      "loss": 0.442,
+      "step": 2682
+    },
+    {
+      "epoch": 0.16488446411012783,
+      "grad_norm": 14.625,
+      "learning_rate": 1.962407846499124e-05,
+      "loss": 0.44,
+      "step": 2683
+    },
+    {
+      "epoch": 0.16494591937069814,
+      "grad_norm": 25.125,
+      "learning_rate": 1.9623446142897123e-05,
+      "loss": 0.5404,
+      "step": 2684
+    },
+    {
+      "epoch": 0.16500737463126844,
+      "grad_norm": 42.0,
+      "learning_rate": 1.9622813299652904e-05,
+      "loss": 0.554,
+      "step": 2685
+    },
+    {
+      "epoch": 0.16506882989183874,
+      "grad_norm": 6.0,
+      "learning_rate": 1.9622179935292855e-05,
+      "loss": 0.2637,
+      "step": 2686
+    },
+    {
+      "epoch": 0.16513028515240905,
+      "grad_norm": 17.5,
+      "learning_rate": 1.9621546049851275e-05,
+      "loss": 0.478,
+      "step": 2687
+    },
+    {
+      "epoch": 0.16519174041297935,
+      "grad_norm": 5.34375,
+      "learning_rate": 1.9620911643362493e-05,
+      "loss": 0.2741,
+      "step": 2688
+    },
+    {
+      "epoch": 0.16525319567354965,
+      "grad_norm": 46.0,
+      "learning_rate": 1.962027671586086e-05,
+      "loss": 0.729,
+      "step": 2689
+    },
+    {
+      "epoch": 0.16531465093411996,
+      "grad_norm": 22.125,
+      "learning_rate": 1.9619641267380766e-05,
+      "loss": 0.6168,
+      "step": 2690
+    },
+    {
+      "epoch": 0.16537610619469026,
+      "grad_norm": 24.625,
+      "learning_rate": 1.9619005297956623e-05,
+      "loss": 0.7338,
+      "step": 2691
+    },
+    {
+      "epoch": 0.16543756145526056,
+      "grad_norm": 19.25,
+      "learning_rate": 1.9618368807622863e-05,
+      "loss": 0.4829,
+      "step": 2692
+    },
+    {
+      "epoch": 0.16549901671583087,
+      "grad_norm": 33.5,
+      "learning_rate": 1.9617731796413962e-05,
+      "loss": 0.5661,
+      "step": 2693
+    },
+    {
+      "epoch": 0.16556047197640117,
+      "grad_norm": 17.375,
+      "learning_rate": 1.9617094264364417e-05,
+      "loss": 0.5333,
+      "step": 2694
+    },
+    {
+      "epoch": 0.16562192723697147,
+      "grad_norm": 20.375,
+      "learning_rate": 1.9616456211508756e-05,
+      "loss": 0.6185,
+      "step": 2695
+    },
+    {
+      "epoch": 0.16568338249754178,
+      "grad_norm": 20.25,
+      "learning_rate": 1.961581763788152e-05,
+      "loss": 0.5673,
+      "step": 2696
+    },
+    {
+      "epoch": 0.1657448377581121,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.9615178543517303e-05,
+      "loss": 0.4576,
+      "step": 2697
+    },
+    {
+      "epoch": 0.1658062930186824,
+      "grad_norm": 16.375,
+      "learning_rate": 1.961453892845071e-05,
+      "loss": 0.4411,
+      "step": 2698
+    },
+    {
+      "epoch": 0.1658677482792527,
+      "grad_norm": 14.625,
+      "learning_rate": 1.9613898792716377e-05,
+      "loss": 0.4168,
+      "step": 2699
+    },
+    {
+      "epoch": 0.16592920353982302,
+      "grad_norm": 15.0,
+      "learning_rate": 1.9613258136348973e-05,
+      "loss": 0.4792,
+      "step": 2700
+    },
+    {
+      "epoch": 0.16599065880039332,
+      "grad_norm": 10.875,
+      "learning_rate": 1.961261695938319e-05,
+      "loss": 0.4795,
+      "step": 2701
+    },
+    {
+      "epoch": 0.16605211406096362,
+      "grad_norm": 21.125,
+      "learning_rate": 1.9611975261853754e-05,
+      "loss": 0.5371,
+      "step": 2702
+    },
+    {
+      "epoch": 0.16611356932153393,
+      "grad_norm": 13.0,
+      "learning_rate": 1.961133304379541e-05,
+      "loss": 0.4912,
+      "step": 2703
+    },
+    {
+      "epoch": 0.16617502458210423,
+      "grad_norm": 17.75,
+      "learning_rate": 1.961069030524294e-05,
+      "loss": 0.5157,
+      "step": 2704
+    },
+    {
+      "epoch": 0.16623647984267453,
+      "grad_norm": 24.875,
+      "learning_rate": 1.9610047046231152e-05,
+      "loss": 0.5773,
+      "step": 2705
+    },
+    {
+      "epoch": 0.16629793510324484,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.960940326679488e-05,
+      "loss": 0.4499,
+      "step": 2706
+    },
+    {
+      "epoch": 0.16635939036381514,
+      "grad_norm": 7.125,
+      "learning_rate": 1.9608758966968987e-05,
+      "loss": 0.408,
+      "step": 2707
+    },
+    {
+      "epoch": 0.16642084562438544,
+      "grad_norm": 108.0,
+      "learning_rate": 1.9608114146788365e-05,
+      "loss": 0.5437,
+      "step": 2708
+    },
+    {
+      "epoch": 0.16648230088495575,
+      "grad_norm": 21.375,
+      "learning_rate": 1.9607468806287936e-05,
+      "loss": 0.5083,
+      "step": 2709
+    },
+    {
+      "epoch": 0.16654375614552605,
+      "grad_norm": 21.25,
+      "learning_rate": 1.9606822945502642e-05,
+      "loss": 0.4602,
+      "step": 2710
+    },
+    {
+      "epoch": 0.16660521140609635,
+      "grad_norm": 15.25,
+      "learning_rate": 1.9606176564467465e-05,
+      "loss": 0.3758,
+      "step": 2711
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.9605529663217406e-05,
+      "loss": 0.408,
+      "step": 2712
+    },
+    {
+      "epoch": 0.16672812192723696,
+      "grad_norm": 22.875,
+      "learning_rate": 1.96048822417875e-05,
+      "loss": 0.4985,
+      "step": 2713
+    },
+    {
+      "epoch": 0.16678957718780726,
+      "grad_norm": 22.375,
+      "learning_rate": 1.9604234300212804e-05,
+      "loss": 0.5584,
+      "step": 2714
+    },
+    {
+      "epoch": 0.16685103244837757,
+      "grad_norm": 24.0,
+      "learning_rate": 1.960358583852841e-05,
+      "loss": 0.5244,
+      "step": 2715
+    },
+    {
+      "epoch": 0.1669124877089479,
+      "grad_norm": 20.625,
+      "learning_rate": 1.9602936856769432e-05,
+      "loss": 0.3737,
+      "step": 2716
+    },
+    {
+      "epoch": 0.1669739429695182,
+      "grad_norm": 22.5,
+      "learning_rate": 1.9602287354971018e-05,
+      "loss": 0.3998,
+      "step": 2717
+    },
+    {
+      "epoch": 0.1670353982300885,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.960163733316834e-05,
+      "loss": 0.3581,
+      "step": 2718
+    },
+    {
+      "epoch": 0.1670968534906588,
+      "grad_norm": 26.0,
+      "learning_rate": 1.96009867913966e-05,
+      "loss": 0.3753,
+      "step": 2719
+    },
+    {
+      "epoch": 0.1671583087512291,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.9600335729691025e-05,
+      "loss": 0.2619,
+      "step": 2720
+    },
+    {
+      "epoch": 0.16721976401179942,
+      "grad_norm": 34.0,
+      "learning_rate": 1.9599684148086876e-05,
+      "loss": 0.6084,
+      "step": 2721
+    },
+    {
+      "epoch": 0.16728121927236972,
+      "grad_norm": 19.75,
+      "learning_rate": 1.9599032046619437e-05,
+      "loss": 0.5304,
+      "step": 2722
+    },
+    {
+      "epoch": 0.16734267453294002,
+      "grad_norm": 26.125,
+      "learning_rate": 1.9598379425324023e-05,
+      "loss": 0.7238,
+      "step": 2723
+    },
+    {
+      "epoch": 0.16740412979351033,
+      "grad_norm": 30.75,
+      "learning_rate": 1.9597726284235978e-05,
+      "loss": 0.4796,
+      "step": 2724
+    },
+    {
+      "epoch": 0.16746558505408063,
+      "grad_norm": 34.5,
+      "learning_rate": 1.9597072623390668e-05,
+      "loss": 0.4942,
+      "step": 2725
+    },
+    {
+      "epoch": 0.16752704031465093,
+      "grad_norm": 46.75,
+      "learning_rate": 1.9596418442823495e-05,
+      "loss": 0.5364,
+      "step": 2726
+    },
+    {
+      "epoch": 0.16758849557522124,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9595763742569883e-05,
+      "loss": 0.4821,
+      "step": 2727
+    },
+    {
+      "epoch": 0.16764995083579154,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.959510852266529e-05,
+      "loss": 0.4277,
+      "step": 2728
+    },
+    {
+      "epoch": 0.16771140609636184,
+      "grad_norm": 22.25,
+      "learning_rate": 1.959445278314519e-05,
+      "loss": 0.5351,
+      "step": 2729
+    },
+    {
+      "epoch": 0.16777286135693215,
+      "grad_norm": 26.0,
+      "learning_rate": 1.959379652404511e-05,
+      "loss": 0.4489,
+      "step": 2730
+    },
+    {
+      "epoch": 0.16783431661750245,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.9593139745400575e-05,
+      "loss": 0.4203,
+      "step": 2731
+    },
+    {
+      "epoch": 0.16789577187807275,
+      "grad_norm": 7.5,
+      "learning_rate": 1.9592482447247162e-05,
+      "loss": 0.4846,
+      "step": 2732
+    },
+    {
+      "epoch": 0.16795722713864306,
+      "grad_norm": 22.75,
+      "learning_rate": 1.959182462962046e-05,
+      "loss": 0.362,
+      "step": 2733
+    },
+    {
+      "epoch": 0.16801868239921336,
+      "grad_norm": 40.0,
+      "learning_rate": 1.9591166292556093e-05,
+      "loss": 0.5135,
+      "step": 2734
+    },
+    {
+      "epoch": 0.1680801376597837,
+      "grad_norm": 5.59375,
+      "learning_rate": 1.9590507436089713e-05,
+      "loss": 0.3596,
+      "step": 2735
+    },
+    {
+      "epoch": 0.168141592920354,
+      "grad_norm": 14.5,
+      "learning_rate": 1.9589848060257007e-05,
+      "loss": 0.3923,
+      "step": 2736
+    },
+    {
+      "epoch": 0.1682030481809243,
+      "grad_norm": 38.25,
+      "learning_rate": 1.958918816509367e-05,
+      "loss": 0.747,
+      "step": 2737
+    },
+    {
+      "epoch": 0.1682645034414946,
+      "grad_norm": 11.625,
+      "learning_rate": 1.958852775063545e-05,
+      "loss": 0.4815,
+      "step": 2738
+    },
+    {
+      "epoch": 0.1683259587020649,
+      "grad_norm": 18.625,
+      "learning_rate": 1.9587866816918107e-05,
+      "loss": 0.4905,
+      "step": 2739
+    },
+    {
+      "epoch": 0.1683874139626352,
+      "grad_norm": 38.0,
+      "learning_rate": 1.9587205363977428e-05,
+      "loss": 0.8357,
+      "step": 2740
+    },
+    {
+      "epoch": 0.1684488692232055,
+      "grad_norm": 29.5,
+      "learning_rate": 1.9586543391849243e-05,
+      "loss": 0.8035,
+      "step": 2741
+    },
+    {
+      "epoch": 0.16851032448377581,
+      "grad_norm": 36.5,
+      "learning_rate": 1.9585880900569397e-05,
+      "loss": 0.5401,
+      "step": 2742
+    },
+    {
+      "epoch": 0.16857177974434612,
+      "grad_norm": 6.4375,
+      "learning_rate": 1.958521789017376e-05,
+      "loss": 0.4603,
+      "step": 2743
+    },
+    {
+      "epoch": 0.16863323500491642,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.9584554360698247e-05,
+      "loss": 0.3537,
+      "step": 2744
+    },
+    {
+      "epoch": 0.16869469026548672,
+      "grad_norm": 24.125,
+      "learning_rate": 1.958389031217878e-05,
+      "loss": 0.5323,
+      "step": 2745
+    },
+    {
+      "epoch": 0.16875614552605703,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.9583225744651334e-05,
+      "loss": 0.4884,
+      "step": 2746
+    },
+    {
+      "epoch": 0.16881760078662733,
+      "grad_norm": 15.875,
+      "learning_rate": 1.9582560658151886e-05,
+      "loss": 0.4831,
+      "step": 2747
+    },
+    {
+      "epoch": 0.16887905604719763,
+      "grad_norm": 10.25,
+      "learning_rate": 1.958189505271646e-05,
+      "loss": 0.5194,
+      "step": 2748
+    },
+    {
+      "epoch": 0.16894051130776794,
+      "grad_norm": 22.5,
+      "learning_rate": 1.95812289283811e-05,
+      "loss": 0.5502,
+      "step": 2749
+    },
+    {
+      "epoch": 0.16900196656833824,
+      "grad_norm": 17.375,
+      "learning_rate": 1.9580562285181875e-05,
+      "loss": 0.5657,
+      "step": 2750
+    },
+    {
+      "epoch": 0.16906342182890854,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.957989512315489e-05,
+      "loss": 0.4717,
+      "step": 2751
+    },
+    {
+      "epoch": 0.16912487708947885,
+      "grad_norm": 21.75,
+      "learning_rate": 1.9579227442336276e-05,
+      "loss": 0.3677,
+      "step": 2752
+    },
+    {
+      "epoch": 0.16918633235004915,
+      "grad_norm": 11.25,
+      "learning_rate": 1.957855924276219e-05,
+      "loss": 0.2888,
+      "step": 2753
+    },
+    {
+      "epoch": 0.16924778761061948,
+      "grad_norm": 22.0,
+      "learning_rate": 1.9577890524468815e-05,
+      "loss": 0.3348,
+      "step": 2754
+    },
+    {
+      "epoch": 0.16930924287118979,
+      "grad_norm": 19.25,
+      "learning_rate": 1.9577221287492368e-05,
+      "loss": 0.4311,
+      "step": 2755
+    },
+    {
+      "epoch": 0.1693706981317601,
+      "grad_norm": 8.625,
+      "learning_rate": 1.9576551531869092e-05,
+      "loss": 0.5632,
+      "step": 2756
+    },
+    {
+      "epoch": 0.1694321533923304,
+      "grad_norm": 29.25,
+      "learning_rate": 1.9575881257635254e-05,
+      "loss": 0.4393,
+      "step": 2757
+    },
+    {
+      "epoch": 0.1694936086529007,
+      "grad_norm": 4.71875,
+      "learning_rate": 1.957521046482715e-05,
+      "loss": 0.2928,
+      "step": 2758
+    },
+    {
+      "epoch": 0.169555063913471,
+      "grad_norm": 29.375,
+      "learning_rate": 1.9574539153481113e-05,
+      "loss": 0.4267,
+      "step": 2759
+    },
+    {
+      "epoch": 0.1696165191740413,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.9573867323633493e-05,
+      "loss": 0.3317,
+      "step": 2760
+    },
+    {
+      "epoch": 0.1696779744346116,
+      "grad_norm": 18.875,
+      "learning_rate": 1.9573194975320672e-05,
+      "loss": 0.404,
+      "step": 2761
+    },
+    {
+      "epoch": 0.1697394296951819,
+      "grad_norm": 33.0,
+      "learning_rate": 1.9572522108579063e-05,
+      "loss": 0.5028,
+      "step": 2762
+    },
+    {
+      "epoch": 0.1698008849557522,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.9571848723445107e-05,
+      "loss": 0.5406,
+      "step": 2763
+    },
+    {
+      "epoch": 0.16986234021632252,
+      "grad_norm": 46.25,
+      "learning_rate": 1.9571174819955264e-05,
+      "loss": 0.3832,
+      "step": 2764
+    },
+    {
+      "epoch": 0.16992379547689282,
+      "grad_norm": 26.875,
+      "learning_rate": 1.957050039814603e-05,
+      "loss": 0.4258,
+      "step": 2765
+    },
+    {
+      "epoch": 0.16998525073746312,
+      "grad_norm": 21.875,
+      "learning_rate": 1.956982545805393e-05,
+      "loss": 0.4427,
+      "step": 2766
+    },
+    {
+      "epoch": 0.17004670599803343,
+      "grad_norm": 36.75,
+      "learning_rate": 1.9569149999715514e-05,
+      "loss": 0.4482,
+      "step": 2767
+    },
+    {
+      "epoch": 0.17010816125860373,
+      "grad_norm": 18.25,
+      "learning_rate": 1.9568474023167364e-05,
+      "loss": 0.3737,
+      "step": 2768
+    },
+    {
+      "epoch": 0.17016961651917403,
+      "grad_norm": 21.75,
+      "learning_rate": 1.9567797528446083e-05,
+      "loss": 0.5789,
+      "step": 2769
+    },
+    {
+      "epoch": 0.17023107177974434,
+      "grad_norm": 23.25,
+      "learning_rate": 1.9567120515588307e-05,
+      "loss": 0.5185,
+      "step": 2770
+    },
+    {
+      "epoch": 0.17029252704031464,
+      "grad_norm": 20.0,
+      "learning_rate": 1.95664429846307e-05,
+      "loss": 0.6249,
+      "step": 2771
+    },
+    {
+      "epoch": 0.17035398230088494,
+      "grad_norm": 32.75,
+      "learning_rate": 1.956576493560995e-05,
+      "loss": 0.6219,
+      "step": 2772
+    },
+    {
+      "epoch": 0.17041543756145527,
+      "grad_norm": 46.5,
+      "learning_rate": 1.956508636856278e-05,
+      "loss": 0.4436,
+      "step": 2773
+    },
+    {
+      "epoch": 0.17047689282202558,
+      "grad_norm": 44.25,
+      "learning_rate": 1.9564407283525937e-05,
+      "loss": 0.473,
+      "step": 2774
+    },
+    {
+      "epoch": 0.17053834808259588,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.9563727680536195e-05,
+      "loss": 0.4315,
+      "step": 2775
+    },
+    {
+      "epoch": 0.17059980334316618,
+      "grad_norm": 39.25,
+      "learning_rate": 1.9563047559630356e-05,
+      "loss": 0.5703,
+      "step": 2776
+    },
+    {
+      "epoch": 0.1706612586037365,
+      "grad_norm": 42.5,
+      "learning_rate": 1.9562366920845256e-05,
+      "loss": 0.483,
+      "step": 2777
+    },
+    {
+      "epoch": 0.1707227138643068,
+      "grad_norm": 35.75,
+      "learning_rate": 1.956168576421775e-05,
+      "loss": 0.5965,
+      "step": 2778
+    },
+    {
+      "epoch": 0.1707841691248771,
+      "grad_norm": 22.375,
+      "learning_rate": 1.9561004089784726e-05,
+      "loss": 0.5092,
+      "step": 2779
+    },
+    {
+      "epoch": 0.1708456243854474,
+      "grad_norm": 28.25,
+      "learning_rate": 1.95603218975831e-05,
+      "loss": 0.4146,
+      "step": 2780
+    },
+    {
+      "epoch": 0.1709070796460177,
+      "grad_norm": 24.125,
+      "learning_rate": 1.9559639187649817e-05,
+      "loss": 0.4344,
+      "step": 2781
+    },
+    {
+      "epoch": 0.170968534906588,
+      "grad_norm": 31.25,
+      "learning_rate": 1.9558955960021847e-05,
+      "loss": 0.4477,
+      "step": 2782
+    },
+    {
+      "epoch": 0.1710299901671583,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.9558272214736193e-05,
+      "loss": 0.3064,
+      "step": 2783
+    },
+    {
+      "epoch": 0.1710914454277286,
+      "grad_norm": 44.75,
+      "learning_rate": 1.955758795182988e-05,
+      "loss": 0.4247,
+      "step": 2784
+    },
+    {
+      "epoch": 0.17115290068829891,
+      "grad_norm": 76.5,
+      "learning_rate": 1.9556903171339963e-05,
+      "loss": 0.5167,
+      "step": 2785
+    },
+    {
+      "epoch": 0.17121435594886922,
+      "grad_norm": 22.125,
+      "learning_rate": 1.9556217873303526e-05,
+      "loss": 0.7476,
+      "step": 2786
+    },
+    {
+      "epoch": 0.17127581120943952,
+      "grad_norm": 45.75,
+      "learning_rate": 1.955553205775768e-05,
+      "loss": 0.7909,
+      "step": 2787
+    },
+    {
+      "epoch": 0.17133726647000982,
+      "grad_norm": 38.75,
+      "learning_rate": 1.9554845724739565e-05,
+      "loss": 0.6302,
+      "step": 2788
+    },
+    {
+      "epoch": 0.17139872173058013,
+      "grad_norm": 38.5,
+      "learning_rate": 1.9554158874286353e-05,
+      "loss": 0.7298,
+      "step": 2789
+    },
+    {
+      "epoch": 0.17146017699115043,
+      "grad_norm": 23.25,
+      "learning_rate": 1.9553471506435235e-05,
+      "loss": 0.5963,
+      "step": 2790
+    },
+    {
+      "epoch": 0.17152163225172073,
+      "grad_norm": 17.75,
+      "learning_rate": 1.9552783621223437e-05,
+      "loss": 0.5503,
+      "step": 2791
+    },
+    {
+      "epoch": 0.17158308751229107,
+      "grad_norm": 42.25,
+      "learning_rate": 1.955209521868821e-05,
+      "loss": 0.5867,
+      "step": 2792
+    },
+    {
+      "epoch": 0.17164454277286137,
+      "grad_norm": 32.0,
+      "learning_rate": 1.9551406298866838e-05,
+      "loss": 0.7864,
+      "step": 2793
+    },
+    {
+      "epoch": 0.17170599803343167,
+      "grad_norm": 41.0,
+      "learning_rate": 1.9550716861796623e-05,
+      "loss": 0.5497,
+      "step": 2794
+    },
+    {
+      "epoch": 0.17176745329400198,
+      "grad_norm": 24.375,
+      "learning_rate": 1.9550026907514903e-05,
+      "loss": 0.5008,
+      "step": 2795
+    },
+    {
+      "epoch": 0.17182890855457228,
+      "grad_norm": 30.5,
+      "learning_rate": 1.954933643605904e-05,
+      "loss": 0.4837,
+      "step": 2796
+    },
+    {
+      "epoch": 0.17189036381514258,
+      "grad_norm": 32.0,
+      "learning_rate": 1.9548645447466433e-05,
+      "loss": 0.6915,
+      "step": 2797
+    },
+    {
+      "epoch": 0.1719518190757129,
+      "grad_norm": 22.875,
+      "learning_rate": 1.9547953941774495e-05,
+      "loss": 0.4997,
+      "step": 2798
+    },
+    {
+      "epoch": 0.1720132743362832,
+      "grad_norm": 48.0,
+      "learning_rate": 1.9547261919020673e-05,
+      "loss": 0.7398,
+      "step": 2799
+    },
+    {
+      "epoch": 0.1720747295968535,
+      "grad_norm": 18.875,
+      "learning_rate": 1.9546569379242446e-05,
+      "loss": 0.464,
+      "step": 2800
+    },
+    {
+      "epoch": 0.1721361848574238,
+      "grad_norm": 30.875,
+      "learning_rate": 1.954587632247732e-05,
+      "loss": 0.5287,
+      "step": 2801
+    },
+    {
+      "epoch": 0.1721976401179941,
+      "grad_norm": 28.875,
+      "learning_rate": 1.9545182748762825e-05,
+      "loss": 0.6245,
+      "step": 2802
+    },
+    {
+      "epoch": 0.1722590953785644,
+      "grad_norm": 19.75,
+      "learning_rate": 1.9544488658136522e-05,
+      "loss": 0.6787,
+      "step": 2803
+    },
+    {
+      "epoch": 0.1723205506391347,
+      "grad_norm": 22.0,
+      "learning_rate": 1.9543794050635997e-05,
+      "loss": 0.4474,
+      "step": 2804
+    },
+    {
+      "epoch": 0.172382005899705,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.9543098926298864e-05,
+      "loss": 0.435,
+      "step": 2805
+    },
+    {
+      "epoch": 0.1724434611602753,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.954240328516277e-05,
+      "loss": 0.4557,
+      "step": 2806
+    },
+    {
+      "epoch": 0.17250491642084562,
+      "grad_norm": 13.75,
+      "learning_rate": 1.9541707127265388e-05,
+      "loss": 0.4515,
+      "step": 2807
+    },
+    {
+      "epoch": 0.17256637168141592,
+      "grad_norm": 25.75,
+      "learning_rate": 1.9541010452644414e-05,
+      "loss": 0.4023,
+      "step": 2808
+    },
+    {
+      "epoch": 0.17262782694198622,
+      "grad_norm": 18.375,
+      "learning_rate": 1.954031326133758e-05,
+      "loss": 0.5423,
+      "step": 2809
+    },
+    {
+      "epoch": 0.17268928220255653,
+      "grad_norm": 14.625,
+      "learning_rate": 1.953961555338264e-05,
+      "loss": 0.5389,
+      "step": 2810
+    },
+    {
+      "epoch": 0.17275073746312683,
+      "grad_norm": 20.875,
+      "learning_rate": 1.9538917328817377e-05,
+      "loss": 0.4475,
+      "step": 2811
+    },
+    {
+      "epoch": 0.17281219272369716,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.9538218587679605e-05,
+      "loss": 0.4602,
+      "step": 2812
+    },
+    {
+      "epoch": 0.17287364798426746,
+      "grad_norm": 16.375,
+      "learning_rate": 1.9537519330007163e-05,
+      "loss": 0.4164,
+      "step": 2813
+    },
+    {
+      "epoch": 0.17293510324483777,
+      "grad_norm": 19.0,
+      "learning_rate": 1.9536819555837915e-05,
+      "loss": 0.4013,
+      "step": 2814
+    },
+    {
+      "epoch": 0.17299655850540807,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9536119265209763e-05,
+      "loss": 0.4106,
+      "step": 2815
+    },
+    {
+      "epoch": 0.17305801376597837,
+      "grad_norm": 18.75,
+      "learning_rate": 1.9535418458160625e-05,
+      "loss": 0.6183,
+      "step": 2816
+    },
+    {
+      "epoch": 0.17311946902654868,
+      "grad_norm": 33.5,
+      "learning_rate": 1.9534717134728455e-05,
+      "loss": 0.4225,
+      "step": 2817
+    },
+    {
+      "epoch": 0.17318092428711898,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.9534015294951235e-05,
+      "loss": 0.5711,
+      "step": 2818
+    },
+    {
+      "epoch": 0.17324237954768928,
+      "grad_norm": 19.25,
+      "learning_rate": 1.953331293886697e-05,
+      "loss": 0.4941,
+      "step": 2819
+    },
+    {
+      "epoch": 0.1733038348082596,
+      "grad_norm": 16.25,
+      "learning_rate": 1.9532610066513693e-05,
+      "loss": 0.5831,
+      "step": 2820
+    },
+    {
+      "epoch": 0.1733652900688299,
+      "grad_norm": 29.625,
+      "learning_rate": 1.9531906677929472e-05,
+      "loss": 0.6078,
+      "step": 2821
+    },
+    {
+      "epoch": 0.1734267453294002,
+      "grad_norm": 26.875,
+      "learning_rate": 1.9531202773152395e-05,
+      "loss": 0.6546,
+      "step": 2822
+    },
+    {
+      "epoch": 0.1734882005899705,
+      "grad_norm": 18.25,
+      "learning_rate": 1.953049835222059e-05,
+      "loss": 0.4316,
+      "step": 2823
+    },
+    {
+      "epoch": 0.1735496558505408,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.952979341517219e-05,
+      "loss": 0.3824,
+      "step": 2824
+    },
+    {
+      "epoch": 0.1736111111111111,
+      "grad_norm": 26.0,
+      "learning_rate": 1.952908796204538e-05,
+      "loss": 0.7611,
+      "step": 2825
+    },
+    {
+      "epoch": 0.1736725663716814,
+      "grad_norm": 44.0,
+      "learning_rate": 1.9528381992878362e-05,
+      "loss": 0.7742,
+      "step": 2826
+    },
+    {
+      "epoch": 0.1737340216322517,
+      "grad_norm": 22.5,
+      "learning_rate": 1.9527675507709368e-05,
+      "loss": 0.669,
+      "step": 2827
+    },
+    {
+      "epoch": 0.17379547689282202,
+      "grad_norm": 31.875,
+      "learning_rate": 1.9526968506576654e-05,
+      "loss": 0.6578,
+      "step": 2828
+    },
+    {
+      "epoch": 0.17385693215339232,
+      "grad_norm": 14.25,
+      "learning_rate": 1.9526260989518506e-05,
+      "loss": 0.4874,
+      "step": 2829
+    },
+    {
+      "epoch": 0.17391838741396262,
+      "grad_norm": 27.25,
+      "learning_rate": 1.9525552956573244e-05,
+      "loss": 0.7209,
+      "step": 2830
+    },
+    {
+      "epoch": 0.17397984267453295,
+      "grad_norm": 14.125,
+      "learning_rate": 1.9524844407779208e-05,
+      "loss": 0.4477,
+      "step": 2831
+    },
+    {
+      "epoch": 0.17404129793510326,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.952413534317477e-05,
+      "loss": 0.4057,
+      "step": 2832
+    },
+    {
+      "epoch": 0.17410275319567356,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.9523425762798328e-05,
+      "loss": 0.5297,
+      "step": 2833
+    },
+    {
+      "epoch": 0.17416420845624386,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.952271566668831e-05,
+      "loss": 0.47,
+      "step": 2834
+    },
+    {
+      "epoch": 0.17422566371681417,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.9522005054883172e-05,
+      "loss": 0.3851,
+      "step": 2835
+    },
+    {
+      "epoch": 0.17428711897738447,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.9521293927421388e-05,
+      "loss": 0.4601,
+      "step": 2836
+    },
+    {
+      "epoch": 0.17434857423795477,
+      "grad_norm": 16.5,
+      "learning_rate": 1.9520582284341478e-05,
+      "loss": 0.5451,
+      "step": 2837
+    },
+    {
+      "epoch": 0.17441002949852508,
+      "grad_norm": 17.75,
+      "learning_rate": 1.951987012568198e-05,
+      "loss": 0.4489,
+      "step": 2838
+    },
+    {
+      "epoch": 0.17447148475909538,
+      "grad_norm": 16.25,
+      "learning_rate": 1.9519157451481453e-05,
+      "loss": 0.3555,
+      "step": 2839
+    },
+    {
+      "epoch": 0.17453294001966568,
+      "grad_norm": 5.75,
+      "learning_rate": 1.95184442617785e-05,
+      "loss": 0.4051,
+      "step": 2840
+    },
+    {
+      "epoch": 0.174594395280236,
+      "grad_norm": 21.5,
+      "learning_rate": 1.951773055661174e-05,
+      "loss": 0.5001,
+      "step": 2841
+    },
+    {
+      "epoch": 0.1746558505408063,
+      "grad_norm": 23.625,
+      "learning_rate": 1.9517016336019817e-05,
+      "loss": 0.4908,
+      "step": 2842
+    },
+    {
+      "epoch": 0.1747173058013766,
+      "grad_norm": 30.0,
+      "learning_rate": 1.9516301600041417e-05,
+      "loss": 0.5732,
+      "step": 2843
+    },
+    {
+      "epoch": 0.1747787610619469,
+      "grad_norm": 27.75,
+      "learning_rate": 1.9515586348715246e-05,
+      "loss": 0.5956,
+      "step": 2844
+    },
+    {
+      "epoch": 0.1748402163225172,
+      "grad_norm": 19.75,
+      "learning_rate": 1.951487058208003e-05,
+      "loss": 0.5303,
+      "step": 2845
+    },
+    {
+      "epoch": 0.1749016715830875,
+      "grad_norm": 29.0,
+      "learning_rate": 1.9514154300174542e-05,
+      "loss": 0.4475,
+      "step": 2846
+    },
+    {
+      "epoch": 0.1749631268436578,
+      "grad_norm": 6.75,
+      "learning_rate": 1.9513437503037563e-05,
+      "loss": 0.3651,
+      "step": 2847
+    },
+    {
+      "epoch": 0.1750245821042281,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9512720190707915e-05,
+      "loss": 0.3686,
+      "step": 2848
+    },
+    {
+      "epoch": 0.1750860373647984,
+      "grad_norm": 11.5,
+      "learning_rate": 1.951200236322444e-05,
+      "loss": 0.5353,
+      "step": 2849
+    },
+    {
+      "epoch": 0.17514749262536874,
+      "grad_norm": 19.875,
+      "learning_rate": 1.9511284020626015e-05,
+      "loss": 0.4094,
+      "step": 2850
+    },
+    {
+      "epoch": 0.17520894788593905,
+      "grad_norm": 24.625,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 0.4881,
+      "step": 2851
+    },
+    {
+      "epoch": 0.17527040314650935,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.9509845790239938e-05,
+      "loss": 0.3908,
+      "step": 2852
+    },
+    {
+      "epoch": 0.17533185840707965,
+      "grad_norm": 25.125,
+      "learning_rate": 1.9509125902530178e-05,
+      "loss": 0.6116,
+      "step": 2853
+    },
+    {
+      "epoch": 0.17539331366764996,
+      "grad_norm": 19.25,
+      "learning_rate": 1.9508405499861235e-05,
+      "loss": 0.3593,
+      "step": 2854
+    },
+    {
+      "epoch": 0.17545476892822026,
+      "grad_norm": 8.875,
+      "learning_rate": 1.9507684582272128e-05,
+      "loss": 0.4842,
+      "step": 2855
+    },
+    {
+      "epoch": 0.17551622418879056,
+      "grad_norm": 11.1875,
+      "learning_rate": 1.9506963149801894e-05,
+      "loss": 0.3666,
+      "step": 2856
+    },
+    {
+      "epoch": 0.17557767944936087,
+      "grad_norm": 21.5,
+      "learning_rate": 1.95062412024896e-05,
+      "loss": 0.6243,
+      "step": 2857
+    },
+    {
+      "epoch": 0.17563913470993117,
+      "grad_norm": 16.25,
+      "learning_rate": 1.950551874037435e-05,
+      "loss": 0.5342,
+      "step": 2858
+    },
+    {
+      "epoch": 0.17570058997050148,
+      "grad_norm": 21.25,
+      "learning_rate": 1.950479576349526e-05,
+      "loss": 0.4998,
+      "step": 2859
+    },
+    {
+      "epoch": 0.17576204523107178,
+      "grad_norm": 22.875,
+      "learning_rate": 1.9504072271891486e-05,
+      "loss": 0.463,
+      "step": 2860
+    },
+    {
+      "epoch": 0.17582350049164208,
+      "grad_norm": 35.25,
+      "learning_rate": 1.9503348265602212e-05,
+      "loss": 0.597,
+      "step": 2861
+    },
+    {
+      "epoch": 0.17588495575221239,
+      "grad_norm": 19.125,
+      "learning_rate": 1.950262374466664e-05,
+      "loss": 0.7796,
+      "step": 2862
+    },
+    {
+      "epoch": 0.1759464110127827,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.950189870912401e-05,
+      "loss": 0.5042,
+      "step": 2863
+    },
+    {
+      "epoch": 0.176007866273353,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.950117315901358e-05,
+      "loss": 0.4301,
+      "step": 2864
+    },
+    {
+      "epoch": 0.1760693215339233,
+      "grad_norm": 41.25,
+      "learning_rate": 1.9500447094374645e-05,
+      "loss": 0.5648,
+      "step": 2865
+    },
+    {
+      "epoch": 0.1761307767944936,
+      "grad_norm": 11.625,
+      "learning_rate": 1.9499720515246524e-05,
+      "loss": 0.4141,
+      "step": 2866
+    },
+    {
+      "epoch": 0.1761922320550639,
+      "grad_norm": 9.0,
+      "learning_rate": 1.949899342166857e-05,
+      "loss": 0.4188,
+      "step": 2867
+    },
+    {
+      "epoch": 0.1762536873156342,
+      "grad_norm": 42.75,
+      "learning_rate": 1.949826581368015e-05,
+      "loss": 0.5076,
+      "step": 2868
+    },
+    {
+      "epoch": 0.17631514257620454,
+      "grad_norm": 50.25,
+      "learning_rate": 1.949753769132067e-05,
+      "loss": 0.5167,
+      "step": 2869
+    },
+    {
+      "epoch": 0.17637659783677484,
+      "grad_norm": 13.5,
+      "learning_rate": 1.9496809054629563e-05,
+      "loss": 0.2912,
+      "step": 2870
+    },
+    {
+      "epoch": 0.17643805309734514,
+      "grad_norm": 11.5,
+      "learning_rate": 1.9496079903646282e-05,
+      "loss": 0.4768,
+      "step": 2871
+    },
+    {
+      "epoch": 0.17649950835791545,
+      "grad_norm": 22.375,
+      "learning_rate": 1.949535023841032e-05,
+      "loss": 0.4446,
+      "step": 2872
+    },
+    {
+      "epoch": 0.17656096361848575,
+      "grad_norm": 237.0,
+      "learning_rate": 1.949462005896119e-05,
+      "loss": 0.6436,
+      "step": 2873
+    },
+    {
+      "epoch": 0.17662241887905605,
+      "grad_norm": 24.375,
+      "learning_rate": 1.949388936533843e-05,
+      "loss": 0.4062,
+      "step": 2874
+    },
+    {
+      "epoch": 0.17668387413962636,
+      "grad_norm": 12.875,
+      "learning_rate": 1.9493158157581617e-05,
+      "loss": 0.345,
+      "step": 2875
+    },
+    {
+      "epoch": 0.17674532940019666,
+      "grad_norm": 42.75,
+      "learning_rate": 1.949242643573034e-05,
+      "loss": 0.3885,
+      "step": 2876
+    },
+    {
+      "epoch": 0.17680678466076696,
+      "grad_norm": 31.125,
+      "learning_rate": 1.9491694199824232e-05,
+      "loss": 0.3725,
+      "step": 2877
+    },
+    {
+      "epoch": 0.17686823992133727,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.9490961449902946e-05,
+      "loss": 0.5013,
+      "step": 2878
+    },
+    {
+      "epoch": 0.17692969518190757,
+      "grad_norm": 21.125,
+      "learning_rate": 1.9490228186006164e-05,
+      "loss": 0.4623,
+      "step": 2879
+    },
+    {
+      "epoch": 0.17699115044247787,
+      "grad_norm": 32.75,
+      "learning_rate": 1.948949440817359e-05,
+      "loss": 0.6061,
+      "step": 2880
+    },
+    {
+      "epoch": 0.17705260570304818,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.9488760116444966e-05,
+      "loss": 0.5194,
+      "step": 2881
+    },
+    {
+      "epoch": 0.17711406096361848,
+      "grad_norm": 28.25,
+      "learning_rate": 1.9488025310860057e-05,
+      "loss": 0.7012,
+      "step": 2882
+    },
+    {
+      "epoch": 0.17717551622418878,
+      "grad_norm": 24.25,
+      "learning_rate": 1.9487289991458653e-05,
+      "loss": 0.396,
+      "step": 2883
+    },
+    {
+      "epoch": 0.1772369714847591,
+      "grad_norm": 28.875,
+      "learning_rate": 1.9486554158280576e-05,
+      "loss": 0.3765,
+      "step": 2884
+    },
+    {
+      "epoch": 0.1772984267453294,
+      "grad_norm": 15.25,
+      "learning_rate": 1.9485817811365676e-05,
+      "loss": 0.5094,
+      "step": 2885
+    },
+    {
+      "epoch": 0.1773598820058997,
+      "grad_norm": 15.75,
+      "learning_rate": 1.948508095075383e-05,
+      "loss": 0.4745,
+      "step": 2886
+    },
+    {
+      "epoch": 0.17742133726647,
+      "grad_norm": 21.25,
+      "learning_rate": 1.9484343576484935e-05,
+      "loss": 0.4205,
+      "step": 2887
+    },
+    {
+      "epoch": 0.17748279252704033,
+      "grad_norm": 8.0,
+      "learning_rate": 1.948360568859893e-05,
+      "loss": 0.4788,
+      "step": 2888
+    },
+    {
+      "epoch": 0.17754424778761063,
+      "grad_norm": 21.375,
+      "learning_rate": 1.9482867287135772e-05,
+      "loss": 0.6695,
+      "step": 2889
+    },
+    {
+      "epoch": 0.17760570304818094,
+      "grad_norm": 17.5,
+      "learning_rate": 1.9482128372135446e-05,
+      "loss": 0.6709,
+      "step": 2890
+    },
+    {
+      "epoch": 0.17766715830875124,
+      "grad_norm": 16.0,
+      "learning_rate": 1.9481388943637976e-05,
+      "loss": 0.7337,
+      "step": 2891
+    },
+    {
+      "epoch": 0.17772861356932154,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.94806490016834e-05,
+      "loss": 0.7453,
+      "step": 2892
+    },
+    {
+      "epoch": 0.17779006882989185,
+      "grad_norm": 29.75,
+      "learning_rate": 1.9479908546311783e-05,
+      "loss": 0.4285,
+      "step": 2893
+    },
+    {
+      "epoch": 0.17785152409046215,
+      "grad_norm": 14.625,
+      "learning_rate": 1.9479167577563233e-05,
+      "loss": 0.5007,
+      "step": 2894
+    },
+    {
+      "epoch": 0.17791297935103245,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.9478426095477877e-05,
+      "loss": 0.3671,
+      "step": 2895
+    },
+    {
+      "epoch": 0.17797443461160276,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.947768410009586e-05,
+      "loss": 0.426,
+      "step": 2896
+    },
+    {
+      "epoch": 0.17803588987217306,
+      "grad_norm": 25.75,
+      "learning_rate": 1.947694159145737e-05,
+      "loss": 0.5182,
+      "step": 2897
+    },
+    {
+      "epoch": 0.17809734513274336,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.9476198569602616e-05,
+      "loss": 0.4622,
+      "step": 2898
+    },
+    {
+      "epoch": 0.17815880039331367,
+      "grad_norm": 29.625,
+      "learning_rate": 1.947545503457184e-05,
+      "loss": 0.5383,
+      "step": 2899
+    },
+    {
+      "epoch": 0.17822025565388397,
+      "grad_norm": 4.96875,
+      "learning_rate": 1.94747109864053e-05,
+      "loss": 0.4186,
+      "step": 2900
+    },
+    {
+      "epoch": 0.17828171091445427,
+      "grad_norm": 13.625,
+      "learning_rate": 1.9473966425143292e-05,
+      "loss": 0.396,
+      "step": 2901
+    },
+    {
+      "epoch": 0.17834316617502458,
+      "grad_norm": 10.75,
+      "learning_rate": 1.9473221350826145e-05,
+      "loss": 0.36,
+      "step": 2902
+    },
+    {
+      "epoch": 0.17840462143559488,
+      "grad_norm": 15.5,
+      "learning_rate": 1.9472475763494196e-05,
+      "loss": 0.3239,
+      "step": 2903
+    },
+    {
+      "epoch": 0.17846607669616518,
+      "grad_norm": 10.3125,
+      "learning_rate": 1.947172966318783e-05,
+      "loss": 0.2852,
+      "step": 2904
+    },
+    {
+      "epoch": 0.17852753195673549,
+      "grad_norm": 48.5,
+      "learning_rate": 1.9470983049947446e-05,
+      "loss": 0.675,
+      "step": 2905
+    },
+    {
+      "epoch": 0.1785889872173058,
+      "grad_norm": 18.5,
+      "learning_rate": 1.947023592381348e-05,
+      "loss": 0.5563,
+      "step": 2906
+    },
+    {
+      "epoch": 0.1786504424778761,
+      "grad_norm": 38.25,
+      "learning_rate": 1.9469488284826386e-05,
+      "loss": 0.5168,
+      "step": 2907
+    },
+    {
+      "epoch": 0.17871189773844642,
+      "grad_norm": 29.625,
+      "learning_rate": 1.946874013302666e-05,
+      "loss": 0.6391,
+      "step": 2908
+    },
+    {
+      "epoch": 0.17877335299901673,
+      "grad_norm": 23.375,
+      "learning_rate": 1.9467991468454817e-05,
+      "loss": 0.4871,
+      "step": 2909
+    },
+    {
+      "epoch": 0.17883480825958703,
+      "grad_norm": 14.375,
+      "learning_rate": 1.9467242291151394e-05,
+      "loss": 0.4639,
+      "step": 2910
+    },
+    {
+      "epoch": 0.17889626352015733,
+      "grad_norm": 13.75,
+      "learning_rate": 1.9466492601156964e-05,
+      "loss": 0.5274,
+      "step": 2911
+    },
+    {
+      "epoch": 0.17895771878072764,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.946574239851213e-05,
+      "loss": 0.335,
+      "step": 2912
+    },
+    {
+      "epoch": 0.17901917404129794,
+      "grad_norm": 30.125,
+      "learning_rate": 1.9464991683257514e-05,
+      "loss": 0.6474,
+      "step": 2913
+    },
+    {
+      "epoch": 0.17908062930186824,
+      "grad_norm": 23.25,
+      "learning_rate": 1.9464240455433775e-05,
+      "loss": 0.5869,
+      "step": 2914
+    },
+    {
+      "epoch": 0.17914208456243855,
+      "grad_norm": 18.0,
+      "learning_rate": 1.946348871508159e-05,
+      "loss": 0.4296,
+      "step": 2915
+    },
+    {
+      "epoch": 0.17920353982300885,
+      "grad_norm": 24.25,
+      "learning_rate": 1.9462736462241672e-05,
+      "loss": 0.3919,
+      "step": 2916
+    },
+    {
+      "epoch": 0.17926499508357915,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.946198369695476e-05,
+      "loss": 0.3655,
+      "step": 2917
+    },
+    {
+      "epoch": 0.17932645034414946,
+      "grad_norm": 22.75,
+      "learning_rate": 1.9461230419261615e-05,
+      "loss": 0.4272,
+      "step": 2918
+    },
+    {
+      "epoch": 0.17938790560471976,
+      "grad_norm": 17.125,
+      "learning_rate": 1.9460476629203035e-05,
+      "loss": 0.4657,
+      "step": 2919
+    },
+    {
+      "epoch": 0.17944936086529006,
+      "grad_norm": 19.5,
+      "learning_rate": 1.945972232681984e-05,
+      "loss": 0.4281,
+      "step": 2920
+    },
+    {
+      "epoch": 0.17951081612586037,
+      "grad_norm": 18.625,
+      "learning_rate": 1.9458967512152872e-05,
+      "loss": 0.4025,
+      "step": 2921
+    },
+    {
+      "epoch": 0.17957227138643067,
+      "grad_norm": 5.21875,
+      "learning_rate": 1.9458212185243015e-05,
+      "loss": 0.49,
+      "step": 2922
+    },
+    {
+      "epoch": 0.17963372664700097,
+      "grad_norm": 6.25,
+      "learning_rate": 1.945745634613117e-05,
+      "loss": 0.4526,
+      "step": 2923
+    },
+    {
+      "epoch": 0.17969518190757128,
+      "grad_norm": 14.75,
+      "learning_rate": 1.945669999485827e-05,
+      "loss": 0.4609,
+      "step": 2924
+    },
+    {
+      "epoch": 0.17975663716814158,
+      "grad_norm": 35.25,
+      "learning_rate": 1.9455943131465272e-05,
+      "loss": 0.3328,
+      "step": 2925
+    },
+    {
+      "epoch": 0.17981809242871188,
+      "grad_norm": 19.25,
+      "learning_rate": 1.945518575599317e-05,
+      "loss": 0.5412,
+      "step": 2926
+    },
+    {
+      "epoch": 0.17987954768928222,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.945442786848297e-05,
+      "loss": 0.3348,
+      "step": 2927
+    },
+    {
+      "epoch": 0.17994100294985252,
+      "grad_norm": 39.75,
+      "learning_rate": 1.9453669468975724e-05,
+      "loss": 0.3417,
+      "step": 2928
+    },
+    {
+      "epoch": 0.18000245821042282,
+      "grad_norm": 32.5,
+      "learning_rate": 1.9452910557512497e-05,
+      "loss": 0.553,
+      "step": 2929
+    },
+    {
+      "epoch": 0.18006391347099313,
+      "grad_norm": 28.875,
+      "learning_rate": 1.9452151134134385e-05,
+      "loss": 0.4277,
+      "step": 2930
+    },
+    {
+      "epoch": 0.18012536873156343,
+      "grad_norm": 28.0,
+      "learning_rate": 1.945139119888252e-05,
+      "loss": 0.5471,
+      "step": 2931
+    },
+    {
+      "epoch": 0.18018682399213373,
+      "grad_norm": 17.625,
+      "learning_rate": 1.945063075179805e-05,
+      "loss": 0.4159,
+      "step": 2932
+    },
+    {
+      "epoch": 0.18024827925270404,
+      "grad_norm": 27.625,
+      "learning_rate": 1.9449869792922162e-05,
+      "loss": 0.4285,
+      "step": 2933
+    },
+    {
+      "epoch": 0.18030973451327434,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.944910832229606e-05,
+      "loss": 0.3986,
+      "step": 2934
+    },
+    {
+      "epoch": 0.18037118977384464,
+      "grad_norm": 39.5,
+      "learning_rate": 1.9448346339960984e-05,
+      "loss": 0.4442,
+      "step": 2935
+    },
+    {
+      "epoch": 0.18043264503441495,
+      "grad_norm": 12.5,
+      "learning_rate": 1.9447583845958198e-05,
+      "loss": 0.3368,
+      "step": 2936
+    },
+    {
+      "epoch": 0.18049410029498525,
+      "grad_norm": 20.25,
+      "learning_rate": 1.9446820840328993e-05,
+      "loss": 0.341,
+      "step": 2937
+    },
+    {
+      "epoch": 0.18055555555555555,
+      "grad_norm": 23.375,
+      "learning_rate": 1.944605732311469e-05,
+      "loss": 0.6011,
+      "step": 2938
+    },
+    {
+      "epoch": 0.18061701081612586,
+      "grad_norm": 33.0,
+      "learning_rate": 1.944529329435664e-05,
+      "loss": 0.4328,
+      "step": 2939
+    },
+    {
+      "epoch": 0.18067846607669616,
+      "grad_norm": 8.0,
+      "learning_rate": 1.944452875409621e-05,
+      "loss": 0.5344,
+      "step": 2940
+    },
+    {
+      "epoch": 0.18073992133726646,
+      "grad_norm": 22.75,
+      "learning_rate": 1.944376370237481e-05,
+      "loss": 0.3518,
+      "step": 2941
+    },
+    {
+      "epoch": 0.18080137659783677,
+      "grad_norm": 19.875,
+      "learning_rate": 1.944299813923387e-05,
+      "loss": 0.5395,
+      "step": 2942
+    },
+    {
+      "epoch": 0.18086283185840707,
+      "grad_norm": 10.125,
+      "learning_rate": 1.9442232064714848e-05,
+      "loss": 0.5081,
+      "step": 2943
+    },
+    {
+      "epoch": 0.18092428711897737,
+      "grad_norm": 17.0,
+      "learning_rate": 1.944146547885923e-05,
+      "loss": 0.2984,
+      "step": 2944
+    },
+    {
+      "epoch": 0.18098574237954768,
+      "grad_norm": 61.75,
+      "learning_rate": 1.9440698381708526e-05,
+      "loss": 0.5799,
+      "step": 2945
+    },
+    {
+      "epoch": 0.181047197640118,
+      "grad_norm": 25.5,
+      "learning_rate": 1.9439930773304284e-05,
+      "loss": 0.4806,
+      "step": 2946
+    },
+    {
+      "epoch": 0.1811086529006883,
+      "grad_norm": 24.625,
+      "learning_rate": 1.9439162653688066e-05,
+      "loss": 0.5856,
+      "step": 2947
+    },
+    {
+      "epoch": 0.18117010816125861,
+      "grad_norm": 27.0,
+      "learning_rate": 1.9438394022901477e-05,
+      "loss": 0.558,
+      "step": 2948
+    },
+    {
+      "epoch": 0.18123156342182892,
+      "grad_norm": 31.375,
+      "learning_rate": 1.9437624880986135e-05,
+      "loss": 0.7091,
+      "step": 2949
+    },
+    {
+      "epoch": 0.18129301868239922,
+      "grad_norm": 20.0,
+      "learning_rate": 1.9436855227983695e-05,
+      "loss": 0.6281,
+      "step": 2950
+    },
+    {
+      "epoch": 0.18135447394296952,
+      "grad_norm": 29.875,
+      "learning_rate": 1.9436085063935837e-05,
+      "loss": 0.5779,
+      "step": 2951
+    },
+    {
+      "epoch": 0.18141592920353983,
+      "grad_norm": 15.125,
+      "learning_rate": 1.9435314388884266e-05,
+      "loss": 0.5043,
+      "step": 2952
+    },
+    {
+      "epoch": 0.18147738446411013,
+      "grad_norm": 22.375,
+      "learning_rate": 1.9434543202870726e-05,
+      "loss": 0.4537,
+      "step": 2953
+    },
+    {
+      "epoch": 0.18153883972468043,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.9433771505936968e-05,
+      "loss": 0.4489,
+      "step": 2954
+    },
+    {
+      "epoch": 0.18160029498525074,
+      "grad_norm": 22.0,
+      "learning_rate": 1.943299929812479e-05,
+      "loss": 0.483,
+      "step": 2955
+    },
+    {
+      "epoch": 0.18166175024582104,
+      "grad_norm": 22.5,
+      "learning_rate": 1.943222657947601e-05,
+      "loss": 0.7295,
+      "step": 2956
+    },
+    {
+      "epoch": 0.18172320550639134,
+      "grad_norm": 35.0,
+      "learning_rate": 1.9431453350032476e-05,
+      "loss": 0.495,
+      "step": 2957
+    },
+    {
+      "epoch": 0.18178466076696165,
+      "grad_norm": 21.25,
+      "learning_rate": 1.943067960983605e-05,
+      "loss": 0.3872,
+      "step": 2958
+    },
+    {
+      "epoch": 0.18184611602753195,
+      "grad_norm": 24.0,
+      "learning_rate": 1.9429905358928648e-05,
+      "loss": 0.3747,
+      "step": 2959
+    },
+    {
+      "epoch": 0.18190757128810225,
+      "grad_norm": 13.125,
+      "learning_rate": 1.9429130597352187e-05,
+      "loss": 0.3277,
+      "step": 2960
+    },
+    {
+      "epoch": 0.18196902654867256,
+      "grad_norm": 34.25,
+      "learning_rate": 1.9428355325148632e-05,
+      "loss": 0.5526,
+      "step": 2961
+    },
+    {
+      "epoch": 0.18203048180924286,
+      "grad_norm": 31.875,
+      "learning_rate": 1.9427579542359966e-05,
+      "loss": 0.5878,
+      "step": 2962
+    },
+    {
+      "epoch": 0.18209193706981316,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.9426803249028198e-05,
+      "loss": 0.5358,
+      "step": 2963
+    },
+    {
+      "epoch": 0.18215339233038347,
+      "grad_norm": 18.25,
+      "learning_rate": 1.9426026445195365e-05,
+      "loss": 0.6044,
+      "step": 2964
+    },
+    {
+      "epoch": 0.1822148475909538,
+      "grad_norm": 21.75,
+      "learning_rate": 1.9425249130903544e-05,
+      "loss": 0.5308,
+      "step": 2965
+    },
+    {
+      "epoch": 0.1822763028515241,
+      "grad_norm": 31.0,
+      "learning_rate": 1.9424471306194822e-05,
+      "loss": 0.5396,
+      "step": 2966
+    },
+    {
+      "epoch": 0.1823377581120944,
+      "grad_norm": 38.5,
+      "learning_rate": 1.942369297111132e-05,
+      "loss": 0.3954,
+      "step": 2967
+    },
+    {
+      "epoch": 0.1823992133726647,
+      "grad_norm": 13.75,
+      "learning_rate": 1.942291412569519e-05,
+      "loss": 0.4964,
+      "step": 2968
+    },
+    {
+      "epoch": 0.182460668633235,
+      "grad_norm": 27.0,
+      "learning_rate": 1.9422134769988617e-05,
+      "loss": 0.5155,
+      "step": 2969
+    },
+    {
+      "epoch": 0.18252212389380532,
+      "grad_norm": 22.375,
+      "learning_rate": 1.9421354904033797e-05,
+      "loss": 0.7103,
+      "step": 2970
+    },
+    {
+      "epoch": 0.18258357915437562,
+      "grad_norm": 13.375,
+      "learning_rate": 1.942057452787297e-05,
+      "loss": 0.628,
+      "step": 2971
+    },
+    {
+      "epoch": 0.18264503441494592,
+      "grad_norm": 21.5,
+      "learning_rate": 1.941979364154839e-05,
+      "loss": 0.4804,
+      "step": 2972
+    },
+    {
+      "epoch": 0.18270648967551623,
+      "grad_norm": 21.625,
+      "learning_rate": 1.9419012245102347e-05,
+      "loss": 0.6384,
+      "step": 2973
+    },
+    {
+      "epoch": 0.18276794493608653,
+      "grad_norm": 17.5,
+      "learning_rate": 1.9418230338577164e-05,
+      "loss": 0.6131,
+      "step": 2974
+    },
+    {
+      "epoch": 0.18282940019665683,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.9417447922015174e-05,
+      "loss": 0.4896,
+      "step": 2975
+    },
+    {
+      "epoch": 0.18289085545722714,
+      "grad_norm": 24.625,
+      "learning_rate": 1.9416664995458756e-05,
+      "loss": 0.5091,
+      "step": 2976
+    },
+    {
+      "epoch": 0.18295231071779744,
+      "grad_norm": 13.375,
+      "learning_rate": 1.9415881558950302e-05,
+      "loss": 0.5236,
+      "step": 2977
+    },
+    {
+      "epoch": 0.18301376597836774,
+      "grad_norm": 17.875,
+      "learning_rate": 1.9415097612532247e-05,
+      "loss": 0.4612,
+      "step": 2978
+    },
+    {
+      "epoch": 0.18307522123893805,
+      "grad_norm": 18.875,
+      "learning_rate": 1.9414313156247035e-05,
+      "loss": 0.481,
+      "step": 2979
+    },
+    {
+      "epoch": 0.18313667649950835,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9413528190137158e-05,
+      "loss": 0.3282,
+      "step": 2980
+    },
+    {
+      "epoch": 0.18319813176007865,
+      "grad_norm": 18.125,
+      "learning_rate": 1.941274271424512e-05,
+      "loss": 0.4691,
+      "step": 2981
+    },
+    {
+      "epoch": 0.18325958702064896,
+      "grad_norm": 37.75,
+      "learning_rate": 1.941195672861345e-05,
+      "loss": 0.4697,
+      "step": 2982
+    },
+    {
+      "epoch": 0.18332104228121926,
+      "grad_norm": 10.8125,
+      "learning_rate": 1.9411170233284728e-05,
+      "loss": 0.4417,
+      "step": 2983
+    },
+    {
+      "epoch": 0.1833824975417896,
+      "grad_norm": 15.5,
+      "learning_rate": 1.9410383228301536e-05,
+      "loss": 0.3815,
+      "step": 2984
+    },
+    {
+      "epoch": 0.1834439528023599,
+      "grad_norm": 15.0,
+      "learning_rate": 1.9409595713706496e-05,
+      "loss": 0.4851,
+      "step": 2985
+    },
+    {
+      "epoch": 0.1835054080629302,
+      "grad_norm": 28.125,
+      "learning_rate": 1.9408807689542257e-05,
+      "loss": 0.4497,
+      "step": 2986
+    },
+    {
+      "epoch": 0.1835668633235005,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.940801915585149e-05,
+      "loss": 0.4029,
+      "step": 2987
+    },
+    {
+      "epoch": 0.1836283185840708,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.9407230112676898e-05,
+      "loss": 0.3591,
+      "step": 2988
+    },
+    {
+      "epoch": 0.1836897738446411,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.9406440560061214e-05,
+      "loss": 0.2735,
+      "step": 2989
+    },
+    {
+      "epoch": 0.1837512291052114,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.9405650498047193e-05,
+      "loss": 0.2801,
+      "step": 2990
+    },
+    {
+      "epoch": 0.18381268436578171,
+      "grad_norm": 20.375,
+      "learning_rate": 1.9404859926677625e-05,
+      "loss": 0.3876,
+      "step": 2991
+    },
+    {
+      "epoch": 0.18387413962635202,
+      "grad_norm": 6.53125,
+      "learning_rate": 1.9404068845995317e-05,
+      "loss": 0.3238,
+      "step": 2992
+    },
+    {
+      "epoch": 0.18393559488692232,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.940327725604311e-05,
+      "loss": 0.3296,
+      "step": 2993
+    },
+    {
+      "epoch": 0.18399705014749262,
+      "grad_norm": 10.25,
+      "learning_rate": 1.9402485156863876e-05,
+      "loss": 0.3617,
+      "step": 2994
+    },
+    {
+      "epoch": 0.18405850540806293,
+      "grad_norm": 30.125,
+      "learning_rate": 1.9401692548500504e-05,
+      "loss": 0.5456,
+      "step": 2995
+    },
+    {
+      "epoch": 0.18411996066863323,
+      "grad_norm": 51.5,
+      "learning_rate": 1.9400899430995923e-05,
+      "loss": 0.5129,
+      "step": 2996
+    },
+    {
+      "epoch": 0.18418141592920353,
+      "grad_norm": 27.25,
+      "learning_rate": 1.9400105804393083e-05,
+      "loss": 0.5718,
+      "step": 2997
+    },
+    {
+      "epoch": 0.18424287118977384,
+      "grad_norm": 20.875,
+      "learning_rate": 1.9399311668734957e-05,
+      "loss": 0.5517,
+      "step": 2998
+    },
+    {
+      "epoch": 0.18430432645034414,
+      "grad_norm": 32.25,
+      "learning_rate": 1.9398517024064556e-05,
+      "loss": 0.2847,
+      "step": 2999
+    },
+    {
+      "epoch": 0.18436578171091444,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.939772187042491e-05,
+      "loss": 0.3404,
+      "step": 3000
+    },
+    {
+      "epoch": 0.18442723697148475,
+      "grad_norm": 33.0,
+      "learning_rate": 1.9396926207859085e-05,
+      "loss": 0.6463,
+      "step": 3001
+    },
+    {
+      "epoch": 0.18448869223205505,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9396130036410166e-05,
+      "loss": 0.7697,
+      "step": 3002
+    },
+    {
+      "epoch": 0.18455014749262535,
+      "grad_norm": 12.125,
+      "learning_rate": 1.9395333356121268e-05,
+      "loss": 0.4747,
+      "step": 3003
+    },
+    {
+      "epoch": 0.18461160275319569,
+      "grad_norm": 29.625,
+      "learning_rate": 1.9394536167035535e-05,
+      "loss": 0.5613,
+      "step": 3004
+    },
+    {
+      "epoch": 0.184673058013766,
+      "grad_norm": 30.625,
+      "learning_rate": 1.939373846919614e-05,
+      "loss": 0.7578,
+      "step": 3005
+    },
+    {
+      "epoch": 0.1847345132743363,
+      "grad_norm": 17.75,
+      "learning_rate": 1.9392940262646284e-05,
+      "loss": 0.5454,
+      "step": 3006
+    },
+    {
+      "epoch": 0.1847959685349066,
+      "grad_norm": 27.0,
+      "learning_rate": 1.9392141547429183e-05,
+      "loss": 0.4712,
+      "step": 3007
+    },
+    {
+      "epoch": 0.1848574237954769,
+      "grad_norm": 21.0,
+      "learning_rate": 1.9391342323588104e-05,
+      "loss": 0.4669,
+      "step": 3008
+    },
+    {
+      "epoch": 0.1849188790560472,
+      "grad_norm": 19.875,
+      "learning_rate": 1.939054259116632e-05,
+      "loss": 0.3443,
+      "step": 3009
+    },
+    {
+      "epoch": 0.1849803343166175,
+      "grad_norm": 18.125,
+      "learning_rate": 1.938974235020714e-05,
+      "loss": 0.4148,
+      "step": 3010
+    },
+    {
+      "epoch": 0.1850417895771878,
+      "grad_norm": 21.5,
+      "learning_rate": 1.9388941600753902e-05,
+      "loss": 0.4424,
+      "step": 3011
+    },
+    {
+      "epoch": 0.1851032448377581,
+      "grad_norm": 10.75,
+      "learning_rate": 1.9388140342849976e-05,
+      "loss": 0.3094,
+      "step": 3012
+    },
+    {
+      "epoch": 0.18516470009832842,
+      "grad_norm": 15.125,
+      "learning_rate": 1.9387338576538743e-05,
+      "loss": 0.5617,
+      "step": 3013
+    },
+    {
+      "epoch": 0.18522615535889872,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.9386536301863626e-05,
+      "loss": 0.2456,
+      "step": 3014
+    },
+    {
+      "epoch": 0.18528761061946902,
+      "grad_norm": 10.8125,
+      "learning_rate": 1.9385733518868074e-05,
+      "loss": 0.2986,
+      "step": 3015
+    },
+    {
+      "epoch": 0.18534906588003933,
+      "grad_norm": 4.0,
+      "learning_rate": 1.938493022759556e-05,
+      "loss": 0.2952,
+      "step": 3016
+    },
+    {
+      "epoch": 0.18541052114060963,
+      "grad_norm": 45.25,
+      "learning_rate": 1.938412642808959e-05,
+      "loss": 0.6612,
+      "step": 3017
+    },
+    {
+      "epoch": 0.18547197640117993,
+      "grad_norm": 50.5,
+      "learning_rate": 1.938332212039368e-05,
+      "loss": 0.5555,
+      "step": 3018
+    },
+    {
+      "epoch": 0.18553343166175024,
+      "grad_norm": 34.25,
+      "learning_rate": 1.9382517304551397e-05,
+      "loss": 0.5172,
+      "step": 3019
+    },
+    {
+      "epoch": 0.18559488692232054,
+      "grad_norm": 37.0,
+      "learning_rate": 1.938171198060632e-05,
+      "loss": 0.4086,
+      "step": 3020
+    },
+    {
+      "epoch": 0.18565634218289084,
+      "grad_norm": 54.0,
+      "learning_rate": 1.938090614860207e-05,
+      "loss": 0.4724,
+      "step": 3021
+    },
+    {
+      "epoch": 0.18571779744346115,
+      "grad_norm": 19.5,
+      "learning_rate": 1.9380099808582278e-05,
+      "loss": 0.3088,
+      "step": 3022
+    },
+    {
+      "epoch": 0.18577925270403148,
+      "grad_norm": 41.25,
+      "learning_rate": 1.937929296059061e-05,
+      "loss": 0.4072,
+      "step": 3023
+    },
+    {
+      "epoch": 0.18584070796460178,
+      "grad_norm": 15.4375,
+      "learning_rate": 1.9378485604670766e-05,
+      "loss": 0.5105,
+      "step": 3024
+    },
+    {
+      "epoch": 0.18590216322517208,
+      "grad_norm": 19.875,
+      "learning_rate": 1.937767774086646e-05,
+      "loss": 0.5776,
+      "step": 3025
+    },
+    {
+      "epoch": 0.1859636184857424,
+      "grad_norm": 21.75,
+      "learning_rate": 1.937686936922145e-05,
+      "loss": 0.5936,
+      "step": 3026
+    },
+    {
+      "epoch": 0.1860250737463127,
+      "grad_norm": 43.25,
+      "learning_rate": 1.937606048977951e-05,
+      "loss": 0.5969,
+      "step": 3027
+    },
+    {
+      "epoch": 0.186086529006883,
+      "grad_norm": 21.125,
+      "learning_rate": 1.9375251102584438e-05,
+      "loss": 0.6166,
+      "step": 3028
+    },
+    {
+      "epoch": 0.1861479842674533,
+      "grad_norm": 17.125,
+      "learning_rate": 1.937444120768007e-05,
+      "loss": 0.5478,
+      "step": 3029
+    },
+    {
+      "epoch": 0.1862094395280236,
+      "grad_norm": 26.875,
+      "learning_rate": 1.9373630805110267e-05,
+      "loss": 0.3757,
+      "step": 3030
+    },
+    {
+      "epoch": 0.1862708947885939,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.937281989491892e-05,
+      "loss": 0.3527,
+      "step": 3031
+    },
+    {
+      "epoch": 0.1863323500491642,
+      "grad_norm": 14.375,
+      "learning_rate": 1.9372008477149928e-05,
+      "loss": 0.4788,
+      "step": 3032
+    },
+    {
+      "epoch": 0.1863938053097345,
+      "grad_norm": 35.5,
+      "learning_rate": 1.9371196551847248e-05,
+      "loss": 0.5742,
+      "step": 3033
+    },
+    {
+      "epoch": 0.18645526057030481,
+      "grad_norm": 24.5,
+      "learning_rate": 1.937038411905484e-05,
+      "loss": 0.5256,
+      "step": 3034
+    },
+    {
+      "epoch": 0.18651671583087512,
+      "grad_norm": 34.75,
+      "learning_rate": 1.9369571178816708e-05,
+      "loss": 0.4345,
+      "step": 3035
+    },
+    {
+      "epoch": 0.18657817109144542,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.936875773117687e-05,
+      "loss": 0.3542,
+      "step": 3036
+    },
+    {
+      "epoch": 0.18663962635201573,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.936794377617938e-05,
+      "loss": 0.3539,
+      "step": 3037
+    },
+    {
+      "epoch": 0.18670108161258603,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.936712931386832e-05,
+      "loss": 0.3697,
+      "step": 3038
+    },
+    {
+      "epoch": 0.18676253687315633,
+      "grad_norm": 16.5,
+      "learning_rate": 1.9366314344287786e-05,
+      "loss": 0.3797,
+      "step": 3039
+    },
+    {
+      "epoch": 0.18682399213372664,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.9365498867481926e-05,
+      "loss": 0.3806,
+      "step": 3040
+    },
+    {
+      "epoch": 0.18688544739429694,
+      "grad_norm": 39.75,
+      "learning_rate": 1.9364682883494892e-05,
+      "loss": 0.4992,
+      "step": 3041
+    },
+    {
+      "epoch": 0.18694690265486727,
+      "grad_norm": 26.625,
+      "learning_rate": 1.9363866392370875e-05,
+      "loss": 0.4558,
+      "step": 3042
+    },
+    {
+      "epoch": 0.18700835791543757,
+      "grad_norm": 13.375,
+      "learning_rate": 1.9363049394154095e-05,
+      "loss": 0.4716,
+      "step": 3043
+    },
+    {
+      "epoch": 0.18706981317600788,
+      "grad_norm": 11.25,
+      "learning_rate": 1.9362231888888793e-05,
+      "loss": 0.3768,
+      "step": 3044
+    },
+    {
+      "epoch": 0.18713126843657818,
+      "grad_norm": 26.25,
+      "learning_rate": 1.936141387661924e-05,
+      "loss": 0.5961,
+      "step": 3045
+    },
+    {
+      "epoch": 0.18719272369714848,
+      "grad_norm": 10.5,
+      "learning_rate": 1.9360595357389735e-05,
+      "loss": 0.3634,
+      "step": 3046
+    },
+    {
+      "epoch": 0.1872541789577188,
+      "grad_norm": 24.875,
+      "learning_rate": 1.9359776331244606e-05,
+      "loss": 0.3673,
+      "step": 3047
+    },
+    {
+      "epoch": 0.1873156342182891,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.9358956798228205e-05,
+      "loss": 0.3366,
+      "step": 3048
+    },
+    {
+      "epoch": 0.1873770894788594,
+      "grad_norm": 19.625,
+      "learning_rate": 1.935813675838491e-05,
+      "loss": 0.6899,
+      "step": 3049
+    },
+    {
+      "epoch": 0.1874385447394297,
+      "grad_norm": 12.0,
+      "learning_rate": 1.9357316211759137e-05,
+      "loss": 0.4356,
+      "step": 3050
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 29.875,
+      "learning_rate": 1.9356495158395317e-05,
+      "loss": 0.412,
+      "step": 3051
+    },
+    {
+      "epoch": 0.1875614552605703,
+      "grad_norm": 34.0,
+      "learning_rate": 1.9355673598337916e-05,
+      "loss": 0.4612,
+      "step": 3052
+    },
+    {
+      "epoch": 0.1876229105211406,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.935485153163142e-05,
+      "loss": 0.4639,
+      "step": 3053
+    },
+    {
+      "epoch": 0.1876843657817109,
+      "grad_norm": 15.125,
+      "learning_rate": 1.9354028958320352e-05,
+      "loss": 0.3553,
+      "step": 3054
+    },
+    {
+      "epoch": 0.1877458210422812,
+      "grad_norm": 26.375,
+      "learning_rate": 1.935320587844926e-05,
+      "loss": 0.4161,
+      "step": 3055
+    },
+    {
+      "epoch": 0.18780727630285152,
+      "grad_norm": 21.25,
+      "learning_rate": 1.9352382292062712e-05,
+      "loss": 0.3405,
+      "step": 3056
+    },
+    {
+      "epoch": 0.18786873156342182,
+      "grad_norm": 25.875,
+      "learning_rate": 1.935155819920531e-05,
+      "loss": 0.4924,
+      "step": 3057
+    },
+    {
+      "epoch": 0.18793018682399212,
+      "grad_norm": 13.25,
+      "learning_rate": 1.9350733599921684e-05,
+      "loss": 0.417,
+      "step": 3058
+    },
+    {
+      "epoch": 0.18799164208456243,
+      "grad_norm": 46.0,
+      "learning_rate": 1.934990849425649e-05,
+      "loss": 0.5806,
+      "step": 3059
+    },
+    {
+      "epoch": 0.18805309734513273,
+      "grad_norm": 87.0,
+      "learning_rate": 1.9349082882254405e-05,
+      "loss": 0.4922,
+      "step": 3060
+    },
+    {
+      "epoch": 0.18811455260570306,
+      "grad_norm": 27.375,
+      "learning_rate": 1.9348256763960146e-05,
+      "loss": 0.3161,
+      "step": 3061
+    },
+    {
+      "epoch": 0.18817600786627336,
+      "grad_norm": 29.0,
+      "learning_rate": 1.9347430139418448e-05,
+      "loss": 0.4742,
+      "step": 3062
+    },
+    {
+      "epoch": 0.18823746312684367,
+      "grad_norm": 40.75,
+      "learning_rate": 1.9346603008674075e-05,
+      "loss": 0.3815,
+      "step": 3063
+    },
+    {
+      "epoch": 0.18829891838741397,
+      "grad_norm": 10.75,
+      "learning_rate": 1.9345775371771826e-05,
+      "loss": 0.5304,
+      "step": 3064
+    },
+    {
+      "epoch": 0.18836037364798427,
+      "grad_norm": 30.5,
+      "learning_rate": 1.934494722875651e-05,
+      "loss": 0.6936,
+      "step": 3065
+    },
+    {
+      "epoch": 0.18842182890855458,
+      "grad_norm": 21.625,
+      "learning_rate": 1.9344118579672987e-05,
+      "loss": 0.5219,
+      "step": 3066
+    },
+    {
+      "epoch": 0.18848328416912488,
+      "grad_norm": 46.75,
+      "learning_rate": 1.9343289424566122e-05,
+      "loss": 0.4943,
+      "step": 3067
+    },
+    {
+      "epoch": 0.18854473942969519,
+      "grad_norm": 28.0,
+      "learning_rate": 1.9342459763480822e-05,
+      "loss": 0.5983,
+      "step": 3068
+    },
+    {
+      "epoch": 0.1886061946902655,
+      "grad_norm": 18.0,
+      "learning_rate": 1.9341629596462015e-05,
+      "loss": 0.4858,
+      "step": 3069
+    },
+    {
+      "epoch": 0.1886676499508358,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.9340798923554657e-05,
+      "loss": 0.4553,
+      "step": 3070
+    },
+    {
+      "epoch": 0.1887291052114061,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.9339967744803735e-05,
+      "loss": 0.4467,
+      "step": 3071
+    },
+    {
+      "epoch": 0.1887905604719764,
+      "grad_norm": 27.0,
+      "learning_rate": 1.933913606025426e-05,
+      "loss": 0.3906,
+      "step": 3072
+    },
+    {
+      "epoch": 0.1888520157325467,
+      "grad_norm": 30.75,
+      "learning_rate": 1.933830386995127e-05,
+      "loss": 0.6018,
+      "step": 3073
+    },
+    {
+      "epoch": 0.188913470993117,
+      "grad_norm": 42.0,
+      "learning_rate": 1.933747117393983e-05,
+      "loss": 0.6741,
+      "step": 3074
+    },
+    {
+      "epoch": 0.1889749262536873,
+      "grad_norm": 36.5,
+      "learning_rate": 1.9336637972265042e-05,
+      "loss": 0.4624,
+      "step": 3075
+    },
+    {
+      "epoch": 0.1890363815142576,
+      "grad_norm": 63.5,
+      "learning_rate": 1.9335804264972018e-05,
+      "loss": 0.521,
+      "step": 3076
+    },
+    {
+      "epoch": 0.18909783677482792,
+      "grad_norm": 36.0,
+      "learning_rate": 1.933497005210591e-05,
+      "loss": 0.3893,
+      "step": 3077
+    },
+    {
+      "epoch": 0.18915929203539822,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.93341353337119e-05,
+      "loss": 0.388,
+      "step": 3078
+    },
+    {
+      "epoch": 0.18922074729596852,
+      "grad_norm": 37.25,
+      "learning_rate": 1.9333300109835182e-05,
+      "loss": 0.4786,
+      "step": 3079
+    },
+    {
+      "epoch": 0.18928220255653885,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.9332464380520993e-05,
+      "loss": 0.3302,
+      "step": 3080
+    },
+    {
+      "epoch": 0.18934365781710916,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.9331628145814587e-05,
+      "loss": 0.4529,
+      "step": 3081
+    },
+    {
+      "epoch": 0.18940511307767946,
+      "grad_norm": 17.5,
+      "learning_rate": 1.9330791405761254e-05,
+      "loss": 0.4259,
+      "step": 3082
+    },
+    {
+      "epoch": 0.18946656833824976,
+      "grad_norm": 38.75,
+      "learning_rate": 1.9329954160406306e-05,
+      "loss": 0.6097,
+      "step": 3083
+    },
+    {
+      "epoch": 0.18952802359882007,
+      "grad_norm": 16.625,
+      "learning_rate": 1.932911640979508e-05,
+      "loss": 0.4654,
+      "step": 3084
+    },
+    {
+      "epoch": 0.18958947885939037,
+      "grad_norm": 47.75,
+      "learning_rate": 1.9328278153972947e-05,
+      "loss": 0.4684,
+      "step": 3085
+    },
+    {
+      "epoch": 0.18965093411996067,
+      "grad_norm": 22.625,
+      "learning_rate": 1.93274393929853e-05,
+      "loss": 0.4892,
+      "step": 3086
+    },
+    {
+      "epoch": 0.18971238938053098,
+      "grad_norm": 26.625,
+      "learning_rate": 1.9326600126877566e-05,
+      "loss": 0.4493,
+      "step": 3087
+    },
+    {
+      "epoch": 0.18977384464110128,
+      "grad_norm": 29.125,
+      "learning_rate": 1.932576035569519e-05,
+      "loss": 0.3125,
+      "step": 3088
+    },
+    {
+      "epoch": 0.18983529990167158,
+      "grad_norm": 33.0,
+      "learning_rate": 1.9324920079483647e-05,
+      "loss": 0.4074,
+      "step": 3089
+    },
+    {
+      "epoch": 0.1898967551622419,
+      "grad_norm": 4.53125,
+      "learning_rate": 1.932407929828845e-05,
+      "loss": 0.3623,
+      "step": 3090
+    },
+    {
+      "epoch": 0.1899582104228122,
+      "grad_norm": 16.75,
+      "learning_rate": 1.9323238012155125e-05,
+      "loss": 0.5931,
+      "step": 3091
+    },
+    {
+      "epoch": 0.1900196656833825,
+      "grad_norm": 23.25,
+      "learning_rate": 1.932239622112923e-05,
+      "loss": 0.4467,
+      "step": 3092
+    },
+    {
+      "epoch": 0.1900811209439528,
+      "grad_norm": 20.25,
+      "learning_rate": 1.9321553925256355e-05,
+      "loss": 0.3878,
+      "step": 3093
+    },
+    {
+      "epoch": 0.1901425762045231,
+      "grad_norm": 10.6875,
+      "learning_rate": 1.932071112458211e-05,
+      "loss": 0.2578,
+      "step": 3094
+    },
+    {
+      "epoch": 0.1902040314650934,
+      "grad_norm": 24.25,
+      "learning_rate": 1.9319867819152144e-05,
+      "loss": 0.5495,
+      "step": 3095
+    },
+    {
+      "epoch": 0.1902654867256637,
+      "grad_norm": 37.0,
+      "learning_rate": 1.9319024009012114e-05,
+      "loss": 0.4262,
+      "step": 3096
+    },
+    {
+      "epoch": 0.190326941986234,
+      "grad_norm": 25.0,
+      "learning_rate": 1.9318179694207726e-05,
+      "loss": 0.3919,
+      "step": 3097
+    },
+    {
+      "epoch": 0.19038839724680431,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.93173348747847e-05,
+      "loss": 0.5086,
+      "step": 3098
+    },
+    {
+      "epoch": 0.19044985250737465,
+      "grad_norm": 11.5625,
+      "learning_rate": 1.9316489550788785e-05,
+      "loss": 0.4526,
+      "step": 3099
+    },
+    {
+      "epoch": 0.19051130776794495,
+      "grad_norm": 32.0,
+      "learning_rate": 1.931564372226576e-05,
+      "loss": 0.5635,
+      "step": 3100
+    },
+    {
+      "epoch": 0.19057276302851525,
+      "grad_norm": 33.75,
+      "learning_rate": 1.9314797389261426e-05,
+      "loss": 0.4972,
+      "step": 3101
+    },
+    {
+      "epoch": 0.19063421828908556,
+      "grad_norm": 19.5,
+      "learning_rate": 1.931395055182162e-05,
+      "loss": 0.3919,
+      "step": 3102
+    },
+    {
+      "epoch": 0.19069567354965586,
+      "grad_norm": 16.25,
+      "learning_rate": 1.9313103209992205e-05,
+      "loss": 0.3942,
+      "step": 3103
+    },
+    {
+      "epoch": 0.19075712881022616,
+      "grad_norm": 28.25,
+      "learning_rate": 1.931225536381906e-05,
+      "loss": 0.3897,
+      "step": 3104
+    },
+    {
+      "epoch": 0.19081858407079647,
+      "grad_norm": 21.25,
+      "learning_rate": 1.931140701334811e-05,
+      "loss": 0.3934,
+      "step": 3105
+    },
+    {
+      "epoch": 0.19088003933136677,
+      "grad_norm": 19.5,
+      "learning_rate": 1.9310558158625286e-05,
+      "loss": 0.475,
+      "step": 3106
+    },
+    {
+      "epoch": 0.19094149459193707,
+      "grad_norm": 22.375,
+      "learning_rate": 1.930970879969656e-05,
+      "loss": 0.5783,
+      "step": 3107
+    },
+    {
+      "epoch": 0.19100294985250738,
+      "grad_norm": 36.75,
+      "learning_rate": 1.9308858936607934e-05,
+      "loss": 0.5278,
+      "step": 3108
+    },
+    {
+      "epoch": 0.19106440511307768,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9308008569405424e-05,
+      "loss": 0.3673,
+      "step": 3109
+    },
+    {
+      "epoch": 0.19112586037364798,
+      "grad_norm": 29.625,
+      "learning_rate": 1.9307157698135085e-05,
+      "loss": 0.3967,
+      "step": 3110
+    },
+    {
+      "epoch": 0.19118731563421829,
+      "grad_norm": 16.0,
+      "learning_rate": 1.9306306322842994e-05,
+      "loss": 0.4106,
+      "step": 3111
+    },
+    {
+      "epoch": 0.1912487708947886,
+      "grad_norm": 16.0,
+      "learning_rate": 1.930545444357526e-05,
+      "loss": 0.4621,
+      "step": 3112
+    },
+    {
+      "epoch": 0.1913102261553589,
+      "grad_norm": 8.25,
+      "learning_rate": 1.9304602060378013e-05,
+      "loss": 0.4185,
+      "step": 3113
+    },
+    {
+      "epoch": 0.1913716814159292,
+      "grad_norm": 18.125,
+      "learning_rate": 1.9303749173297412e-05,
+      "loss": 0.4325,
+      "step": 3114
+    },
+    {
+      "epoch": 0.1914331366764995,
+      "grad_norm": 30.0,
+      "learning_rate": 1.9302895782379648e-05,
+      "loss": 0.5012,
+      "step": 3115
+    },
+    {
+      "epoch": 0.1914945919370698,
+      "grad_norm": 19.5,
+      "learning_rate": 1.930204188767093e-05,
+      "loss": 0.3983,
+      "step": 3116
+    },
+    {
+      "epoch": 0.1915560471976401,
+      "grad_norm": 31.125,
+      "learning_rate": 1.9301187489217508e-05,
+      "loss": 0.4154,
+      "step": 3117
+    },
+    {
+      "epoch": 0.1916175024582104,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9300332587065644e-05,
+      "loss": 0.3312,
+      "step": 3118
+    },
+    {
+      "epoch": 0.19167895771878074,
+      "grad_norm": 10.875,
+      "learning_rate": 1.9299477181261633e-05,
+      "loss": 0.3302,
+      "step": 3119
+    },
+    {
+      "epoch": 0.19174041297935104,
+      "grad_norm": 12.25,
+      "learning_rate": 1.929862127185181e-05,
+      "loss": 0.3188,
+      "step": 3120
+    },
+    {
+      "epoch": 0.19180186823992135,
+      "grad_norm": 19.375,
+      "learning_rate": 1.9297764858882516e-05,
+      "loss": 0.5155,
+      "step": 3121
+    },
+    {
+      "epoch": 0.19186332350049165,
+      "grad_norm": 17.125,
+      "learning_rate": 1.9296907942400133e-05,
+      "loss": 0.3994,
+      "step": 3122
+    },
+    {
+      "epoch": 0.19192477876106195,
+      "grad_norm": 10.3125,
+      "learning_rate": 1.9296050522451067e-05,
+      "loss": 0.3074,
+      "step": 3123
+    },
+    {
+      "epoch": 0.19198623402163226,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.9295192599081747e-05,
+      "loss": 0.3895,
+      "step": 3124
+    },
+    {
+      "epoch": 0.19204768928220256,
+      "grad_norm": 25.25,
+      "learning_rate": 1.9294334172338637e-05,
+      "loss": 0.4576,
+      "step": 3125
+    },
+    {
+      "epoch": 0.19210914454277286,
+      "grad_norm": 17.0,
+      "learning_rate": 1.9293475242268224e-05,
+      "loss": 0.2665,
+      "step": 3126
+    },
+    {
+      "epoch": 0.19217059980334317,
+      "grad_norm": 11.5625,
+      "learning_rate": 1.9292615808917027e-05,
+      "loss": 0.4694,
+      "step": 3127
+    },
+    {
+      "epoch": 0.19223205506391347,
+      "grad_norm": 15.125,
+      "learning_rate": 1.9291755872331576e-05,
+      "loss": 0.3566,
+      "step": 3128
+    },
+    {
+      "epoch": 0.19229351032448377,
+      "grad_norm": 30.0,
+      "learning_rate": 1.929089543255845e-05,
+      "loss": 0.4633,
+      "step": 3129
+    },
+    {
+      "epoch": 0.19235496558505408,
+      "grad_norm": 22.5,
+      "learning_rate": 1.9290034489644247e-05,
+      "loss": 0.5858,
+      "step": 3130
+    },
+    {
+      "epoch": 0.19241642084562438,
+      "grad_norm": 20.0,
+      "learning_rate": 1.9289173043635584e-05,
+      "loss": 0.4043,
+      "step": 3131
+    },
+    {
+      "epoch": 0.19247787610619468,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.9288311094579112e-05,
+      "loss": 0.4139,
+      "step": 3132
+    },
+    {
+      "epoch": 0.192539331366765,
+      "grad_norm": 12.0,
+      "learning_rate": 1.9287448642521513e-05,
+      "loss": 0.3883,
+      "step": 3133
+    },
+    {
+      "epoch": 0.1926007866273353,
+      "grad_norm": 4.0,
+      "learning_rate": 1.9286585687509492e-05,
+      "loss": 0.4217,
+      "step": 3134
+    },
+    {
+      "epoch": 0.1926622418879056,
+      "grad_norm": 17.875,
+      "learning_rate": 1.9285722229589782e-05,
+      "loss": 0.2882,
+      "step": 3135
+    },
+    {
+      "epoch": 0.1927236971484759,
+      "grad_norm": 38.5,
+      "learning_rate": 1.9284858268809135e-05,
+      "loss": 0.4638,
+      "step": 3136
+    },
+    {
+      "epoch": 0.1927851524090462,
+      "grad_norm": 27.125,
+      "learning_rate": 1.9283993805214353e-05,
+      "loss": 0.6033,
+      "step": 3137
+    },
+    {
+      "epoch": 0.19284660766961653,
+      "grad_norm": 49.25,
+      "learning_rate": 1.9283128838852237e-05,
+      "loss": 0.5463,
+      "step": 3138
+    },
+    {
+      "epoch": 0.19290806293018684,
+      "grad_norm": 23.5,
+      "learning_rate": 1.9282263369769633e-05,
+      "loss": 0.5622,
+      "step": 3139
+    },
+    {
+      "epoch": 0.19296951819075714,
+      "grad_norm": 20.0,
+      "learning_rate": 1.9281397398013416e-05,
+      "loss": 0.5249,
+      "step": 3140
+    },
+    {
+      "epoch": 0.19303097345132744,
+      "grad_norm": 50.75,
+      "learning_rate": 1.928053092363047e-05,
+      "loss": 0.4112,
+      "step": 3141
+    },
+    {
+      "epoch": 0.19309242871189775,
+      "grad_norm": 36.25,
+      "learning_rate": 1.927966394666773e-05,
+      "loss": 0.4306,
+      "step": 3142
+    },
+    {
+      "epoch": 0.19315388397246805,
+      "grad_norm": 27.125,
+      "learning_rate": 1.9278796467172137e-05,
+      "loss": 0.3881,
+      "step": 3143
+    },
+    {
+      "epoch": 0.19321533923303835,
+      "grad_norm": 32.75,
+      "learning_rate": 1.9277928485190674e-05,
+      "loss": 0.3317,
+      "step": 3144
+    },
+    {
+      "epoch": 0.19327679449360866,
+      "grad_norm": 36.75,
+      "learning_rate": 1.9277060000770342e-05,
+      "loss": 0.7708,
+      "step": 3145
+    },
+    {
+      "epoch": 0.19333824975417896,
+      "grad_norm": 31.25,
+      "learning_rate": 1.927619101395818e-05,
+      "loss": 0.595,
+      "step": 3146
+    },
+    {
+      "epoch": 0.19339970501474926,
+      "grad_norm": 63.0,
+      "learning_rate": 1.927532152480124e-05,
+      "loss": 0.7295,
+      "step": 3147
+    },
+    {
+      "epoch": 0.19346116027531957,
+      "grad_norm": 35.5,
+      "learning_rate": 1.9274451533346617e-05,
+      "loss": 0.6259,
+      "step": 3148
+    },
+    {
+      "epoch": 0.19352261553588987,
+      "grad_norm": 30.375,
+      "learning_rate": 1.9273581039641412e-05,
+      "loss": 0.4116,
+      "step": 3149
+    },
+    {
+      "epoch": 0.19358407079646017,
+      "grad_norm": 15.25,
+      "learning_rate": 1.9272710043732775e-05,
+      "loss": 0.3198,
+      "step": 3150
+    },
+    {
+      "epoch": 0.19364552605703048,
+      "grad_norm": 16.25,
+      "learning_rate": 1.9271838545667876e-05,
+      "loss": 0.3557,
+      "step": 3151
+    },
+    {
+      "epoch": 0.19370698131760078,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.9270966545493903e-05,
+      "loss": 0.4748,
+      "step": 3152
+    },
+    {
+      "epoch": 0.19376843657817108,
+      "grad_norm": 24.0,
+      "learning_rate": 1.927009404325808e-05,
+      "loss": 0.5944,
+      "step": 3153
+    },
+    {
+      "epoch": 0.19382989183874139,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9269221039007666e-05,
+      "loss": 0.6175,
+      "step": 3154
+    },
+    {
+      "epoch": 0.1938913470993117,
+      "grad_norm": 23.125,
+      "learning_rate": 1.9268347532789926e-05,
+      "loss": 0.4974,
+      "step": 3155
+    },
+    {
+      "epoch": 0.193952802359882,
+      "grad_norm": 26.25,
+      "learning_rate": 1.9267473524652168e-05,
+      "loss": 0.4146,
+      "step": 3156
+    },
+    {
+      "epoch": 0.19401425762045232,
+      "grad_norm": 29.25,
+      "learning_rate": 1.9266599014641724e-05,
+      "loss": 0.4516,
+      "step": 3157
+    },
+    {
+      "epoch": 0.19407571288102263,
+      "grad_norm": 44.25,
+      "learning_rate": 1.9265724002805952e-05,
+      "loss": 0.4985,
+      "step": 3158
+    },
+    {
+      "epoch": 0.19413716814159293,
+      "grad_norm": 5.0625,
+      "learning_rate": 1.9264848489192237e-05,
+      "loss": 0.3678,
+      "step": 3159
+    },
+    {
+      "epoch": 0.19419862340216323,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.9263972473847995e-05,
+      "loss": 0.3074,
+      "step": 3160
+    },
+    {
+      "epoch": 0.19426007866273354,
+      "grad_norm": 29.125,
+      "learning_rate": 1.926309595682066e-05,
+      "loss": 0.5712,
+      "step": 3161
+    },
+    {
+      "epoch": 0.19432153392330384,
+      "grad_norm": 26.625,
+      "learning_rate": 1.9262218938157703e-05,
+      "loss": 0.4782,
+      "step": 3162
+    },
+    {
+      "epoch": 0.19438298918387414,
+      "grad_norm": 33.0,
+      "learning_rate": 1.9261341417906622e-05,
+      "loss": 0.4408,
+      "step": 3163
+    },
+    {
+      "epoch": 0.19444444444444445,
+      "grad_norm": 27.625,
+      "learning_rate": 1.926046339611493e-05,
+      "loss": 0.3974,
+      "step": 3164
+    },
+    {
+      "epoch": 0.19450589970501475,
+      "grad_norm": 26.375,
+      "learning_rate": 1.925958487283018e-05,
+      "loss": 0.4634,
+      "step": 3165
+    },
+    {
+      "epoch": 0.19456735496558505,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.925870584809995e-05,
+      "loss": 0.3452,
+      "step": 3166
+    },
+    {
+      "epoch": 0.19462881022615536,
+      "grad_norm": 13.625,
+      "learning_rate": 1.9257826321971838e-05,
+      "loss": 0.3272,
+      "step": 3167
+    },
+    {
+      "epoch": 0.19469026548672566,
+      "grad_norm": 23.375,
+      "learning_rate": 1.9256946294493475e-05,
+      "loss": 0.266,
+      "step": 3168
+    },
+    {
+      "epoch": 0.19475172074729596,
+      "grad_norm": 34.75,
+      "learning_rate": 1.9256065765712524e-05,
+      "loss": 0.5639,
+      "step": 3169
+    },
+    {
+      "epoch": 0.19481317600786627,
+      "grad_norm": 29.75,
+      "learning_rate": 1.925518473567666e-05,
+      "loss": 0.7866,
+      "step": 3170
+    },
+    {
+      "epoch": 0.19487463126843657,
+      "grad_norm": 31.625,
+      "learning_rate": 1.9254303204433602e-05,
+      "loss": 0.5783,
+      "step": 3171
+    },
+    {
+      "epoch": 0.19493608652900687,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.9253421172031086e-05,
+      "loss": 0.5233,
+      "step": 3172
+    },
+    {
+      "epoch": 0.19499754178957718,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9252538638516878e-05,
+      "loss": 0.4656,
+      "step": 3173
+    },
+    {
+      "epoch": 0.19505899705014748,
+      "grad_norm": 19.875,
+      "learning_rate": 1.925165560393877e-05,
+      "loss": 0.5181,
+      "step": 3174
+    },
+    {
+      "epoch": 0.19512045231071778,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.925077206834458e-05,
+      "loss": 0.3447,
+      "step": 3175
+    },
+    {
+      "epoch": 0.19518190757128812,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.924988803178216e-05,
+      "loss": 0.3941,
+      "step": 3176
+    },
+    {
+      "epoch": 0.19524336283185842,
+      "grad_norm": 25.375,
+      "learning_rate": 1.9249003494299385e-05,
+      "loss": 0.7266,
+      "step": 3177
+    },
+    {
+      "epoch": 0.19530481809242872,
+      "grad_norm": 25.25,
+      "learning_rate": 1.9248118455944153e-05,
+      "loss": 0.5372,
+      "step": 3178
+    },
+    {
+      "epoch": 0.19536627335299903,
+      "grad_norm": 11.4375,
+      "learning_rate": 1.9247232916764388e-05,
+      "loss": 0.4471,
+      "step": 3179
+    },
+    {
+      "epoch": 0.19542772861356933,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9246346876808054e-05,
+      "loss": 0.4106,
+      "step": 3180
+    },
+    {
+      "epoch": 0.19548918387413963,
+      "grad_norm": 13.125,
+      "learning_rate": 1.9245460336123136e-05,
+      "loss": 0.6621,
+      "step": 3181
+    },
+    {
+      "epoch": 0.19555063913470994,
+      "grad_norm": 16.875,
+      "learning_rate": 1.924457329475763e-05,
+      "loss": 0.4883,
+      "step": 3182
+    },
+    {
+      "epoch": 0.19561209439528024,
+      "grad_norm": 11.5,
+      "learning_rate": 1.9243685752759585e-05,
+      "loss": 0.5566,
+      "step": 3183
+    },
+    {
+      "epoch": 0.19567354965585054,
+      "grad_norm": 20.875,
+      "learning_rate": 1.924279771017706e-05,
+      "loss": 0.541,
+      "step": 3184
+    },
+    {
+      "epoch": 0.19573500491642085,
+      "grad_norm": 20.125,
+      "learning_rate": 1.9241909167058154e-05,
+      "loss": 0.58,
+      "step": 3185
+    },
+    {
+      "epoch": 0.19579646017699115,
+      "grad_norm": 27.875,
+      "learning_rate": 1.9241020123450972e-05,
+      "loss": 0.569,
+      "step": 3186
+    },
+    {
+      "epoch": 0.19585791543756145,
+      "grad_norm": 16.375,
+      "learning_rate": 1.924013057940367e-05,
+      "loss": 0.4874,
+      "step": 3187
+    },
+    {
+      "epoch": 0.19591937069813176,
+      "grad_norm": 21.75,
+      "learning_rate": 1.9239240534964417e-05,
+      "loss": 0.4432,
+      "step": 3188
+    },
+    {
+      "epoch": 0.19598082595870206,
+      "grad_norm": 12.25,
+      "learning_rate": 1.923834999018141e-05,
+      "loss": 0.5229,
+      "step": 3189
+    },
+    {
+      "epoch": 0.19604228121927236,
+      "grad_norm": 45.25,
+      "learning_rate": 1.923745894510288e-05,
+      "loss": 0.493,
+      "step": 3190
+    },
+    {
+      "epoch": 0.19610373647984267,
+      "grad_norm": 8.5,
+      "learning_rate": 1.9236567399777086e-05,
+      "loss": 0.3939,
+      "step": 3191
+    },
+    {
+      "epoch": 0.19616519174041297,
+      "grad_norm": 8.5,
+      "learning_rate": 1.9235675354252293e-05,
+      "loss": 0.3573,
+      "step": 3192
+    },
+    {
+      "epoch": 0.19622664700098327,
+      "grad_norm": 49.75,
+      "learning_rate": 1.9234782808576823e-05,
+      "loss": 0.442,
+      "step": 3193
+    },
+    {
+      "epoch": 0.19628810226155358,
+      "grad_norm": 48.5,
+      "learning_rate": 1.9233889762799006e-05,
+      "loss": 0.4996,
+      "step": 3194
+    },
+    {
+      "epoch": 0.1963495575221239,
+      "grad_norm": 6.21875,
+      "learning_rate": 1.9232996216967204e-05,
+      "loss": 0.2889,
+      "step": 3195
+    },
+    {
+      "epoch": 0.1964110127826942,
+      "grad_norm": 37.5,
+      "learning_rate": 1.923210217112981e-05,
+      "loss": 0.3482,
+      "step": 3196
+    },
+    {
+      "epoch": 0.19647246804326451,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.9231207625335235e-05,
+      "loss": 0.4698,
+      "step": 3197
+    },
+    {
+      "epoch": 0.19653392330383482,
+      "grad_norm": 4.375,
+      "learning_rate": 1.9230312579631928e-05,
+      "loss": 0.2884,
+      "step": 3198
+    },
+    {
+      "epoch": 0.19659537856440512,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.9229417034068352e-05,
+      "loss": 0.226,
+      "step": 3199
+    },
+    {
+      "epoch": 0.19665683382497542,
+      "grad_norm": 5.34375,
+      "learning_rate": 1.922852098869301e-05,
+      "loss": 0.3132,
+      "step": 3200
+    },
+    {
+      "epoch": 0.19671828908554573,
+      "grad_norm": 24.625,
+      "learning_rate": 1.9227624443554425e-05,
+      "loss": 0.4517,
+      "step": 3201
+    },
+    {
+      "epoch": 0.19677974434611603,
+      "grad_norm": 34.0,
+      "learning_rate": 1.922672739870115e-05,
+      "loss": 0.4096,
+      "step": 3202
+    },
+    {
+      "epoch": 0.19684119960668633,
+      "grad_norm": 10.3125,
+      "learning_rate": 1.9225829854181764e-05,
+      "loss": 0.353,
+      "step": 3203
+    },
+    {
+      "epoch": 0.19690265486725664,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.922493181004487e-05,
+      "loss": 0.2855,
+      "step": 3204
+    },
+    {
+      "epoch": 0.19696411012782694,
+      "grad_norm": 28.875,
+      "learning_rate": 1.9224033266339103e-05,
+      "loss": 0.3964,
+      "step": 3205
+    },
+    {
+      "epoch": 0.19702556538839724,
+      "grad_norm": 11.5,
+      "learning_rate": 1.9223134223113122e-05,
+      "loss": 0.3643,
+      "step": 3206
+    },
+    {
+      "epoch": 0.19708702064896755,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.9222234680415614e-05,
+      "loss": 0.2807,
+      "step": 3207
+    },
+    {
+      "epoch": 0.19714847590953785,
+      "grad_norm": 10.25,
+      "learning_rate": 1.9221334638295296e-05,
+      "loss": 0.36,
+      "step": 3208
+    },
+    {
+      "epoch": 0.19720993117010815,
+      "grad_norm": 43.75,
+      "learning_rate": 1.9220434096800903e-05,
+      "loss": 0.73,
+      "step": 3209
+    },
+    {
+      "epoch": 0.19727138643067846,
+      "grad_norm": 51.5,
+      "learning_rate": 1.921953305598121e-05,
+      "loss": 0.4778,
+      "step": 3210
+    },
+    {
+      "epoch": 0.19733284169124876,
+      "grad_norm": 34.0,
+      "learning_rate": 1.9218631515885007e-05,
+      "loss": 0.564,
+      "step": 3211
+    },
+    {
+      "epoch": 0.19739429695181906,
+      "grad_norm": 14.125,
+      "learning_rate": 1.921772947656112e-05,
+      "loss": 0.4783,
+      "step": 3212
+    },
+    {
+      "epoch": 0.19745575221238937,
+      "grad_norm": 9.75,
+      "learning_rate": 1.9216826938058394e-05,
+      "loss": 0.4605,
+      "step": 3213
+    },
+    {
+      "epoch": 0.19751720747295967,
+      "grad_norm": 18.875,
+      "learning_rate": 1.921592390042571e-05,
+      "loss": 0.4696,
+      "step": 3214
+    },
+    {
+      "epoch": 0.19757866273353,
+      "grad_norm": 55.75,
+      "learning_rate": 1.9215020363711966e-05,
+      "loss": 0.6129,
+      "step": 3215
+    },
+    {
+      "epoch": 0.1976401179941003,
+      "grad_norm": 25.25,
+      "learning_rate": 1.9214116327966095e-05,
+      "loss": 0.4371,
+      "step": 3216
+    },
+    {
+      "epoch": 0.1977015732546706,
+      "grad_norm": 14.125,
+      "learning_rate": 1.9213211793237056e-05,
+      "loss": 0.2968,
+      "step": 3217
+    },
+    {
+      "epoch": 0.1977630285152409,
+      "grad_norm": 16.875,
+      "learning_rate": 1.921230675957383e-05,
+      "loss": 0.4517,
+      "step": 3218
+    },
+    {
+      "epoch": 0.19782448377581122,
+      "grad_norm": 13.75,
+      "learning_rate": 1.9211401227025433e-05,
+      "loss": 0.2866,
+      "step": 3219
+    },
+    {
+      "epoch": 0.19788593903638152,
+      "grad_norm": 17.75,
+      "learning_rate": 1.9210495195640895e-05,
+      "loss": 0.5418,
+      "step": 3220
+    },
+    {
+      "epoch": 0.19794739429695182,
+      "grad_norm": 10.6875,
+      "learning_rate": 1.9209588665469294e-05,
+      "loss": 0.4282,
+      "step": 3221
+    },
+    {
+      "epoch": 0.19800884955752213,
+      "grad_norm": 26.375,
+      "learning_rate": 1.9208681636559713e-05,
+      "loss": 0.3204,
+      "step": 3222
+    },
+    {
+      "epoch": 0.19807030481809243,
+      "grad_norm": 31.0,
+      "learning_rate": 1.9207774108961273e-05,
+      "loss": 0.4528,
+      "step": 3223
+    },
+    {
+      "epoch": 0.19813176007866273,
+      "grad_norm": 5.53125,
+      "learning_rate": 1.9206866082723122e-05,
+      "loss": 0.3949,
+      "step": 3224
+    },
+    {
+      "epoch": 0.19819321533923304,
+      "grad_norm": 14.875,
+      "learning_rate": 1.9205957557894433e-05,
+      "loss": 0.352,
+      "step": 3225
+    },
+    {
+      "epoch": 0.19825467059980334,
+      "grad_norm": 6.53125,
+      "learning_rate": 1.9205048534524405e-05,
+      "loss": 0.2267,
+      "step": 3226
+    },
+    {
+      "epoch": 0.19831612586037364,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9204139012662266e-05,
+      "loss": 0.3691,
+      "step": 3227
+    },
+    {
+      "epoch": 0.19837758112094395,
+      "grad_norm": 20.5,
+      "learning_rate": 1.9203228992357275e-05,
+      "loss": 0.3678,
+      "step": 3228
+    },
+    {
+      "epoch": 0.19843903638151425,
+      "grad_norm": 18.625,
+      "learning_rate": 1.9202318473658707e-05,
+      "loss": 0.39,
+      "step": 3229
+    },
+    {
+      "epoch": 0.19850049164208455,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.920140745661587e-05,
+      "loss": 0.2961,
+      "step": 3230
+    },
+    {
+      "epoch": 0.19856194690265486,
+      "grad_norm": 24.75,
+      "learning_rate": 1.9200495941278105e-05,
+      "loss": 0.3088,
+      "step": 3231
+    },
+    {
+      "epoch": 0.19862340216322516,
+      "grad_norm": 8.75,
+      "learning_rate": 1.9199583927694775e-05,
+      "loss": 0.2132,
+      "step": 3232
+    },
+    {
+      "epoch": 0.19868485742379546,
+      "grad_norm": 41.25,
+      "learning_rate": 1.919867141591526e-05,
+      "loss": 0.4536,
+      "step": 3233
+    },
+    {
+      "epoch": 0.1987463126843658,
+      "grad_norm": 25.625,
+      "learning_rate": 1.9197758405988986e-05,
+      "loss": 0.597,
+      "step": 3234
+    },
+    {
+      "epoch": 0.1988077679449361,
+      "grad_norm": 21.75,
+      "learning_rate": 1.9196844897965393e-05,
+      "loss": 0.4037,
+      "step": 3235
+    },
+    {
+      "epoch": 0.1988692232055064,
+      "grad_norm": 25.875,
+      "learning_rate": 1.9195930891893946e-05,
+      "loss": 0.3735,
+      "step": 3236
+    },
+    {
+      "epoch": 0.1989306784660767,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.9195016387824154e-05,
+      "loss": 0.4605,
+      "step": 3237
+    },
+    {
+      "epoch": 0.198992133726647,
+      "grad_norm": 30.5,
+      "learning_rate": 1.919410138580553e-05,
+      "loss": 0.3089,
+      "step": 3238
+    },
+    {
+      "epoch": 0.1990535889872173,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.9193185885887633e-05,
+      "loss": 0.3694,
+      "step": 3239
+    },
+    {
+      "epoch": 0.19911504424778761,
+      "grad_norm": 8.375,
+      "learning_rate": 1.9192269888120036e-05,
+      "loss": 0.3337,
+      "step": 3240
+    },
+    {
+      "epoch": 0.19917649950835792,
+      "grad_norm": 21.625,
+      "learning_rate": 1.9191353392552346e-05,
+      "loss": 0.6298,
+      "step": 3241
+    },
+    {
+      "epoch": 0.19923795476892822,
+      "grad_norm": 49.0,
+      "learning_rate": 1.9190436399234194e-05,
+      "loss": 0.826,
+      "step": 3242
+    },
+    {
+      "epoch": 0.19929941002949852,
+      "grad_norm": 40.75,
+      "learning_rate": 1.9189518908215244e-05,
+      "loss": 0.8753,
+      "step": 3243
+    },
+    {
+      "epoch": 0.19936086529006883,
+      "grad_norm": 43.0,
+      "learning_rate": 1.9188600919545176e-05,
+      "loss": 0.5461,
+      "step": 3244
+    },
+    {
+      "epoch": 0.19942232055063913,
+      "grad_norm": 14.5,
+      "learning_rate": 1.9187682433273704e-05,
+      "loss": 0.5653,
+      "step": 3245
+    },
+    {
+      "epoch": 0.19948377581120944,
+      "grad_norm": 38.25,
+      "learning_rate": 1.9186763449450572e-05,
+      "loss": 0.5067,
+      "step": 3246
+    },
+    {
+      "epoch": 0.19954523107177974,
+      "grad_norm": 19.75,
+      "learning_rate": 1.9185843968125543e-05,
+      "loss": 0.515,
+      "step": 3247
+    },
+    {
+      "epoch": 0.19960668633235004,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.918492398934841e-05,
+      "loss": 0.6222,
+      "step": 3248
+    },
+    {
+      "epoch": 0.19966814159292035,
+      "grad_norm": 35.0,
+      "learning_rate": 1.9184003513168995e-05,
+      "loss": 0.58,
+      "step": 3249
+    },
+    {
+      "epoch": 0.19972959685349065,
+      "grad_norm": 29.625,
+      "learning_rate": 1.918308253963715e-05,
+      "loss": 0.4658,
+      "step": 3250
+    },
+    {
+      "epoch": 0.19979105211406095,
+      "grad_norm": 25.875,
+      "learning_rate": 1.9182161068802742e-05,
+      "loss": 0.4651,
+      "step": 3251
+    },
+    {
+      "epoch": 0.19985250737463126,
+      "grad_norm": 23.125,
+      "learning_rate": 1.9181239100715675e-05,
+      "loss": 0.4433,
+      "step": 3252
+    },
+    {
+      "epoch": 0.1999139626352016,
+      "grad_norm": 32.5,
+      "learning_rate": 1.9180316635425883e-05,
+      "loss": 0.3804,
+      "step": 3253
+    },
+    {
+      "epoch": 0.1999754178957719,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.9179393672983315e-05,
+      "loss": 0.3226,
+      "step": 3254
+    },
+    {
+      "epoch": 0.2000368731563422,
+      "grad_norm": 21.625,
+      "learning_rate": 1.9178470213437955e-05,
+      "loss": 0.3348,
+      "step": 3255
+    },
+    {
+      "epoch": 0.2000983284169125,
+      "grad_norm": 10.5,
+      "learning_rate": 1.9177546256839814e-05,
+      "loss": 0.4113,
+      "step": 3256
+    },
+    {
+      "epoch": 0.2001597836774828,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.9176621803238924e-05,
+      "loss": 0.5429,
+      "step": 3257
+    },
+    {
+      "epoch": 0.2002212389380531,
+      "grad_norm": 18.875,
+      "learning_rate": 1.9175696852685353e-05,
+      "loss": 0.5471,
+      "step": 3258
+    },
+    {
+      "epoch": 0.2002826941986234,
+      "grad_norm": 25.875,
+      "learning_rate": 1.9174771405229187e-05,
+      "loss": 0.5805,
+      "step": 3259
+    },
+    {
+      "epoch": 0.2003441494591937,
+      "grad_norm": 27.0,
+      "learning_rate": 1.9173845460920546e-05,
+      "loss": 0.5394,
+      "step": 3260
+    },
+    {
+      "epoch": 0.200405604719764,
+      "grad_norm": 14.5,
+      "learning_rate": 1.9172919019809572e-05,
+      "loss": 0.4442,
+      "step": 3261
+    },
+    {
+      "epoch": 0.20046705998033432,
+      "grad_norm": 30.625,
+      "learning_rate": 1.9171992081946436e-05,
+      "loss": 0.3667,
+      "step": 3262
+    },
+    {
+      "epoch": 0.20052851524090462,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.9171064647381334e-05,
+      "loss": 0.44,
+      "step": 3263
+    },
+    {
+      "epoch": 0.20058997050147492,
+      "grad_norm": 7.875,
+      "learning_rate": 1.9170136716164497e-05,
+      "loss": 0.351,
+      "step": 3264
+    },
+    {
+      "epoch": 0.20065142576204523,
+      "grad_norm": 48.75,
+      "learning_rate": 1.9169208288346168e-05,
+      "loss": 0.5977,
+      "step": 3265
+    },
+    {
+      "epoch": 0.20071288102261553,
+      "grad_norm": 16.625,
+      "learning_rate": 1.9168279363976627e-05,
+      "loss": 0.5756,
+      "step": 3266
+    },
+    {
+      "epoch": 0.20077433628318583,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.916734994310618e-05,
+      "loss": 0.5098,
+      "step": 3267
+    },
+    {
+      "epoch": 0.20083579154375614,
+      "grad_norm": 33.25,
+      "learning_rate": 1.9166420025785165e-05,
+      "loss": 0.4842,
+      "step": 3268
+    },
+    {
+      "epoch": 0.20089724680432644,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9165489612063932e-05,
+      "loss": 0.5166,
+      "step": 3269
+    },
+    {
+      "epoch": 0.20095870206489674,
+      "grad_norm": 19.125,
+      "learning_rate": 1.9164558701992875e-05,
+      "loss": 0.4537,
+      "step": 3270
+    },
+    {
+      "epoch": 0.20102015732546705,
+      "grad_norm": 38.25,
+      "learning_rate": 1.9163627295622397e-05,
+      "loss": 0.5321,
+      "step": 3271
+    },
+    {
+      "epoch": 0.20108161258603738,
+      "grad_norm": 13.25,
+      "learning_rate": 1.9162695393002947e-05,
+      "loss": 0.4797,
+      "step": 3272
+    },
+    {
+      "epoch": 0.20114306784660768,
+      "grad_norm": 20.125,
+      "learning_rate": 1.9161762994184986e-05,
+      "loss": 0.5974,
+      "step": 3273
+    },
+    {
+      "epoch": 0.20120452310717798,
+      "grad_norm": 9.75,
+      "learning_rate": 1.9160830099219007e-05,
+      "loss": 0.5726,
+      "step": 3274
+    },
+    {
+      "epoch": 0.2012659783677483,
+      "grad_norm": 27.0,
+      "learning_rate": 1.9159896708155534e-05,
+      "loss": 0.6421,
+      "step": 3275
+    },
+    {
+      "epoch": 0.2013274336283186,
+      "grad_norm": 33.25,
+      "learning_rate": 1.9158962821045113e-05,
+      "loss": 0.6594,
+      "step": 3276
+    },
+    {
+      "epoch": 0.2013888888888889,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.9158028437938316e-05,
+      "loss": 0.4718,
+      "step": 3277
+    },
+    {
+      "epoch": 0.2014503441494592,
+      "grad_norm": 21.25,
+      "learning_rate": 1.9157093558885743e-05,
+      "loss": 0.5557,
+      "step": 3278
+    },
+    {
+      "epoch": 0.2015117994100295,
+      "grad_norm": 20.5,
+      "learning_rate": 1.9156158183938026e-05,
+      "loss": 0.5225,
+      "step": 3279
+    },
+    {
+      "epoch": 0.2015732546705998,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.9155222313145817e-05,
+      "loss": 0.4436,
+      "step": 3280
+    },
+    {
+      "epoch": 0.2016347099311701,
+      "grad_norm": 10.75,
+      "learning_rate": 1.9154285946559792e-05,
+      "loss": 0.5134,
+      "step": 3281
+    },
+    {
+      "epoch": 0.2016961651917404,
+      "grad_norm": 44.75,
+      "learning_rate": 1.9153349084230666e-05,
+      "loss": 0.6567,
+      "step": 3282
+    },
+    {
+      "epoch": 0.20175762045231072,
+      "grad_norm": 21.5,
+      "learning_rate": 1.9152411726209176e-05,
+      "loss": 0.589,
+      "step": 3283
+    },
+    {
+      "epoch": 0.20181907571288102,
+      "grad_norm": 23.375,
+      "learning_rate": 1.9151473872546082e-05,
+      "loss": 0.4641,
+      "step": 3284
+    },
+    {
+      "epoch": 0.20188053097345132,
+      "grad_norm": 24.875,
+      "learning_rate": 1.9150535523292165e-05,
+      "loss": 0.4986,
+      "step": 3285
+    },
+    {
+      "epoch": 0.20194198623402163,
+      "grad_norm": 24.625,
+      "learning_rate": 1.914959667849825e-05,
+      "loss": 0.64,
+      "step": 3286
+    },
+    {
+      "epoch": 0.20200344149459193,
+      "grad_norm": 41.25,
+      "learning_rate": 1.9148657338215175e-05,
+      "loss": 0.6606,
+      "step": 3287
+    },
+    {
+      "epoch": 0.20206489675516223,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.914771750249381e-05,
+      "loss": 0.3861,
+      "step": 3288
+    },
+    {
+      "epoch": 0.20212635201573254,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.914677717138505e-05,
+      "loss": 0.4235,
+      "step": 3289
+    },
+    {
+      "epoch": 0.20218780727630284,
+      "grad_norm": 13.5,
+      "learning_rate": 1.9145836344939823e-05,
+      "loss": 0.3992,
+      "step": 3290
+    },
+    {
+      "epoch": 0.20224926253687317,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.9144895023209072e-05,
+      "loss": 0.3946,
+      "step": 3291
+    },
+    {
+      "epoch": 0.20231071779744347,
+      "grad_norm": 36.5,
+      "learning_rate": 1.9143953206243778e-05,
+      "loss": 0.468,
+      "step": 3292
+    },
+    {
+      "epoch": 0.20237217305801378,
+      "grad_norm": 16.125,
+      "learning_rate": 1.9143010894094938e-05,
+      "loss": 0.394,
+      "step": 3293
+    },
+    {
+      "epoch": 0.20243362831858408,
+      "grad_norm": 20.75,
+      "learning_rate": 1.9142068086813593e-05,
+      "loss": 0.3944,
+      "step": 3294
+    },
+    {
+      "epoch": 0.20249508357915438,
+      "grad_norm": 9.5,
+      "learning_rate": 1.914112478445079e-05,
+      "loss": 0.3043,
+      "step": 3295
+    },
+    {
+      "epoch": 0.2025565388397247,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.914018098705762e-05,
+      "loss": 0.3684,
+      "step": 3296
+    },
+    {
+      "epoch": 0.202617994100295,
+      "grad_norm": 45.75,
+      "learning_rate": 1.9139236694685184e-05,
+      "loss": 0.7134,
+      "step": 3297
+    },
+    {
+      "epoch": 0.2026794493608653,
+      "grad_norm": 50.5,
+      "learning_rate": 1.9138291907384632e-05,
+      "loss": 0.7402,
+      "step": 3298
+    },
+    {
+      "epoch": 0.2027409046214356,
+      "grad_norm": 33.25,
+      "learning_rate": 1.913734662520712e-05,
+      "loss": 0.6951,
+      "step": 3299
+    },
+    {
+      "epoch": 0.2028023598820059,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9136400848203837e-05,
+      "loss": 0.6361,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2028638151425762,
+      "grad_norm": 42.25,
+      "learning_rate": 1.913545457642601e-05,
+      "loss": 0.5913,
+      "step": 3301
+    },
+    {
+      "epoch": 0.2029252704031465,
+      "grad_norm": 29.5,
+      "learning_rate": 1.9134507809924878e-05,
+      "loss": 0.6491,
+      "step": 3302
+    },
+    {
+      "epoch": 0.2029867256637168,
+      "grad_norm": 22.0,
+      "learning_rate": 1.913356054875171e-05,
+      "loss": 0.5775,
+      "step": 3303
+    },
+    {
+      "epoch": 0.2030481809242871,
+      "grad_norm": 24.875,
+      "learning_rate": 1.9132612792957808e-05,
+      "loss": 0.5075,
+      "step": 3304
+    },
+    {
+      "epoch": 0.20310963618485742,
+      "grad_norm": 45.75,
+      "learning_rate": 1.9131664542594496e-05,
+      "loss": 0.673,
+      "step": 3305
+    },
+    {
+      "epoch": 0.20317109144542772,
+      "grad_norm": 30.625,
+      "learning_rate": 1.9130715797713123e-05,
+      "loss": 0.6118,
+      "step": 3306
+    },
+    {
+      "epoch": 0.20323254670599802,
+      "grad_norm": 25.0,
+      "learning_rate": 1.9129766558365076e-05,
+      "loss": 0.6173,
+      "step": 3307
+    },
+    {
+      "epoch": 0.20329400196656833,
+      "grad_norm": 16.375,
+      "learning_rate": 1.912881682460175e-05,
+      "loss": 0.4625,
+      "step": 3308
+    },
+    {
+      "epoch": 0.20335545722713863,
+      "grad_norm": 10.25,
+      "learning_rate": 1.912786659647458e-05,
+      "loss": 0.4507,
+      "step": 3309
+    },
+    {
+      "epoch": 0.20341691248770893,
+      "grad_norm": 42.75,
+      "learning_rate": 1.912691587403503e-05,
+      "loss": 0.5495,
+      "step": 3310
+    },
+    {
+      "epoch": 0.20347836774827927,
+      "grad_norm": 17.75,
+      "learning_rate": 1.912596465733458e-05,
+      "loss": 0.4086,
+      "step": 3311
+    },
+    {
+      "epoch": 0.20353982300884957,
+      "grad_norm": 6.1875,
+      "learning_rate": 1.9125012946424742e-05,
+      "loss": 0.4421,
+      "step": 3312
+    },
+    {
+      "epoch": 0.20360127826941987,
+      "grad_norm": 16.125,
+      "learning_rate": 1.9124060741357065e-05,
+      "loss": 0.4439,
+      "step": 3313
+    },
+    {
+      "epoch": 0.20366273352999018,
+      "grad_norm": 24.125,
+      "learning_rate": 1.91231080421831e-05,
+      "loss": 0.4542,
+      "step": 3314
+    },
+    {
+      "epoch": 0.20372418879056048,
+      "grad_norm": 17.0,
+      "learning_rate": 1.912215484895445e-05,
+      "loss": 0.5242,
+      "step": 3315
+    },
+    {
+      "epoch": 0.20378564405113078,
+      "grad_norm": 11.5625,
+      "learning_rate": 1.9121201161722732e-05,
+      "loss": 0.4114,
+      "step": 3316
+    },
+    {
+      "epoch": 0.20384709931170109,
+      "grad_norm": 17.75,
+      "learning_rate": 1.912024698053959e-05,
+      "loss": 0.2947,
+      "step": 3317
+    },
+    {
+      "epoch": 0.2039085545722714,
+      "grad_norm": 40.0,
+      "learning_rate": 1.91192923054567e-05,
+      "loss": 0.5266,
+      "step": 3318
+    },
+    {
+      "epoch": 0.2039700098328417,
+      "grad_norm": 22.75,
+      "learning_rate": 1.911833713652576e-05,
+      "loss": 0.4076,
+      "step": 3319
+    },
+    {
+      "epoch": 0.204031465093412,
+      "grad_norm": 35.5,
+      "learning_rate": 1.9117381473798496e-05,
+      "loss": 0.4072,
+      "step": 3320
+    },
+    {
+      "epoch": 0.2040929203539823,
+      "grad_norm": 24.875,
+      "learning_rate": 1.911642531732666e-05,
+      "loss": 0.404,
+      "step": 3321
+    },
+    {
+      "epoch": 0.2041543756145526,
+      "grad_norm": 5.15625,
+      "learning_rate": 1.9115468667162038e-05,
+      "loss": 0.4201,
+      "step": 3322
+    },
+    {
+      "epoch": 0.2042158308751229,
+      "grad_norm": 17.875,
+      "learning_rate": 1.9114511523356428e-05,
+      "loss": 0.377,
+      "step": 3323
+    },
+    {
+      "epoch": 0.2042772861356932,
+      "grad_norm": 13.25,
+      "learning_rate": 1.9113553885961674e-05,
+      "loss": 0.4704,
+      "step": 3324
+    },
+    {
+      "epoch": 0.2043387413962635,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.9112595755029625e-05,
+      "loss": 0.4726,
+      "step": 3325
+    },
+    {
+      "epoch": 0.20440019665683382,
+      "grad_norm": 18.875,
+      "learning_rate": 1.9111637130612172e-05,
+      "loss": 0.4006,
+      "step": 3326
+    },
+    {
+      "epoch": 0.20446165191740412,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.911067801276123e-05,
+      "loss": 0.4242,
+      "step": 3327
+    },
+    {
+      "epoch": 0.20452310717797442,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.9109718401528742e-05,
+      "loss": 0.323,
+      "step": 3328
+    },
+    {
+      "epoch": 0.20458456243854473,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.9108758296966666e-05,
+      "loss": 0.435,
+      "step": 3329
+    },
+    {
+      "epoch": 0.20464601769911506,
+      "grad_norm": 11.875,
+      "learning_rate": 1.9107797699127004e-05,
+      "loss": 0.4746,
+      "step": 3330
+    },
+    {
+      "epoch": 0.20470747295968536,
+      "grad_norm": 17.0,
+      "learning_rate": 1.910683660806177e-05,
+      "loss": 0.4663,
+      "step": 3331
+    },
+    {
+      "epoch": 0.20476892822025566,
+      "grad_norm": 23.625,
+      "learning_rate": 1.9105875023823023e-05,
+      "loss": 0.3944,
+      "step": 3332
+    },
+    {
+      "epoch": 0.20483038348082597,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.910491294646282e-05,
+      "loss": 0.3831,
+      "step": 3333
+    },
+    {
+      "epoch": 0.20489183874139627,
+      "grad_norm": 31.875,
+      "learning_rate": 1.9103950376033276e-05,
+      "loss": 0.4428,
+      "step": 3334
+    },
+    {
+      "epoch": 0.20495329400196657,
+      "grad_norm": 19.125,
+      "learning_rate": 1.910298731258651e-05,
+      "loss": 0.3606,
+      "step": 3335
+    },
+    {
+      "epoch": 0.20501474926253688,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.9102023756174675e-05,
+      "loss": 0.3203,
+      "step": 3336
+    },
+    {
+      "epoch": 0.20507620452310718,
+      "grad_norm": 21.5,
+      "learning_rate": 1.9101059706849957e-05,
+      "loss": 0.5632,
+      "step": 3337
+    },
+    {
+      "epoch": 0.20513765978367748,
+      "grad_norm": 21.375,
+      "learning_rate": 1.9100095164664564e-05,
+      "loss": 0.515,
+      "step": 3338
+    },
+    {
+      "epoch": 0.2051991150442478,
+      "grad_norm": 36.0,
+      "learning_rate": 1.9099130129670726e-05,
+      "loss": 0.6015,
+      "step": 3339
+    },
+    {
+      "epoch": 0.2052605703048181,
+      "grad_norm": 9.375,
+      "learning_rate": 1.9098164601920702e-05,
+      "loss": 0.5169,
+      "step": 3340
+    },
+    {
+      "epoch": 0.2053220255653884,
+      "grad_norm": 36.75,
+      "learning_rate": 1.9097198581466785e-05,
+      "loss": 0.4809,
+      "step": 3341
+    },
+    {
+      "epoch": 0.2053834808259587,
+      "grad_norm": 35.0,
+      "learning_rate": 1.9096232068361288e-05,
+      "loss": 0.4912,
+      "step": 3342
+    },
+    {
+      "epoch": 0.205444936086529,
+      "grad_norm": 24.5,
+      "learning_rate": 1.9095265062656546e-05,
+      "loss": 0.328,
+      "step": 3343
+    },
+    {
+      "epoch": 0.2055063913470993,
+      "grad_norm": 14.375,
+      "learning_rate": 1.9094297564404934e-05,
+      "loss": 0.323,
+      "step": 3344
+    },
+    {
+      "epoch": 0.2055678466076696,
+      "grad_norm": 40.5,
+      "learning_rate": 1.909332957365884e-05,
+      "loss": 0.4303,
+      "step": 3345
+    },
+    {
+      "epoch": 0.2056293018682399,
+      "grad_norm": 24.625,
+      "learning_rate": 1.9092361090470688e-05,
+      "loss": 0.5611,
+      "step": 3346
+    },
+    {
+      "epoch": 0.20569075712881021,
+      "grad_norm": 20.75,
+      "learning_rate": 1.9091392114892928e-05,
+      "loss": 0.3901,
+      "step": 3347
+    },
+    {
+      "epoch": 0.20575221238938052,
+      "grad_norm": 25.5,
+      "learning_rate": 1.9090422646978025e-05,
+      "loss": 0.4598,
+      "step": 3348
+    },
+    {
+      "epoch": 0.20581366764995085,
+      "grad_norm": 20.875,
+      "learning_rate": 1.908945268677849e-05,
+      "loss": 0.4094,
+      "step": 3349
+    },
+    {
+      "epoch": 0.20587512291052115,
+      "grad_norm": 19.0,
+      "learning_rate": 1.9088482234346843e-05,
+      "loss": 0.4999,
+      "step": 3350
+    },
+    {
+      "epoch": 0.20593657817109146,
+      "grad_norm": 23.125,
+      "learning_rate": 1.9087511289735646e-05,
+      "loss": 0.3852,
+      "step": 3351
+    },
+    {
+      "epoch": 0.20599803343166176,
+      "grad_norm": 24.25,
+      "learning_rate": 1.908653985299747e-05,
+      "loss": 0.3966,
+      "step": 3352
+    },
+    {
+      "epoch": 0.20605948869223206,
+      "grad_norm": 62.75,
+      "learning_rate": 1.9085567924184932e-05,
+      "loss": 0.5909,
+      "step": 3353
+    },
+    {
+      "epoch": 0.20612094395280237,
+      "grad_norm": 25.0,
+      "learning_rate": 1.9084595503350657e-05,
+      "loss": 0.7012,
+      "step": 3354
+    },
+    {
+      "epoch": 0.20618239921337267,
+      "grad_norm": 41.0,
+      "learning_rate": 1.9083622590547313e-05,
+      "loss": 0.8123,
+      "step": 3355
+    },
+    {
+      "epoch": 0.20624385447394297,
+      "grad_norm": 40.75,
+      "learning_rate": 1.9082649185827583e-05,
+      "loss": 0.7931,
+      "step": 3356
+    },
+    {
+      "epoch": 0.20630530973451328,
+      "grad_norm": 29.75,
+      "learning_rate": 1.9081675289244183e-05,
+      "loss": 0.7985,
+      "step": 3357
+    },
+    {
+      "epoch": 0.20636676499508358,
+      "grad_norm": 16.625,
+      "learning_rate": 1.9080700900849855e-05,
+      "loss": 0.4981,
+      "step": 3358
+    },
+    {
+      "epoch": 0.20642822025565388,
+      "grad_norm": 28.125,
+      "learning_rate": 1.907972602069736e-05,
+      "loss": 0.5605,
+      "step": 3359
+    },
+    {
+      "epoch": 0.20648967551622419,
+      "grad_norm": 19.125,
+      "learning_rate": 1.9078750648839497e-05,
+      "loss": 0.4373,
+      "step": 3360
+    },
+    {
+      "epoch": 0.2065511307767945,
+      "grad_norm": 20.0,
+      "learning_rate": 1.907777478532909e-05,
+      "loss": 0.357,
+      "step": 3361
+    },
+    {
+      "epoch": 0.2066125860373648,
+      "grad_norm": 13.875,
+      "learning_rate": 1.9076798430218975e-05,
+      "loss": 0.3645,
+      "step": 3362
+    },
+    {
+      "epoch": 0.2066740412979351,
+      "grad_norm": 34.75,
+      "learning_rate": 1.9075821583562036e-05,
+      "loss": 0.323,
+      "step": 3363
+    },
+    {
+      "epoch": 0.2067354965585054,
+      "grad_norm": 14.25,
+      "learning_rate": 1.907484424541117e-05,
+      "loss": 0.4351,
+      "step": 3364
+    },
+    {
+      "epoch": 0.2067969518190757,
+      "grad_norm": 28.25,
+      "learning_rate": 1.9073866415819303e-05,
+      "loss": 0.43,
+      "step": 3365
+    },
+    {
+      "epoch": 0.206858407079646,
+      "grad_norm": 21.625,
+      "learning_rate": 1.907288809483939e-05,
+      "loss": 0.4072,
+      "step": 3366
+    },
+    {
+      "epoch": 0.2069198623402163,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.907190928252441e-05,
+      "loss": 0.3202,
+      "step": 3367
+    },
+    {
+      "epoch": 0.20698131760078664,
+      "grad_norm": 11.5625,
+      "learning_rate": 1.907092997892737e-05,
+      "loss": 0.3836,
+      "step": 3368
+    },
+    {
+      "epoch": 0.20704277286135694,
+      "grad_norm": 18.25,
+      "learning_rate": 1.9069950184101305e-05,
+      "loss": 0.3967,
+      "step": 3369
+    },
+    {
+      "epoch": 0.20710422812192725,
+      "grad_norm": 29.125,
+      "learning_rate": 1.906896989809927e-05,
+      "loss": 0.5302,
+      "step": 3370
+    },
+    {
+      "epoch": 0.20716568338249755,
+      "grad_norm": 24.625,
+      "learning_rate": 1.906798912097436e-05,
+      "loss": 0.6367,
+      "step": 3371
+    },
+    {
+      "epoch": 0.20722713864306785,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.9067007852779684e-05,
+      "loss": 0.4452,
+      "step": 3372
+    },
+    {
+      "epoch": 0.20728859390363816,
+      "grad_norm": 31.375,
+      "learning_rate": 1.906602609356838e-05,
+      "loss": 0.4258,
+      "step": 3373
+    },
+    {
+      "epoch": 0.20735004916420846,
+      "grad_norm": 8.25,
+      "learning_rate": 1.9065043843393615e-05,
+      "loss": 0.3305,
+      "step": 3374
+    },
+    {
+      "epoch": 0.20741150442477876,
+      "grad_norm": 10.4375,
+      "learning_rate": 1.906406110230858e-05,
+      "loss": 0.3517,
+      "step": 3375
+    },
+    {
+      "epoch": 0.20747295968534907,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.9063077870366504e-05,
+      "loss": 0.3475,
+      "step": 3376
+    },
+    {
+      "epoch": 0.20753441494591937,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.906209414762062e-05,
+      "loss": 0.7558,
+      "step": 3377
+    },
+    {
+      "epoch": 0.20759587020648967,
+      "grad_norm": 36.0,
+      "learning_rate": 1.9061109934124208e-05,
+      "loss": 0.6361,
+      "step": 3378
+    },
+    {
+      "epoch": 0.20765732546705998,
+      "grad_norm": 48.0,
+      "learning_rate": 1.9060125229930572e-05,
+      "loss": 0.5517,
+      "step": 3379
+    },
+    {
+      "epoch": 0.20771878072763028,
+      "grad_norm": 18.0,
+      "learning_rate": 1.9059140035093024e-05,
+      "loss": 0.6132,
+      "step": 3380
+    },
+    {
+      "epoch": 0.20778023598820058,
+      "grad_norm": 12.125,
+      "learning_rate": 1.9058154349664932e-05,
+      "loss": 0.5487,
+      "step": 3381
+    },
+    {
+      "epoch": 0.2078416912487709,
+      "grad_norm": 40.0,
+      "learning_rate": 1.9057168173699664e-05,
+      "loss": 0.4196,
+      "step": 3382
+    },
+    {
+      "epoch": 0.2079031465093412,
+      "grad_norm": 40.0,
+      "learning_rate": 1.905618150725063e-05,
+      "loss": 0.6362,
+      "step": 3383
+    },
+    {
+      "epoch": 0.2079646017699115,
+      "grad_norm": 38.75,
+      "learning_rate": 1.9055194350371262e-05,
+      "loss": 0.4734,
+      "step": 3384
+    },
+    {
+      "epoch": 0.2080260570304818,
+      "grad_norm": 16.625,
+      "learning_rate": 1.905420670311502e-05,
+      "loss": 0.5246,
+      "step": 3385
+    },
+    {
+      "epoch": 0.2080875122910521,
+      "grad_norm": 49.25,
+      "learning_rate": 1.9053218565535383e-05,
+      "loss": 0.5238,
+      "step": 3386
+    },
+    {
+      "epoch": 0.20814896755162243,
+      "grad_norm": 48.5,
+      "learning_rate": 1.9052229937685872e-05,
+      "loss": 0.4325,
+      "step": 3387
+    },
+    {
+      "epoch": 0.20821042281219274,
+      "grad_norm": 17.25,
+      "learning_rate": 1.9051240819620018e-05,
+      "loss": 0.3485,
+      "step": 3388
+    },
+    {
+      "epoch": 0.20827187807276304,
+      "grad_norm": 8.625,
+      "learning_rate": 1.905025121139139e-05,
+      "loss": 0.5452,
+      "step": 3389
+    },
+    {
+      "epoch": 0.20833333333333334,
+      "grad_norm": 23.875,
+      "learning_rate": 1.9049261113053576e-05,
+      "loss": 0.4885,
+      "step": 3390
+    },
+    {
+      "epoch": 0.20839478859390365,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.9048270524660197e-05,
+      "loss": 0.3394,
+      "step": 3391
+    },
+    {
+      "epoch": 0.20845624385447395,
+      "grad_norm": 5.09375,
+      "learning_rate": 1.9047279446264898e-05,
+      "loss": 0.4145,
+      "step": 3392
+    },
+    {
+      "epoch": 0.20851769911504425,
+      "grad_norm": 36.5,
+      "learning_rate": 1.9046287877921344e-05,
+      "loss": 0.6575,
+      "step": 3393
+    },
+    {
+      "epoch": 0.20857915437561456,
+      "grad_norm": 27.375,
+      "learning_rate": 1.904529581968324e-05,
+      "loss": 0.5768,
+      "step": 3394
+    },
+    {
+      "epoch": 0.20864060963618486,
+      "grad_norm": 33.0,
+      "learning_rate": 1.904430327160431e-05,
+      "loss": 0.7122,
+      "step": 3395
+    },
+    {
+      "epoch": 0.20870206489675516,
+      "grad_norm": 16.875,
+      "learning_rate": 1.90433102337383e-05,
+      "loss": 0.7786,
+      "step": 3396
+    },
+    {
+      "epoch": 0.20876352015732547,
+      "grad_norm": 36.75,
+      "learning_rate": 1.9042316706138987e-05,
+      "loss": 0.586,
+      "step": 3397
+    },
+    {
+      "epoch": 0.20882497541789577,
+      "grad_norm": 33.25,
+      "learning_rate": 1.904132268886018e-05,
+      "loss": 0.5464,
+      "step": 3398
+    },
+    {
+      "epoch": 0.20888643067846607,
+      "grad_norm": 25.5,
+      "learning_rate": 1.9040328181955706e-05,
+      "loss": 0.5645,
+      "step": 3399
+    },
+    {
+      "epoch": 0.20894788593903638,
+      "grad_norm": 15.0,
+      "learning_rate": 1.903933318547942e-05,
+      "loss": 0.4591,
+      "step": 3400
+    },
+    {
+      "epoch": 0.20900934119960668,
+      "grad_norm": 21.75,
+      "learning_rate": 1.9038337699485207e-05,
+      "loss": 0.5846,
+      "step": 3401
+    },
+    {
+      "epoch": 0.20907079646017698,
+      "grad_norm": 22.875,
+      "learning_rate": 1.9037341724026975e-05,
+      "loss": 0.5218,
+      "step": 3402
+    },
+    {
+      "epoch": 0.2091322517207473,
+      "grad_norm": 12.875,
+      "learning_rate": 1.9036345259158667e-05,
+      "loss": 0.3924,
+      "step": 3403
+    },
+    {
+      "epoch": 0.2091937069813176,
+      "grad_norm": 23.25,
+      "learning_rate": 1.9035348304934238e-05,
+      "loss": 0.5038,
+      "step": 3404
+    },
+    {
+      "epoch": 0.2092551622418879,
+      "grad_norm": 18.625,
+      "learning_rate": 1.903435086140768e-05,
+      "loss": 0.5567,
+      "step": 3405
+    },
+    {
+      "epoch": 0.20931661750245822,
+      "grad_norm": 30.625,
+      "learning_rate": 1.903335292863301e-05,
+      "loss": 0.531,
+      "step": 3406
+    },
+    {
+      "epoch": 0.20937807276302853,
+      "grad_norm": 24.625,
+      "learning_rate": 1.903235450666427e-05,
+      "loss": 0.4959,
+      "step": 3407
+    },
+    {
+      "epoch": 0.20943952802359883,
+      "grad_norm": 23.25,
+      "learning_rate": 1.9031355595555525e-05,
+      "loss": 0.6077,
+      "step": 3408
+    },
+    {
+      "epoch": 0.20950098328416913,
+      "grad_norm": 19.5,
+      "learning_rate": 1.9030356195360875e-05,
+      "loss": 0.706,
+      "step": 3409
+    },
+    {
+      "epoch": 0.20956243854473944,
+      "grad_norm": 35.75,
+      "learning_rate": 1.9029356306134435e-05,
+      "loss": 0.7165,
+      "step": 3410
+    },
+    {
+      "epoch": 0.20962389380530974,
+      "grad_norm": 16.125,
+      "learning_rate": 1.9028355927930363e-05,
+      "loss": 0.5179,
+      "step": 3411
+    },
+    {
+      "epoch": 0.20968534906588004,
+      "grad_norm": 16.625,
+      "learning_rate": 1.902735506080283e-05,
+      "loss": 0.5954,
+      "step": 3412
+    },
+    {
+      "epoch": 0.20974680432645035,
+      "grad_norm": 10.5625,
+      "learning_rate": 1.9026353704806032e-05,
+      "loss": 0.3825,
+      "step": 3413
+    },
+    {
+      "epoch": 0.20980825958702065,
+      "grad_norm": 23.125,
+      "learning_rate": 1.9025351859994203e-05,
+      "loss": 0.442,
+      "step": 3414
+    },
+    {
+      "epoch": 0.20986971484759095,
+      "grad_norm": 17.75,
+      "learning_rate": 1.9024349526421596e-05,
+      "loss": 0.5192,
+      "step": 3415
+    },
+    {
+      "epoch": 0.20993117010816126,
+      "grad_norm": 18.875,
+      "learning_rate": 1.9023346704142488e-05,
+      "loss": 0.3977,
+      "step": 3416
+    },
+    {
+      "epoch": 0.20999262536873156,
+      "grad_norm": 37.25,
+      "learning_rate": 1.902234339321119e-05,
+      "loss": 0.3841,
+      "step": 3417
+    },
+    {
+      "epoch": 0.21005408062930186,
+      "grad_norm": 12.75,
+      "learning_rate": 1.902133959368203e-05,
+      "loss": 0.3731,
+      "step": 3418
+    },
+    {
+      "epoch": 0.21011553588987217,
+      "grad_norm": 9.5,
+      "learning_rate": 1.9020335305609373e-05,
+      "loss": 0.3999,
+      "step": 3419
+    },
+    {
+      "epoch": 0.21017699115044247,
+      "grad_norm": 8.875,
+      "learning_rate": 1.9019330529047605e-05,
+      "loss": 0.3533,
+      "step": 3420
+    },
+    {
+      "epoch": 0.21023844641101277,
+      "grad_norm": 10.5,
+      "learning_rate": 1.901832526405114e-05,
+      "loss": 0.3358,
+      "step": 3421
+    },
+    {
+      "epoch": 0.21029990167158308,
+      "grad_norm": 16.375,
+      "learning_rate": 1.9017319510674413e-05,
+      "loss": 0.3219,
+      "step": 3422
+    },
+    {
+      "epoch": 0.21036135693215338,
+      "grad_norm": 10.3125,
+      "learning_rate": 1.9016313268971892e-05,
+      "loss": 0.3167,
+      "step": 3423
+    },
+    {
+      "epoch": 0.21042281219272368,
+      "grad_norm": 17.5,
+      "learning_rate": 1.901530653899807e-05,
+      "loss": 0.2909,
+      "step": 3424
+    },
+    {
+      "epoch": 0.210484267453294,
+      "grad_norm": 17.125,
+      "learning_rate": 1.9014299320807467e-05,
+      "loss": 0.4499,
+      "step": 3425
+    },
+    {
+      "epoch": 0.21054572271386432,
+      "grad_norm": 29.375,
+      "learning_rate": 1.9013291614454622e-05,
+      "loss": 0.6185,
+      "step": 3426
+    },
+    {
+      "epoch": 0.21060717797443462,
+      "grad_norm": 22.875,
+      "learning_rate": 1.9012283419994115e-05,
+      "loss": 0.4255,
+      "step": 3427
+    },
+    {
+      "epoch": 0.21066863323500493,
+      "grad_norm": 21.125,
+      "learning_rate": 1.9011274737480536e-05,
+      "loss": 0.404,
+      "step": 3428
+    },
+    {
+      "epoch": 0.21073008849557523,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.9010265566968518e-05,
+      "loss": 0.4227,
+      "step": 3429
+    },
+    {
+      "epoch": 0.21079154375614553,
+      "grad_norm": 9.75,
+      "learning_rate": 1.9009255908512704e-05,
+      "loss": 0.2796,
+      "step": 3430
+    },
+    {
+      "epoch": 0.21085299901671584,
+      "grad_norm": 13.625,
+      "learning_rate": 1.9008245762167773e-05,
+      "loss": 0.2448,
+      "step": 3431
+    },
+    {
+      "epoch": 0.21091445427728614,
+      "grad_norm": 7.75,
+      "learning_rate": 1.900723512798843e-05,
+      "loss": 0.369,
+      "step": 3432
+    },
+    {
+      "epoch": 0.21097590953785644,
+      "grad_norm": 19.375,
+      "learning_rate": 1.9006224006029404e-05,
+      "loss": 0.4336,
+      "step": 3433
+    },
+    {
+      "epoch": 0.21103736479842675,
+      "grad_norm": 31.875,
+      "learning_rate": 1.9005212396345458e-05,
+      "loss": 0.3779,
+      "step": 3434
+    },
+    {
+      "epoch": 0.21109882005899705,
+      "grad_norm": 17.375,
+      "learning_rate": 1.900420029899136e-05,
+      "loss": 0.4318,
+      "step": 3435
+    },
+    {
+      "epoch": 0.21116027531956735,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.9003187714021936e-05,
+      "loss": 0.4246,
+      "step": 3436
+    },
+    {
+      "epoch": 0.21122173058013766,
+      "grad_norm": 49.5,
+      "learning_rate": 1.9002174641492015e-05,
+      "loss": 0.446,
+      "step": 3437
+    },
+    {
+      "epoch": 0.21128318584070796,
+      "grad_norm": 38.5,
+      "learning_rate": 1.9001161081456455e-05,
+      "loss": 0.3657,
+      "step": 3438
+    },
+    {
+      "epoch": 0.21134464110127826,
+      "grad_norm": 60.75,
+      "learning_rate": 1.9000147033970148e-05,
+      "loss": 0.3347,
+      "step": 3439
+    },
+    {
+      "epoch": 0.21140609636184857,
+      "grad_norm": 35.5,
+      "learning_rate": 1.899913249908801e-05,
+      "loss": 0.3616,
+      "step": 3440
+    },
+    {
+      "epoch": 0.21146755162241887,
+      "grad_norm": 18.125,
+      "learning_rate": 1.8998117476864984e-05,
+      "loss": 0.5344,
+      "step": 3441
+    },
+    {
+      "epoch": 0.21152900688298917,
+      "grad_norm": 18.75,
+      "learning_rate": 1.899710196735603e-05,
+      "loss": 0.3985,
+      "step": 3442
+    },
+    {
+      "epoch": 0.21159046214355948,
+      "grad_norm": 36.5,
+      "learning_rate": 1.899608597061615e-05,
+      "loss": 0.4223,
+      "step": 3443
+    },
+    {
+      "epoch": 0.21165191740412978,
+      "grad_norm": 64.0,
+      "learning_rate": 1.899506948670036e-05,
+      "loss": 0.5576,
+      "step": 3444
+    },
+    {
+      "epoch": 0.2117133726647001,
+      "grad_norm": 23.375,
+      "learning_rate": 1.899405251566371e-05,
+      "loss": 0.5121,
+      "step": 3445
+    },
+    {
+      "epoch": 0.21177482792527041,
+      "grad_norm": 27.375,
+      "learning_rate": 1.8993035057561274e-05,
+      "loss": 0.4471,
+      "step": 3446
+    },
+    {
+      "epoch": 0.21183628318584072,
+      "grad_norm": 17.375,
+      "learning_rate": 1.899201711244815e-05,
+      "loss": 0.5354,
+      "step": 3447
+    },
+    {
+      "epoch": 0.21189773844641102,
+      "grad_norm": 27.375,
+      "learning_rate": 1.8990998680379458e-05,
+      "loss": 0.4154,
+      "step": 3448
+    },
+    {
+      "epoch": 0.21195919370698132,
+      "grad_norm": 65.5,
+      "learning_rate": 1.8989979761410356e-05,
+      "loss": 0.8404,
+      "step": 3449
+    },
+    {
+      "epoch": 0.21202064896755163,
+      "grad_norm": 50.0,
+      "learning_rate": 1.8988960355596027e-05,
+      "loss": 0.7976,
+      "step": 3450
+    },
+    {
+      "epoch": 0.21208210422812193,
+      "grad_norm": 25.125,
+      "learning_rate": 1.8987940462991673e-05,
+      "loss": 0.6242,
+      "step": 3451
+    },
+    {
+      "epoch": 0.21214355948869223,
+      "grad_norm": 52.0,
+      "learning_rate": 1.898692008365252e-05,
+      "loss": 0.436,
+      "step": 3452
+    },
+    {
+      "epoch": 0.21220501474926254,
+      "grad_norm": 27.125,
+      "learning_rate": 1.898589921763383e-05,
+      "loss": 0.502,
+      "step": 3453
+    },
+    {
+      "epoch": 0.21226647000983284,
+      "grad_norm": 31.375,
+      "learning_rate": 1.8984877864990888e-05,
+      "loss": 0.5196,
+      "step": 3454
+    },
+    {
+      "epoch": 0.21232792527040315,
+      "grad_norm": 17.375,
+      "learning_rate": 1.8983856025779007e-05,
+      "loss": 0.4961,
+      "step": 3455
+    },
+    {
+      "epoch": 0.21238938053097345,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.8982833700053518e-05,
+      "loss": 0.3489,
+      "step": 3456
+    },
+    {
+      "epoch": 0.21245083579154375,
+      "grad_norm": 30.875,
+      "learning_rate": 1.8981810887869784e-05,
+      "loss": 0.2947,
+      "step": 3457
+    },
+    {
+      "epoch": 0.21251229105211406,
+      "grad_norm": 4.96875,
+      "learning_rate": 1.8980787589283204e-05,
+      "loss": 0.3005,
+      "step": 3458
+    },
+    {
+      "epoch": 0.21257374631268436,
+      "grad_norm": 23.875,
+      "learning_rate": 1.897976380434918e-05,
+      "loss": 0.3242,
+      "step": 3459
+    },
+    {
+      "epoch": 0.21263520157325466,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.897873953312317e-05,
+      "loss": 0.2953,
+      "step": 3460
+    },
+    {
+      "epoch": 0.21269665683382497,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.897771477566063e-05,
+      "loss": 0.3113,
+      "step": 3461
+    },
+    {
+      "epoch": 0.21275811209439527,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.8976689532017055e-05,
+      "loss": 0.3601,
+      "step": 3462
+    },
+    {
+      "epoch": 0.21281956735496557,
+      "grad_norm": 41.0,
+      "learning_rate": 1.8975663802247978e-05,
+      "loss": 0.2737,
+      "step": 3463
+    },
+    {
+      "epoch": 0.2128810226155359,
+      "grad_norm": 9.625,
+      "learning_rate": 1.8974637586408933e-05,
+      "loss": 0.3694,
+      "step": 3464
+    },
+    {
+      "epoch": 0.2129424778761062,
+      "grad_norm": 43.5,
+      "learning_rate": 1.8973610884555505e-05,
+      "loss": 0.5624,
+      "step": 3465
+    },
+    {
+      "epoch": 0.2130039331366765,
+      "grad_norm": 45.75,
+      "learning_rate": 1.8972583696743284e-05,
+      "loss": 0.7223,
+      "step": 3466
+    },
+    {
+      "epoch": 0.2130653883972468,
+      "grad_norm": 48.0,
+      "learning_rate": 1.8971556023027905e-05,
+      "loss": 0.5296,
+      "step": 3467
+    },
+    {
+      "epoch": 0.21312684365781712,
+      "grad_norm": 35.5,
+      "learning_rate": 1.897052786346502e-05,
+      "loss": 0.5389,
+      "step": 3468
+    },
+    {
+      "epoch": 0.21318829891838742,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.8969499218110302e-05,
+      "loss": 0.579,
+      "step": 3469
+    },
+    {
+      "epoch": 0.21324975417895772,
+      "grad_norm": 56.75,
+      "learning_rate": 1.896847008701946e-05,
+      "loss": 0.474,
+      "step": 3470
+    },
+    {
+      "epoch": 0.21331120943952803,
+      "grad_norm": 21.75,
+      "learning_rate": 1.8967440470248227e-05,
+      "loss": 0.3857,
+      "step": 3471
+    },
+    {
+      "epoch": 0.21337266470009833,
+      "grad_norm": 21.0,
+      "learning_rate": 1.896641036785236e-05,
+      "loss": 0.3608,
+      "step": 3472
+    },
+    {
+      "epoch": 0.21343411996066863,
+      "grad_norm": 16.125,
+      "learning_rate": 1.896537977988764e-05,
+      "loss": 0.4179,
+      "step": 3473
+    },
+    {
+      "epoch": 0.21349557522123894,
+      "grad_norm": 26.375,
+      "learning_rate": 1.896434870640989e-05,
+      "loss": 0.4416,
+      "step": 3474
+    },
+    {
+      "epoch": 0.21355703048180924,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.896331714747493e-05,
+      "loss": 0.3183,
+      "step": 3475
+    },
+    {
+      "epoch": 0.21361848574237954,
+      "grad_norm": 11.125,
+      "learning_rate": 1.8962285103138637e-05,
+      "loss": 0.3087,
+      "step": 3476
+    },
+    {
+      "epoch": 0.21367994100294985,
+      "grad_norm": 13.125,
+      "learning_rate": 1.896125257345689e-05,
+      "loss": 0.3474,
+      "step": 3477
+    },
+    {
+      "epoch": 0.21374139626352015,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.896021955848561e-05,
+      "loss": 0.3544,
+      "step": 3478
+    },
+    {
+      "epoch": 0.21380285152409045,
+      "grad_norm": 12.125,
+      "learning_rate": 1.8959186058280744e-05,
+      "loss": 0.3563,
+      "step": 3479
+    },
+    {
+      "epoch": 0.21386430678466076,
+      "grad_norm": 22.75,
+      "learning_rate": 1.895815207289825e-05,
+      "loss": 0.3235,
+      "step": 3480
+    },
+    {
+      "epoch": 0.21392576204523106,
+      "grad_norm": 33.5,
+      "learning_rate": 1.895711760239413e-05,
+      "loss": 0.5054,
+      "step": 3481
+    },
+    {
+      "epoch": 0.21398721730580136,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.89560826468244e-05,
+      "loss": 0.4144,
+      "step": 3482
+    },
+    {
+      "epoch": 0.2140486725663717,
+      "grad_norm": 20.75,
+      "learning_rate": 1.8955047206245115e-05,
+      "loss": 0.5112,
+      "step": 3483
+    },
+    {
+      "epoch": 0.214110127826942,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.895401128071234e-05,
+      "loss": 0.3733,
+      "step": 3484
+    },
+    {
+      "epoch": 0.2141715830875123,
+      "grad_norm": 39.0,
+      "learning_rate": 1.895297487028218e-05,
+      "loss": 0.3263,
+      "step": 3485
+    },
+    {
+      "epoch": 0.2142330383480826,
+      "grad_norm": 28.5,
+      "learning_rate": 1.895193797501076e-05,
+      "loss": 0.5216,
+      "step": 3486
+    },
+    {
+      "epoch": 0.2142944936086529,
+      "grad_norm": 12.5,
+      "learning_rate": 1.8950900594954226e-05,
+      "loss": 0.3861,
+      "step": 3487
+    },
+    {
+      "epoch": 0.2143559488692232,
+      "grad_norm": 10.3125,
+      "learning_rate": 1.8949862730168766e-05,
+      "loss": 0.3608,
+      "step": 3488
+    },
+    {
+      "epoch": 0.21441740412979352,
+      "grad_norm": 22.5,
+      "learning_rate": 1.8948824380710582e-05,
+      "loss": 0.3415,
+      "step": 3489
+    },
+    {
+      "epoch": 0.21447885939036382,
+      "grad_norm": 44.75,
+      "learning_rate": 1.8947785546635905e-05,
+      "loss": 0.4517,
+      "step": 3490
+    },
+    {
+      "epoch": 0.21454031465093412,
+      "grad_norm": 35.25,
+      "learning_rate": 1.8946746228000987e-05,
+      "loss": 0.3534,
+      "step": 3491
+    },
+    {
+      "epoch": 0.21460176991150443,
+      "grad_norm": 16.125,
+      "learning_rate": 1.8945706424862115e-05,
+      "loss": 0.485,
+      "step": 3492
+    },
+    {
+      "epoch": 0.21466322517207473,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.89446661372756e-05,
+      "loss": 0.365,
+      "step": 3493
+    },
+    {
+      "epoch": 0.21472468043264503,
+      "grad_norm": 21.0,
+      "learning_rate": 1.8943625365297776e-05,
+      "loss": 0.3745,
+      "step": 3494
+    },
+    {
+      "epoch": 0.21478613569321534,
+      "grad_norm": 28.625,
+      "learning_rate": 1.894258410898501e-05,
+      "loss": 0.3674,
+      "step": 3495
+    },
+    {
+      "epoch": 0.21484759095378564,
+      "grad_norm": 34.0,
+      "learning_rate": 1.8941542368393683e-05,
+      "loss": 0.3151,
+      "step": 3496
+    },
+    {
+      "epoch": 0.21490904621435594,
+      "grad_norm": 44.0,
+      "learning_rate": 1.8940500143580215e-05,
+      "loss": 0.6101,
+      "step": 3497
+    },
+    {
+      "epoch": 0.21497050147492625,
+      "grad_norm": 55.25,
+      "learning_rate": 1.8939457434601044e-05,
+      "loss": 0.5991,
+      "step": 3498
+    },
+    {
+      "epoch": 0.21503195673549655,
+      "grad_norm": 21.625,
+      "learning_rate": 1.893841424151264e-05,
+      "loss": 0.5392,
+      "step": 3499
+    },
+    {
+      "epoch": 0.21509341199606685,
+      "grad_norm": 38.0,
+      "learning_rate": 1.893737056437149e-05,
+      "loss": 0.4825,
+      "step": 3500
+    },
+    {
+      "epoch": 0.21515486725663716,
+      "grad_norm": 34.0,
+      "learning_rate": 1.8936326403234125e-05,
+      "loss": 0.4443,
+      "step": 3501
+    },
+    {
+      "epoch": 0.2152163225172075,
+      "grad_norm": 23.125,
+      "learning_rate": 1.893528175815708e-05,
+      "loss": 0.3608,
+      "step": 3502
+    },
+    {
+      "epoch": 0.2152777777777778,
+      "grad_norm": 38.75,
+      "learning_rate": 1.8934236629196928e-05,
+      "loss": 0.337,
+      "step": 3503
+    },
+    {
+      "epoch": 0.2153392330383481,
+      "grad_norm": 14.375,
+      "learning_rate": 1.8933191016410273e-05,
+      "loss": 0.283,
+      "step": 3504
+    },
+    {
+      "epoch": 0.2154006882989184,
+      "grad_norm": 16.625,
+      "learning_rate": 1.893214491985374e-05,
+      "loss": 0.524,
+      "step": 3505
+    },
+    {
+      "epoch": 0.2154621435594887,
+      "grad_norm": 24.75,
+      "learning_rate": 1.893109833958397e-05,
+      "loss": 0.4749,
+      "step": 3506
+    },
+    {
+      "epoch": 0.215523598820059,
+      "grad_norm": 33.75,
+      "learning_rate": 1.8930051275657643e-05,
+      "loss": 0.6193,
+      "step": 3507
+    },
+    {
+      "epoch": 0.2155850540806293,
+      "grad_norm": 19.625,
+      "learning_rate": 1.892900372813147e-05,
+      "loss": 0.3183,
+      "step": 3508
+    },
+    {
+      "epoch": 0.2156465093411996,
+      "grad_norm": 45.25,
+      "learning_rate": 1.8927955697062172e-05,
+      "loss": 0.676,
+      "step": 3509
+    },
+    {
+      "epoch": 0.2157079646017699,
+      "grad_norm": 16.5,
+      "learning_rate": 1.8926907182506505e-05,
+      "loss": 0.5075,
+      "step": 3510
+    },
+    {
+      "epoch": 0.21576941986234022,
+      "grad_norm": 8.375,
+      "learning_rate": 1.892585818452126e-05,
+      "loss": 0.3383,
+      "step": 3511
+    },
+    {
+      "epoch": 0.21583087512291052,
+      "grad_norm": 6.4375,
+      "learning_rate": 1.8924808703163226e-05,
+      "loss": 0.4292,
+      "step": 3512
+    },
+    {
+      "epoch": 0.21589233038348082,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.892375873848925e-05,
+      "loss": 0.506,
+      "step": 3513
+    },
+    {
+      "epoch": 0.21595378564405113,
+      "grad_norm": 10.5625,
+      "learning_rate": 1.8922708290556197e-05,
+      "loss": 0.4999,
+      "step": 3514
+    },
+    {
+      "epoch": 0.21601524090462143,
+      "grad_norm": 16.25,
+      "learning_rate": 1.892165735942094e-05,
+      "loss": 0.5492,
+      "step": 3515
+    },
+    {
+      "epoch": 0.21607669616519173,
+      "grad_norm": 27.375,
+      "learning_rate": 1.8920605945140396e-05,
+      "loss": 0.369,
+      "step": 3516
+    },
+    {
+      "epoch": 0.21613815142576204,
+      "grad_norm": 13.375,
+      "learning_rate": 1.8919554047771508e-05,
+      "loss": 0.5076,
+      "step": 3517
+    },
+    {
+      "epoch": 0.21619960668633234,
+      "grad_norm": 51.0,
+      "learning_rate": 1.891850166737124e-05,
+      "loss": 0.4182,
+      "step": 3518
+    },
+    {
+      "epoch": 0.21626106194690264,
+      "grad_norm": 40.0,
+      "learning_rate": 1.8917448803996573e-05,
+      "loss": 0.42,
+      "step": 3519
+    },
+    {
+      "epoch": 0.21632251720747295,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.8916395457704536e-05,
+      "loss": 0.3401,
+      "step": 3520
+    },
+    {
+      "epoch": 0.21638397246804325,
+      "grad_norm": 34.25,
+      "learning_rate": 1.8915341628552166e-05,
+      "loss": 0.4286,
+      "step": 3521
+    },
+    {
+      "epoch": 0.21644542772861358,
+      "grad_norm": 4.875,
+      "learning_rate": 1.8914287316596532e-05,
+      "loss": 0.2178,
+      "step": 3522
+    },
+    {
+      "epoch": 0.21650688298918389,
+      "grad_norm": 10.5625,
+      "learning_rate": 1.8913232521894734e-05,
+      "loss": 0.3303,
+      "step": 3523
+    },
+    {
+      "epoch": 0.2165683382497542,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.891217724450389e-05,
+      "loss": 0.3541,
+      "step": 3524
+    },
+    {
+      "epoch": 0.2166297935103245,
+      "grad_norm": 23.375,
+      "learning_rate": 1.8911121484481145e-05,
+      "loss": 0.4064,
+      "step": 3525
+    },
+    {
+      "epoch": 0.2166912487708948,
+      "grad_norm": 39.25,
+      "learning_rate": 1.891006524188368e-05,
+      "loss": 0.4571,
+      "step": 3526
+    },
+    {
+      "epoch": 0.2167527040314651,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.890900851676869e-05,
+      "loss": 0.2396,
+      "step": 3527
+    },
+    {
+      "epoch": 0.2168141592920354,
+      "grad_norm": 5.125,
+      "learning_rate": 1.89079513091934e-05,
+      "loss": 0.2493,
+      "step": 3528
+    },
+    {
+      "epoch": 0.2168756145526057,
+      "grad_norm": 37.75,
+      "learning_rate": 1.890689361921507e-05,
+      "loss": 0.5346,
+      "step": 3529
+    },
+    {
+      "epoch": 0.216937069813176,
+      "grad_norm": 38.25,
+      "learning_rate": 1.8905835446890968e-05,
+      "loss": 0.6114,
+      "step": 3530
+    },
+    {
+      "epoch": 0.2169985250737463,
+      "grad_norm": 31.625,
+      "learning_rate": 1.8904776792278403e-05,
+      "loss": 0.5315,
+      "step": 3531
+    },
+    {
+      "epoch": 0.21705998033431662,
+      "grad_norm": 32.0,
+      "learning_rate": 1.8903717655434708e-05,
+      "loss": 0.456,
+      "step": 3532
+    },
+    {
+      "epoch": 0.21712143559488692,
+      "grad_norm": 22.5,
+      "learning_rate": 1.8902658036417236e-05,
+      "loss": 0.5363,
+      "step": 3533
+    },
+    {
+      "epoch": 0.21718289085545722,
+      "grad_norm": 25.125,
+      "learning_rate": 1.8901597935283373e-05,
+      "loss": 0.4296,
+      "step": 3534
+    },
+    {
+      "epoch": 0.21724434611602753,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.8900537352090523e-05,
+      "loss": 0.3458,
+      "step": 3535
+    },
+    {
+      "epoch": 0.21730580137659783,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.889947628689613e-05,
+      "loss": 0.3885,
+      "step": 3536
+    },
+    {
+      "epoch": 0.21736725663716813,
+      "grad_norm": 53.75,
+      "learning_rate": 1.8898414739757648e-05,
+      "loss": 0.5621,
+      "step": 3537
+    },
+    {
+      "epoch": 0.21742871189773844,
+      "grad_norm": 13.125,
+      "learning_rate": 1.8897352710732564e-05,
+      "loss": 0.4764,
+      "step": 3538
+    },
+    {
+      "epoch": 0.21749016715830874,
+      "grad_norm": 36.75,
+      "learning_rate": 1.8896290199878393e-05,
+      "loss": 0.4968,
+      "step": 3539
+    },
+    {
+      "epoch": 0.21755162241887904,
+      "grad_norm": 30.0,
+      "learning_rate": 1.8895227207252674e-05,
+      "loss": 0.3151,
+      "step": 3540
+    },
+    {
+      "epoch": 0.21761307767944937,
+      "grad_norm": 20.625,
+      "learning_rate": 1.889416373291298e-05,
+      "loss": 0.3569,
+      "step": 3541
+    },
+    {
+      "epoch": 0.21767453294001968,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.889309977691689e-05,
+      "loss": 0.4138,
+      "step": 3542
+    },
+    {
+      "epoch": 0.21773598820058998,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.8892035339322023e-05,
+      "loss": 0.4461,
+      "step": 3543
+    },
+    {
+      "epoch": 0.21779744346116028,
+      "grad_norm": 28.25,
+      "learning_rate": 1.8890970420186035e-05,
+      "loss": 0.3911,
+      "step": 3544
+    },
+    {
+      "epoch": 0.2178588987217306,
+      "grad_norm": 29.625,
+      "learning_rate": 1.8889905019566584e-05,
+      "loss": 0.5838,
+      "step": 3545
+    },
+    {
+      "epoch": 0.2179203539823009,
+      "grad_norm": 17.5,
+      "learning_rate": 1.888883913752137e-05,
+      "loss": 0.5225,
+      "step": 3546
+    },
+    {
+      "epoch": 0.2179818092428712,
+      "grad_norm": 17.125,
+      "learning_rate": 1.8887772774108116e-05,
+      "loss": 0.3765,
+      "step": 3547
+    },
+    {
+      "epoch": 0.2180432645034415,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.888670592938457e-05,
+      "loss": 0.3852,
+      "step": 3548
+    },
+    {
+      "epoch": 0.2181047197640118,
+      "grad_norm": 29.125,
+      "learning_rate": 1.8885638603408502e-05,
+      "loss": 0.4373,
+      "step": 3549
+    },
+    {
+      "epoch": 0.2181661750245821,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.888457079623772e-05,
+      "loss": 0.4385,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2182276302851524,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.8883502507930044e-05,
+      "loss": 0.3072,
+      "step": 3551
+    },
+    {
+      "epoch": 0.2182890855457227,
+      "grad_norm": 15.625,
+      "learning_rate": 1.8882433738543325e-05,
+      "loss": 0.4408,
+      "step": 3552
+    },
+    {
+      "epoch": 0.21835054080629301,
+      "grad_norm": 17.375,
+      "learning_rate": 1.8881364488135448e-05,
+      "loss": 0.4208,
+      "step": 3553
+    },
+    {
+      "epoch": 0.21841199606686332,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.888029475676431e-05,
+      "loss": 0.5357,
+      "step": 3554
+    },
+    {
+      "epoch": 0.21847345132743362,
+      "grad_norm": 12.0,
+      "learning_rate": 1.887922454448785e-05,
+      "loss": 0.3352,
+      "step": 3555
+    },
+    {
+      "epoch": 0.21853490658800392,
+      "grad_norm": 5.4375,
+      "learning_rate": 1.8878153851364013e-05,
+      "loss": 0.2896,
+      "step": 3556
+    },
+    {
+      "epoch": 0.21859636184857423,
+      "grad_norm": 13.5,
+      "learning_rate": 1.8877082677450795e-05,
+      "loss": 0.3957,
+      "step": 3557
+    },
+    {
+      "epoch": 0.21865781710914453,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8876011022806193e-05,
+      "loss": 0.3333,
+      "step": 3558
+    },
+    {
+      "epoch": 0.21871927236971483,
+      "grad_norm": 5.875,
+      "learning_rate": 1.887493888748825e-05,
+      "loss": 0.2617,
+      "step": 3559
+    },
+    {
+      "epoch": 0.21878072763028517,
+      "grad_norm": 10.25,
+      "learning_rate": 1.8873866271555022e-05,
+      "loss": 0.2539,
+      "step": 3560
+    },
+    {
+      "epoch": 0.21884218289085547,
+      "grad_norm": 18.875,
+      "learning_rate": 1.8872793175064594e-05,
+      "loss": 0.3132,
+      "step": 3561
+    },
+    {
+      "epoch": 0.21890363815142577,
+      "grad_norm": 78.0,
+      "learning_rate": 1.8871719598075083e-05,
+      "loss": 0.4044,
+      "step": 3562
+    },
+    {
+      "epoch": 0.21896509341199608,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.8870645540644625e-05,
+      "loss": 0.3298,
+      "step": 3563
+    },
+    {
+      "epoch": 0.21902654867256638,
+      "grad_norm": 32.75,
+      "learning_rate": 1.8869571002831388e-05,
+      "loss": 0.5178,
+      "step": 3564
+    },
+    {
+      "epoch": 0.21908800393313668,
+      "grad_norm": 56.0,
+      "learning_rate": 1.886849598469356e-05,
+      "loss": 0.3818,
+      "step": 3565
+    },
+    {
+      "epoch": 0.21914945919370699,
+      "grad_norm": 27.0,
+      "learning_rate": 1.886742048628936e-05,
+      "loss": 0.3706,
+      "step": 3566
+    },
+    {
+      "epoch": 0.2192109144542773,
+      "grad_norm": 73.0,
+      "learning_rate": 1.8866344507677023e-05,
+      "loss": 0.3682,
+      "step": 3567
+    },
+    {
+      "epoch": 0.2192723697148476,
+      "grad_norm": 49.25,
+      "learning_rate": 1.8865268048914828e-05,
+      "loss": 0.4882,
+      "step": 3568
+    },
+    {
+      "epoch": 0.2193338249754179,
+      "grad_norm": 59.5,
+      "learning_rate": 1.8864191110061066e-05,
+      "loss": 0.4289,
+      "step": 3569
+    },
+    {
+      "epoch": 0.2193952802359882,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.8863113691174058e-05,
+      "loss": 0.3772,
+      "step": 3570
+    },
+    {
+      "epoch": 0.2194567354965585,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.8862035792312148e-05,
+      "loss": 0.3209,
+      "step": 3571
+    },
+    {
+      "epoch": 0.2195181907571288,
+      "grad_norm": 9.5,
+      "learning_rate": 1.8860957413533713e-05,
+      "loss": 0.4487,
+      "step": 3572
+    },
+    {
+      "epoch": 0.2195796460176991,
+      "grad_norm": 10.5625,
+      "learning_rate": 1.8859878554897147e-05,
+      "loss": 0.2822,
+      "step": 3573
+    },
+    {
+      "epoch": 0.2196411012782694,
+      "grad_norm": 5.28125,
+      "learning_rate": 1.8858799216460883e-05,
+      "loss": 0.3355,
+      "step": 3574
+    },
+    {
+      "epoch": 0.21970255653883972,
+      "grad_norm": 17.375,
+      "learning_rate": 1.8857719398283358e-05,
+      "loss": 0.3819,
+      "step": 3575
+    },
+    {
+      "epoch": 0.21976401179941002,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.8856639100423062e-05,
+      "loss": 0.2435,
+      "step": 3576
+    },
+    {
+      "epoch": 0.21982546705998032,
+      "grad_norm": 9.625,
+      "learning_rate": 1.8855558322938492e-05,
+      "loss": 0.2443,
+      "step": 3577
+    },
+    {
+      "epoch": 0.21988692232055063,
+      "grad_norm": 34.25,
+      "learning_rate": 1.8854477065888176e-05,
+      "loss": 0.6694,
+      "step": 3578
+    },
+    {
+      "epoch": 0.21994837758112096,
+      "grad_norm": 17.0,
+      "learning_rate": 1.8853395329330667e-05,
+      "loss": 0.3439,
+      "step": 3579
+    },
+    {
+      "epoch": 0.22000983284169126,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.8852313113324553e-05,
+      "loss": 0.3224,
+      "step": 3580
+    },
+    {
+      "epoch": 0.22007128810226156,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.8851230417928433e-05,
+      "loss": 0.4,
+      "step": 3581
+    },
+    {
+      "epoch": 0.22013274336283187,
+      "grad_norm": 11.875,
+      "learning_rate": 1.8850147243200946e-05,
+      "loss": 0.4037,
+      "step": 3582
+    },
+    {
+      "epoch": 0.22019419862340217,
+      "grad_norm": 68.5,
+      "learning_rate": 1.8849063589200744e-05,
+      "loss": 0.4553,
+      "step": 3583
+    },
+    {
+      "epoch": 0.22025565388397247,
+      "grad_norm": 54.5,
+      "learning_rate": 1.8847979455986515e-05,
+      "loss": 0.5413,
+      "step": 3584
+    },
+    {
+      "epoch": 0.22031710914454278,
+      "grad_norm": 42.25,
+      "learning_rate": 1.884689484361697e-05,
+      "loss": 0.5473,
+      "step": 3585
+    },
+    {
+      "epoch": 0.22037856440511308,
+      "grad_norm": 17.25,
+      "learning_rate": 1.884580975215084e-05,
+      "loss": 0.4447,
+      "step": 3586
+    },
+    {
+      "epoch": 0.22044001966568338,
+      "grad_norm": 32.25,
+      "learning_rate": 1.8844724181646895e-05,
+      "loss": 0.5432,
+      "step": 3587
+    },
+    {
+      "epoch": 0.2205014749262537,
+      "grad_norm": 16.25,
+      "learning_rate": 1.8843638132163917e-05,
+      "loss": 0.3536,
+      "step": 3588
+    },
+    {
+      "epoch": 0.220562930186824,
+      "grad_norm": 15.625,
+      "learning_rate": 1.8842551603760725e-05,
+      "loss": 0.4254,
+      "step": 3589
+    },
+    {
+      "epoch": 0.2206243854473943,
+      "grad_norm": 42.5,
+      "learning_rate": 1.8841464596496154e-05,
+      "loss": 0.4676,
+      "step": 3590
+    },
+    {
+      "epoch": 0.2206858407079646,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.8840377110429075e-05,
+      "loss": 0.3577,
+      "step": 3591
+    },
+    {
+      "epoch": 0.2207472959685349,
+      "grad_norm": 84.5,
+      "learning_rate": 1.8839289145618378e-05,
+      "loss": 0.3929,
+      "step": 3592
+    },
+    {
+      "epoch": 0.2208087512291052,
+      "grad_norm": 23.0,
+      "learning_rate": 1.883820070212298e-05,
+      "loss": 0.3703,
+      "step": 3593
+    },
+    {
+      "epoch": 0.2208702064896755,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8837111780001822e-05,
+      "loss": 0.3,
+      "step": 3594
+    },
+    {
+      "epoch": 0.2209316617502458,
+      "grad_norm": 21.0,
+      "learning_rate": 1.8836022379313884e-05,
+      "loss": 0.4689,
+      "step": 3595
+    },
+    {
+      "epoch": 0.22099311701081611,
+      "grad_norm": 5.78125,
+      "learning_rate": 1.8834932500118148e-05,
+      "loss": 0.2945,
+      "step": 3596
+    },
+    {
+      "epoch": 0.22105457227138642,
+      "grad_norm": 19.0,
+      "learning_rate": 1.8833842142473643e-05,
+      "loss": 0.4116,
+      "step": 3597
+    },
+    {
+      "epoch": 0.22111602753195675,
+      "grad_norm": 59.0,
+      "learning_rate": 1.883275130643942e-05,
+      "loss": 0.4224,
+      "step": 3598
+    },
+    {
+      "epoch": 0.22117748279252705,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.8831659992074545e-05,
+      "loss": 0.4355,
+      "step": 3599
+    },
+    {
+      "epoch": 0.22123893805309736,
+      "grad_norm": 70.0,
+      "learning_rate": 1.883056819943812e-05,
+      "loss": 0.3068,
+      "step": 3600
+    },
+    {
+      "epoch": 0.22130039331366766,
+      "grad_norm": 11.25,
+      "learning_rate": 1.8829475928589272e-05,
+      "loss": 0.4112,
+      "step": 3601
+    },
+    {
+      "epoch": 0.22136184857423796,
+      "grad_norm": 34.0,
+      "learning_rate": 1.8828383179587147e-05,
+      "loss": 0.4113,
+      "step": 3602
+    },
+    {
+      "epoch": 0.22142330383480827,
+      "grad_norm": 23.0,
+      "learning_rate": 1.882728995249093e-05,
+      "loss": 0.4657,
+      "step": 3603
+    },
+    {
+      "epoch": 0.22148475909537857,
+      "grad_norm": 23.25,
+      "learning_rate": 1.882619624735982e-05,
+      "loss": 0.4101,
+      "step": 3604
+    },
+    {
+      "epoch": 0.22154621435594887,
+      "grad_norm": 18.75,
+      "learning_rate": 1.882510206425304e-05,
+      "loss": 0.3494,
+      "step": 3605
+    },
+    {
+      "epoch": 0.22160766961651918,
+      "grad_norm": 17.125,
+      "learning_rate": 1.8824007403229852e-05,
+      "loss": 0.3571,
+      "step": 3606
+    },
+    {
+      "epoch": 0.22166912487708948,
+      "grad_norm": 10.875,
+      "learning_rate": 1.8822912264349535e-05,
+      "loss": 0.293,
+      "step": 3607
+    },
+    {
+      "epoch": 0.22173058013765978,
+      "grad_norm": 53.5,
+      "learning_rate": 1.8821816647671394e-05,
+      "loss": 0.3555,
+      "step": 3608
+    },
+    {
+      "epoch": 0.2217920353982301,
+      "grad_norm": 4.8125,
+      "learning_rate": 1.882072055325476e-05,
+      "loss": 0.4772,
+      "step": 3609
+    },
+    {
+      "epoch": 0.2218534906588004,
+      "grad_norm": 14.125,
+      "learning_rate": 1.8819623981158996e-05,
+      "loss": 0.4706,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2219149459193707,
+      "grad_norm": 41.0,
+      "learning_rate": 1.881852693144348e-05,
+      "loss": 0.2986,
+      "step": 3611
+    },
+    {
+      "epoch": 0.221976401179941,
+      "grad_norm": 34.0,
+      "learning_rate": 1.881742940416763e-05,
+      "loss": 0.3609,
+      "step": 3612
+    },
+    {
+      "epoch": 0.2220378564405113,
+      "grad_norm": 24.625,
+      "learning_rate": 1.881633139939087e-05,
+      "loss": 0.5729,
+      "step": 3613
+    },
+    {
+      "epoch": 0.2220993117010816,
+      "grad_norm": 38.0,
+      "learning_rate": 1.8815232917172672e-05,
+      "loss": 0.3309,
+      "step": 3614
+    },
+    {
+      "epoch": 0.2221607669616519,
+      "grad_norm": 22.125,
+      "learning_rate": 1.881413395757252e-05,
+      "loss": 0.3461,
+      "step": 3615
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.8813034520649923e-05,
+      "loss": 0.4716,
+      "step": 3616
+    },
+    {
+      "epoch": 0.2222836774827925,
+      "grad_norm": 35.5,
+      "learning_rate": 1.8811934606464425e-05,
+      "loss": 0.4695,
+      "step": 3617
+    },
+    {
+      "epoch": 0.22234513274336284,
+      "grad_norm": 43.25,
+      "learning_rate": 1.8810834215075593e-05,
+      "loss": 0.7914,
+      "step": 3618
+    },
+    {
+      "epoch": 0.22240658800393315,
+      "grad_norm": 26.5,
+      "learning_rate": 1.8809733346543013e-05,
+      "loss": 0.5603,
+      "step": 3619
+    },
+    {
+      "epoch": 0.22246804326450345,
+      "grad_norm": 26.25,
+      "learning_rate": 1.8808632000926304e-05,
+      "loss": 0.4051,
+      "step": 3620
+    },
+    {
+      "epoch": 0.22252949852507375,
+      "grad_norm": 33.0,
+      "learning_rate": 1.880753017828511e-05,
+      "loss": 0.4801,
+      "step": 3621
+    },
+    {
+      "epoch": 0.22259095378564406,
+      "grad_norm": 17.75,
+      "learning_rate": 1.880642787867909e-05,
+      "loss": 0.4233,
+      "step": 3622
+    },
+    {
+      "epoch": 0.22265240904621436,
+      "grad_norm": 20.25,
+      "learning_rate": 1.880532510216795e-05,
+      "loss": 0.5419,
+      "step": 3623
+    },
+    {
+      "epoch": 0.22271386430678466,
+      "grad_norm": 40.5,
+      "learning_rate": 1.8804221848811408e-05,
+      "loss": 0.5003,
+      "step": 3624
+    },
+    {
+      "epoch": 0.22277531956735497,
+      "grad_norm": 37.25,
+      "learning_rate": 1.8803118118669203e-05,
+      "loss": 0.5687,
+      "step": 3625
+    },
+    {
+      "epoch": 0.22283677482792527,
+      "grad_norm": 4.5625,
+      "learning_rate": 1.880201391180111e-05,
+      "loss": 0.304,
+      "step": 3626
+    },
+    {
+      "epoch": 0.22289823008849557,
+      "grad_norm": 18.625,
+      "learning_rate": 1.8800909228266932e-05,
+      "loss": 0.4378,
+      "step": 3627
+    },
+    {
+      "epoch": 0.22295968534906588,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.8799804068126487e-05,
+      "loss": 0.4157,
+      "step": 3628
+    },
+    {
+      "epoch": 0.22302114060963618,
+      "grad_norm": 18.625,
+      "learning_rate": 1.879869843143962e-05,
+      "loss": 0.3903,
+      "step": 3629
+    },
+    {
+      "epoch": 0.22308259587020648,
+      "grad_norm": 30.125,
+      "learning_rate": 1.8797592318266215e-05,
+      "loss": 0.4224,
+      "step": 3630
+    },
+    {
+      "epoch": 0.2231440511307768,
+      "grad_norm": 28.625,
+      "learning_rate": 1.879648572866617e-05,
+      "loss": 0.3015,
+      "step": 3631
+    },
+    {
+      "epoch": 0.2232055063913471,
+      "grad_norm": 34.75,
+      "learning_rate": 1.8795378662699404e-05,
+      "loss": 0.4788,
+      "step": 3632
+    },
+    {
+      "epoch": 0.2232669616519174,
+      "grad_norm": 30.875,
+      "learning_rate": 1.8794271120425874e-05,
+      "loss": 0.6278,
+      "step": 3633
+    },
+    {
+      "epoch": 0.2233284169124877,
+      "grad_norm": 66.0,
+      "learning_rate": 1.8793163101905562e-05,
+      "loss": 0.7065,
+      "step": 3634
+    },
+    {
+      "epoch": 0.223389872173058,
+      "grad_norm": 25.375,
+      "learning_rate": 1.8792054607198472e-05,
+      "loss": 0.6876,
+      "step": 3635
+    },
+    {
+      "epoch": 0.2234513274336283,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8790945636364628e-05,
+      "loss": 0.6306,
+      "step": 3636
+    },
+    {
+      "epoch": 0.22351278269419864,
+      "grad_norm": 24.375,
+      "learning_rate": 1.878983618946409e-05,
+      "loss": 0.5529,
+      "step": 3637
+    },
+    {
+      "epoch": 0.22357423795476894,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.878872626655693e-05,
+      "loss": 0.4752,
+      "step": 3638
+    },
+    {
+      "epoch": 0.22363569321533924,
+      "grad_norm": 25.5,
+      "learning_rate": 1.878761586770327e-05,
+      "loss": 0.4684,
+      "step": 3639
+    },
+    {
+      "epoch": 0.22369714847590955,
+      "grad_norm": 31.875,
+      "learning_rate": 1.878650499296323e-05,
+      "loss": 0.4198,
+      "step": 3640
+    },
+    {
+      "epoch": 0.22375860373647985,
+      "grad_norm": 46.5,
+      "learning_rate": 1.8785393642396976e-05,
+      "loss": 0.7929,
+      "step": 3641
+    },
+    {
+      "epoch": 0.22382005899705015,
+      "grad_norm": 32.5,
+      "learning_rate": 1.8784281816064688e-05,
+      "loss": 0.6527,
+      "step": 3642
+    },
+    {
+      "epoch": 0.22388151425762046,
+      "grad_norm": 25.375,
+      "learning_rate": 1.878316951402658e-05,
+      "loss": 0.5229,
+      "step": 3643
+    },
+    {
+      "epoch": 0.22394296951819076,
+      "grad_norm": 10.25,
+      "learning_rate": 1.8782056736342886e-05,
+      "loss": 0.4397,
+      "step": 3644
+    },
+    {
+      "epoch": 0.22400442477876106,
+      "grad_norm": 22.25,
+      "learning_rate": 1.8780943483073867e-05,
+      "loss": 0.4247,
+      "step": 3645
+    },
+    {
+      "epoch": 0.22406588003933137,
+      "grad_norm": 16.875,
+      "learning_rate": 1.8779829754279806e-05,
+      "loss": 0.3937,
+      "step": 3646
+    },
+    {
+      "epoch": 0.22412733529990167,
+      "grad_norm": 26.125,
+      "learning_rate": 1.8778715550021025e-05,
+      "loss": 0.5254,
+      "step": 3647
+    },
+    {
+      "epoch": 0.22418879056047197,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.877760087035786e-05,
+      "loss": 0.2696,
+      "step": 3648
+    },
+    {
+      "epoch": 0.22425024582104228,
+      "grad_norm": 11.75,
+      "learning_rate": 1.8776485715350672e-05,
+      "loss": 0.3185,
+      "step": 3649
+    },
+    {
+      "epoch": 0.22431170108161258,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.8775370085059853e-05,
+      "loss": 0.3613,
+      "step": 3650
+    },
+    {
+      "epoch": 0.22437315634218288,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.877425397954582e-05,
+      "loss": 0.3302,
+      "step": 3651
+    },
+    {
+      "epoch": 0.2244346116027532,
+      "grad_norm": 7.0,
+      "learning_rate": 1.8773137398869017e-05,
+      "loss": 0.2979,
+      "step": 3652
+    },
+    {
+      "epoch": 0.2244960668633235,
+      "grad_norm": 9.0,
+      "learning_rate": 1.8772020343089906e-05,
+      "loss": 0.2752,
+      "step": 3653
+    },
+    {
+      "epoch": 0.2245575221238938,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.8770902812268983e-05,
+      "loss": 0.3746,
+      "step": 3654
+    },
+    {
+      "epoch": 0.2246189773844641,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.8769784806466768e-05,
+      "loss": 0.2763,
+      "step": 3655
+    },
+    {
+      "epoch": 0.22468043264503443,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.876866632574381e-05,
+      "loss": 0.3241,
+      "step": 3656
+    },
+    {
+      "epoch": 0.22474188790560473,
+      "grad_norm": 27.875,
+      "learning_rate": 1.8767547370160667e-05,
+      "loss": 0.4575,
+      "step": 3657
+    },
+    {
+      "epoch": 0.22480334316617503,
+      "grad_norm": 16.125,
+      "learning_rate": 1.8766427939777943e-05,
+      "loss": 0.4181,
+      "step": 3658
+    },
+    {
+      "epoch": 0.22486479842674534,
+      "grad_norm": 43.25,
+      "learning_rate": 1.8765308034656265e-05,
+      "loss": 0.5262,
+      "step": 3659
+    },
+    {
+      "epoch": 0.22492625368731564,
+      "grad_norm": 22.0,
+      "learning_rate": 1.876418765485627e-05,
+      "loss": 0.4021,
+      "step": 3660
+    },
+    {
+      "epoch": 0.22498770894788594,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.8763066800438638e-05,
+      "loss": 0.3508,
+      "step": 3661
+    },
+    {
+      "epoch": 0.22504916420845625,
+      "grad_norm": 12.25,
+      "learning_rate": 1.8761945471464065e-05,
+      "loss": 0.2922,
+      "step": 3662
+    },
+    {
+      "epoch": 0.22511061946902655,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.8760823667993277e-05,
+      "loss": 0.2793,
+      "step": 3663
+    },
+    {
+      "epoch": 0.22517207472959685,
+      "grad_norm": 16.5,
+      "learning_rate": 1.8759701390087026e-05,
+      "loss": 0.2548,
+      "step": 3664
+    },
+    {
+      "epoch": 0.22523352999016716,
+      "grad_norm": 29.75,
+      "learning_rate": 1.875857863780609e-05,
+      "loss": 0.4239,
+      "step": 3665
+    },
+    {
+      "epoch": 0.22529498525073746,
+      "grad_norm": 16.375,
+      "learning_rate": 1.875745541121126e-05,
+      "loss": 0.3824,
+      "step": 3666
+    },
+    {
+      "epoch": 0.22535644051130777,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.8756331710363375e-05,
+      "loss": 0.2915,
+      "step": 3667
+    },
+    {
+      "epoch": 0.22541789577187807,
+      "grad_norm": 17.0,
+      "learning_rate": 1.875520753532328e-05,
+      "loss": 0.3001,
+      "step": 3668
+    },
+    {
+      "epoch": 0.22547935103244837,
+      "grad_norm": 48.75,
+      "learning_rate": 1.8754082886151857e-05,
+      "loss": 0.4074,
+      "step": 3669
+    },
+    {
+      "epoch": 0.22554080629301868,
+      "grad_norm": 43.75,
+      "learning_rate": 1.8752957762910016e-05,
+      "loss": 0.3504,
+      "step": 3670
+    },
+    {
+      "epoch": 0.22560226155358898,
+      "grad_norm": 10.125,
+      "learning_rate": 1.8751832165658682e-05,
+      "loss": 0.3128,
+      "step": 3671
+    },
+    {
+      "epoch": 0.22566371681415928,
+      "grad_norm": 26.0,
+      "learning_rate": 1.8750706094458807e-05,
+      "loss": 0.3254,
+      "step": 3672
+    },
+    {
+      "epoch": 0.22572517207472959,
+      "grad_norm": 29.75,
+      "learning_rate": 1.874957954937138e-05,
+      "loss": 0.5327,
+      "step": 3673
+    },
+    {
+      "epoch": 0.2257866273352999,
+      "grad_norm": 32.75,
+      "learning_rate": 1.8748452530457403e-05,
+      "loss": 0.5976,
+      "step": 3674
+    },
+    {
+      "epoch": 0.22584808259587022,
+      "grad_norm": 38.5,
+      "learning_rate": 1.874732503777791e-05,
+      "loss": 0.4392,
+      "step": 3675
+    },
+    {
+      "epoch": 0.22590953785644052,
+      "grad_norm": 28.875,
+      "learning_rate": 1.874619707139396e-05,
+      "loss": 0.5262,
+      "step": 3676
+    },
+    {
+      "epoch": 0.22597099311701083,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.874506863136664e-05,
+      "loss": 0.4742,
+      "step": 3677
+    },
+    {
+      "epoch": 0.22603244837758113,
+      "grad_norm": 19.375,
+      "learning_rate": 1.8743939717757053e-05,
+      "loss": 0.3324,
+      "step": 3678
+    },
+    {
+      "epoch": 0.22609390363815143,
+      "grad_norm": 9.0,
+      "learning_rate": 1.8742810330626338e-05,
+      "loss": 0.3878,
+      "step": 3679
+    },
+    {
+      "epoch": 0.22615535889872174,
+      "grad_norm": 44.25,
+      "learning_rate": 1.8741680470035658e-05,
+      "loss": 0.4829,
+      "step": 3680
+    },
+    {
+      "epoch": 0.22621681415929204,
+      "grad_norm": 30.5,
+      "learning_rate": 1.8740550136046195e-05,
+      "loss": 0.4587,
+      "step": 3681
+    },
+    {
+      "epoch": 0.22627826941986234,
+      "grad_norm": 6.21875,
+      "learning_rate": 1.873941932871917e-05,
+      "loss": 0.4167,
+      "step": 3682
+    },
+    {
+      "epoch": 0.22633972468043265,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8738288048115812e-05,
+      "loss": 0.4226,
+      "step": 3683
+    },
+    {
+      "epoch": 0.22640117994100295,
+      "grad_norm": 8.375,
+      "learning_rate": 1.873715629429739e-05,
+      "loss": 0.3196,
+      "step": 3684
+    },
+    {
+      "epoch": 0.22646263520157325,
+      "grad_norm": 16.5,
+      "learning_rate": 1.8736024067325188e-05,
+      "loss": 0.3357,
+      "step": 3685
+    },
+    {
+      "epoch": 0.22652409046214356,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.8734891367260528e-05,
+      "loss": 0.2814,
+      "step": 3686
+    },
+    {
+      "epoch": 0.22658554572271386,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.873375819416474e-05,
+      "loss": 0.3215,
+      "step": 3687
+    },
+    {
+      "epoch": 0.22664700098328416,
+      "grad_norm": 5.9375,
+      "learning_rate": 1.8732624548099204e-05,
+      "loss": 0.32,
+      "step": 3688
+    },
+    {
+      "epoch": 0.22670845624385447,
+      "grad_norm": 5.9375,
+      "learning_rate": 1.87314904291253e-05,
+      "loss": 0.3085,
+      "step": 3689
+    },
+    {
+      "epoch": 0.22676991150442477,
+      "grad_norm": 29.0,
+      "learning_rate": 1.873035583730445e-05,
+      "loss": 0.5647,
+      "step": 3690
+    },
+    {
+      "epoch": 0.22683136676499507,
+      "grad_norm": 25.0,
+      "learning_rate": 1.8729220772698096e-05,
+      "loss": 0.4393,
+      "step": 3691
+    },
+    {
+      "epoch": 0.22689282202556538,
+      "grad_norm": 20.625,
+      "learning_rate": 1.872808523536771e-05,
+      "loss": 0.4552,
+      "step": 3692
+    },
+    {
+      "epoch": 0.22695427728613568,
+      "grad_norm": 13.375,
+      "learning_rate": 1.8726949225374782e-05,
+      "loss": 0.4206,
+      "step": 3693
+    },
+    {
+      "epoch": 0.227015732546706,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.8725812742780832e-05,
+      "loss": 0.2847,
+      "step": 3694
+    },
+    {
+      "epoch": 0.22707718780727632,
+      "grad_norm": 12.75,
+      "learning_rate": 1.8724675787647405e-05,
+      "loss": 0.3417,
+      "step": 3695
+    },
+    {
+      "epoch": 0.22713864306784662,
+      "grad_norm": 6.875,
+      "learning_rate": 1.872353836003608e-05,
+      "loss": 0.2654,
+      "step": 3696
+    },
+    {
+      "epoch": 0.22720009832841692,
+      "grad_norm": 18.875,
+      "learning_rate": 1.8722400460008437e-05,
+      "loss": 0.5002,
+      "step": 3697
+    },
+    {
+      "epoch": 0.22726155358898723,
+      "grad_norm": 17.625,
+      "learning_rate": 1.8721262087626113e-05,
+      "loss": 0.4191,
+      "step": 3698
+    },
+    {
+      "epoch": 0.22732300884955753,
+      "grad_norm": 16.125,
+      "learning_rate": 1.872012324295075e-05,
+      "loss": 0.3985,
+      "step": 3699
+    },
+    {
+      "epoch": 0.22738446411012783,
+      "grad_norm": 4.3125,
+      "learning_rate": 1.871898392604402e-05,
+      "loss": 0.3253,
+      "step": 3700
+    },
+    {
+      "epoch": 0.22744591937069814,
+      "grad_norm": 21.125,
+      "learning_rate": 1.8717844136967626e-05,
+      "loss": 0.3465,
+      "step": 3701
+    },
+    {
+      "epoch": 0.22750737463126844,
+      "grad_norm": 33.75,
+      "learning_rate": 1.8716703875783287e-05,
+      "loss": 0.4148,
+      "step": 3702
+    },
+    {
+      "epoch": 0.22756882989183874,
+      "grad_norm": 4.96875,
+      "learning_rate": 1.8715563142552758e-05,
+      "loss": 0.4097,
+      "step": 3703
+    },
+    {
+      "epoch": 0.22763028515240905,
+      "grad_norm": 7.75,
+      "learning_rate": 1.8714421937337812e-05,
+      "loss": 0.277,
+      "step": 3704
+    },
+    {
+      "epoch": 0.22769174041297935,
+      "grad_norm": 14.0,
+      "learning_rate": 1.8713280260200245e-05,
+      "loss": 0.3533,
+      "step": 3705
+    },
+    {
+      "epoch": 0.22775319567354965,
+      "grad_norm": 15.75,
+      "learning_rate": 1.8712138111201898e-05,
+      "loss": 0.3028,
+      "step": 3706
+    },
+    {
+      "epoch": 0.22781465093411996,
+      "grad_norm": 18.625,
+      "learning_rate": 1.871099549040461e-05,
+      "loss": 0.3031,
+      "step": 3707
+    },
+    {
+      "epoch": 0.22787610619469026,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.870985239787026e-05,
+      "loss": 0.4144,
+      "step": 3708
+    },
+    {
+      "epoch": 0.22793756145526056,
+      "grad_norm": 19.125,
+      "learning_rate": 1.8708708833660755e-05,
+      "loss": 0.3397,
+      "step": 3709
+    },
+    {
+      "epoch": 0.22799901671583087,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.8707564797838024e-05,
+      "loss": 0.2869,
+      "step": 3710
+    },
+    {
+      "epoch": 0.22806047197640117,
+      "grad_norm": 18.5,
+      "learning_rate": 1.870642029046402e-05,
+      "loss": 0.3242,
+      "step": 3711
+    },
+    {
+      "epoch": 0.22812192723697147,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.8705275311600724e-05,
+      "loss": 0.2222,
+      "step": 3712
+    },
+    {
+      "epoch": 0.22818338249754178,
+      "grad_norm": 38.25,
+      "learning_rate": 1.8704129861310137e-05,
+      "loss": 0.5253,
+      "step": 3713
+    },
+    {
+      "epoch": 0.2282448377581121,
+      "grad_norm": 12.125,
+      "learning_rate": 1.8702983939654296e-05,
+      "loss": 0.4688,
+      "step": 3714
+    },
+    {
+      "epoch": 0.2283062930186824,
+      "grad_norm": 28.25,
+      "learning_rate": 1.870183754669526e-05,
+      "loss": 0.4871,
+      "step": 3715
+    },
+    {
+      "epoch": 0.2283677482792527,
+      "grad_norm": 32.75,
+      "learning_rate": 1.87006906824951e-05,
+      "loss": 0.508,
+      "step": 3716
+    },
+    {
+      "epoch": 0.22842920353982302,
+      "grad_norm": 24.25,
+      "learning_rate": 1.8699543347115932e-05,
+      "loss": 0.7204,
+      "step": 3717
+    },
+    {
+      "epoch": 0.22849065880039332,
+      "grad_norm": 18.125,
+      "learning_rate": 1.8698395540619883e-05,
+      "loss": 0.4022,
+      "step": 3718
+    },
+    {
+      "epoch": 0.22855211406096362,
+      "grad_norm": 30.125,
+      "learning_rate": 1.8697247263069117e-05,
+      "loss": 0.4663,
+      "step": 3719
+    },
+    {
+      "epoch": 0.22861356932153393,
+      "grad_norm": 33.0,
+      "learning_rate": 1.8696098514525817e-05,
+      "loss": 0.611,
+      "step": 3720
+    },
+    {
+      "epoch": 0.22867502458210423,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.869494929505219e-05,
+      "loss": 0.4505,
+      "step": 3721
+    },
+    {
+      "epoch": 0.22873647984267453,
+      "grad_norm": 46.0,
+      "learning_rate": 1.8693799604710476e-05,
+      "loss": 0.3797,
+      "step": 3722
+    },
+    {
+      "epoch": 0.22879793510324484,
+      "grad_norm": 10.625,
+      "learning_rate": 1.869264944356293e-05,
+      "loss": 0.3709,
+      "step": 3723
+    },
+    {
+      "epoch": 0.22885939036381514,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.869149881167184e-05,
+      "loss": 0.419,
+      "step": 3724
+    },
+    {
+      "epoch": 0.22892084562438544,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.8690347709099522e-05,
+      "loss": 0.3723,
+      "step": 3725
+    },
+    {
+      "epoch": 0.22898230088495575,
+      "grad_norm": 25.25,
+      "learning_rate": 1.8689196135908303e-05,
+      "loss": 0.4392,
+      "step": 3726
+    },
+    {
+      "epoch": 0.22904375614552605,
+      "grad_norm": 23.25,
+      "learning_rate": 1.8688044092160554e-05,
+      "loss": 0.326,
+      "step": 3727
+    },
+    {
+      "epoch": 0.22910521140609635,
+      "grad_norm": 18.625,
+      "learning_rate": 1.868689157791866e-05,
+      "loss": 0.291,
+      "step": 3728
+    },
+    {
+      "epoch": 0.22916666666666666,
+      "grad_norm": 17.75,
+      "learning_rate": 1.8685738593245038e-05,
+      "loss": 0.4228,
+      "step": 3729
+    },
+    {
+      "epoch": 0.22922812192723696,
+      "grad_norm": 24.5,
+      "learning_rate": 1.8684585138202122e-05,
+      "loss": 0.3606,
+      "step": 3730
+    },
+    {
+      "epoch": 0.22928957718780726,
+      "grad_norm": 22.25,
+      "learning_rate": 1.868343121285238e-05,
+      "loss": 0.4152,
+      "step": 3731
+    },
+    {
+      "epoch": 0.22935103244837757,
+      "grad_norm": 20.375,
+      "learning_rate": 1.86822768172583e-05,
+      "loss": 0.4731,
+      "step": 3732
+    },
+    {
+      "epoch": 0.2294124877089479,
+      "grad_norm": 18.5,
+      "learning_rate": 1.8681121951482397e-05,
+      "loss": 0.4204,
+      "step": 3733
+    },
+    {
+      "epoch": 0.2294739429695182,
+      "grad_norm": 30.5,
+      "learning_rate": 1.867996661558721e-05,
+      "loss": 0.4416,
+      "step": 3734
+    },
+    {
+      "epoch": 0.2295353982300885,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.867881080963531e-05,
+      "loss": 0.3793,
+      "step": 3735
+    },
+    {
+      "epoch": 0.2295968534906588,
+      "grad_norm": 35.75,
+      "learning_rate": 1.8677654533689287e-05,
+      "loss": 0.5674,
+      "step": 3736
+    },
+    {
+      "epoch": 0.2296583087512291,
+      "grad_norm": 17.125,
+      "learning_rate": 1.867649778781176e-05,
+      "loss": 0.2961,
+      "step": 3737
+    },
+    {
+      "epoch": 0.22971976401179942,
+      "grad_norm": 26.25,
+      "learning_rate": 1.8675340572065365e-05,
+      "loss": 0.4206,
+      "step": 3738
+    },
+    {
+      "epoch": 0.22978121927236972,
+      "grad_norm": 5.875,
+      "learning_rate": 1.8674182886512776e-05,
+      "loss": 0.2942,
+      "step": 3739
+    },
+    {
+      "epoch": 0.22984267453294002,
+      "grad_norm": 5.96875,
+      "learning_rate": 1.8673024731216687e-05,
+      "loss": 0.2738,
+      "step": 3740
+    },
+    {
+      "epoch": 0.22990412979351033,
+      "grad_norm": 44.25,
+      "learning_rate": 1.8671866106239812e-05,
+      "loss": 0.274,
+      "step": 3741
+    },
+    {
+      "epoch": 0.22996558505408063,
+      "grad_norm": 32.5,
+      "learning_rate": 1.86707070116449e-05,
+      "loss": 0.3128,
+      "step": 3742
+    },
+    {
+      "epoch": 0.23002704031465093,
+      "grad_norm": 17.75,
+      "learning_rate": 1.8669547447494723e-05,
+      "loss": 0.2851,
+      "step": 3743
+    },
+    {
+      "epoch": 0.23008849557522124,
+      "grad_norm": 9.875,
+      "learning_rate": 1.866838741385207e-05,
+      "loss": 0.3147,
+      "step": 3744
+    },
+    {
+      "epoch": 0.23014995083579154,
+      "grad_norm": 19.125,
+      "learning_rate": 1.8667226910779767e-05,
+      "loss": 0.4185,
+      "step": 3745
+    },
+    {
+      "epoch": 0.23021140609636184,
+      "grad_norm": 18.875,
+      "learning_rate": 1.866606593834065e-05,
+      "loss": 0.3311,
+      "step": 3746
+    },
+    {
+      "epoch": 0.23027286135693215,
+      "grad_norm": 38.0,
+      "learning_rate": 1.8664904496597605e-05,
+      "loss": 0.4375,
+      "step": 3747
+    },
+    {
+      "epoch": 0.23033431661750245,
+      "grad_norm": 25.75,
+      "learning_rate": 1.866374258561352e-05,
+      "loss": 0.4427,
+      "step": 3748
+    },
+    {
+      "epoch": 0.23039577187807275,
+      "grad_norm": 12.125,
+      "learning_rate": 1.866258020545132e-05,
+      "loss": 0.3772,
+      "step": 3749
+    },
+    {
+      "epoch": 0.23045722713864306,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.866141735617395e-05,
+      "loss": 0.3149,
+      "step": 3750
+    },
+    {
+      "epoch": 0.23051868239921336,
+      "grad_norm": 10.875,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 0.3991,
+      "step": 3751
+    },
+    {
+      "epoch": 0.2305801376597837,
+      "grad_norm": 21.25,
+      "learning_rate": 1.865909025052563e-05,
+      "loss": 0.254,
+      "step": 3752
+    },
+    {
+      "epoch": 0.230641592920354,
+      "grad_norm": 22.875,
+      "learning_rate": 1.86579259942807e-05,
+      "loss": 0.4425,
+      "step": 3753
+    },
+    {
+      "epoch": 0.2307030481809243,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.8656761269172645e-05,
+      "loss": 0.6698,
+      "step": 3754
+    },
+    {
+      "epoch": 0.2307645034414946,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.8655596075264543e-05,
+      "loss": 0.3844,
+      "step": 3755
+    },
+    {
+      "epoch": 0.2308259587020649,
+      "grad_norm": 17.5,
+      "learning_rate": 1.8654430412619494e-05,
+      "loss": 0.5303,
+      "step": 3756
+    },
+    {
+      "epoch": 0.2308874139626352,
+      "grad_norm": 28.875,
+      "learning_rate": 1.8653264281300622e-05,
+      "loss": 0.4399,
+      "step": 3757
+    },
+    {
+      "epoch": 0.2309488692232055,
+      "grad_norm": 41.75,
+      "learning_rate": 1.8652097681371076e-05,
+      "loss": 0.5223,
+      "step": 3758
+    },
+    {
+      "epoch": 0.23101032448377581,
+      "grad_norm": 3.96875,
+      "learning_rate": 1.865093061289404e-05,
+      "loss": 0.2942,
+      "step": 3759
+    },
+    {
+      "epoch": 0.23107177974434612,
+      "grad_norm": 15.0,
+      "learning_rate": 1.864976307593271e-05,
+      "loss": 0.3056,
+      "step": 3760
+    },
+    {
+      "epoch": 0.23113323500491642,
+      "grad_norm": 28.0,
+      "learning_rate": 1.8648595070550312e-05,
+      "loss": 0.4433,
+      "step": 3761
+    },
+    {
+      "epoch": 0.23119469026548672,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.86474265968101e-05,
+      "loss": 0.4442,
+      "step": 3762
+    },
+    {
+      "epoch": 0.23125614552605703,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.864625765477535e-05,
+      "loss": 0.3193,
+      "step": 3763
+    },
+    {
+      "epoch": 0.23131760078662733,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.864508824450937e-05,
+      "loss": 0.3706,
+      "step": 3764
+    },
+    {
+      "epoch": 0.23137905604719763,
+      "grad_norm": 10.0,
+      "learning_rate": 1.8643918366075485e-05,
+      "loss": 0.337,
+      "step": 3765
+    },
+    {
+      "epoch": 0.23144051130776794,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.864274801953705e-05,
+      "loss": 0.4285,
+      "step": 3766
+    },
+    {
+      "epoch": 0.23150196656833824,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.864157720495744e-05,
+      "loss": 0.439,
+      "step": 3767
+    },
+    {
+      "epoch": 0.23156342182890854,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.8640405922400068e-05,
+      "loss": 0.3328,
+      "step": 3768
+    },
+    {
+      "epoch": 0.23162487708947885,
+      "grad_norm": 23.875,
+      "learning_rate": 1.8639234171928355e-05,
+      "loss": 0.4916,
+      "step": 3769
+    },
+    {
+      "epoch": 0.23168633235004915,
+      "grad_norm": 17.625,
+      "learning_rate": 1.8638061953605762e-05,
+      "loss": 0.4495,
+      "step": 3770
+    },
+    {
+      "epoch": 0.23174778761061948,
+      "grad_norm": 11.0625,
+      "learning_rate": 1.8636889267495767e-05,
+      "loss": 0.4177,
+      "step": 3771
+    },
+    {
+      "epoch": 0.23180924287118979,
+      "grad_norm": 14.0,
+      "learning_rate": 1.8635716113661876e-05,
+      "loss": 0.3727,
+      "step": 3772
+    },
+    {
+      "epoch": 0.2318706981317601,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.863454249216762e-05,
+      "loss": 0.3415,
+      "step": 3773
+    },
+    {
+      "epoch": 0.2319321533923304,
+      "grad_norm": 23.625,
+      "learning_rate": 1.8633368403076558e-05,
+      "loss": 0.4788,
+      "step": 3774
+    },
+    {
+      "epoch": 0.2319936086529007,
+      "grad_norm": 15.875,
+      "learning_rate": 1.863219384645227e-05,
+      "loss": 0.4249,
+      "step": 3775
+    },
+    {
+      "epoch": 0.232055063913471,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.8631018822358363e-05,
+      "loss": 0.288,
+      "step": 3776
+    },
+    {
+      "epoch": 0.2321165191740413,
+      "grad_norm": 27.25,
+      "learning_rate": 1.8629843330858472e-05,
+      "loss": 0.6041,
+      "step": 3777
+    },
+    {
+      "epoch": 0.2321779744346116,
+      "grad_norm": 20.25,
+      "learning_rate": 1.862866737201625e-05,
+      "loss": 0.3561,
+      "step": 3778
+    },
+    {
+      "epoch": 0.2322394296951819,
+      "grad_norm": 9.125,
+      "learning_rate": 1.8627490945895387e-05,
+      "loss": 0.6297,
+      "step": 3779
+    },
+    {
+      "epoch": 0.2323008849557522,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.8626314052559583e-05,
+      "loss": 0.611,
+      "step": 3780
+    },
+    {
+      "epoch": 0.23236234021632252,
+      "grad_norm": 27.5,
+      "learning_rate": 1.8625136692072577e-05,
+      "loss": 0.4662,
+      "step": 3781
+    },
+    {
+      "epoch": 0.23242379547689282,
+      "grad_norm": 30.5,
+      "learning_rate": 1.8623958864498125e-05,
+      "loss": 0.4833,
+      "step": 3782
+    },
+    {
+      "epoch": 0.23248525073746312,
+      "grad_norm": 26.875,
+      "learning_rate": 1.8622780569900016e-05,
+      "loss": 0.4608,
+      "step": 3783
+    },
+    {
+      "epoch": 0.23254670599803343,
+      "grad_norm": 17.25,
+      "learning_rate": 1.862160180834206e-05,
+      "loss": 0.6771,
+      "step": 3784
+    },
+    {
+      "epoch": 0.23260816125860373,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.8620422579888082e-05,
+      "loss": 0.4398,
+      "step": 3785
+    },
+    {
+      "epoch": 0.23266961651917403,
+      "grad_norm": 35.5,
+      "learning_rate": 1.8619242884601953e-05,
+      "loss": 0.5837,
+      "step": 3786
+    },
+    {
+      "epoch": 0.23273107177974434,
+      "grad_norm": 10.125,
+      "learning_rate": 1.861806272254755e-05,
+      "loss": 0.5081,
+      "step": 3787
+    },
+    {
+      "epoch": 0.23279252704031464,
+      "grad_norm": 16.0,
+      "learning_rate": 1.8616882093788792e-05,
+      "loss": 0.3637,
+      "step": 3788
+    },
+    {
+      "epoch": 0.23285398230088494,
+      "grad_norm": 14.5,
+      "learning_rate": 1.8615700998389612e-05,
+      "loss": 0.3568,
+      "step": 3789
+    },
+    {
+      "epoch": 0.23291543756145527,
+      "grad_norm": 21.5,
+      "learning_rate": 1.8614519436413968e-05,
+      "loss": 0.4398,
+      "step": 3790
+    },
+    {
+      "epoch": 0.23297689282202558,
+      "grad_norm": 28.125,
+      "learning_rate": 1.8613337407925854e-05,
+      "loss": 0.3344,
+      "step": 3791
+    },
+    {
+      "epoch": 0.23303834808259588,
+      "grad_norm": 19.125,
+      "learning_rate": 1.861215491298927e-05,
+      "loss": 0.3715,
+      "step": 3792
+    },
+    {
+      "epoch": 0.23309980334316618,
+      "grad_norm": 5.40625,
+      "learning_rate": 1.8610971951668265e-05,
+      "loss": 0.3784,
+      "step": 3793
+    },
+    {
+      "epoch": 0.2331612586037365,
+      "grad_norm": 44.75,
+      "learning_rate": 1.8609788524026894e-05,
+      "loss": 0.6965,
+      "step": 3794
+    },
+    {
+      "epoch": 0.2332227138643068,
+      "grad_norm": 21.875,
+      "learning_rate": 1.8608604630129247e-05,
+      "loss": 0.4643,
+      "step": 3795
+    },
+    {
+      "epoch": 0.2332841691248771,
+      "grad_norm": 37.0,
+      "learning_rate": 1.860742027003944e-05,
+      "loss": 0.5774,
+      "step": 3796
+    },
+    {
+      "epoch": 0.2333456243854474,
+      "grad_norm": 17.5,
+      "learning_rate": 1.8606235443821602e-05,
+      "loss": 0.428,
+      "step": 3797
+    },
+    {
+      "epoch": 0.2334070796460177,
+      "grad_norm": 30.0,
+      "learning_rate": 1.8605050151539906e-05,
+      "loss": 0.3946,
+      "step": 3798
+    },
+    {
+      "epoch": 0.233468534906588,
+      "grad_norm": 10.625,
+      "learning_rate": 1.8603864393258534e-05,
+      "loss": 0.4832,
+      "step": 3799
+    },
+    {
+      "epoch": 0.2335299901671583,
+      "grad_norm": 26.625,
+      "learning_rate": 1.8602678169041705e-05,
+      "loss": 0.3553,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2335914454277286,
+      "grad_norm": 13.25,
+      "learning_rate": 1.860149147895366e-05,
+      "loss": 0.3103,
+      "step": 3801
+    },
+    {
+      "epoch": 0.23365290068829891,
+      "grad_norm": 22.75,
+      "learning_rate": 1.860030432305865e-05,
+      "loss": 0.4174,
+      "step": 3802
+    },
+    {
+      "epoch": 0.23371435594886922,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.8599116701420974e-05,
+      "loss": 0.3957,
+      "step": 3803
+    },
+    {
+      "epoch": 0.23377581120943952,
+      "grad_norm": 12.875,
+      "learning_rate": 1.859792861410495e-05,
+      "loss": 0.2933,
+      "step": 3804
+    },
+    {
+      "epoch": 0.23383726647000982,
+      "grad_norm": 29.125,
+      "learning_rate": 1.8596740061174912e-05,
+      "loss": 0.3773,
+      "step": 3805
+    },
+    {
+      "epoch": 0.23389872173058013,
+      "grad_norm": 5.59375,
+      "learning_rate": 1.859555104269523e-05,
+      "loss": 0.3295,
+      "step": 3806
+    },
+    {
+      "epoch": 0.23396017699115043,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8594361558730287e-05,
+      "loss": 0.2904,
+      "step": 3807
+    },
+    {
+      "epoch": 0.23402163225172073,
+      "grad_norm": 25.5,
+      "learning_rate": 1.8593171609344505e-05,
+      "loss": 0.485,
+      "step": 3808
+    },
+    {
+      "epoch": 0.23408308751229107,
+      "grad_norm": 10.25,
+      "learning_rate": 1.8591981194602322e-05,
+      "loss": 0.4362,
+      "step": 3809
+    },
+    {
+      "epoch": 0.23414454277286137,
+      "grad_norm": 34.0,
+      "learning_rate": 1.8590790314568207e-05,
+      "loss": 0.5425,
+      "step": 3810
+    },
+    {
+      "epoch": 0.23420599803343167,
+      "grad_norm": 18.875,
+      "learning_rate": 1.8589598969306646e-05,
+      "loss": 0.396,
+      "step": 3811
+    },
+    {
+      "epoch": 0.23426745329400198,
+      "grad_norm": 21.375,
+      "learning_rate": 1.8588407158882157e-05,
+      "loss": 0.3493,
+      "step": 3812
+    },
+    {
+      "epoch": 0.23432890855457228,
+      "grad_norm": 23.375,
+      "learning_rate": 1.858721488335929e-05,
+      "loss": 0.3823,
+      "step": 3813
+    },
+    {
+      "epoch": 0.23439036381514258,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.8586022142802597e-05,
+      "loss": 0.3937,
+      "step": 3814
+    },
+    {
+      "epoch": 0.2344518190757129,
+      "grad_norm": 26.375,
+      "learning_rate": 1.858482893727668e-05,
+      "loss": 0.3148,
+      "step": 3815
+    },
+    {
+      "epoch": 0.2345132743362832,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.8583635266846155e-05,
+      "loss": 0.3068,
+      "step": 3816
+    },
+    {
+      "epoch": 0.2345747295968535,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.8582441131575658e-05,
+      "loss": 0.3567,
+      "step": 3817
+    },
+    {
+      "epoch": 0.2346361848574238,
+      "grad_norm": 17.125,
+      "learning_rate": 1.8581246531529868e-05,
+      "loss": 0.3456,
+      "step": 3818
+    },
+    {
+      "epoch": 0.2346976401179941,
+      "grad_norm": 41.25,
+      "learning_rate": 1.8580051466773465e-05,
+      "loss": 0.3157,
+      "step": 3819
+    },
+    {
+      "epoch": 0.2347590953785644,
+      "grad_norm": 5.15625,
+      "learning_rate": 1.8578855937371176e-05,
+      "loss": 0.2712,
+      "step": 3820
+    },
+    {
+      "epoch": 0.2348205506391347,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.8577659943387737e-05,
+      "loss": 0.362,
+      "step": 3821
+    },
+    {
+      "epoch": 0.234882005899705,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.8576463484887927e-05,
+      "loss": 0.2734,
+      "step": 3822
+    },
+    {
+      "epoch": 0.2349434611602753,
+      "grad_norm": 4.90625,
+      "learning_rate": 1.8575266561936526e-05,
+      "loss": 0.2876,
+      "step": 3823
+    },
+    {
+      "epoch": 0.23500491642084562,
+      "grad_norm": 16.125,
+      "learning_rate": 1.857406917459836e-05,
+      "loss": 0.3983,
+      "step": 3824
+    },
+    {
+      "epoch": 0.23506637168141592,
+      "grad_norm": 46.0,
+      "learning_rate": 1.8572871322938267e-05,
+      "loss": 0.5058,
+      "step": 3825
+    },
+    {
+      "epoch": 0.23512782694198622,
+      "grad_norm": 36.5,
+      "learning_rate": 1.8571673007021124e-05,
+      "loss": 0.5802,
+      "step": 3826
+    },
+    {
+      "epoch": 0.23518928220255653,
+      "grad_norm": 41.0,
+      "learning_rate": 1.857047422691182e-05,
+      "loss": 0.4912,
+      "step": 3827
+    },
+    {
+      "epoch": 0.23525073746312683,
+      "grad_norm": 10.4375,
+      "learning_rate": 1.8569274982675273e-05,
+      "loss": 0.5207,
+      "step": 3828
+    },
+    {
+      "epoch": 0.23531219272369716,
+      "grad_norm": 13.375,
+      "learning_rate": 1.856807527437643e-05,
+      "loss": 0.4101,
+      "step": 3829
+    },
+    {
+      "epoch": 0.23537364798426746,
+      "grad_norm": 39.75,
+      "learning_rate": 1.856687510208026e-05,
+      "loss": 0.6383,
+      "step": 3830
+    },
+    {
+      "epoch": 0.23543510324483777,
+      "grad_norm": 45.0,
+      "learning_rate": 1.8565674465851753e-05,
+      "loss": 0.4806,
+      "step": 3831
+    },
+    {
+      "epoch": 0.23549655850540807,
+      "grad_norm": 10.75,
+      "learning_rate": 1.8564473365755936e-05,
+      "loss": 0.4865,
+      "step": 3832
+    },
+    {
+      "epoch": 0.23555801376597837,
+      "grad_norm": 18.75,
+      "learning_rate": 1.8563271801857844e-05,
+      "loss": 0.3519,
+      "step": 3833
+    },
+    {
+      "epoch": 0.23561946902654868,
+      "grad_norm": 39.0,
+      "learning_rate": 1.8562069774222556e-05,
+      "loss": 0.5569,
+      "step": 3834
+    },
+    {
+      "epoch": 0.23568092428711898,
+      "grad_norm": 24.375,
+      "learning_rate": 1.8560867282915164e-05,
+      "loss": 0.2704,
+      "step": 3835
+    },
+    {
+      "epoch": 0.23574237954768928,
+      "grad_norm": 23.375,
+      "learning_rate": 1.8559664328000782e-05,
+      "loss": 0.3785,
+      "step": 3836
+    },
+    {
+      "epoch": 0.2358038348082596,
+      "grad_norm": 19.0,
+      "learning_rate": 1.8558460909544564e-05,
+      "loss": 0.6539,
+      "step": 3837
+    },
+    {
+      "epoch": 0.2358652900688299,
+      "grad_norm": 17.25,
+      "learning_rate": 1.8557257027611677e-05,
+      "loss": 0.4072,
+      "step": 3838
+    },
+    {
+      "epoch": 0.2359267453294002,
+      "grad_norm": 9.375,
+      "learning_rate": 1.8556052682267314e-05,
+      "loss": 0.2761,
+      "step": 3839
+    },
+    {
+      "epoch": 0.2359882005899705,
+      "grad_norm": 22.375,
+      "learning_rate": 1.8554847873576696e-05,
+      "loss": 0.3507,
+      "step": 3840
+    },
+    {
+      "epoch": 0.2360496558505408,
+      "grad_norm": 32.5,
+      "learning_rate": 1.855364260160507e-05,
+      "loss": 0.4804,
+      "step": 3841
+    },
+    {
+      "epoch": 0.2361111111111111,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.8552436866417703e-05,
+      "loss": 0.3391,
+      "step": 3842
+    },
+    {
+      "epoch": 0.2361725663716814,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.8551230668079895e-05,
+      "loss": 0.4324,
+      "step": 3843
+    },
+    {
+      "epoch": 0.2362340216322517,
+      "grad_norm": 25.0,
+      "learning_rate": 1.8550024006656967e-05,
+      "loss": 0.5259,
+      "step": 3844
+    },
+    {
+      "epoch": 0.23629547689282202,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8548816882214262e-05,
+      "loss": 0.4221,
+      "step": 3845
+    },
+    {
+      "epoch": 0.23635693215339232,
+      "grad_norm": 17.125,
+      "learning_rate": 1.854760929481715e-05,
+      "loss": 0.4661,
+      "step": 3846
+    },
+    {
+      "epoch": 0.23641838741396262,
+      "grad_norm": 7.875,
+      "learning_rate": 1.854640124453103e-05,
+      "loss": 0.3656,
+      "step": 3847
+    },
+    {
+      "epoch": 0.23647984267453295,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.8545192731421323e-05,
+      "loss": 0.5368,
+      "step": 3848
+    },
+    {
+      "epoch": 0.23654129793510326,
+      "grad_norm": 21.5,
+      "learning_rate": 1.8543983755553474e-05,
+      "loss": 0.4362,
+      "step": 3849
+    },
+    {
+      "epoch": 0.23660275319567356,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.8542774316992953e-05,
+      "loss": 0.3799,
+      "step": 3850
+    },
+    {
+      "epoch": 0.23666420845624386,
+      "grad_norm": 18.375,
+      "learning_rate": 1.854156441580526e-05,
+      "loss": 0.4987,
+      "step": 3851
+    },
+    {
+      "epoch": 0.23672566371681417,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8540354052055908e-05,
+      "loss": 0.4421,
+      "step": 3852
+    },
+    {
+      "epoch": 0.23678711897738447,
+      "grad_norm": 5.125,
+      "learning_rate": 1.8539143225810453e-05,
+      "loss": 0.3247,
+      "step": 3853
+    },
+    {
+      "epoch": 0.23684857423795477,
+      "grad_norm": 4.6875,
+      "learning_rate": 1.853793193713446e-05,
+      "loss": 0.2877,
+      "step": 3854
+    },
+    {
+      "epoch": 0.23691002949852508,
+      "grad_norm": 22.0,
+      "learning_rate": 1.8536720186093528e-05,
+      "loss": 0.2818,
+      "step": 3855
+    },
+    {
+      "epoch": 0.23697148475909538,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.8535507972753275e-05,
+      "loss": 0.2693,
+      "step": 3856
+    },
+    {
+      "epoch": 0.23703294001966568,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8534295297179354e-05,
+      "loss": 0.3294,
+      "step": 3857
+    },
+    {
+      "epoch": 0.237094395280236,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.8533082159437433e-05,
+      "loss": 0.4744,
+      "step": 3858
+    },
+    {
+      "epoch": 0.2371558505408063,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8531868559593205e-05,
+      "loss": 0.3113,
+      "step": 3859
+    },
+    {
+      "epoch": 0.2372173058013766,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8530654497712395e-05,
+      "loss": 0.4113,
+      "step": 3860
+    },
+    {
+      "epoch": 0.2372787610619469,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.852943997386075e-05,
+      "loss": 0.3171,
+      "step": 3861
+    },
+    {
+      "epoch": 0.2373402163225172,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.8528224988104044e-05,
+      "loss": 0.2275,
+      "step": 3862
+    },
+    {
+      "epoch": 0.2374016715830875,
+      "grad_norm": 21.625,
+      "learning_rate": 1.852700954050807e-05,
+      "loss": 0.4057,
+      "step": 3863
+    },
+    {
+      "epoch": 0.2374631268436578,
+      "grad_norm": 36.75,
+      "learning_rate": 1.8525793631138647e-05,
+      "loss": 0.3626,
+      "step": 3864
+    },
+    {
+      "epoch": 0.2375245821042281,
+      "grad_norm": 24.625,
+      "learning_rate": 1.8524577260061628e-05,
+      "loss": 0.3797,
+      "step": 3865
+    },
+    {
+      "epoch": 0.2375860373647984,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.8523360427342877e-05,
+      "loss": 0.6008,
+      "step": 3866
+    },
+    {
+      "epoch": 0.23764749262536874,
+      "grad_norm": 8.625,
+      "learning_rate": 1.8522143133048298e-05,
+      "loss": 0.3107,
+      "step": 3867
+    },
+    {
+      "epoch": 0.23770894788593905,
+      "grad_norm": 33.75,
+      "learning_rate": 1.8520925377243812e-05,
+      "loss": 0.3394,
+      "step": 3868
+    },
+    {
+      "epoch": 0.23777040314650935,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.851970715999536e-05,
+      "loss": 0.3601,
+      "step": 3869
+    },
+    {
+      "epoch": 0.23783185840707965,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.8518488481368915e-05,
+      "loss": 0.2946,
+      "step": 3870
+    },
+    {
+      "epoch": 0.23789331366764996,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.851726934143048e-05,
+      "loss": 0.27,
+      "step": 3871
+    },
+    {
+      "epoch": 0.23795476892822026,
+      "grad_norm": 5.28125,
+      "learning_rate": 1.8516049740246068e-05,
+      "loss": 0.361,
+      "step": 3872
+    },
+    {
+      "epoch": 0.23801622418879056,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.851482967788173e-05,
+      "loss": 0.3298,
+      "step": 3873
+    },
+    {
+      "epoch": 0.23807767944936087,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.8513609154403535e-05,
+      "loss": 0.3937,
+      "step": 3874
+    },
+    {
+      "epoch": 0.23813913470993117,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.8512388169877587e-05,
+      "loss": 0.4005,
+      "step": 3875
+    },
+    {
+      "epoch": 0.23820058997050148,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.8511166724369997e-05,
+      "loss": 0.3274,
+      "step": 3876
+    },
+    {
+      "epoch": 0.23826204523107178,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.850994481794692e-05,
+      "loss": 0.3153,
+      "step": 3877
+    },
+    {
+      "epoch": 0.23832350049164208,
+      "grad_norm": 11.375,
+      "learning_rate": 1.850872245067452e-05,
+      "loss": 0.3425,
+      "step": 3878
+    },
+    {
+      "epoch": 0.23838495575221239,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.8507499622619e-05,
+      "loss": 0.2811,
+      "step": 3879
+    },
+    {
+      "epoch": 0.2384464110127827,
+      "grad_norm": 12.25,
+      "learning_rate": 1.850627633384658e-05,
+      "loss": 0.4582,
+      "step": 3880
+    },
+    {
+      "epoch": 0.238507866273353,
+      "grad_norm": 4.125,
+      "learning_rate": 1.85050525844235e-05,
+      "loss": 0.3322,
+      "step": 3881
+    },
+    {
+      "epoch": 0.2385693215339233,
+      "grad_norm": 55.5,
+      "learning_rate": 1.8503828374416038e-05,
+      "loss": 0.3364,
+      "step": 3882
+    },
+    {
+      "epoch": 0.2386307767944936,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.8502603703890488e-05,
+      "loss": 0.3438,
+      "step": 3883
+    },
+    {
+      "epoch": 0.2386922320550639,
+      "grad_norm": 42.75,
+      "learning_rate": 1.850137857291317e-05,
+      "loss": 0.3413,
+      "step": 3884
+    },
+    {
+      "epoch": 0.2387536873156342,
+      "grad_norm": 5.375,
+      "learning_rate": 1.850015298155043e-05,
+      "loss": 0.3213,
+      "step": 3885
+    },
+    {
+      "epoch": 0.23881514257620454,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.849892692986864e-05,
+      "loss": 0.3287,
+      "step": 3886
+    },
+    {
+      "epoch": 0.23887659783677484,
+      "grad_norm": 34.75,
+      "learning_rate": 1.8497700417934195e-05,
+      "loss": 0.3921,
+      "step": 3887
+    },
+    {
+      "epoch": 0.23893805309734514,
+      "grad_norm": 35.25,
+      "learning_rate": 1.849647344581352e-05,
+      "loss": 0.3449,
+      "step": 3888
+    },
+    {
+      "epoch": 0.23899950835791545,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.8495246013573057e-05,
+      "loss": 0.4264,
+      "step": 3889
+    },
+    {
+      "epoch": 0.23906096361848575,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.8494018121279274e-05,
+      "loss": 0.5112,
+      "step": 3890
+    },
+    {
+      "epoch": 0.23912241887905605,
+      "grad_norm": 34.25,
+      "learning_rate": 1.8492789768998668e-05,
+      "loss": 0.5012,
+      "step": 3891
+    },
+    {
+      "epoch": 0.23918387413962636,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.8491560956797766e-05,
+      "loss": 0.5751,
+      "step": 3892
+    },
+    {
+      "epoch": 0.23924532940019666,
+      "grad_norm": 19.5,
+      "learning_rate": 1.8490331684743105e-05,
+      "loss": 0.4268,
+      "step": 3893
+    },
+    {
+      "epoch": 0.23930678466076696,
+      "grad_norm": 20.875,
+      "learning_rate": 1.8489101952901256e-05,
+      "loss": 0.5277,
+      "step": 3894
+    },
+    {
+      "epoch": 0.23936823992133727,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.848787176133882e-05,
+      "loss": 0.3717,
+      "step": 3895
+    },
+    {
+      "epoch": 0.23942969518190757,
+      "grad_norm": 17.25,
+      "learning_rate": 1.848664111012241e-05,
+      "loss": 0.3938,
+      "step": 3896
+    },
+    {
+      "epoch": 0.23949115044247787,
+      "grad_norm": 20.5,
+      "learning_rate": 1.848540999931868e-05,
+      "loss": 0.4175,
+      "step": 3897
+    },
+    {
+      "epoch": 0.23955260570304818,
+      "grad_norm": 4.84375,
+      "learning_rate": 1.848417842899429e-05,
+      "loss": 0.2745,
+      "step": 3898
+    },
+    {
+      "epoch": 0.23961406096361848,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8482946399215943e-05,
+      "loss": 0.481,
+      "step": 3899
+    },
+    {
+      "epoch": 0.23967551622418878,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.848171391005035e-05,
+      "loss": 0.4082,
+      "step": 3900
+    },
+    {
+      "epoch": 0.2397369714847591,
+      "grad_norm": 29.25,
+      "learning_rate": 1.848048096156426e-05,
+      "loss": 0.6365,
+      "step": 3901
+    },
+    {
+      "epoch": 0.2397984267453294,
+      "grad_norm": 38.25,
+      "learning_rate": 1.8479247553824445e-05,
+      "loss": 0.3512,
+      "step": 3902
+    },
+    {
+      "epoch": 0.2398598820058997,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.8478013686897696e-05,
+      "loss": 0.4087,
+      "step": 3903
+    },
+    {
+      "epoch": 0.23992133726647,
+      "grad_norm": 4.8125,
+      "learning_rate": 1.8476779360850833e-05,
+      "loss": 0.3764,
+      "step": 3904
+    },
+    {
+      "epoch": 0.23998279252704033,
+      "grad_norm": 4.4375,
+      "learning_rate": 1.84755445757507e-05,
+      "loss": 0.4335,
+      "step": 3905
+    },
+    {
+      "epoch": 0.24004424778761063,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.8474309331664165e-05,
+      "loss": 0.4527,
+      "step": 3906
+    },
+    {
+      "epoch": 0.24010570304818094,
+      "grad_norm": 9.0,
+      "learning_rate": 1.8473073628658123e-05,
+      "loss": 0.5094,
+      "step": 3907
+    },
+    {
+      "epoch": 0.24016715830875124,
+      "grad_norm": 30.375,
+      "learning_rate": 1.847183746679949e-05,
+      "loss": 0.5736,
+      "step": 3908
+    },
+    {
+      "epoch": 0.24022861356932154,
+      "grad_norm": 17.75,
+      "learning_rate": 1.8470600846155213e-05,
+      "loss": 0.4153,
+      "step": 3909
+    },
+    {
+      "epoch": 0.24029006882989185,
+      "grad_norm": 10.0,
+      "learning_rate": 1.8469363766792258e-05,
+      "loss": 0.4624,
+      "step": 3910
+    },
+    {
+      "epoch": 0.24035152409046215,
+      "grad_norm": 24.5,
+      "learning_rate": 1.8468126228777617e-05,
+      "loss": 0.5461,
+      "step": 3911
+    },
+    {
+      "epoch": 0.24041297935103245,
+      "grad_norm": 51.25,
+      "learning_rate": 1.8466888232178312e-05,
+      "loss": 0.4496,
+      "step": 3912
+    },
+    {
+      "epoch": 0.24047443461160276,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8465649777061377e-05,
+      "loss": 0.5879,
+      "step": 3913
+    },
+    {
+      "epoch": 0.24053588987217306,
+      "grad_norm": 26.875,
+      "learning_rate": 1.846441086349389e-05,
+      "loss": 0.6312,
+      "step": 3914
+    },
+    {
+      "epoch": 0.24059734513274336,
+      "grad_norm": 21.875,
+      "learning_rate": 1.846317149154294e-05,
+      "loss": 0.4866,
+      "step": 3915
+    },
+    {
+      "epoch": 0.24065880039331367,
+      "grad_norm": 39.0,
+      "learning_rate": 1.8461931661275642e-05,
+      "loss": 0.5921,
+      "step": 3916
+    },
+    {
+      "epoch": 0.24072025565388397,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.846069137275914e-05,
+      "loss": 0.4138,
+      "step": 3917
+    },
+    {
+      "epoch": 0.24078171091445427,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.8459450626060596e-05,
+      "loss": 0.4337,
+      "step": 3918
+    },
+    {
+      "epoch": 0.24084316617502458,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8458209421247208e-05,
+      "loss": 0.4422,
+      "step": 3919
+    },
+    {
+      "epoch": 0.24090462143559488,
+      "grad_norm": 25.75,
+      "learning_rate": 1.8456967758386192e-05,
+      "loss": 0.5354,
+      "step": 3920
+    },
+    {
+      "epoch": 0.24096607669616518,
+      "grad_norm": 50.0,
+      "learning_rate": 1.8455725637544784e-05,
+      "loss": 0.4677,
+      "step": 3921
+    },
+    {
+      "epoch": 0.24102753195673549,
+      "grad_norm": 32.25,
+      "learning_rate": 1.8454483058790254e-05,
+      "loss": 0.5111,
+      "step": 3922
+    },
+    {
+      "epoch": 0.2410889872173058,
+      "grad_norm": 26.875,
+      "learning_rate": 1.8453240022189896e-05,
+      "loss": 0.3977,
+      "step": 3923
+    },
+    {
+      "epoch": 0.2411504424778761,
+      "grad_norm": 19.25,
+      "learning_rate": 1.8451996527811018e-05,
+      "loss": 0.4372,
+      "step": 3924
+    },
+    {
+      "epoch": 0.24121189773844642,
+      "grad_norm": 9.5,
+      "learning_rate": 1.8450752575720967e-05,
+      "loss": 0.4313,
+      "step": 3925
+    },
+    {
+      "epoch": 0.24127335299901673,
+      "grad_norm": 28.5,
+      "learning_rate": 1.8449508165987106e-05,
+      "loss": 0.4289,
+      "step": 3926
+    },
+    {
+      "epoch": 0.24133480825958703,
+      "grad_norm": 38.25,
+      "learning_rate": 1.844826329867682e-05,
+      "loss": 0.3315,
+      "step": 3927
+    },
+    {
+      "epoch": 0.24139626352015733,
+      "grad_norm": 9.25,
+      "learning_rate": 1.844701797385753e-05,
+      "loss": 0.2797,
+      "step": 3928
+    },
+    {
+      "epoch": 0.24145771878072764,
+      "grad_norm": 23.375,
+      "learning_rate": 1.8445772191596676e-05,
+      "loss": 0.3894,
+      "step": 3929
+    },
+    {
+      "epoch": 0.24151917404129794,
+      "grad_norm": 19.75,
+      "learning_rate": 1.844452595196172e-05,
+      "loss": 0.5158,
+      "step": 3930
+    },
+    {
+      "epoch": 0.24158062930186824,
+      "grad_norm": 34.5,
+      "learning_rate": 1.8443279255020153e-05,
+      "loss": 0.5255,
+      "step": 3931
+    },
+    {
+      "epoch": 0.24164208456243855,
+      "grad_norm": 12.125,
+      "learning_rate": 1.8442032100839484e-05,
+      "loss": 0.4037,
+      "step": 3932
+    },
+    {
+      "epoch": 0.24170353982300885,
+      "grad_norm": 5.21875,
+      "learning_rate": 1.8440784489487256e-05,
+      "loss": 0.3674,
+      "step": 3933
+    },
+    {
+      "epoch": 0.24176499508357915,
+      "grad_norm": 27.375,
+      "learning_rate": 1.8439536421031035e-05,
+      "loss": 0.3846,
+      "step": 3934
+    },
+    {
+      "epoch": 0.24182645034414946,
+      "grad_norm": 20.125,
+      "learning_rate": 1.84382878955384e-05,
+      "loss": 0.4185,
+      "step": 3935
+    },
+    {
+      "epoch": 0.24188790560471976,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8437038913076974e-05,
+      "loss": 0.4124,
+      "step": 3936
+    },
+    {
+      "epoch": 0.24194936086529006,
+      "grad_norm": 21.5,
+      "learning_rate": 1.843578947371439e-05,
+      "loss": 0.4238,
+      "step": 3937
+    },
+    {
+      "epoch": 0.24201081612586037,
+      "grad_norm": 16.5,
+      "learning_rate": 1.843453957751831e-05,
+      "loss": 0.3941,
+      "step": 3938
+    },
+    {
+      "epoch": 0.24207227138643067,
+      "grad_norm": 16.875,
+      "learning_rate": 1.843328922455642e-05,
+      "loss": 0.4682,
+      "step": 3939
+    },
+    {
+      "epoch": 0.24213372664700097,
+      "grad_norm": 17.625,
+      "learning_rate": 1.8432038414896432e-05,
+      "loss": 0.5307,
+      "step": 3940
+    },
+    {
+      "epoch": 0.24219518190757128,
+      "grad_norm": 14.75,
+      "learning_rate": 1.8430787148606087e-05,
+      "loss": 0.3972,
+      "step": 3941
+    },
+    {
+      "epoch": 0.24225663716814158,
+      "grad_norm": 20.0,
+      "learning_rate": 1.8429535425753145e-05,
+      "loss": 0.4482,
+      "step": 3942
+    },
+    {
+      "epoch": 0.24231809242871188,
+      "grad_norm": 23.625,
+      "learning_rate": 1.842828324640539e-05,
+      "loss": 0.5109,
+      "step": 3943
+    },
+    {
+      "epoch": 0.24237954768928222,
+      "grad_norm": 29.125,
+      "learning_rate": 1.842703061063063e-05,
+      "loss": 0.4073,
+      "step": 3944
+    },
+    {
+      "epoch": 0.24244100294985252,
+      "grad_norm": 22.25,
+      "learning_rate": 1.8425777518496707e-05,
+      "loss": 0.5443,
+      "step": 3945
+    },
+    {
+      "epoch": 0.24250245821042282,
+      "grad_norm": 25.25,
+      "learning_rate": 1.842452397007148e-05,
+      "loss": 0.4738,
+      "step": 3946
+    },
+    {
+      "epoch": 0.24256391347099313,
+      "grad_norm": 22.875,
+      "learning_rate": 1.8423269965422826e-05,
+      "loss": 0.3923,
+      "step": 3947
+    },
+    {
+      "epoch": 0.24262536873156343,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.8422015504618664e-05,
+      "loss": 0.3796,
+      "step": 3948
+    },
+    {
+      "epoch": 0.24268682399213373,
+      "grad_norm": 25.625,
+      "learning_rate": 1.8420760587726925e-05,
+      "loss": 0.6177,
+      "step": 3949
+    },
+    {
+      "epoch": 0.24274827925270404,
+      "grad_norm": 4.9375,
+      "learning_rate": 1.8419505214815567e-05,
+      "loss": 0.3117,
+      "step": 3950
+    },
+    {
+      "epoch": 0.24280973451327434,
+      "grad_norm": 37.75,
+      "learning_rate": 1.8418249385952575e-05,
+      "loss": 0.3569,
+      "step": 3951
+    },
+    {
+      "epoch": 0.24287118977384464,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.8416993101205957e-05,
+      "loss": 0.4344,
+      "step": 3952
+    },
+    {
+      "epoch": 0.24293264503441495,
+      "grad_norm": 22.25,
+      "learning_rate": 1.841573636064375e-05,
+      "loss": 0.4031,
+      "step": 3953
+    },
+    {
+      "epoch": 0.24299410029498525,
+      "grad_norm": 31.875,
+      "learning_rate": 1.8414479164334004e-05,
+      "loss": 0.4428,
+      "step": 3954
+    },
+    {
+      "epoch": 0.24305555555555555,
+      "grad_norm": 16.375,
+      "learning_rate": 1.8413221512344805e-05,
+      "loss": 0.3358,
+      "step": 3955
+    },
+    {
+      "epoch": 0.24311701081612586,
+      "grad_norm": 4.90625,
+      "learning_rate": 1.8411963404744263e-05,
+      "loss": 0.308,
+      "step": 3956
+    },
+    {
+      "epoch": 0.24317846607669616,
+      "grad_norm": 18.0,
+      "learning_rate": 1.8410704841600506e-05,
+      "loss": 0.301,
+      "step": 3957
+    },
+    {
+      "epoch": 0.24323992133726646,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.8409445822981694e-05,
+      "loss": 0.4239,
+      "step": 3958
+    },
+    {
+      "epoch": 0.24330137659783677,
+      "grad_norm": 25.875,
+      "learning_rate": 1.8408186348956e-05,
+      "loss": 0.3616,
+      "step": 3959
+    },
+    {
+      "epoch": 0.24336283185840707,
+      "grad_norm": 6.3125,
+      "learning_rate": 1.8406926419591644e-05,
+      "loss": 0.3402,
+      "step": 3960
+    },
+    {
+      "epoch": 0.24342428711897737,
+      "grad_norm": 13.25,
+      "learning_rate": 1.8405666034956842e-05,
+      "loss": 0.326,
+      "step": 3961
+    },
+    {
+      "epoch": 0.24348574237954768,
+      "grad_norm": 17.25,
+      "learning_rate": 1.840440519511986e-05,
+      "loss": 0.4571,
+      "step": 3962
+    },
+    {
+      "epoch": 0.243547197640118,
+      "grad_norm": 17.25,
+      "learning_rate": 1.840314390014897e-05,
+      "loss": 0.6504,
+      "step": 3963
+    },
+    {
+      "epoch": 0.2436086529006883,
+      "grad_norm": 21.625,
+      "learning_rate": 1.8401882150112485e-05,
+      "loss": 0.5625,
+      "step": 3964
+    },
+    {
+      "epoch": 0.24367010816125861,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.8400619945078725e-05,
+      "loss": 0.3691,
+      "step": 3965
+    },
+    {
+      "epoch": 0.24373156342182892,
+      "grad_norm": 25.625,
+      "learning_rate": 1.8399357285116045e-05,
+      "loss": 0.3519,
+      "step": 3966
+    },
+    {
+      "epoch": 0.24379301868239922,
+      "grad_norm": 6.3125,
+      "learning_rate": 1.839809417029283e-05,
+      "loss": 0.331,
+      "step": 3967
+    },
+    {
+      "epoch": 0.24385447394296952,
+      "grad_norm": 8.0,
+      "learning_rate": 1.839683060067748e-05,
+      "loss": 0.2673,
+      "step": 3968
+    },
+    {
+      "epoch": 0.24391592920353983,
+      "grad_norm": 34.5,
+      "learning_rate": 1.8395566576338416e-05,
+      "loss": 0.4243,
+      "step": 3969
+    },
+    {
+      "epoch": 0.24397738446411013,
+      "grad_norm": 41.25,
+      "learning_rate": 1.8394302097344103e-05,
+      "loss": 0.5283,
+      "step": 3970
+    },
+    {
+      "epoch": 0.24403883972468043,
+      "grad_norm": 37.5,
+      "learning_rate": 1.8393037163763005e-05,
+      "loss": 0.4097,
+      "step": 3971
+    },
+    {
+      "epoch": 0.24410029498525074,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.8391771775663628e-05,
+      "loss": 0.3138,
+      "step": 3972
+    },
+    {
+      "epoch": 0.24416175024582104,
+      "grad_norm": 53.75,
+      "learning_rate": 1.8390505933114503e-05,
+      "loss": 0.4512,
+      "step": 3973
+    },
+    {
+      "epoch": 0.24422320550639134,
+      "grad_norm": 6.34375,
+      "learning_rate": 1.8389239636184177e-05,
+      "loss": 0.3777,
+      "step": 3974
+    },
+    {
+      "epoch": 0.24428466076696165,
+      "grad_norm": 5.6875,
+      "learning_rate": 1.838797288494122e-05,
+      "loss": 0.5333,
+      "step": 3975
+    },
+    {
+      "epoch": 0.24434611602753195,
+      "grad_norm": 14.75,
+      "learning_rate": 1.8386705679454243e-05,
+      "loss": 0.5531,
+      "step": 3976
+    },
+    {
+      "epoch": 0.24440757128810225,
+      "grad_norm": 43.25,
+      "learning_rate": 1.8385438019791862e-05,
+      "loss": 0.4105,
+      "step": 3977
+    },
+    {
+      "epoch": 0.24446902654867256,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.838416990602273e-05,
+      "loss": 0.3496,
+      "step": 3978
+    },
+    {
+      "epoch": 0.24453048180924286,
+      "grad_norm": 38.5,
+      "learning_rate": 1.8382901338215515e-05,
+      "loss": 0.3609,
+      "step": 3979
+    },
+    {
+      "epoch": 0.24459193706981316,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.8381632316438923e-05,
+      "loss": 0.2488,
+      "step": 3980
+    },
+    {
+      "epoch": 0.24465339233038347,
+      "grad_norm": 4.90625,
+      "learning_rate": 1.8380362840761675e-05,
+      "loss": 0.3516,
+      "step": 3981
+    },
+    {
+      "epoch": 0.2447148475909538,
+      "grad_norm": 18.25,
+      "learning_rate": 1.8379092911252515e-05,
+      "loss": 0.334,
+      "step": 3982
+    },
+    {
+      "epoch": 0.2447763028515241,
+      "grad_norm": 8.125,
+      "learning_rate": 1.8377822527980216e-05,
+      "loss": 0.2432,
+      "step": 3983
+    },
+    {
+      "epoch": 0.2448377581120944,
+      "grad_norm": 21.875,
+      "learning_rate": 1.8376551691013577e-05,
+      "loss": 0.473,
+      "step": 3984
+    },
+    {
+      "epoch": 0.2448992133726647,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.837528040042142e-05,
+      "loss": 0.2835,
+      "step": 3985
+    },
+    {
+      "epoch": 0.244960668633235,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.8374008656272585e-05,
+      "loss": 0.3384,
+      "step": 3986
+    },
+    {
+      "epoch": 0.24502212389380532,
+      "grad_norm": 20.75,
+      "learning_rate": 1.8372736458635947e-05,
+      "loss": 0.442,
+      "step": 3987
+    },
+    {
+      "epoch": 0.24508357915437562,
+      "grad_norm": 41.5,
+      "learning_rate": 1.83714638075804e-05,
+      "loss": 0.5617,
+      "step": 3988
+    },
+    {
+      "epoch": 0.24514503441494592,
+      "grad_norm": 25.25,
+      "learning_rate": 1.8370190703174862e-05,
+      "loss": 0.326,
+      "step": 3989
+    },
+    {
+      "epoch": 0.24520648967551623,
+      "grad_norm": 44.5,
+      "learning_rate": 1.836891714548828e-05,
+      "loss": 0.3708,
+      "step": 3990
+    },
+    {
+      "epoch": 0.24526794493608653,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.836764313458962e-05,
+      "loss": 0.322,
+      "step": 3991
+    },
+    {
+      "epoch": 0.24532940019665683,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.8366368670547873e-05,
+      "loss": 0.3048,
+      "step": 3992
+    },
+    {
+      "epoch": 0.24539085545722714,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.836509375343206e-05,
+      "loss": 0.3057,
+      "step": 3993
+    },
+    {
+      "epoch": 0.24545231071779744,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.8363818383311226e-05,
+      "loss": 0.3365,
+      "step": 3994
+    },
+    {
+      "epoch": 0.24551376597836774,
+      "grad_norm": 20.5,
+      "learning_rate": 1.836254256025443e-05,
+      "loss": 0.4896,
+      "step": 3995
+    },
+    {
+      "epoch": 0.24557522123893805,
+      "grad_norm": 5.90625,
+      "learning_rate": 1.836126628433077e-05,
+      "loss": 0.2832,
+      "step": 3996
+    },
+    {
+      "epoch": 0.24563667649950835,
+      "grad_norm": 17.25,
+      "learning_rate": 1.8359989555609355e-05,
+      "loss": 0.3137,
+      "step": 3997
+    },
+    {
+      "epoch": 0.24569813176007865,
+      "grad_norm": 19.5,
+      "learning_rate": 1.8358712374159332e-05,
+      "loss": 0.2704,
+      "step": 3998
+    },
+    {
+      "epoch": 0.24575958702064896,
+      "grad_norm": 24.0,
+      "learning_rate": 1.835743474004986e-05,
+      "loss": 0.5175,
+      "step": 3999
+    },
+    {
+      "epoch": 0.24582104228121926,
+      "grad_norm": 5.125,
+      "learning_rate": 1.8356156653350138e-05,
+      "loss": 0.2654,
+      "step": 4000
+    },
+    {
+      "epoch": 0.2458824975417896,
+      "grad_norm": 25.0,
+      "learning_rate": 1.8354878114129368e-05,
+      "loss": 0.4489,
+      "step": 4001
+    },
+    {
+      "epoch": 0.2459439528023599,
+      "grad_norm": 26.5,
+      "learning_rate": 1.835359912245679e-05,
+      "loss": 0.5935,
+      "step": 4002
+    },
+    {
+      "epoch": 0.2460054080629302,
+      "grad_norm": 53.25,
+      "learning_rate": 1.8352319678401677e-05,
+      "loss": 0.5617,
+      "step": 4003
+    },
+    {
+      "epoch": 0.2460668633235005,
+      "grad_norm": 29.625,
+      "learning_rate": 1.8351039782033303e-05,
+      "loss": 0.5743,
+      "step": 4004
+    },
+    {
+      "epoch": 0.2461283185840708,
+      "grad_norm": 30.0,
+      "learning_rate": 1.834975943342099e-05,
+      "loss": 0.4428,
+      "step": 4005
+    },
+    {
+      "epoch": 0.2461897738446411,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.8348478632634067e-05,
+      "loss": 0.3116,
+      "step": 4006
+    },
+    {
+      "epoch": 0.2462512291052114,
+      "grad_norm": 24.75,
+      "learning_rate": 1.83471973797419e-05,
+      "loss": 0.3855,
+      "step": 4007
+    },
+    {
+      "epoch": 0.24631268436578171,
+      "grad_norm": 24.75,
+      "learning_rate": 1.834591567481387e-05,
+      "loss": 0.3111,
+      "step": 4008
+    },
+    {
+      "epoch": 0.24637413962635202,
+      "grad_norm": 19.5,
+      "learning_rate": 1.834463351791939e-05,
+      "loss": 0.3486,
+      "step": 4009
+    },
+    {
+      "epoch": 0.24643559488692232,
+      "grad_norm": 17.375,
+      "learning_rate": 1.8343350909127897e-05,
+      "loss": 0.3523,
+      "step": 4010
+    },
+    {
+      "epoch": 0.24649705014749262,
+      "grad_norm": 25.375,
+      "learning_rate": 1.8342067848508843e-05,
+      "loss": 0.3483,
+      "step": 4011
+    },
+    {
+      "epoch": 0.24655850540806293,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.8340784336131715e-05,
+      "loss": 0.313,
+      "step": 4012
+    },
+    {
+      "epoch": 0.24661996066863323,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8339500372066018e-05,
+      "loss": 0.2793,
+      "step": 4013
+    },
+    {
+      "epoch": 0.24668141592920353,
+      "grad_norm": 36.25,
+      "learning_rate": 1.8338215956381283e-05,
+      "loss": 0.3837,
+      "step": 4014
+    },
+    {
+      "epoch": 0.24674287118977384,
+      "grad_norm": 11.375,
+      "learning_rate": 1.8336931089147076e-05,
+      "loss": 0.2733,
+      "step": 4015
+    },
+    {
+      "epoch": 0.24680432645034414,
+      "grad_norm": 31.875,
+      "learning_rate": 1.8335645770432963e-05,
+      "loss": 0.3363,
+      "step": 4016
+    },
+    {
+      "epoch": 0.24686578171091444,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8334360000308563e-05,
+      "loss": 0.5316,
+      "step": 4017
+    },
+    {
+      "epoch": 0.24692723697148475,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.83330737788435e-05,
+      "loss": 0.3169,
+      "step": 4018
+    },
+    {
+      "epoch": 0.24698869223205505,
+      "grad_norm": 3.5,
+      "learning_rate": 1.8331787106107425e-05,
+      "loss": 0.3207,
+      "step": 4019
+    },
+    {
+      "epoch": 0.24705014749262535,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.8330499982170024e-05,
+      "loss": 0.2851,
+      "step": 4020
+    },
+    {
+      "epoch": 0.24711160275319569,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.8329212407100996e-05,
+      "loss": 0.3365,
+      "step": 4021
+    },
+    {
+      "epoch": 0.247173058013766,
+      "grad_norm": 5.125,
+      "learning_rate": 1.8327924380970066e-05,
+      "loss": 0.3067,
+      "step": 4022
+    },
+    {
+      "epoch": 0.2472345132743363,
+      "grad_norm": 3.96875,
+      "learning_rate": 1.8326635903846995e-05,
+      "loss": 0.3348,
+      "step": 4023
+    },
+    {
+      "epoch": 0.2472959685349066,
+      "grad_norm": 21.5,
+      "learning_rate": 1.832534697580155e-05,
+      "loss": 0.384,
+      "step": 4024
+    },
+    {
+      "epoch": 0.2473574237954769,
+      "grad_norm": 11.25,
+      "learning_rate": 1.8324057596903534e-05,
+      "loss": 0.3548,
+      "step": 4025
+    },
+    {
+      "epoch": 0.2474188790560472,
+      "grad_norm": 49.0,
+      "learning_rate": 1.832276776722278e-05,
+      "loss": 0.588,
+      "step": 4026
+    },
+    {
+      "epoch": 0.2474803343166175,
+      "grad_norm": 16.125,
+      "learning_rate": 1.8321477486829128e-05,
+      "loss": 0.3635,
+      "step": 4027
+    },
+    {
+      "epoch": 0.2475417895771878,
+      "grad_norm": 4.8125,
+      "learning_rate": 1.8320186755792452e-05,
+      "loss": 0.274,
+      "step": 4028
+    },
+    {
+      "epoch": 0.2476032448377581,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.8318895574182657e-05,
+      "loss": 0.3442,
+      "step": 4029
+    },
+    {
+      "epoch": 0.24766470009832842,
+      "grad_norm": 16.25,
+      "learning_rate": 1.8317603942069665e-05,
+      "loss": 0.3387,
+      "step": 4030
+    },
+    {
+      "epoch": 0.24772615535889872,
+      "grad_norm": 26.375,
+      "learning_rate": 1.831631185952342e-05,
+      "loss": 0.4472,
+      "step": 4031
+    },
+    {
+      "epoch": 0.24778761061946902,
+      "grad_norm": 28.125,
+      "learning_rate": 1.8315019326613897e-05,
+      "loss": 0.4723,
+      "step": 4032
+    },
+    {
+      "epoch": 0.24784906588003933,
+      "grad_norm": 24.25,
+      "learning_rate": 1.8313726343411085e-05,
+      "loss": 0.5292,
+      "step": 4033
+    },
+    {
+      "epoch": 0.24791052114060963,
+      "grad_norm": 38.0,
+      "learning_rate": 1.8312432909985017e-05,
+      "loss": 0.4845,
+      "step": 4034
+    },
+    {
+      "epoch": 0.24797197640117993,
+      "grad_norm": 24.375,
+      "learning_rate": 1.8311139026405725e-05,
+      "loss": 0.3009,
+      "step": 4035
+    },
+    {
+      "epoch": 0.24803343166175024,
+      "grad_norm": 34.75,
+      "learning_rate": 1.8309844692743283e-05,
+      "loss": 0.4395,
+      "step": 4036
+    },
+    {
+      "epoch": 0.24809488692232054,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.830854990906779e-05,
+      "loss": 0.3267,
+      "step": 4037
+    },
+    {
+      "epoch": 0.24815634218289084,
+      "grad_norm": 6.15625,
+      "learning_rate": 1.8307254675449355e-05,
+      "loss": 0.394,
+      "step": 4038
+    },
+    {
+      "epoch": 0.24821779744346115,
+      "grad_norm": 18.0,
+      "learning_rate": 1.830595899195813e-05,
+      "loss": 0.2889,
+      "step": 4039
+    },
+    {
+      "epoch": 0.24827925270403148,
+      "grad_norm": 21.5,
+      "learning_rate": 1.8304662858664272e-05,
+      "loss": 0.3714,
+      "step": 4040
+    },
+    {
+      "epoch": 0.24834070796460178,
+      "grad_norm": 37.75,
+      "learning_rate": 1.8303366275637977e-05,
+      "loss": 0.3317,
+      "step": 4041
+    },
+    {
+      "epoch": 0.24840216322517208,
+      "grad_norm": 24.625,
+      "learning_rate": 1.830206924294946e-05,
+      "loss": 0.378,
+      "step": 4042
+    },
+    {
+      "epoch": 0.2484636184857424,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.830077176066896e-05,
+      "loss": 0.3003,
+      "step": 4043
+    },
+    {
+      "epoch": 0.2485250737463127,
+      "grad_norm": 25.25,
+      "learning_rate": 1.8299473828866746e-05,
+      "loss": 0.3986,
+      "step": 4044
+    },
+    {
+      "epoch": 0.248586529006883,
+      "grad_norm": 21.5,
+      "learning_rate": 1.82981754476131e-05,
+      "loss": 0.6326,
+      "step": 4045
+    },
+    {
+      "epoch": 0.2486479842674533,
+      "grad_norm": 24.75,
+      "learning_rate": 1.8296876616978337e-05,
+      "loss": 0.4619,
+      "step": 4046
+    },
+    {
+      "epoch": 0.2487094395280236,
+      "grad_norm": 13.0,
+      "learning_rate": 1.8295577337032792e-05,
+      "loss": 0.3393,
+      "step": 4047
+    },
+    {
+      "epoch": 0.2487708947885939,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.8294277607846834e-05,
+      "loss": 0.4538,
+      "step": 4048
+    },
+    {
+      "epoch": 0.2488323500491642,
+      "grad_norm": 31.5,
+      "learning_rate": 1.829297742949084e-05,
+      "loss": 0.6041,
+      "step": 4049
+    },
+    {
+      "epoch": 0.2488938053097345,
+      "grad_norm": 20.0,
+      "learning_rate": 1.829167680203522e-05,
+      "loss": 0.4036,
+      "step": 4050
+    },
+    {
+      "epoch": 0.24895526057030481,
+      "grad_norm": 19.75,
+      "learning_rate": 1.8290375725550417e-05,
+      "loss": 0.466,
+      "step": 4051
+    },
+    {
+      "epoch": 0.24901671583087512,
+      "grad_norm": 23.125,
+      "learning_rate": 1.8289074200106885e-05,
+      "loss": 0.4013,
+      "step": 4052
+    },
+    {
+      "epoch": 0.24907817109144542,
+      "grad_norm": 8.375,
+      "learning_rate": 1.8287772225775105e-05,
+      "loss": 0.3244,
+      "step": 4053
+    },
+    {
+      "epoch": 0.24913962635201573,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.828646980262559e-05,
+      "loss": 0.3141,
+      "step": 4054
+    },
+    {
+      "epoch": 0.24920108161258603,
+      "grad_norm": 17.375,
+      "learning_rate": 1.8285166930728865e-05,
+      "loss": 0.3037,
+      "step": 4055
+    },
+    {
+      "epoch": 0.24926253687315633,
+      "grad_norm": 20.875,
+      "learning_rate": 1.828386361015549e-05,
+      "loss": 0.3363,
+      "step": 4056
+    },
+    {
+      "epoch": 0.24932399213372664,
+      "grad_norm": 24.125,
+      "learning_rate": 1.8282559840976043e-05,
+      "loss": 0.498,
+      "step": 4057
+    },
+    {
+      "epoch": 0.24938544739429694,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.8281255623261128e-05,
+      "loss": 0.4123,
+      "step": 4058
+    },
+    {
+      "epoch": 0.24944690265486727,
+      "grad_norm": 16.25,
+      "learning_rate": 1.827995095708138e-05,
+      "loss": 0.312,
+      "step": 4059
+    },
+    {
+      "epoch": 0.24950835791543757,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.8278645842507448e-05,
+      "loss": 0.4186,
+      "step": 4060
+    },
+    {
+      "epoch": 0.24956981317600788,
+      "grad_norm": 9.875,
+      "learning_rate": 1.827734027961001e-05,
+      "loss": 0.5515,
+      "step": 4061
+    },
+    {
+      "epoch": 0.24963126843657818,
+      "grad_norm": 29.875,
+      "learning_rate": 1.8276034268459763e-05,
+      "loss": 0.3916,
+      "step": 4062
+    },
+    {
+      "epoch": 0.24969272369714848,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.827472780912744e-05,
+      "loss": 0.4083,
+      "step": 4063
+    },
+    {
+      "epoch": 0.2497541789577188,
+      "grad_norm": 14.875,
+      "learning_rate": 1.8273420901683786e-05,
+      "loss": 0.4185,
+      "step": 4064
+    },
+    {
+      "epoch": 0.2498156342182891,
+      "grad_norm": 21.625,
+      "learning_rate": 1.827211354619958e-05,
+      "loss": 0.3801,
+      "step": 4065
+    },
+    {
+      "epoch": 0.2498770894788594,
+      "grad_norm": 32.0,
+      "learning_rate": 1.827080574274562e-05,
+      "loss": 0.6807,
+      "step": 4066
+    },
+    {
+      "epoch": 0.2499385447394297,
+      "grad_norm": 28.875,
+      "learning_rate": 1.8269497491392724e-05,
+      "loss": 0.4956,
+      "step": 4067
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8268188792211748e-05,
+      "loss": 0.6116,
+      "step": 4068
+    },
+    {
+      "epoch": 0.25006145526057033,
+      "grad_norm": 17.25,
+      "learning_rate": 1.8266879645273557e-05,
+      "loss": 0.4544,
+      "step": 4069
+    },
+    {
+      "epoch": 0.2501229105211406,
+      "grad_norm": 21.5,
+      "learning_rate": 1.8265570050649048e-05,
+      "loss": 0.4382,
+      "step": 4070
+    },
+    {
+      "epoch": 0.25018436578171094,
+      "grad_norm": 34.0,
+      "learning_rate": 1.8264260008409138e-05,
+      "loss": 0.3984,
+      "step": 4071
+    },
+    {
+      "epoch": 0.2502458210422812,
+      "grad_norm": 21.875,
+      "learning_rate": 1.826294951862478e-05,
+      "loss": 0.367,
+      "step": 4072
+    },
+    {
+      "epoch": 0.25030727630285154,
+      "grad_norm": 31.5,
+      "learning_rate": 1.8261638581366937e-05,
+      "loss": 0.4182,
+      "step": 4073
+    },
+    {
+      "epoch": 0.2503687315634218,
+      "grad_norm": 21.0,
+      "learning_rate": 1.8260327196706596e-05,
+      "loss": 0.4689,
+      "step": 4074
+    },
+    {
+      "epoch": 0.25043018682399215,
+      "grad_norm": 31.625,
+      "learning_rate": 1.8259015364714786e-05,
+      "loss": 0.4647,
+      "step": 4075
+    },
+    {
+      "epoch": 0.2504916420845624,
+      "grad_norm": 11.625,
+      "learning_rate": 1.8257703085462542e-05,
+      "loss": 0.3539,
+      "step": 4076
+    },
+    {
+      "epoch": 0.25055309734513276,
+      "grad_norm": 23.25,
+      "learning_rate": 1.825639035902093e-05,
+      "loss": 0.3953,
+      "step": 4077
+    },
+    {
+      "epoch": 0.25061455260570303,
+      "grad_norm": 20.875,
+      "learning_rate": 1.825507718546104e-05,
+      "loss": 0.4242,
+      "step": 4078
+    },
+    {
+      "epoch": 0.25067600786627336,
+      "grad_norm": 28.125,
+      "learning_rate": 1.825376356485398e-05,
+      "loss": 0.4373,
+      "step": 4079
+    },
+    {
+      "epoch": 0.25073746312684364,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.82524494972709e-05,
+      "loss": 0.3859,
+      "step": 4080
+    },
+    {
+      "epoch": 0.25079891838741397,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.8251134982782952e-05,
+      "loss": 0.3157,
+      "step": 4081
+    },
+    {
+      "epoch": 0.25086037364798425,
+      "grad_norm": 16.5,
+      "learning_rate": 1.824982002146133e-05,
+      "loss": 0.4286,
+      "step": 4082
+    },
+    {
+      "epoch": 0.2509218289085546,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.8248504613377236e-05,
+      "loss": 0.3369,
+      "step": 4083
+    },
+    {
+      "epoch": 0.25098328416912485,
+      "grad_norm": 29.5,
+      "learning_rate": 1.8247188758601912e-05,
+      "loss": 0.349,
+      "step": 4084
+    },
+    {
+      "epoch": 0.2510447394296952,
+      "grad_norm": 22.0,
+      "learning_rate": 1.8245872457206614e-05,
+      "loss": 0.2813,
+      "step": 4085
+    },
+    {
+      "epoch": 0.25110619469026546,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.8244555709262627e-05,
+      "loss": 0.3769,
+      "step": 4086
+    },
+    {
+      "epoch": 0.2511676499508358,
+      "grad_norm": 33.5,
+      "learning_rate": 1.824323851484126e-05,
+      "loss": 0.395,
+      "step": 4087
+    },
+    {
+      "epoch": 0.2512291052114061,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.824192087401384e-05,
+      "loss": 0.2735,
+      "step": 4088
+    },
+    {
+      "epoch": 0.2512905604719764,
+      "grad_norm": 22.875,
+      "learning_rate": 1.8240602786851726e-05,
+      "loss": 0.4715,
+      "step": 4089
+    },
+    {
+      "epoch": 0.25135201573254673,
+      "grad_norm": 25.125,
+      "learning_rate": 1.8239284253426294e-05,
+      "loss": 0.438,
+      "step": 4090
+    },
+    {
+      "epoch": 0.251413470993117,
+      "grad_norm": 28.5,
+      "learning_rate": 1.823796527380895e-05,
+      "loss": 0.4423,
+      "step": 4091
+    },
+    {
+      "epoch": 0.25147492625368734,
+      "grad_norm": 22.75,
+      "learning_rate": 1.823664584807113e-05,
+      "loss": 0.377,
+      "step": 4092
+    },
+    {
+      "epoch": 0.2515363815142576,
+      "grad_norm": 13.125,
+      "learning_rate": 1.8235325976284276e-05,
+      "loss": 0.3212,
+      "step": 4093
+    },
+    {
+      "epoch": 0.25159783677482794,
+      "grad_norm": 12.75,
+      "learning_rate": 1.823400565851987e-05,
+      "loss": 0.3056,
+      "step": 4094
+    },
+    {
+      "epoch": 0.2516592920353982,
+      "grad_norm": 11.25,
+      "learning_rate": 1.823268489484941e-05,
+      "loss": 0.3318,
+      "step": 4095
+    },
+    {
+      "epoch": 0.25172074729596855,
+      "grad_norm": 19.0,
+      "learning_rate": 1.8231363685344422e-05,
+      "loss": 0.3109,
+      "step": 4096
+    },
+    {
+      "epoch": 0.2517822025565388,
+      "grad_norm": 60.75,
+      "learning_rate": 1.8230042030076454e-05,
+      "loss": 0.4796,
+      "step": 4097
+    },
+    {
+      "epoch": 0.25184365781710916,
+      "grad_norm": 43.75,
+      "learning_rate": 1.8228719929117083e-05,
+      "loss": 0.4281,
+      "step": 4098
+    },
+    {
+      "epoch": 0.25190511307767943,
+      "grad_norm": 7.5,
+      "learning_rate": 1.82273973825379e-05,
+      "loss": 0.3413,
+      "step": 4099
+    },
+    {
+      "epoch": 0.25196656833824976,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8226074390410534e-05,
+      "loss": 0.355,
+      "step": 4100
+    },
+    {
+      "epoch": 0.25202802359882004,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.8224750952806626e-05,
+      "loss": 0.3679,
+      "step": 4101
+    },
+    {
+      "epoch": 0.25208947885939037,
+      "grad_norm": 31.0,
+      "learning_rate": 1.8223427069797845e-05,
+      "loss": 0.4405,
+      "step": 4102
+    },
+    {
+      "epoch": 0.25215093411996065,
+      "grad_norm": 6.3125,
+      "learning_rate": 1.8222102741455886e-05,
+      "loss": 0.3148,
+      "step": 4103
+    },
+    {
+      "epoch": 0.252212389380531,
+      "grad_norm": 9.625,
+      "learning_rate": 1.8220777967852467e-05,
+      "loss": 0.307,
+      "step": 4104
+    },
+    {
+      "epoch": 0.25227384464110125,
+      "grad_norm": 18.25,
+      "learning_rate": 1.8219452749059332e-05,
+      "loss": 0.6263,
+      "step": 4105
+    },
+    {
+      "epoch": 0.2523352999016716,
+      "grad_norm": 11.25,
+      "learning_rate": 1.8218127085148246e-05,
+      "loss": 0.4393,
+      "step": 4106
+    },
+    {
+      "epoch": 0.2523967551622419,
+      "grad_norm": 25.0,
+      "learning_rate": 1.8216800976191e-05,
+      "loss": 0.5935,
+      "step": 4107
+    },
+    {
+      "epoch": 0.2524582104228122,
+      "grad_norm": 36.0,
+      "learning_rate": 1.8215474422259403e-05,
+      "loss": 0.4726,
+      "step": 4108
+    },
+    {
+      "epoch": 0.2525196656833825,
+      "grad_norm": 18.875,
+      "learning_rate": 1.82141474234253e-05,
+      "loss": 0.4493,
+      "step": 4109
+    },
+    {
+      "epoch": 0.2525811209439528,
+      "grad_norm": 18.5,
+      "learning_rate": 1.821281997976055e-05,
+      "loss": 0.4508,
+      "step": 4110
+    },
+    {
+      "epoch": 0.25264257620452313,
+      "grad_norm": 16.375,
+      "learning_rate": 1.821149209133704e-05,
+      "loss": 0.4377,
+      "step": 4111
+    },
+    {
+      "epoch": 0.2527040314650934,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.8210163758226685e-05,
+      "loss": 0.3957,
+      "step": 4112
+    },
+    {
+      "epoch": 0.25276548672566373,
+      "grad_norm": 19.625,
+      "learning_rate": 1.8208834980501417e-05,
+      "loss": 0.4855,
+      "step": 4113
+    },
+    {
+      "epoch": 0.252826941986234,
+      "grad_norm": 21.125,
+      "learning_rate": 1.820750575823319e-05,
+      "loss": 0.4922,
+      "step": 4114
+    },
+    {
+      "epoch": 0.25288839724680434,
+      "grad_norm": 10.6875,
+      "learning_rate": 1.8206176091493995e-05,
+      "loss": 0.5336,
+      "step": 4115
+    },
+    {
+      "epoch": 0.2529498525073746,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.8204845980355834e-05,
+      "loss": 0.4201,
+      "step": 4116
+    },
+    {
+      "epoch": 0.25301130776794495,
+      "grad_norm": 33.75,
+      "learning_rate": 1.8203515424890738e-05,
+      "loss": 0.5766,
+      "step": 4117
+    },
+    {
+      "epoch": 0.2530727630285152,
+      "grad_norm": 22.875,
+      "learning_rate": 1.8202184425170766e-05,
+      "loss": 0.5656,
+      "step": 4118
+    },
+    {
+      "epoch": 0.25313421828908556,
+      "grad_norm": 20.375,
+      "learning_rate": 1.820085298126799e-05,
+      "loss": 0.5482,
+      "step": 4119
+    },
+    {
+      "epoch": 0.25319567354965583,
+      "grad_norm": 21.75,
+      "learning_rate": 1.8199521093254524e-05,
+      "loss": 0.5552,
+      "step": 4120
+    },
+    {
+      "epoch": 0.25325712881022616,
+      "grad_norm": 25.125,
+      "learning_rate": 1.8198188761202487e-05,
+      "loss": 0.5031,
+      "step": 4121
+    },
+    {
+      "epoch": 0.25331858407079644,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.8196855985184033e-05,
+      "loss": 0.4674,
+      "step": 4122
+    },
+    {
+      "epoch": 0.25338003933136677,
+      "grad_norm": 32.25,
+      "learning_rate": 1.819552276527134e-05,
+      "loss": 0.4217,
+      "step": 4123
+    },
+    {
+      "epoch": 0.25344149459193704,
+      "grad_norm": 15.75,
+      "learning_rate": 1.8194189101536603e-05,
+      "loss": 0.4838,
+      "step": 4124
+    },
+    {
+      "epoch": 0.2535029498525074,
+      "grad_norm": 10.5,
+      "learning_rate": 1.8192854994052047e-05,
+      "loss": 0.3384,
+      "step": 4125
+    },
+    {
+      "epoch": 0.2535644051130777,
+      "grad_norm": 12.75,
+      "learning_rate": 1.819152044288992e-05,
+      "loss": 0.4161,
+      "step": 4126
+    },
+    {
+      "epoch": 0.253625860373648,
+      "grad_norm": 18.125,
+      "learning_rate": 1.8190185448122496e-05,
+      "loss": 0.4645,
+      "step": 4127
+    },
+    {
+      "epoch": 0.2536873156342183,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.8188850009822063e-05,
+      "loss": 0.3758,
+      "step": 4128
+    },
+    {
+      "epoch": 0.2537487708947886,
+      "grad_norm": 15.4375,
+      "learning_rate": 1.8187514128060946e-05,
+      "loss": 0.6049,
+      "step": 4129
+    },
+    {
+      "epoch": 0.2538102261553589,
+      "grad_norm": 23.125,
+      "learning_rate": 1.818617780291149e-05,
+      "loss": 0.4706,
+      "step": 4130
+    },
+    {
+      "epoch": 0.2538716814159292,
+      "grad_norm": 18.5,
+      "learning_rate": 1.818484103444606e-05,
+      "loss": 0.6852,
+      "step": 4131
+    },
+    {
+      "epoch": 0.2539331366764995,
+      "grad_norm": 13.625,
+      "learning_rate": 1.818350382273705e-05,
+      "loss": 0.4768,
+      "step": 4132
+    },
+    {
+      "epoch": 0.2539945919370698,
+      "grad_norm": 22.5,
+      "learning_rate": 1.8182166167856875e-05,
+      "loss": 0.4391,
+      "step": 4133
+    },
+    {
+      "epoch": 0.25405604719764013,
+      "grad_norm": 15.125,
+      "learning_rate": 1.8180828069877972e-05,
+      "loss": 0.4925,
+      "step": 4134
+    },
+    {
+      "epoch": 0.2541175024582104,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.8179489528872808e-05,
+      "loss": 0.4346,
+      "step": 4135
+    },
+    {
+      "epoch": 0.25417895771878074,
+      "grad_norm": 5.96875,
+      "learning_rate": 1.8178150544913867e-05,
+      "loss": 0.3729,
+      "step": 4136
+    },
+    {
+      "epoch": 0.254240412979351,
+      "grad_norm": 23.625,
+      "learning_rate": 1.8176811118073664e-05,
+      "loss": 0.4421,
+      "step": 4137
+    },
+    {
+      "epoch": 0.25430186823992135,
+      "grad_norm": 19.625,
+      "learning_rate": 1.817547124842473e-05,
+      "loss": 0.69,
+      "step": 4138
+    },
+    {
+      "epoch": 0.2543633235004916,
+      "grad_norm": 27.125,
+      "learning_rate": 1.817413093603963e-05,
+      "loss": 0.5077,
+      "step": 4139
+    },
+    {
+      "epoch": 0.25442477876106195,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.817279018099095e-05,
+      "loss": 0.3464,
+      "step": 4140
+    },
+    {
+      "epoch": 0.25448623402163223,
+      "grad_norm": 23.625,
+      "learning_rate": 1.8171448983351284e-05,
+      "loss": 0.4602,
+      "step": 4141
+    },
+    {
+      "epoch": 0.25454768928220256,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.817010734319328e-05,
+      "loss": 0.3403,
+      "step": 4142
+    },
+    {
+      "epoch": 0.25460914454277284,
+      "grad_norm": 20.5,
+      "learning_rate": 1.8168765260589583e-05,
+      "loss": 0.5052,
+      "step": 4143
+    },
+    {
+      "epoch": 0.25467059980334317,
+      "grad_norm": 12.625,
+      "learning_rate": 1.8167422735612877e-05,
+      "loss": 0.423,
+      "step": 4144
+    },
+    {
+      "epoch": 0.2547320550639135,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.816607976833586e-05,
+      "loss": 0.363,
+      "step": 4145
+    },
+    {
+      "epoch": 0.2547935103244838,
+      "grad_norm": 5.8125,
+      "learning_rate": 1.8164736358831265e-05,
+      "loss": 0.4327,
+      "step": 4146
+    },
+    {
+      "epoch": 0.2548549655850541,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.816339250717184e-05,
+      "loss": 0.3551,
+      "step": 4147
+    },
+    {
+      "epoch": 0.2549164208456244,
+      "grad_norm": 21.25,
+      "learning_rate": 1.8162048213430362e-05,
+      "loss": 0.3017,
+      "step": 4148
+    },
+    {
+      "epoch": 0.2549778761061947,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.816070347767963e-05,
+      "loss": 0.2681,
+      "step": 4149
+    },
+    {
+      "epoch": 0.255039331366765,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.815935829999247e-05,
+      "loss": 0.2206,
+      "step": 4150
+    },
+    {
+      "epoch": 0.2551007866273353,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.8158012680441723e-05,
+      "loss": 0.275,
+      "step": 4151
+    },
+    {
+      "epoch": 0.2551622418879056,
+      "grad_norm": 10.6875,
+      "learning_rate": 1.8156666619100263e-05,
+      "loss": 0.2664,
+      "step": 4152
+    },
+    {
+      "epoch": 0.2552236971484759,
+      "grad_norm": 10.4375,
+      "learning_rate": 1.8155320116040983e-05,
+      "loss": 0.3529,
+      "step": 4153
+    },
+    {
+      "epoch": 0.2552851524090462,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.8153973171336804e-05,
+      "loss": 0.5122,
+      "step": 4154
+    },
+    {
+      "epoch": 0.25534660766961653,
+      "grad_norm": 20.5,
+      "learning_rate": 1.815262578506067e-05,
+      "loss": 0.3103,
+      "step": 4155
+    },
+    {
+      "epoch": 0.2554080629301868,
+      "grad_norm": 21.375,
+      "learning_rate": 1.815127795728554e-05,
+      "loss": 0.5232,
+      "step": 4156
+    },
+    {
+      "epoch": 0.25546951819075714,
+      "grad_norm": 18.625,
+      "learning_rate": 1.814992968808442e-05,
+      "loss": 0.3637,
+      "step": 4157
+    },
+    {
+      "epoch": 0.2555309734513274,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.814858097753031e-05,
+      "loss": 0.2555,
+      "step": 4158
+    },
+    {
+      "epoch": 0.25559242871189775,
+      "grad_norm": 4.125,
+      "learning_rate": 1.814723182569625e-05,
+      "loss": 0.3357,
+      "step": 4159
+    },
+    {
+      "epoch": 0.255653883972468,
+      "grad_norm": 28.625,
+      "learning_rate": 1.8145882232655308e-05,
+      "loss": 0.3667,
+      "step": 4160
+    },
+    {
+      "epoch": 0.25571533923303835,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.814453219848057e-05,
+      "loss": 0.3146,
+      "step": 4161
+    },
+    {
+      "epoch": 0.25577679449360863,
+      "grad_norm": 12.75,
+      "learning_rate": 1.814318172324514e-05,
+      "loss": 0.4413,
+      "step": 4162
+    },
+    {
+      "epoch": 0.25583824975417896,
+      "grad_norm": 19.25,
+      "learning_rate": 1.814183080702216e-05,
+      "loss": 0.4508,
+      "step": 4163
+    },
+    {
+      "epoch": 0.2558997050147493,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8140479449884778e-05,
+      "loss": 0.4866,
+      "step": 4164
+    },
+    {
+      "epoch": 0.25596116027531957,
+      "grad_norm": 52.25,
+      "learning_rate": 1.8139127651906183e-05,
+      "loss": 0.3641,
+      "step": 4165
+    },
+    {
+      "epoch": 0.2560226155358899,
+      "grad_norm": 22.625,
+      "learning_rate": 1.813777541315958e-05,
+      "loss": 0.334,
+      "step": 4166
+    },
+    {
+      "epoch": 0.2560840707964602,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.8136422733718192e-05,
+      "loss": 0.4829,
+      "step": 4167
+    },
+    {
+      "epoch": 0.2561455260570305,
+      "grad_norm": 21.125,
+      "learning_rate": 1.813506961365528e-05,
+      "loss": 0.3637,
+      "step": 4168
+    },
+    {
+      "epoch": 0.2562069813176008,
+      "grad_norm": 8.625,
+      "learning_rate": 1.8133716053044123e-05,
+      "loss": 0.4434,
+      "step": 4169
+    },
+    {
+      "epoch": 0.2562684365781711,
+      "grad_norm": 54.25,
+      "learning_rate": 1.8132362051958013e-05,
+      "loss": 0.4382,
+      "step": 4170
+    },
+    {
+      "epoch": 0.2563298918387414,
+      "grad_norm": 9.25,
+      "learning_rate": 1.8131007610470278e-05,
+      "loss": 0.2613,
+      "step": 4171
+    },
+    {
+      "epoch": 0.2563913470993117,
+      "grad_norm": 91.0,
+      "learning_rate": 1.812965272865427e-05,
+      "loss": 0.3696,
+      "step": 4172
+    },
+    {
+      "epoch": 0.256452802359882,
+      "grad_norm": 39.25,
+      "learning_rate": 1.812829740658336e-05,
+      "loss": 0.3255,
+      "step": 4173
+    },
+    {
+      "epoch": 0.2565142576204523,
+      "grad_norm": 35.75,
+      "learning_rate": 1.812694164433094e-05,
+      "loss": 0.3954,
+      "step": 4174
+    },
+    {
+      "epoch": 0.2565757128810226,
+      "grad_norm": 70.0,
+      "learning_rate": 1.8125585441970435e-05,
+      "loss": 0.2871,
+      "step": 4175
+    },
+    {
+      "epoch": 0.25663716814159293,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.8124228799575295e-05,
+      "loss": 0.2422,
+      "step": 4176
+    },
+    {
+      "epoch": 0.2566986234021632,
+      "grad_norm": 26.25,
+      "learning_rate": 1.812287171721897e-05,
+      "loss": 0.3589,
+      "step": 4177
+    },
+    {
+      "epoch": 0.25676007866273354,
+      "grad_norm": 47.5,
+      "learning_rate": 1.812151419497497e-05,
+      "loss": 0.3081,
+      "step": 4178
+    },
+    {
+      "epoch": 0.2568215339233038,
+      "grad_norm": 21.25,
+      "learning_rate": 1.8120156232916804e-05,
+      "loss": 0.3336,
+      "step": 4179
+    },
+    {
+      "epoch": 0.25688298918387414,
+      "grad_norm": 14.75,
+      "learning_rate": 1.811879783111801e-05,
+      "loss": 0.4109,
+      "step": 4180
+    },
+    {
+      "epoch": 0.2569444444444444,
+      "grad_norm": 64.0,
+      "learning_rate": 1.811743898965215e-05,
+      "loss": 0.4077,
+      "step": 4181
+    },
+    {
+      "epoch": 0.25700589970501475,
+      "grad_norm": 54.5,
+      "learning_rate": 1.8116079708592815e-05,
+      "loss": 0.2536,
+      "step": 4182
+    },
+    {
+      "epoch": 0.257067354965585,
+      "grad_norm": 21.875,
+      "learning_rate": 1.8114719988013612e-05,
+      "loss": 0.2583,
+      "step": 4183
+    },
+    {
+      "epoch": 0.25712881022615536,
+      "grad_norm": 12.25,
+      "learning_rate": 1.8113359827988177e-05,
+      "loss": 0.298,
+      "step": 4184
+    },
+    {
+      "epoch": 0.2571902654867257,
+      "grad_norm": 26.0,
+      "learning_rate": 1.8111999228590167e-05,
+      "loss": 0.3614,
+      "step": 4185
+    },
+    {
+      "epoch": 0.25725172074729596,
+      "grad_norm": 33.25,
+      "learning_rate": 1.8110638189893267e-05,
+      "loss": 0.5462,
+      "step": 4186
+    },
+    {
+      "epoch": 0.2573131760078663,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.8109276711971183e-05,
+      "loss": 0.3567,
+      "step": 4187
+    },
+    {
+      "epoch": 0.25737463126843657,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.8107914794897644e-05,
+      "loss": 0.4188,
+      "step": 4188
+    },
+    {
+      "epoch": 0.2574360865290069,
+      "grad_norm": 16.25,
+      "learning_rate": 1.81065524387464e-05,
+      "loss": 0.3487,
+      "step": 4189
+    },
+    {
+      "epoch": 0.2574975417895772,
+      "grad_norm": 16.875,
+      "learning_rate": 1.8105189643591233e-05,
+      "loss": 0.3498,
+      "step": 4190
+    },
+    {
+      "epoch": 0.2575589970501475,
+      "grad_norm": 37.25,
+      "learning_rate": 1.8103826409505944e-05,
+      "loss": 0.3238,
+      "step": 4191
+    },
+    {
+      "epoch": 0.2576204523107178,
+      "grad_norm": 52.5,
+      "learning_rate": 1.8102462736564355e-05,
+      "loss": 0.3865,
+      "step": 4192
+    },
+    {
+      "epoch": 0.2576819075712881,
+      "grad_norm": 9.375,
+      "learning_rate": 1.8101098624840318e-05,
+      "loss": 0.3869,
+      "step": 4193
+    },
+    {
+      "epoch": 0.2577433628318584,
+      "grad_norm": 43.75,
+      "learning_rate": 1.80997340744077e-05,
+      "loss": 0.5714,
+      "step": 4194
+    },
+    {
+      "epoch": 0.2578048180924287,
+      "grad_norm": 25.875,
+      "learning_rate": 1.80983690853404e-05,
+      "loss": 0.3951,
+      "step": 4195
+    },
+    {
+      "epoch": 0.257866273352999,
+      "grad_norm": 19.375,
+      "learning_rate": 1.8097003657712343e-05,
+      "loss": 0.418,
+      "step": 4196
+    },
+    {
+      "epoch": 0.25792772861356933,
+      "grad_norm": 13.375,
+      "learning_rate": 1.809563779159746e-05,
+      "loss": 0.4214,
+      "step": 4197
+    },
+    {
+      "epoch": 0.2579891838741396,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.8094271487069733e-05,
+      "loss": 0.2927,
+      "step": 4198
+    },
+    {
+      "epoch": 0.25805063913470994,
+      "grad_norm": 5.96875,
+      "learning_rate": 1.8092904744203146e-05,
+      "loss": 0.3161,
+      "step": 4199
+    },
+    {
+      "epoch": 0.2581120943952802,
+      "grad_norm": 10.875,
+      "learning_rate": 1.8091537563071714e-05,
+      "loss": 0.2786,
+      "step": 4200
+    },
+    {
+      "epoch": 0.25817354965585054,
+      "grad_norm": 25.125,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.4333,
+      "step": 4201
+    },
+    {
+      "epoch": 0.2582350049164208,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.8088801886310493e-05,
+      "loss": 0.4532,
+      "step": 4202
+    },
+    {
+      "epoch": 0.25829646017699115,
+      "grad_norm": 46.75,
+      "learning_rate": 1.8087433390828855e-05,
+      "loss": 0.5425,
+      "step": 4203
+    },
+    {
+      "epoch": 0.2583579154375615,
+      "grad_norm": 56.5,
+      "learning_rate": 1.8086064457378667e-05,
+      "loss": 0.5156,
+      "step": 4204
+    },
+    {
+      "epoch": 0.25841937069813176,
+      "grad_norm": 5.625,
+      "learning_rate": 1.8084695086034064e-05,
+      "loss": 0.4975,
+      "step": 4205
+    },
+    {
+      "epoch": 0.2584808259587021,
+      "grad_norm": 33.0,
+      "learning_rate": 1.8083325276869207e-05,
+      "loss": 0.4064,
+      "step": 4206
+    },
+    {
+      "epoch": 0.25854228121927236,
+      "grad_norm": 28.625,
+      "learning_rate": 1.8081955029958272e-05,
+      "loss": 0.431,
+      "step": 4207
+    },
+    {
+      "epoch": 0.2586037364798427,
+      "grad_norm": 27.5,
+      "learning_rate": 1.8080584345375466e-05,
+      "loss": 0.301,
+      "step": 4208
+    },
+    {
+      "epoch": 0.25866519174041297,
+      "grad_norm": 18.75,
+      "learning_rate": 1.8079213223195018e-05,
+      "loss": 0.4399,
+      "step": 4209
+    },
+    {
+      "epoch": 0.2587266470009833,
+      "grad_norm": 22.0,
+      "learning_rate": 1.8077841663491174e-05,
+      "loss": 0.3813,
+      "step": 4210
+    },
+    {
+      "epoch": 0.2587881022615536,
+      "grad_norm": 32.0,
+      "learning_rate": 1.807646966633822e-05,
+      "loss": 0.4416,
+      "step": 4211
+    },
+    {
+      "epoch": 0.2588495575221239,
+      "grad_norm": 27.125,
+      "learning_rate": 1.807509723181045e-05,
+      "loss": 0.5876,
+      "step": 4212
+    },
+    {
+      "epoch": 0.2589110127826942,
+      "grad_norm": 16.25,
+      "learning_rate": 1.8073724359982184e-05,
+      "loss": 0.5149,
+      "step": 4213
+    },
+    {
+      "epoch": 0.2589724680432645,
+      "grad_norm": 45.25,
+      "learning_rate": 1.8072351050927774e-05,
+      "loss": 0.4984,
+      "step": 4214
+    },
+    {
+      "epoch": 0.2590339233038348,
+      "grad_norm": 25.625,
+      "learning_rate": 1.8070977304721587e-05,
+      "loss": 0.3867,
+      "step": 4215
+    },
+    {
+      "epoch": 0.2590953785644051,
+      "grad_norm": 17.125,
+      "learning_rate": 1.806960312143802e-05,
+      "loss": 0.3835,
+      "step": 4216
+    },
+    {
+      "epoch": 0.2591568338249754,
+      "grad_norm": 30.875,
+      "learning_rate": 1.806822850115149e-05,
+      "loss": 0.5571,
+      "step": 4217
+    },
+    {
+      "epoch": 0.25921828908554573,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.8066853443936435e-05,
+      "loss": 0.2761,
+      "step": 4218
+    },
+    {
+      "epoch": 0.259279744346116,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.8065477949867327e-05,
+      "loss": 0.4148,
+      "step": 4219
+    },
+    {
+      "epoch": 0.25934119960668633,
+      "grad_norm": 18.0,
+      "learning_rate": 1.806410201901865e-05,
+      "loss": 0.5232,
+      "step": 4220
+    },
+    {
+      "epoch": 0.2594026548672566,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.8062725651464913e-05,
+      "loss": 0.2265,
+      "step": 4221
+    },
+    {
+      "epoch": 0.25946411012782694,
+      "grad_norm": 28.875,
+      "learning_rate": 1.806134884728066e-05,
+      "loss": 0.4246,
+      "step": 4222
+    },
+    {
+      "epoch": 0.2595255653883973,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.8059971606540446e-05,
+      "loss": 0.3047,
+      "step": 4223
+    },
+    {
+      "epoch": 0.25958702064896755,
+      "grad_norm": 25.125,
+      "learning_rate": 1.8058593929318855e-05,
+      "loss": 0.3112,
+      "step": 4224
+    },
+    {
+      "epoch": 0.2596484759095379,
+      "grad_norm": 22.5,
+      "learning_rate": 1.8057215815690494e-05,
+      "loss": 0.388,
+      "step": 4225
+    },
+    {
+      "epoch": 0.25970993117010815,
+      "grad_norm": 40.0,
+      "learning_rate": 1.8055837265729996e-05,
+      "loss": 0.3805,
+      "step": 4226
+    },
+    {
+      "epoch": 0.2597713864306785,
+      "grad_norm": 21.625,
+      "learning_rate": 1.8054458279512012e-05,
+      "loss": 0.3759,
+      "step": 4227
+    },
+    {
+      "epoch": 0.25983284169124876,
+      "grad_norm": 10.3125,
+      "learning_rate": 1.8053078857111218e-05,
+      "loss": 0.3041,
+      "step": 4228
+    },
+    {
+      "epoch": 0.2598942969518191,
+      "grad_norm": 24.0,
+      "learning_rate": 1.8051698998602324e-05,
+      "loss": 0.3018,
+      "step": 4229
+    },
+    {
+      "epoch": 0.25995575221238937,
+      "grad_norm": 23.125,
+      "learning_rate": 1.805031870406005e-05,
+      "loss": 0.3549,
+      "step": 4230
+    },
+    {
+      "epoch": 0.2600172074729597,
+      "grad_norm": 27.125,
+      "learning_rate": 1.804893797355914e-05,
+      "loss": 0.2752,
+      "step": 4231
+    },
+    {
+      "epoch": 0.26007866273353,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.8047556807174378e-05,
+      "loss": 0.3193,
+      "step": 4232
+    },
+    {
+      "epoch": 0.2601401179941003,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.804617520498055e-05,
+      "loss": 0.3511,
+      "step": 4233
+    },
+    {
+      "epoch": 0.2602015732546706,
+      "grad_norm": 12.75,
+      "learning_rate": 1.8044793167052476e-05,
+      "loss": 0.5629,
+      "step": 4234
+    },
+    {
+      "epoch": 0.2602630285152409,
+      "grad_norm": 13.125,
+      "learning_rate": 1.8043410693465e-05,
+      "loss": 0.4739,
+      "step": 4235
+    },
+    {
+      "epoch": 0.2603244837758112,
+      "grad_norm": 22.625,
+      "learning_rate": 1.8042027784292998e-05,
+      "loss": 0.4125,
+      "step": 4236
+    },
+    {
+      "epoch": 0.2603859390363815,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.8040644439611348e-05,
+      "loss": 0.4686,
+      "step": 4237
+    },
+    {
+      "epoch": 0.2604473942969518,
+      "grad_norm": 15.25,
+      "learning_rate": 1.803926065949497e-05,
+      "loss": 0.2943,
+      "step": 4238
+    },
+    {
+      "epoch": 0.2605088495575221,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.80378764440188e-05,
+      "loss": 0.3094,
+      "step": 4239
+    },
+    {
+      "epoch": 0.2605703048180924,
+      "grad_norm": 42.25,
+      "learning_rate": 1.80364917932578e-05,
+      "loss": 0.461,
+      "step": 4240
+    },
+    {
+      "epoch": 0.26063176007866273,
+      "grad_norm": 11.4375,
+      "learning_rate": 1.8035106707286957e-05,
+      "loss": 0.2831,
+      "step": 4241
+    },
+    {
+      "epoch": 0.26069321533923306,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.803372118618127e-05,
+      "loss": 0.3005,
+      "step": 4242
+    },
+    {
+      "epoch": 0.26075467059980334,
+      "grad_norm": 26.875,
+      "learning_rate": 1.803233523001578e-05,
+      "loss": 0.4125,
+      "step": 4243
+    },
+    {
+      "epoch": 0.26081612586037367,
+      "grad_norm": 23.625,
+      "learning_rate": 1.8030948838865542e-05,
+      "loss": 0.4211,
+      "step": 4244
+    },
+    {
+      "epoch": 0.26087758112094395,
+      "grad_norm": 24.0,
+      "learning_rate": 1.8029562012805627e-05,
+      "loss": 0.3436,
+      "step": 4245
+    },
+    {
+      "epoch": 0.2609390363815143,
+      "grad_norm": 7.625,
+      "learning_rate": 1.8028174751911147e-05,
+      "loss": 0.3951,
+      "step": 4246
+    },
+    {
+      "epoch": 0.26100049164208455,
+      "grad_norm": 47.75,
+      "learning_rate": 1.802678705625722e-05,
+      "loss": 0.4105,
+      "step": 4247
+    },
+    {
+      "epoch": 0.2610619469026549,
+      "grad_norm": 23.875,
+      "learning_rate": 1.8025398925919006e-05,
+      "loss": 0.289,
+      "step": 4248
+    },
+    {
+      "epoch": 0.26112340216322516,
+      "grad_norm": 5.6875,
+      "learning_rate": 1.802401036097167e-05,
+      "loss": 0.3061,
+      "step": 4249
+    },
+    {
+      "epoch": 0.2611848574237955,
+      "grad_norm": 6.25,
+      "learning_rate": 1.8022621361490406e-05,
+      "loss": 0.4481,
+      "step": 4250
+    },
+    {
+      "epoch": 0.26124631268436577,
+      "grad_norm": 27.125,
+      "learning_rate": 1.802123192755044e-05,
+      "loss": 0.3947,
+      "step": 4251
+    },
+    {
+      "epoch": 0.2613077679449361,
+      "grad_norm": 18.125,
+      "learning_rate": 1.801984205922701e-05,
+      "loss": 0.2871,
+      "step": 4252
+    },
+    {
+      "epoch": 0.2613692232055064,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.8018451756595395e-05,
+      "loss": 0.3546,
+      "step": 4253
+    },
+    {
+      "epoch": 0.2614306784660767,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.8017061019730874e-05,
+      "loss": 0.3179,
+      "step": 4254
+    },
+    {
+      "epoch": 0.261492133726647,
+      "grad_norm": 18.125,
+      "learning_rate": 1.8015669848708768e-05,
+      "loss": 0.2804,
+      "step": 4255
+    },
+    {
+      "epoch": 0.2615535889872173,
+      "grad_norm": 23.25,
+      "learning_rate": 1.8014278243604407e-05,
+      "loss": 0.2787,
+      "step": 4256
+    },
+    {
+      "epoch": 0.2616150442477876,
+      "grad_norm": 28.5,
+      "learning_rate": 1.8012886204493164e-05,
+      "loss": 0.3825,
+      "step": 4257
+    },
+    {
+      "epoch": 0.2616764995083579,
+      "grad_norm": 29.875,
+      "learning_rate": 1.8011493731450412e-05,
+      "loss": 0.5037,
+      "step": 4258
+    },
+    {
+      "epoch": 0.2617379547689282,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.8010100824551563e-05,
+      "loss": 0.3621,
+      "step": 4259
+    },
+    {
+      "epoch": 0.2617994100294985,
+      "grad_norm": 17.625,
+      "learning_rate": 1.8008707483872055e-05,
+      "loss": 0.5608,
+      "step": 4260
+    },
+    {
+      "epoch": 0.26186086529006886,
+      "grad_norm": 12.75,
+      "learning_rate": 1.8007313709487334e-05,
+      "loss": 0.4398,
+      "step": 4261
+    },
+    {
+      "epoch": 0.26192232055063913,
+      "grad_norm": 34.5,
+      "learning_rate": 1.800591950147289e-05,
+      "loss": 0.3544,
+      "step": 4262
+    },
+    {
+      "epoch": 0.26198377581120946,
+      "grad_norm": 11.25,
+      "learning_rate": 1.8004524859904212e-05,
+      "loss": 0.3154,
+      "step": 4263
+    },
+    {
+      "epoch": 0.26204523107177974,
+      "grad_norm": 24.75,
+      "learning_rate": 1.8003129784856832e-05,
+      "loss": 0.3709,
+      "step": 4264
+    },
+    {
+      "epoch": 0.26210668633235007,
+      "grad_norm": 4.375,
+      "learning_rate": 1.8001734276406302e-05,
+      "loss": 0.2653,
+      "step": 4265
+    },
+    {
+      "epoch": 0.26216814159292035,
+      "grad_norm": 8.125,
+      "learning_rate": 1.800033833462819e-05,
+      "loss": 0.3774,
+      "step": 4266
+    },
+    {
+      "epoch": 0.2622295968534907,
+      "grad_norm": 21.75,
+      "learning_rate": 1.7998941959598097e-05,
+      "loss": 0.3589,
+      "step": 4267
+    },
+    {
+      "epoch": 0.26229105211406095,
+      "grad_norm": 6.59375,
+      "learning_rate": 1.7997545151391637e-05,
+      "loss": 0.4866,
+      "step": 4268
+    },
+    {
+      "epoch": 0.2623525073746313,
+      "grad_norm": 16.5,
+      "learning_rate": 1.7996147910084458e-05,
+      "loss": 0.3565,
+      "step": 4269
+    },
+    {
+      "epoch": 0.26241396263520156,
+      "grad_norm": 24.625,
+      "learning_rate": 1.799475023575222e-05,
+      "loss": 0.3473,
+      "step": 4270
+    },
+    {
+      "epoch": 0.2624754178957719,
+      "grad_norm": 25.625,
+      "learning_rate": 1.7993352128470617e-05,
+      "loss": 0.3704,
+      "step": 4271
+    },
+    {
+      "epoch": 0.26253687315634217,
+      "grad_norm": 5.46875,
+      "learning_rate": 1.7991953588315368e-05,
+      "loss": 0.3182,
+      "step": 4272
+    },
+    {
+      "epoch": 0.2625983284169125,
+      "grad_norm": 11.5,
+      "learning_rate": 1.79905546153622e-05,
+      "loss": 0.384,
+      "step": 4273
+    },
+    {
+      "epoch": 0.26265978367748277,
+      "grad_norm": 18.75,
+      "learning_rate": 1.798915520968688e-05,
+      "loss": 0.4202,
+      "step": 4274
+    },
+    {
+      "epoch": 0.2627212389380531,
+      "grad_norm": 22.25,
+      "learning_rate": 1.7987755371365185e-05,
+      "loss": 0.3984,
+      "step": 4275
+    },
+    {
+      "epoch": 0.2627826941986234,
+      "grad_norm": 60.75,
+      "learning_rate": 1.798635510047293e-05,
+      "loss": 0.5458,
+      "step": 4276
+    },
+    {
+      "epoch": 0.2628441494591937,
+      "grad_norm": 8.875,
+      "learning_rate": 1.798495439708594e-05,
+      "loss": 0.3383,
+      "step": 4277
+    },
+    {
+      "epoch": 0.262905604719764,
+      "grad_norm": 26.25,
+      "learning_rate": 1.7983553261280072e-05,
+      "loss": 0.3799,
+      "step": 4278
+    },
+    {
+      "epoch": 0.2629670599803343,
+      "grad_norm": 35.25,
+      "learning_rate": 1.7982151693131206e-05,
+      "loss": 0.5531,
+      "step": 4279
+    },
+    {
+      "epoch": 0.26302851524090465,
+      "grad_norm": 5.84375,
+      "learning_rate": 1.7980749692715233e-05,
+      "loss": 0.3394,
+      "step": 4280
+    },
+    {
+      "epoch": 0.2630899705014749,
+      "grad_norm": 25.5,
+      "learning_rate": 1.7979347260108088e-05,
+      "loss": 0.3616,
+      "step": 4281
+    },
+    {
+      "epoch": 0.26315142576204525,
+      "grad_norm": 24.125,
+      "learning_rate": 1.7977944395385713e-05,
+      "loss": 0.5491,
+      "step": 4282
+    },
+    {
+      "epoch": 0.26321288102261553,
+      "grad_norm": 20.25,
+      "learning_rate": 1.797654109862408e-05,
+      "loss": 0.4295,
+      "step": 4283
+    },
+    {
+      "epoch": 0.26327433628318586,
+      "grad_norm": 8.375,
+      "learning_rate": 1.797513736989918e-05,
+      "loss": 0.5473,
+      "step": 4284
+    },
+    {
+      "epoch": 0.26333579154375614,
+      "grad_norm": 15.125,
+      "learning_rate": 1.7973733209287036e-05,
+      "loss": 0.4334,
+      "step": 4285
+    },
+    {
+      "epoch": 0.26339724680432647,
+      "grad_norm": 15.25,
+      "learning_rate": 1.797232861686369e-05,
+      "loss": 0.3881,
+      "step": 4286
+    },
+    {
+      "epoch": 0.26345870206489674,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.7970923592705203e-05,
+      "loss": 0.3396,
+      "step": 4287
+    },
+    {
+      "epoch": 0.2635201573254671,
+      "grad_norm": 16.375,
+      "learning_rate": 1.7969518136887664e-05,
+      "loss": 0.3157,
+      "step": 4288
+    },
+    {
+      "epoch": 0.26358161258603735,
+      "grad_norm": 19.125,
+      "learning_rate": 1.7968112249487183e-05,
+      "loss": 0.4432,
+      "step": 4289
+    },
+    {
+      "epoch": 0.2636430678466077,
+      "grad_norm": 18.5,
+      "learning_rate": 1.79667059305799e-05,
+      "loss": 0.4313,
+      "step": 4290
+    },
+    {
+      "epoch": 0.26370452310717796,
+      "grad_norm": 33.25,
+      "learning_rate": 1.7965299180241963e-05,
+      "loss": 0.4168,
+      "step": 4291
+    },
+    {
+      "epoch": 0.2637659783677483,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.7963891998549562e-05,
+      "loss": 0.534,
+      "step": 4292
+    },
+    {
+      "epoch": 0.26382743362831856,
+      "grad_norm": 5.34375,
+      "learning_rate": 1.7962484385578904e-05,
+      "loss": 0.2712,
+      "step": 4293
+    },
+    {
+      "epoch": 0.2638888888888889,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.796107634140621e-05,
+      "loss": 0.2788,
+      "step": 4294
+    },
+    {
+      "epoch": 0.26395034414945917,
+      "grad_norm": 55.25,
+      "learning_rate": 1.795966786610773e-05,
+      "loss": 0.4355,
+      "step": 4295
+    },
+    {
+      "epoch": 0.2640117994100295,
+      "grad_norm": 3.125,
+      "learning_rate": 1.7958258959759747e-05,
+      "loss": 0.254,
+      "step": 4296
+    },
+    {
+      "epoch": 0.2640732546705998,
+      "grad_norm": 24.25,
+      "learning_rate": 1.7956849622438554e-05,
+      "loss": 0.305,
+      "step": 4297
+    },
+    {
+      "epoch": 0.2641347099311701,
+      "grad_norm": 17.125,
+      "learning_rate": 1.7955439854220477e-05,
+      "loss": 0.2689,
+      "step": 4298
+    },
+    {
+      "epoch": 0.26419616519174044,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.7954029655181856e-05,
+      "loss": 0.3408,
+      "step": 4299
+    },
+    {
+      "epoch": 0.2642576204523107,
+      "grad_norm": 4.3125,
+      "learning_rate": 1.795261902539906e-05,
+      "loss": 0.2369,
+      "step": 4300
+    },
+    {
+      "epoch": 0.26431907571288105,
+      "grad_norm": 38.75,
+      "learning_rate": 1.795120796494848e-05,
+      "loss": 0.3055,
+      "step": 4301
+    },
+    {
+      "epoch": 0.2643805309734513,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.7949796473906533e-05,
+      "loss": 0.2509,
+      "step": 4302
+    },
+    {
+      "epoch": 0.26444198623402165,
+      "grad_norm": 33.5,
+      "learning_rate": 1.794838455234966e-05,
+      "loss": 0.2663,
+      "step": 4303
+    },
+    {
+      "epoch": 0.26450344149459193,
+      "grad_norm": 20.625,
+      "learning_rate": 1.794697220035431e-05,
+      "loss": 0.2604,
+      "step": 4304
+    },
+    {
+      "epoch": 0.26456489675516226,
+      "grad_norm": 21.125,
+      "learning_rate": 1.7945559417996984e-05,
+      "loss": 0.394,
+      "step": 4305
+    },
+    {
+      "epoch": 0.26462635201573254,
+      "grad_norm": 18.75,
+      "learning_rate": 1.7944146205354182e-05,
+      "loss": 0.4566,
+      "step": 4306
+    },
+    {
+      "epoch": 0.26468780727630287,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.7942732562502435e-05,
+      "loss": 0.2514,
+      "step": 4307
+    },
+    {
+      "epoch": 0.26474926253687314,
+      "grad_norm": 15.375,
+      "learning_rate": 1.79413184895183e-05,
+      "loss": 0.372,
+      "step": 4308
+    },
+    {
+      "epoch": 0.2648107177974435,
+      "grad_norm": 23.25,
+      "learning_rate": 1.7939903986478354e-05,
+      "loss": 0.3144,
+      "step": 4309
+    },
+    {
+      "epoch": 0.26487217305801375,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.7938489053459198e-05,
+      "loss": 0.2951,
+      "step": 4310
+    },
+    {
+      "epoch": 0.2649336283185841,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.793707369053746e-05,
+      "loss": 0.2469,
+      "step": 4311
+    },
+    {
+      "epoch": 0.26499508357915436,
+      "grad_norm": 19.25,
+      "learning_rate": 1.793565789778978e-05,
+      "loss": 0.3618,
+      "step": 4312
+    },
+    {
+      "epoch": 0.2650565388397247,
+      "grad_norm": 18.375,
+      "learning_rate": 1.7934241675292834e-05,
+      "loss": 0.4894,
+      "step": 4313
+    },
+    {
+      "epoch": 0.26511799410029496,
+      "grad_norm": 30.625,
+      "learning_rate": 1.793282502312332e-05,
+      "loss": 0.6266,
+      "step": 4314
+    },
+    {
+      "epoch": 0.2651794493608653,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.793140794135795e-05,
+      "loss": 0.4613,
+      "step": 4315
+    },
+    {
+      "epoch": 0.26524090462143557,
+      "grad_norm": 21.5,
+      "learning_rate": 1.7929990430073463e-05,
+      "loss": 0.5646,
+      "step": 4316
+    },
+    {
+      "epoch": 0.2653023598820059,
+      "grad_norm": 23.125,
+      "learning_rate": 1.792857248934663e-05,
+      "loss": 0.3674,
+      "step": 4317
+    },
+    {
+      "epoch": 0.26536381514257623,
+      "grad_norm": 48.0,
+      "learning_rate": 1.7927154119254234e-05,
+      "loss": 0.6913,
+      "step": 4318
+    },
+    {
+      "epoch": 0.2654252704031465,
+      "grad_norm": 27.25,
+      "learning_rate": 1.792573531987309e-05,
+      "loss": 0.4885,
+      "step": 4319
+    },
+    {
+      "epoch": 0.26548672566371684,
+      "grad_norm": 19.5,
+      "learning_rate": 1.792431609128003e-05,
+      "loss": 0.4729,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2655481809242871,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.792289643355191e-05,
+      "loss": 0.4218,
+      "step": 4321
+    },
+    {
+      "epoch": 0.26560963618485744,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.7921476346765607e-05,
+      "loss": 0.3223,
+      "step": 4322
+    },
+    {
+      "epoch": 0.2656710914454277,
+      "grad_norm": 10.0,
+      "learning_rate": 1.7920055830998032e-05,
+      "loss": 0.3082,
+      "step": 4323
+    },
+    {
+      "epoch": 0.26573254670599805,
+      "grad_norm": 6.5625,
+      "learning_rate": 1.791863488632611e-05,
+      "loss": 0.2957,
+      "step": 4324
+    },
+    {
+      "epoch": 0.2657940019665683,
+      "grad_norm": 24.625,
+      "learning_rate": 1.7917213512826785e-05,
+      "loss": 0.3626,
+      "step": 4325
+    },
+    {
+      "epoch": 0.26585545722713866,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.7915791710577035e-05,
+      "loss": 0.249,
+      "step": 4326
+    },
+    {
+      "epoch": 0.26591691248770893,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.7914369479653858e-05,
+      "loss": 0.3,
+      "step": 4327
+    },
+    {
+      "epoch": 0.26597836774827927,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.7912946820134273e-05,
+      "loss": 0.2719,
+      "step": 4328
+    },
+    {
+      "epoch": 0.26603982300884954,
+      "grad_norm": 14.75,
+      "learning_rate": 1.7911523732095322e-05,
+      "loss": 0.5472,
+      "step": 4329
+    },
+    {
+      "epoch": 0.26610127826941987,
+      "grad_norm": 63.0,
+      "learning_rate": 1.791010021561407e-05,
+      "loss": 0.5495,
+      "step": 4330
+    },
+    {
+      "epoch": 0.26616273352999015,
+      "grad_norm": 21.5,
+      "learning_rate": 1.7908676270767608e-05,
+      "loss": 0.4389,
+      "step": 4331
+    },
+    {
+      "epoch": 0.2662241887905605,
+      "grad_norm": 22.125,
+      "learning_rate": 1.7907251897633052e-05,
+      "loss": 0.6534,
+      "step": 4332
+    },
+    {
+      "epoch": 0.26628564405113075,
+      "grad_norm": 38.25,
+      "learning_rate": 1.7905827096287532e-05,
+      "loss": 0.6959,
+      "step": 4333
+    },
+    {
+      "epoch": 0.2663470993117011,
+      "grad_norm": 26.25,
+      "learning_rate": 1.790440186680821e-05,
+      "loss": 0.4163,
+      "step": 4334
+    },
+    {
+      "epoch": 0.26640855457227136,
+      "grad_norm": 55.25,
+      "learning_rate": 1.7902976209272265e-05,
+      "loss": 0.443,
+      "step": 4335
+    },
+    {
+      "epoch": 0.2664700098328417,
+      "grad_norm": 30.25,
+      "learning_rate": 1.7901550123756906e-05,
+      "loss": 0.3894,
+      "step": 4336
+    },
+    {
+      "epoch": 0.266531465093412,
+      "grad_norm": 15.4375,
+      "learning_rate": 1.790012361033936e-05,
+      "loss": 0.358,
+      "step": 4337
+    },
+    {
+      "epoch": 0.2665929203539823,
+      "grad_norm": 33.5,
+      "learning_rate": 1.789869666909688e-05,
+      "loss": 0.4074,
+      "step": 4338
+    },
+    {
+      "epoch": 0.26665437561455263,
+      "grad_norm": 31.625,
+      "learning_rate": 1.789726930010674e-05,
+      "loss": 0.463,
+      "step": 4339
+    },
+    {
+      "epoch": 0.2667158308751229,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7895841503446237e-05,
+      "loss": 0.2734,
+      "step": 4340
+    },
+    {
+      "epoch": 0.26677728613569324,
+      "grad_norm": 35.25,
+      "learning_rate": 1.7894413279192693e-05,
+      "loss": 0.3907,
+      "step": 4341
+    },
+    {
+      "epoch": 0.2668387413962635,
+      "grad_norm": 16.125,
+      "learning_rate": 1.789298462742345e-05,
+      "loss": 0.4822,
+      "step": 4342
+    },
+    {
+      "epoch": 0.26690019665683384,
+      "grad_norm": 6.125,
+      "learning_rate": 1.789155554821588e-05,
+      "loss": 0.325,
+      "step": 4343
+    },
+    {
+      "epoch": 0.2669616519174041,
+      "grad_norm": 10.25,
+      "learning_rate": 1.7890126041647375e-05,
+      "loss": 0.3348,
+      "step": 4344
+    },
+    {
+      "epoch": 0.26702310717797445,
+      "grad_norm": 25.375,
+      "learning_rate": 1.7888696107795343e-05,
+      "loss": 0.3869,
+      "step": 4345
+    },
+    {
+      "epoch": 0.2670845624385447,
+      "grad_norm": 8.375,
+      "learning_rate": 1.7887265746737224e-05,
+      "loss": 0.4301,
+      "step": 4346
+    },
+    {
+      "epoch": 0.26714601769911506,
+      "grad_norm": 39.25,
+      "learning_rate": 1.7885834958550473e-05,
+      "loss": 0.4198,
+      "step": 4347
+    },
+    {
+      "epoch": 0.26720747295968533,
+      "grad_norm": 16.875,
+      "learning_rate": 1.7884403743312583e-05,
+      "loss": 0.4369,
+      "step": 4348
+    },
+    {
+      "epoch": 0.26726892822025566,
+      "grad_norm": 29.375,
+      "learning_rate": 1.7882972101101054e-05,
+      "loss": 0.4247,
+      "step": 4349
+    },
+    {
+      "epoch": 0.26733038348082594,
+      "grad_norm": 17.25,
+      "learning_rate": 1.7881540031993415e-05,
+      "loss": 0.3474,
+      "step": 4350
+    },
+    {
+      "epoch": 0.26739183874139627,
+      "grad_norm": 17.125,
+      "learning_rate": 1.788010753606722e-05,
+      "loss": 0.3022,
+      "step": 4351
+    },
+    {
+      "epoch": 0.26745329400196655,
+      "grad_norm": 35.5,
+      "learning_rate": 1.787867461340005e-05,
+      "loss": 0.2518,
+      "step": 4352
+    },
+    {
+      "epoch": 0.2675147492625369,
+      "grad_norm": 36.75,
+      "learning_rate": 1.787724126406949e-05,
+      "loss": 0.5584,
+      "step": 4353
+    },
+    {
+      "epoch": 0.26757620452310715,
+      "grad_norm": 31.0,
+      "learning_rate": 1.7875807488153173e-05,
+      "loss": 0.4568,
+      "step": 4354
+    },
+    {
+      "epoch": 0.2676376597836775,
+      "grad_norm": 27.5,
+      "learning_rate": 1.7874373285728742e-05,
+      "loss": 0.5332,
+      "step": 4355
+    },
+    {
+      "epoch": 0.2676991150442478,
+      "grad_norm": 19.875,
+      "learning_rate": 1.7872938656873864e-05,
+      "loss": 0.4849,
+      "step": 4356
+    },
+    {
+      "epoch": 0.2677605703048181,
+      "grad_norm": 22.625,
+      "learning_rate": 1.7871503601666233e-05,
+      "loss": 0.3369,
+      "step": 4357
+    },
+    {
+      "epoch": 0.2678220255653884,
+      "grad_norm": 19.125,
+      "learning_rate": 1.7870068120183563e-05,
+      "loss": 0.4508,
+      "step": 4358
+    },
+    {
+      "epoch": 0.2678834808259587,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.7868632212503583e-05,
+      "loss": 0.2925,
+      "step": 4359
+    },
+    {
+      "epoch": 0.26794493608652903,
+      "grad_norm": 8.625,
+      "learning_rate": 1.7867195878704062e-05,
+      "loss": 0.3755,
+      "step": 4360
+    },
+    {
+      "epoch": 0.2680063913470993,
+      "grad_norm": 25.25,
+      "learning_rate": 1.7865759118862784e-05,
+      "loss": 0.4023,
+      "step": 4361
+    },
+    {
+      "epoch": 0.26806784660766964,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.7864321933057547e-05,
+      "loss": 0.4347,
+      "step": 4362
+    },
+    {
+      "epoch": 0.2681293018682399,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.786288432136619e-05,
+      "loss": 0.618,
+      "step": 4363
+    },
+    {
+      "epoch": 0.26819075712881024,
+      "grad_norm": 6.5625,
+      "learning_rate": 1.7861446283866562e-05,
+      "loss": 0.3377,
+      "step": 4364
+    },
+    {
+      "epoch": 0.2682522123893805,
+      "grad_norm": 19.875,
+      "learning_rate": 1.786000782063654e-05,
+      "loss": 0.4208,
+      "step": 4365
+    },
+    {
+      "epoch": 0.26831366764995085,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.785856893175402e-05,
+      "loss": 0.3504,
+      "step": 4366
+    },
+    {
+      "epoch": 0.2683751229105211,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.7857129617296926e-05,
+      "loss": 0.3679,
+      "step": 4367
+    },
+    {
+      "epoch": 0.26843657817109146,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.7855689877343204e-05,
+      "loss": 0.407,
+      "step": 4368
+    },
+    {
+      "epoch": 0.26849803343166173,
+      "grad_norm": 12.75,
+      "learning_rate": 1.785424971197082e-05,
+      "loss": 0.2408,
+      "step": 4369
+    },
+    {
+      "epoch": 0.26855948869223206,
+      "grad_norm": 20.75,
+      "learning_rate": 1.7852809121257765e-05,
+      "loss": 0.3611,
+      "step": 4370
+    },
+    {
+      "epoch": 0.26862094395280234,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.7851368105282054e-05,
+      "loss": 0.2275,
+      "step": 4371
+    },
+    {
+      "epoch": 0.26868239921337267,
+      "grad_norm": 26.375,
+      "learning_rate": 1.7849926664121726e-05,
+      "loss": 0.3555,
+      "step": 4372
+    },
+    {
+      "epoch": 0.26874385447394294,
+      "grad_norm": 33.25,
+      "learning_rate": 1.7848484797854838e-05,
+      "loss": 0.2765,
+      "step": 4373
+    },
+    {
+      "epoch": 0.2688053097345133,
+      "grad_norm": 30.25,
+      "learning_rate": 1.7847042506559467e-05,
+      "loss": 0.3247,
+      "step": 4374
+    },
+    {
+      "epoch": 0.26886676499508355,
+      "grad_norm": 50.5,
+      "learning_rate": 1.7845599790313735e-05,
+      "loss": 0.3491,
+      "step": 4375
+    },
+    {
+      "epoch": 0.2689282202556539,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.784415664919576e-05,
+      "loss": 0.2743,
+      "step": 4376
+    },
+    {
+      "epoch": 0.2689896755162242,
+      "grad_norm": 15.75,
+      "learning_rate": 1.7842713083283695e-05,
+      "loss": 0.5154,
+      "step": 4377
+    },
+    {
+      "epoch": 0.2690511307767945,
+      "grad_norm": 47.75,
+      "learning_rate": 1.7841269092655714e-05,
+      "loss": 0.5194,
+      "step": 4378
+    },
+    {
+      "epoch": 0.2691125860373648,
+      "grad_norm": 36.5,
+      "learning_rate": 1.7839824677390023e-05,
+      "loss": 0.3639,
+      "step": 4379
+    },
+    {
+      "epoch": 0.2691740412979351,
+      "grad_norm": 42.5,
+      "learning_rate": 1.7838379837564838e-05,
+      "loss": 0.6486,
+      "step": 4380
+    },
+    {
+      "epoch": 0.2692354965585054,
+      "grad_norm": 28.75,
+      "learning_rate": 1.78369345732584e-05,
+      "loss": 0.6015,
+      "step": 4381
+    },
+    {
+      "epoch": 0.2692969518190757,
+      "grad_norm": 5.0625,
+      "learning_rate": 1.7835488884548978e-05,
+      "loss": 0.4801,
+      "step": 4382
+    },
+    {
+      "epoch": 0.26935840707964603,
+      "grad_norm": 4.5,
+      "learning_rate": 1.7834042771514868e-05,
+      "loss": 0.3875,
+      "step": 4383
+    },
+    {
+      "epoch": 0.2694198623402163,
+      "grad_norm": 5.90625,
+      "learning_rate": 1.7832596234234376e-05,
+      "loss": 0.4638,
+      "step": 4384
+    },
+    {
+      "epoch": 0.26948131760078664,
+      "grad_norm": 12.875,
+      "learning_rate": 1.7831149272785843e-05,
+      "loss": 0.2728,
+      "step": 4385
+    },
+    {
+      "epoch": 0.2695427728613569,
+      "grad_norm": 8.625,
+      "learning_rate": 1.7829701887247618e-05,
+      "loss": 0.2833,
+      "step": 4386
+    },
+    {
+      "epoch": 0.26960422812192725,
+      "grad_norm": 24.125,
+      "learning_rate": 1.78282540776981e-05,
+      "loss": 0.3736,
+      "step": 4387
+    },
+    {
+      "epoch": 0.2696656833824975,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.782680584421568e-05,
+      "loss": 0.297,
+      "step": 4388
+    },
+    {
+      "epoch": 0.26972713864306785,
+      "grad_norm": 50.25,
+      "learning_rate": 1.7825357186878788e-05,
+      "loss": 0.3855,
+      "step": 4389
+    },
+    {
+      "epoch": 0.26978859390363813,
+      "grad_norm": 15.625,
+      "learning_rate": 1.7823908105765883e-05,
+      "loss": 0.3442,
+      "step": 4390
+    },
+    {
+      "epoch": 0.26985004916420846,
+      "grad_norm": 9.125,
+      "learning_rate": 1.7822458600955432e-05,
+      "loss": 0.3818,
+      "step": 4391
+    },
+    {
+      "epoch": 0.26991150442477874,
+      "grad_norm": 14.625,
+      "learning_rate": 1.782100867252593e-05,
+      "loss": 0.264,
+      "step": 4392
+    },
+    {
+      "epoch": 0.26997295968534907,
+      "grad_norm": 44.0,
+      "learning_rate": 1.7819558320555902e-05,
+      "loss": 0.2789,
+      "step": 4393
+    },
+    {
+      "epoch": 0.27003441494591934,
+      "grad_norm": 63.0,
+      "learning_rate": 1.781810754512389e-05,
+      "loss": 0.4387,
+      "step": 4394
+    },
+    {
+      "epoch": 0.2700958702064897,
+      "grad_norm": 43.75,
+      "learning_rate": 1.7816656346308453e-05,
+      "loss": 0.3304,
+      "step": 4395
+    },
+    {
+      "epoch": 0.27015732546706,
+      "grad_norm": 17.0,
+      "learning_rate": 1.781520472418819e-05,
+      "loss": 0.3274,
+      "step": 4396
+    },
+    {
+      "epoch": 0.2702187807276303,
+      "grad_norm": 18.625,
+      "learning_rate": 1.7813752678841702e-05,
+      "loss": 0.3602,
+      "step": 4397
+    },
+    {
+      "epoch": 0.2702802359882006,
+      "grad_norm": 27.5,
+      "learning_rate": 1.7812300210347634e-05,
+      "loss": 0.4781,
+      "step": 4398
+    },
+    {
+      "epoch": 0.2703416912487709,
+      "grad_norm": 22.75,
+      "learning_rate": 1.7810847318784632e-05,
+      "loss": 0.3716,
+      "step": 4399
+    },
+    {
+      "epoch": 0.2704031465093412,
+      "grad_norm": 23.25,
+      "learning_rate": 1.7809394004231386e-05,
+      "loss": 0.3729,
+      "step": 4400
+    },
+    {
+      "epoch": 0.2704646017699115,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.7807940266766595e-05,
+      "loss": 0.3171,
+      "step": 4401
+    },
+    {
+      "epoch": 0.2705260570304818,
+      "grad_norm": 70.5,
+      "learning_rate": 1.7806486106468983e-05,
+      "loss": 0.3627,
+      "step": 4402
+    },
+    {
+      "epoch": 0.2705875122910521,
+      "grad_norm": 20.5,
+      "learning_rate": 1.78050315234173e-05,
+      "loss": 0.385,
+      "step": 4403
+    },
+    {
+      "epoch": 0.27064896755162243,
+      "grad_norm": 26.625,
+      "learning_rate": 1.7803576517690323e-05,
+      "loss": 0.2544,
+      "step": 4404
+    },
+    {
+      "epoch": 0.2707104228121927,
+      "grad_norm": 32.25,
+      "learning_rate": 1.780212108936684e-05,
+      "loss": 0.33,
+      "step": 4405
+    },
+    {
+      "epoch": 0.27077187807276304,
+      "grad_norm": 18.125,
+      "learning_rate": 1.780066523852567e-05,
+      "loss": 0.3494,
+      "step": 4406
+    },
+    {
+      "epoch": 0.2708333333333333,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.779920896524565e-05,
+      "loss": 0.2756,
+      "step": 4407
+    },
+    {
+      "epoch": 0.27089478859390365,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.7797752269605654e-05,
+      "loss": 0.2128,
+      "step": 4408
+    },
+    {
+      "epoch": 0.2709562438544739,
+      "grad_norm": 6.125,
+      "learning_rate": 1.779629515168456e-05,
+      "loss": 0.2239,
+      "step": 4409
+    },
+    {
+      "epoch": 0.27101769911504425,
+      "grad_norm": 25.75,
+      "learning_rate": 1.779483761156128e-05,
+      "loss": 0.3208,
+      "step": 4410
+    },
+    {
+      "epoch": 0.27107915437561453,
+      "grad_norm": 4.84375,
+      "learning_rate": 1.7793379649314743e-05,
+      "loss": 0.2531,
+      "step": 4411
+    },
+    {
+      "epoch": 0.27114060963618486,
+      "grad_norm": 43.75,
+      "learning_rate": 1.7791921265023906e-05,
+      "loss": 0.3843,
+      "step": 4412
+    },
+    {
+      "epoch": 0.27120206489675514,
+      "grad_norm": 22.25,
+      "learning_rate": 1.7790462458767747e-05,
+      "loss": 0.2271,
+      "step": 4413
+    },
+    {
+      "epoch": 0.27126352015732547,
+      "grad_norm": 4.59375,
+      "learning_rate": 1.7789003230625266e-05,
+      "loss": 0.2941,
+      "step": 4414
+    },
+    {
+      "epoch": 0.2713249754178958,
+      "grad_norm": 5.375,
+      "learning_rate": 1.7787543580675485e-05,
+      "loss": 0.1836,
+      "step": 4415
+    },
+    {
+      "epoch": 0.2713864306784661,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7786083508997452e-05,
+      "loss": 0.2255,
+      "step": 4416
+    },
+    {
+      "epoch": 0.2714478859390364,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.7784623015670237e-05,
+      "loss": 0.2567,
+      "step": 4417
+    },
+    {
+      "epoch": 0.2715093411996067,
+      "grad_norm": 27.5,
+      "learning_rate": 1.7783162100772928e-05,
+      "loss": 0.342,
+      "step": 4418
+    },
+    {
+      "epoch": 0.271570796460177,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.7781700764384642e-05,
+      "loss": 0.2613,
+      "step": 4419
+    },
+    {
+      "epoch": 0.2716322517207473,
+      "grad_norm": 6.25,
+      "learning_rate": 1.7780239006584515e-05,
+      "loss": 0.329,
+      "step": 4420
+    },
+    {
+      "epoch": 0.2716937069813176,
+      "grad_norm": 44.0,
+      "learning_rate": 1.7778776827451715e-05,
+      "loss": 0.4563,
+      "step": 4421
+    },
+    {
+      "epoch": 0.2717551622418879,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.777731422706541e-05,
+      "loss": 0.275,
+      "step": 4422
+    },
+    {
+      "epoch": 0.2718166175024582,
+      "grad_norm": 4.5625,
+      "learning_rate": 1.7775851205504823e-05,
+      "loss": 0.2362,
+      "step": 4423
+    },
+    {
+      "epoch": 0.2718780727630285,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.777438776284917e-05,
+      "loss": 0.2526,
+      "step": 4424
+    },
+    {
+      "epoch": 0.27193952802359883,
+      "grad_norm": 21.75,
+      "learning_rate": 1.777292389917771e-05,
+      "loss": 0.4178,
+      "step": 4425
+    },
+    {
+      "epoch": 0.2720009832841691,
+      "grad_norm": 34.5,
+      "learning_rate": 1.777145961456971e-05,
+      "loss": 0.4348,
+      "step": 4426
+    },
+    {
+      "epoch": 0.27206243854473944,
+      "grad_norm": 10.6875,
+      "learning_rate": 1.7769994909104475e-05,
+      "loss": 0.5546,
+      "step": 4427
+    },
+    {
+      "epoch": 0.2721238938053097,
+      "grad_norm": 13.25,
+      "learning_rate": 1.776852978286132e-05,
+      "loss": 0.3859,
+      "step": 4428
+    },
+    {
+      "epoch": 0.27218534906588004,
+      "grad_norm": 19.0,
+      "learning_rate": 1.7767064235919594e-05,
+      "loss": 0.5555,
+      "step": 4429
+    },
+    {
+      "epoch": 0.2722468043264503,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.776559826835865e-05,
+      "loss": 0.3213,
+      "step": 4430
+    },
+    {
+      "epoch": 0.27230825958702065,
+      "grad_norm": 30.75,
+      "learning_rate": 1.7764131880257892e-05,
+      "loss": 0.5271,
+      "step": 4431
+    },
+    {
+      "epoch": 0.2723697148475909,
+      "grad_norm": 26.625,
+      "learning_rate": 1.776266507169672e-05,
+      "loss": 0.4243,
+      "step": 4432
+    },
+    {
+      "epoch": 0.27243117010816126,
+      "grad_norm": 23.125,
+      "learning_rate": 1.776119784275457e-05,
+      "loss": 0.4466,
+      "step": 4433
+    },
+    {
+      "epoch": 0.2724926253687316,
+      "grad_norm": 21.75,
+      "learning_rate": 1.7759730193510905e-05,
+      "loss": 0.4243,
+      "step": 4434
+    },
+    {
+      "epoch": 0.27255408062930186,
+      "grad_norm": 18.875,
+      "learning_rate": 1.7758262124045195e-05,
+      "loss": 0.3668,
+      "step": 4435
+    },
+    {
+      "epoch": 0.2726155358898722,
+      "grad_norm": 23.625,
+      "learning_rate": 1.7756793634436947e-05,
+      "loss": 0.3364,
+      "step": 4436
+    },
+    {
+      "epoch": 0.27267699115044247,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.7755324724765688e-05,
+      "loss": 0.4631,
+      "step": 4437
+    },
+    {
+      "epoch": 0.2727384464110128,
+      "grad_norm": 25.875,
+      "learning_rate": 1.775385539511096e-05,
+      "loss": 0.3267,
+      "step": 4438
+    },
+    {
+      "epoch": 0.2727999016715831,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.7752385645552336e-05,
+      "loss": 0.2511,
+      "step": 4439
+    },
+    {
+      "epoch": 0.2728613569321534,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.775091547616941e-05,
+      "loss": 0.3514,
+      "step": 4440
+    },
+    {
+      "epoch": 0.2729228121927237,
+      "grad_norm": 21.75,
+      "learning_rate": 1.7749444887041797e-05,
+      "loss": 0.3792,
+      "step": 4441
+    },
+    {
+      "epoch": 0.272984267453294,
+      "grad_norm": 19.25,
+      "learning_rate": 1.7747973878249137e-05,
+      "loss": 0.2857,
+      "step": 4442
+    },
+    {
+      "epoch": 0.2730457227138643,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.774650244987109e-05,
+      "loss": 0.387,
+      "step": 4443
+    },
+    {
+      "epoch": 0.2731071779744346,
+      "grad_norm": 23.125,
+      "learning_rate": 1.7745030601987338e-05,
+      "loss": 0.5915,
+      "step": 4444
+    },
+    {
+      "epoch": 0.2731686332350049,
+      "grad_norm": 6.0625,
+      "learning_rate": 1.7743558334677594e-05,
+      "loss": 0.2619,
+      "step": 4445
+    },
+    {
+      "epoch": 0.27323008849557523,
+      "grad_norm": 21.0,
+      "learning_rate": 1.774208564802158e-05,
+      "loss": 0.3962,
+      "step": 4446
+    },
+    {
+      "epoch": 0.2732915437561455,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.7740612542099054e-05,
+      "loss": 0.3811,
+      "step": 4447
+    },
+    {
+      "epoch": 0.27335299901671584,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.7739139016989788e-05,
+      "loss": 0.3293,
+      "step": 4448
+    },
+    {
+      "epoch": 0.2734144542772861,
+      "grad_norm": 18.0,
+      "learning_rate": 1.773766507277358e-05,
+      "loss": 0.6534,
+      "step": 4449
+    },
+    {
+      "epoch": 0.27347590953785644,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.773619070953025e-05,
+      "loss": 0.354,
+      "step": 4450
+    },
+    {
+      "epoch": 0.2735373647984267,
+      "grad_norm": 7.75,
+      "learning_rate": 1.7734715927339642e-05,
+      "loss": 0.4767,
+      "step": 4451
+    },
+    {
+      "epoch": 0.27359882005899705,
+      "grad_norm": 5.46875,
+      "learning_rate": 1.773324072628162e-05,
+      "loss": 0.5079,
+      "step": 4452
+    },
+    {
+      "epoch": 0.2736602753195674,
+      "grad_norm": 16.5,
+      "learning_rate": 1.7731765106436073e-05,
+      "loss": 0.2627,
+      "step": 4453
+    },
+    {
+      "epoch": 0.27372173058013766,
+      "grad_norm": 11.625,
+      "learning_rate": 1.7730289067882913e-05,
+      "loss": 0.2655,
+      "step": 4454
+    },
+    {
+      "epoch": 0.273783185840708,
+      "grad_norm": 32.5,
+      "learning_rate": 1.7728812610702076e-05,
+      "loss": 0.2862,
+      "step": 4455
+    },
+    {
+      "epoch": 0.27384464110127826,
+      "grad_norm": 34.0,
+      "learning_rate": 1.7727335734973512e-05,
+      "loss": 0.3525,
+      "step": 4456
+    },
+    {
+      "epoch": 0.2739060963618486,
+      "grad_norm": 13.125,
+      "learning_rate": 1.7725858440777204e-05,
+      "loss": 0.3135,
+      "step": 4457
+    },
+    {
+      "epoch": 0.27396755162241887,
+      "grad_norm": 13.875,
+      "learning_rate": 1.7724380728193152e-05,
+      "loss": 0.39,
+      "step": 4458
+    },
+    {
+      "epoch": 0.2740290068829892,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.7722902597301385e-05,
+      "loss": 0.2811,
+      "step": 4459
+    },
+    {
+      "epoch": 0.2740904621435595,
+      "grad_norm": 6.09375,
+      "learning_rate": 1.7721424048181943e-05,
+      "loss": 0.2941,
+      "step": 4460
+    },
+    {
+      "epoch": 0.2741519174041298,
+      "grad_norm": 4.53125,
+      "learning_rate": 1.7719945080914902e-05,
+      "loss": 0.3535,
+      "step": 4461
+    },
+    {
+      "epoch": 0.2742133726647001,
+      "grad_norm": 16.125,
+      "learning_rate": 1.771846569558035e-05,
+      "loss": 0.3268,
+      "step": 4462
+    },
+    {
+      "epoch": 0.2742748279252704,
+      "grad_norm": 34.0,
+      "learning_rate": 1.7716985892258404e-05,
+      "loss": 0.3882,
+      "step": 4463
+    },
+    {
+      "epoch": 0.2743362831858407,
+      "grad_norm": 18.125,
+      "learning_rate": 1.77155056710292e-05,
+      "loss": 0.3461,
+      "step": 4464
+    },
+    {
+      "epoch": 0.274397738446411,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.7714025031972904e-05,
+      "loss": 0.2612,
+      "step": 4465
+    },
+    {
+      "epoch": 0.2744591937069813,
+      "grad_norm": 15.0,
+      "learning_rate": 1.7712543975169687e-05,
+      "loss": 0.2753,
+      "step": 4466
+    },
+    {
+      "epoch": 0.27452064896755163,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.7711062500699766e-05,
+      "loss": 0.2669,
+      "step": 4467
+    },
+    {
+      "epoch": 0.2745821042281219,
+      "grad_norm": 6.40625,
+      "learning_rate": 1.7709580608643364e-05,
+      "loss": 0.3326,
+      "step": 4468
+    },
+    {
+      "epoch": 0.27464355948869223,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.7708098299080732e-05,
+      "loss": 0.3,
+      "step": 4469
+    },
+    {
+      "epoch": 0.2747050147492625,
+      "grad_norm": 5.21875,
+      "learning_rate": 1.7706615572092146e-05,
+      "loss": 0.3731,
+      "step": 4470
+    },
+    {
+      "epoch": 0.27476647000983284,
+      "grad_norm": 6.15625,
+      "learning_rate": 1.7705132427757895e-05,
+      "loss": 0.3559,
+      "step": 4471
+    },
+    {
+      "epoch": 0.2748279252704032,
+      "grad_norm": 39.0,
+      "learning_rate": 1.7703648866158303e-05,
+      "loss": 0.31,
+      "step": 4472
+    },
+    {
+      "epoch": 0.27488938053097345,
+      "grad_norm": 13.25,
+      "learning_rate": 1.770216488737371e-05,
+      "loss": 0.3331,
+      "step": 4473
+    },
+    {
+      "epoch": 0.2749508357915438,
+      "grad_norm": 30.0,
+      "learning_rate": 1.770068049148448e-05,
+      "loss": 0.4853,
+      "step": 4474
+    },
+    {
+      "epoch": 0.27501229105211406,
+      "grad_norm": 35.25,
+      "learning_rate": 1.7699195678571e-05,
+      "loss": 0.3477,
+      "step": 4475
+    },
+    {
+      "epoch": 0.2750737463126844,
+      "grad_norm": 6.59375,
+      "learning_rate": 1.769771044871368e-05,
+      "loss": 0.2995,
+      "step": 4476
+    },
+    {
+      "epoch": 0.27513520157325466,
+      "grad_norm": 36.25,
+      "learning_rate": 1.7696224801992947e-05,
+      "loss": 0.3232,
+      "step": 4477
+    },
+    {
+      "epoch": 0.275196656833825,
+      "grad_norm": 19.0,
+      "learning_rate": 1.7694738738489255e-05,
+      "loss": 0.2853,
+      "step": 4478
+    },
+    {
+      "epoch": 0.27525811209439527,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.7693252258283088e-05,
+      "loss": 0.3147,
+      "step": 4479
+    },
+    {
+      "epoch": 0.2753195673549656,
+      "grad_norm": 20.5,
+      "learning_rate": 1.769176536145494e-05,
+      "loss": 0.2401,
+      "step": 4480
+    },
+    {
+      "epoch": 0.2753810226155359,
+      "grad_norm": 15.375,
+      "learning_rate": 1.7690278048085327e-05,
+      "loss": 0.2675,
+      "step": 4481
+    },
+    {
+      "epoch": 0.2754424778761062,
+      "grad_norm": 27.875,
+      "learning_rate": 1.7688790318254803e-05,
+      "loss": 0.2239,
+      "step": 4482
+    },
+    {
+      "epoch": 0.2755039331366765,
+      "grad_norm": 32.0,
+      "learning_rate": 1.7687302172043933e-05,
+      "loss": 0.2535,
+      "step": 4483
+    },
+    {
+      "epoch": 0.2755653883972468,
+      "grad_norm": 10.625,
+      "learning_rate": 1.7685813609533303e-05,
+      "loss": 0.2297,
+      "step": 4484
+    },
+    {
+      "epoch": 0.2756268436578171,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.7684324630803523e-05,
+      "loss": 0.2774,
+      "step": 4485
+    },
+    {
+      "epoch": 0.2756882989183874,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.7682835235935236e-05,
+      "loss": 0.3726,
+      "step": 4486
+    },
+    {
+      "epoch": 0.2757497541789577,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.768134542500909e-05,
+      "loss": 0.3561,
+      "step": 4487
+    },
+    {
+      "epoch": 0.275811209439528,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.767985519810577e-05,
+      "loss": 0.17,
+      "step": 4488
+    },
+    {
+      "epoch": 0.2758726647000983,
+      "grad_norm": 10.6875,
+      "learning_rate": 1.767836455530598e-05,
+      "loss": 0.332,
+      "step": 4489
+    },
+    {
+      "epoch": 0.27593411996066863,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.7676873496690433e-05,
+      "loss": 0.4322,
+      "step": 4490
+    },
+    {
+      "epoch": 0.27599557522123896,
+      "grad_norm": 21.375,
+      "learning_rate": 1.767538202233989e-05,
+      "loss": 0.3272,
+      "step": 4491
+    },
+    {
+      "epoch": 0.27605703048180924,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.767389013233511e-05,
+      "loss": 0.539,
+      "step": 4492
+    },
+    {
+      "epoch": 0.27611848574237957,
+      "grad_norm": 12.875,
+      "learning_rate": 1.767239782675689e-05,
+      "loss": 0.3205,
+      "step": 4493
+    },
+    {
+      "epoch": 0.27617994100294985,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.767090510568605e-05,
+      "loss": 0.2875,
+      "step": 4494
+    },
+    {
+      "epoch": 0.2762413962635202,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.7669411969203417e-05,
+      "loss": 0.28,
+      "step": 4495
+    },
+    {
+      "epoch": 0.27630285152409045,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.7667918417389857e-05,
+      "loss": 0.234,
+      "step": 4496
+    },
+    {
+      "epoch": 0.2763643067846608,
+      "grad_norm": 32.5,
+      "learning_rate": 1.766642445032625e-05,
+      "loss": 0.5171,
+      "step": 4497
+    },
+    {
+      "epoch": 0.27642576204523106,
+      "grad_norm": 7.875,
+      "learning_rate": 1.76649300680935e-05,
+      "loss": 0.4516,
+      "step": 4498
+    },
+    {
+      "epoch": 0.2764872173058014,
+      "grad_norm": 29.0,
+      "learning_rate": 1.7663435270772537e-05,
+      "loss": 0.5245,
+      "step": 4499
+    },
+    {
+      "epoch": 0.27654867256637167,
+      "grad_norm": 53.5,
+      "learning_rate": 1.7661940058444305e-05,
+      "loss": 0.7653,
+      "step": 4500
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 15000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}