diff --git "a/checkpoint-3250/trainer_state.json" "b/checkpoint-3250/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-3250/trainer_state.json"
@@ -0,0 +1,22783 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.0325,
+  "eval_steps": 500,
+  "global_step": 3250,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1e-05,
+      "grad_norm": 1.4574347149106512,
+      "learning_rate": 3e-06,
+      "loss": 10.8576,
+      "step": 1
+    },
+    {
+      "epoch": 2e-05,
+      "grad_norm": 1.4482443114713344,
+      "learning_rate": 6e-06,
+      "loss": 10.8575,
+      "step": 2
+    },
+    {
+      "epoch": 3e-05,
+      "grad_norm": 1.4642003089092823,
+      "learning_rate": 9e-06,
+      "loss": 10.8576,
+      "step": 3
+    },
+    {
+      "epoch": 4e-05,
+      "grad_norm": 1.4485584072048276,
+      "learning_rate": 1.2e-05,
+      "loss": 10.8569,
+      "step": 4
+    },
+    {
+      "epoch": 5e-05,
+      "grad_norm": 1.4574321657516995,
+      "learning_rate": 1.5e-05,
+      "loss": 10.8527,
+      "step": 5
+    },
+    {
+      "epoch": 6e-05,
+      "grad_norm": 1.4585537094370684,
+      "learning_rate": 1.8e-05,
+      "loss": 10.8518,
+      "step": 6
+    },
+    {
+      "epoch": 7e-05,
+      "grad_norm": 1.4183780170798466,
+      "learning_rate": 2.1000000000000002e-05,
+      "loss": 10.8383,
+      "step": 7
+    },
+    {
+      "epoch": 8e-05,
+      "grad_norm": 1.286272643239374,
+      "learning_rate": 2.4e-05,
+      "loss": 10.8119,
+      "step": 8
+    },
+    {
+      "epoch": 9e-05,
+      "grad_norm": 1.246364249616181,
+      "learning_rate": 2.7e-05,
+      "loss": 10.8063,
+      "step": 9
+    },
+    {
+      "epoch": 0.0001,
+      "grad_norm": 1.2300728857390288,
+      "learning_rate": 3e-05,
+      "loss": 10.7913,
+      "step": 10
+    },
+    {
+      "epoch": 0.00011,
+      "grad_norm": 1.1639358472437353,
+      "learning_rate": 3.2999999999999996e-05,
+      "loss": 10.7756,
+      "step": 11
+    },
+    {
+      "epoch": 0.00012,
+      "grad_norm": 1.1455069336623074,
+      "learning_rate": 3.6e-05,
+      "loss": 10.7622,
+      "step": 12
+    },
+    {
+      "epoch": 0.00013,
+      "grad_norm": 1.104711768149659,
+      "learning_rate": 3.9e-05,
+      "loss": 10.7415,
+      "step": 13
+    },
+    {
+      "epoch": 0.00014,
+      "grad_norm": 1.0916409179063882,
+      "learning_rate": 4.2000000000000004e-05,
+      "loss": 10.7296,
+      "step": 14
+    },
+    {
+      "epoch": 0.00015,
+      "grad_norm": 1.0723152562842644,
+      "learning_rate": 4.4999999999999996e-05,
+      "loss": 10.7183,
+      "step": 15
+    },
+    {
+      "epoch": 0.00016,
+      "grad_norm": 1.0297043497697835,
+      "learning_rate": 4.8e-05,
+      "loss": 10.7,
+      "step": 16
+    },
+    {
+      "epoch": 0.00017,
+      "grad_norm": 0.9882210863590011,
+      "learning_rate": 5.1000000000000006e-05,
+      "loss": 10.6838,
+      "step": 17
+    },
+    {
+      "epoch": 0.00018,
+      "grad_norm": 0.966791538351231,
+      "learning_rate": 5.4e-05,
+      "loss": 10.6659,
+      "step": 18
+    },
+    {
+      "epoch": 0.00019,
+      "grad_norm": 0.9358769725124266,
+      "learning_rate": 5.7e-05,
+      "loss": 10.6509,
+      "step": 19
+    },
+    {
+      "epoch": 0.0002,
+      "grad_norm": 0.9313245306892226,
+      "learning_rate": 6e-05,
+      "loss": 10.638,
+      "step": 20
+    },
+    {
+      "epoch": 0.00021,
+      "grad_norm": 0.9056933112957348,
+      "learning_rate": 6.3e-05,
+      "loss": 10.6235,
+      "step": 21
+    },
+    {
+      "epoch": 0.00022,
+      "grad_norm": 0.9013277325687112,
+      "learning_rate": 6.599999999999999e-05,
+      "loss": 10.6069,
+      "step": 22
+    },
+    {
+      "epoch": 0.00023,
+      "grad_norm": 0.8982860560109955,
+      "learning_rate": 6.9e-05,
+      "loss": 10.5935,
+      "step": 23
+    },
+    {
+      "epoch": 0.00024,
+      "grad_norm": 0.8941575476865218,
+      "learning_rate": 7.2e-05,
+      "loss": 10.5802,
+      "step": 24
+    },
+    {
+      "epoch": 0.00025,
+      "grad_norm": 0.8955811690787305,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 10.5652,
+      "step": 25
+    },
+    {
+      "epoch": 0.00026,
+      "grad_norm": 0.9005417636666915,
+      "learning_rate": 7.8e-05,
+      "loss": 10.5507,
+      "step": 26
+    },
+    {
+      "epoch": 0.00027,
+      "grad_norm": 0.8929394290279239,
+      "learning_rate": 8.1e-05,
+      "loss": 10.5379,
+      "step": 27
+    },
+    {
+      "epoch": 0.00028,
+      "grad_norm": 0.892893092343029,
+      "learning_rate": 8.400000000000001e-05,
+      "loss": 10.5234,
+      "step": 28
+    },
+    {
+      "epoch": 0.00029,
+      "grad_norm": 0.8951984198637418,
+      "learning_rate": 8.7e-05,
+      "loss": 10.5078,
+      "step": 29
+    },
+    {
+      "epoch": 0.0003,
+      "grad_norm": 0.901037080772758,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 10.4913,
+      "step": 30
+    },
+    {
+      "epoch": 0.00031,
+      "grad_norm": 0.899757879512845,
+      "learning_rate": 9.3e-05,
+      "loss": 10.4759,
+      "step": 31
+    },
+    {
+      "epoch": 0.00032,
+      "grad_norm": 0.8963415000423225,
+      "learning_rate": 9.6e-05,
+      "loss": 10.4612,
+      "step": 32
+    },
+    {
+      "epoch": 0.00033,
+      "grad_norm": 0.8971468155029705,
+      "learning_rate": 9.900000000000001e-05,
+      "loss": 10.4429,
+      "step": 33
+    },
+    {
+      "epoch": 0.00034,
+      "grad_norm": 0.8964368558533553,
+      "learning_rate": 0.00010200000000000001,
+      "loss": 10.426,
+      "step": 34
+    },
+    {
+      "epoch": 0.00035,
+      "grad_norm": 0.9007340868108898,
+      "learning_rate": 0.00010500000000000002,
+      "loss": 10.4083,
+      "step": 35
+    },
+    {
+      "epoch": 0.00036,
+      "grad_norm": 0.8977475222905443,
+      "learning_rate": 0.000108,
+      "loss": 10.3895,
+      "step": 36
+    },
+    {
+      "epoch": 0.00037,
+      "grad_norm": 0.8929896912849768,
+      "learning_rate": 0.000111,
+      "loss": 10.3721,
+      "step": 37
+    },
+    {
+      "epoch": 0.00038,
+      "grad_norm": 0.8925204062685723,
+      "learning_rate": 0.000114,
+      "loss": 10.3515,
+      "step": 38
+    },
+    {
+      "epoch": 0.00039,
+      "grad_norm": 0.8947925451707294,
+      "learning_rate": 0.000117,
+      "loss": 10.3314,
+      "step": 39
+    },
+    {
+      "epoch": 0.0004,
+      "grad_norm": 0.8990961452390619,
+      "learning_rate": 0.00012,
+      "loss": 10.3088,
+      "step": 40
+    },
+    {
+      "epoch": 0.00041,
+      "grad_norm": 0.8951984584897338,
+      "learning_rate": 0.000123,
+      "loss": 10.2891,
+      "step": 41
+    },
+    {
+      "epoch": 0.00042,
+      "grad_norm": 0.8942493459254965,
+      "learning_rate": 0.000126,
+      "loss": 10.2679,
+      "step": 42
+    },
+    {
+      "epoch": 0.00043,
+      "grad_norm": 0.8983684110980745,
+      "learning_rate": 0.000129,
+      "loss": 10.243,
+      "step": 43
+    },
+    {
+      "epoch": 0.00044,
+      "grad_norm": 0.8985869855625094,
+      "learning_rate": 0.00013199999999999998,
+      "loss": 10.2206,
+      "step": 44
+    },
+    {
+      "epoch": 0.00045,
+      "grad_norm": 0.899962207774676,
+      "learning_rate": 0.000135,
+      "loss": 10.1962,
+      "step": 45
+    },
+    {
+      "epoch": 0.00046,
+      "grad_norm": 0.8914153211826606,
+      "learning_rate": 0.000138,
+      "loss": 10.1735,
+      "step": 46
+    },
+    {
+      "epoch": 0.00047,
+      "grad_norm": 0.8980761529388176,
+      "learning_rate": 0.000141,
+      "loss": 10.146,
+      "step": 47
+    },
+    {
+      "epoch": 0.00048,
+      "grad_norm": 0.9007540011988344,
+      "learning_rate": 0.000144,
+      "loss": 10.1222,
+      "step": 48
+    },
+    {
+      "epoch": 0.00049,
+      "grad_norm": 0.896344611975273,
+      "learning_rate": 0.000147,
+      "loss": 10.0974,
+      "step": 49
+    },
+    {
+      "epoch": 0.0005,
+      "grad_norm": 0.9015948414141589,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 10.0706,
+      "step": 50
+    },
+    {
+      "epoch": 0.00051,
+      "grad_norm": 0.9016102183936469,
+      "learning_rate": 0.000153,
+      "loss": 10.0433,
+      "step": 51
+    },
+    {
+      "epoch": 0.00052,
+      "grad_norm": 0.8993585803645682,
+      "learning_rate": 0.000156,
+      "loss": 10.0158,
+      "step": 52
+    },
+    {
+      "epoch": 0.00053,
+      "grad_norm": 0.8901736410301992,
+      "learning_rate": 0.000159,
+      "loss": 9.9917,
+      "step": 53
+    },
+    {
+      "epoch": 0.00054,
+      "grad_norm": 0.9083328832659304,
+      "learning_rate": 0.000162,
+      "loss": 9.9578,
+      "step": 54
+    },
+    {
+      "epoch": 0.00055,
+      "grad_norm": 0.8975570632788401,
+      "learning_rate": 0.000165,
+      "loss": 9.9337,
+      "step": 55
+    },
+    {
+      "epoch": 0.00056,
+      "grad_norm": 0.8924582131156604,
+      "learning_rate": 0.00016800000000000002,
+      "loss": 9.904,
+      "step": 56
+    },
+    {
+      "epoch": 0.00057,
+      "grad_norm": 0.8937117244788118,
+      "learning_rate": 0.000171,
+      "loss": 9.8765,
+      "step": 57
+    },
+    {
+      "epoch": 0.00058,
+      "grad_norm": 0.8912667557674427,
+      "learning_rate": 0.000174,
+      "loss": 9.8514,
+      "step": 58
+    },
+    {
+      "epoch": 0.00059,
+      "grad_norm": 0.899445317105929,
+      "learning_rate": 0.000177,
+      "loss": 9.819,
+      "step": 59
+    },
+    {
+      "epoch": 0.0006,
+      "grad_norm": 0.8827247644946434,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 9.7927,
+      "step": 60
+    },
+    {
+      "epoch": 0.00061,
+      "grad_norm": 0.8856006114384688,
+      "learning_rate": 0.000183,
+      "loss": 9.7645,
+      "step": 61
+    },
+    {
+      "epoch": 0.00062,
+      "grad_norm": 0.8887861476206323,
+      "learning_rate": 0.000186,
+      "loss": 9.7329,
+      "step": 62
+    },
+    {
+      "epoch": 0.00063,
+      "grad_norm": 0.8870948257177573,
+      "learning_rate": 0.000189,
+      "loss": 9.7044,
+      "step": 63
+    },
+    {
+      "epoch": 0.00064,
+      "grad_norm": 0.8854465518366905,
+      "learning_rate": 0.000192,
+      "loss": 9.6741,
+      "step": 64
+    },
+    {
+      "epoch": 0.00065,
+      "grad_norm": 0.8913964558604793,
+      "learning_rate": 0.00019500000000000002,
+      "loss": 9.6417,
+      "step": 65
+    },
+    {
+      "epoch": 0.00066,
+      "grad_norm": 0.8860459843940278,
+      "learning_rate": 0.00019800000000000002,
+      "loss": 9.6181,
+      "step": 66
+    },
+    {
+      "epoch": 0.00067,
+      "grad_norm": 0.8855602986146246,
+      "learning_rate": 0.000201,
+      "loss": 9.5886,
+      "step": 67
+    },
+    {
+      "epoch": 0.00068,
+      "grad_norm": 0.89182613307124,
+      "learning_rate": 0.00020400000000000003,
+      "loss": 9.5554,
+      "step": 68
+    },
+    {
+      "epoch": 0.00069,
+      "grad_norm": 0.8788147221752699,
+      "learning_rate": 0.00020700000000000002,
+      "loss": 9.5276,
+      "step": 69
+    },
+    {
+      "epoch": 0.0007,
+      "grad_norm": 0.89013480384513,
+      "learning_rate": 0.00021000000000000004,
+      "loss": 9.4911,
+      "step": 70
+    },
+    {
+      "epoch": 0.00071,
+      "grad_norm": 0.8832608390392153,
+      "learning_rate": 0.00021299999999999997,
+      "loss": 9.4637,
+      "step": 71
+    },
+    {
+      "epoch": 0.00072,
+      "grad_norm": 0.8850551460752941,
+      "learning_rate": 0.000216,
+      "loss": 9.4371,
+      "step": 72
+    },
+    {
+      "epoch": 0.00073,
+      "grad_norm": 0.8854430272945265,
+      "learning_rate": 0.00021899999999999998,
+      "loss": 9.4029,
+      "step": 73
+    },
+    {
+      "epoch": 0.00074,
+      "grad_norm": 0.881388508551569,
+      "learning_rate": 0.000222,
+      "loss": 9.3766,
+      "step": 74
+    },
+    {
+      "epoch": 0.00075,
+      "grad_norm": 0.8804734413831777,
+      "learning_rate": 0.000225,
+      "loss": 9.3413,
+      "step": 75
+    },
+    {
+      "epoch": 0.00076,
+      "grad_norm": 0.891470306594604,
+      "learning_rate": 0.000228,
+      "loss": 9.3136,
+      "step": 76
+    },
+    {
+      "epoch": 0.00077,
+      "grad_norm": 0.8835806398314079,
+      "learning_rate": 0.000231,
+      "loss": 9.2843,
+      "step": 77
+    },
+    {
+      "epoch": 0.00078,
+      "grad_norm": 0.8778470434340718,
+      "learning_rate": 0.000234,
+      "loss": 9.2534,
+      "step": 78
+    },
+    {
+      "epoch": 0.00079,
+      "grad_norm": 0.8849142747864956,
+      "learning_rate": 0.00023700000000000001,
+      "loss": 9.2174,
+      "step": 79
+    },
+    {
+      "epoch": 0.0008,
+      "grad_norm": 0.8831859916332898,
+      "learning_rate": 0.00024,
+      "loss": 9.1835,
+      "step": 80
+    },
+    {
+      "epoch": 0.00081,
+      "grad_norm": 0.8883671155155716,
+      "learning_rate": 0.00024300000000000002,
+      "loss": 9.1516,
+      "step": 81
+    },
+    {
+      "epoch": 0.00082,
+      "grad_norm": 0.8919005998406658,
+      "learning_rate": 0.000246,
+      "loss": 9.1278,
+      "step": 82
+    },
+    {
+      "epoch": 0.00083,
+      "grad_norm": 0.8904109612127737,
+      "learning_rate": 0.00024900000000000004,
+      "loss": 9.0926,
+      "step": 83
+    },
+    {
+      "epoch": 0.00084,
+      "grad_norm": 0.8922919840670517,
+      "learning_rate": 0.000252,
+      "loss": 9.0627,
+      "step": 84
+    },
+    {
+      "epoch": 0.00085,
+      "grad_norm": 0.8851770394662277,
+      "learning_rate": 0.000255,
+      "loss": 9.0389,
+      "step": 85
+    },
+    {
+      "epoch": 0.00086,
+      "grad_norm": 0.892287423580011,
+      "learning_rate": 0.000258,
+      "loss": 9.0029,
+      "step": 86
+    },
+    {
+      "epoch": 0.00087,
+      "grad_norm": 0.8819038715566697,
+      "learning_rate": 0.000261,
+      "loss": 8.9819,
+      "step": 87
+    },
+    {
+      "epoch": 0.00088,
+      "grad_norm": 0.8825670273484024,
+      "learning_rate": 0.00026399999999999997,
+      "loss": 8.9452,
+      "step": 88
+    },
+    {
+      "epoch": 0.00089,
+      "grad_norm": 0.87734677010751,
+      "learning_rate": 0.000267,
+      "loss": 8.92,
+      "step": 89
+    },
+    {
+      "epoch": 0.0009,
+      "grad_norm": 0.8800982265590005,
+      "learning_rate": 0.00027,
+      "loss": 8.8935,
+      "step": 90
+    },
+    {
+      "epoch": 0.00091,
+      "grad_norm": 0.8751738333241339,
+      "learning_rate": 0.000273,
+      "loss": 8.8635,
+      "step": 91
+    },
+    {
+      "epoch": 0.00092,
+      "grad_norm": 0.8805112062017327,
+      "learning_rate": 0.000276,
+      "loss": 8.8313,
+      "step": 92
+    },
+    {
+      "epoch": 0.00093,
+      "grad_norm": 0.8782110873043902,
+      "learning_rate": 0.000279,
+      "loss": 8.8018,
+      "step": 93
+    },
+    {
+      "epoch": 0.00094,
+      "grad_norm": 0.8793675998346636,
+      "learning_rate": 0.000282,
+      "loss": 8.7769,
+      "step": 94
+    },
+    {
+      "epoch": 0.00095,
+      "grad_norm": 0.8739437022686155,
+      "learning_rate": 0.000285,
+      "loss": 8.7468,
+      "step": 95
+    },
+    {
+      "epoch": 0.00096,
+      "grad_norm": 0.8792093745975903,
+      "learning_rate": 0.000288,
+      "loss": 8.7137,
+      "step": 96
+    },
+    {
+      "epoch": 0.00097,
+      "grad_norm": 0.8733827683228169,
+      "learning_rate": 0.000291,
+      "loss": 8.6904,
+      "step": 97
+    },
+    {
+      "epoch": 0.00098,
+      "grad_norm": 0.8720585520116785,
+      "learning_rate": 0.000294,
+      "loss": 8.6531,
+      "step": 98
+    },
+    {
+      "epoch": 0.00099,
+      "grad_norm": 0.8706078195741852,
+      "learning_rate": 0.000297,
+      "loss": 8.6333,
+      "step": 99
+    },
+    {
+      "epoch": 0.001,
+      "grad_norm": 0.868963072522692,
+      "learning_rate": 0.00030000000000000003,
+      "loss": 8.6078,
+      "step": 100
+    },
+    {
+      "epoch": 0.00101,
+      "grad_norm": 0.8733428125714966,
+      "learning_rate": 0.00030300000000000005,
+      "loss": 8.5742,
+      "step": 101
+    },
+    {
+      "epoch": 0.00102,
+      "grad_norm": 0.8594222895452354,
+      "learning_rate": 0.000306,
+      "loss": 8.5555,
+      "step": 102
+    },
+    {
+      "epoch": 0.00103,
+      "grad_norm": 0.8663526773106932,
+      "learning_rate": 0.000309,
+      "loss": 8.5314,
+      "step": 103
+    },
+    {
+      "epoch": 0.00104,
+      "grad_norm": 0.8588890352637412,
+      "learning_rate": 0.000312,
+      "loss": 8.5113,
+      "step": 104
+    },
+    {
+      "epoch": 0.00105,
+      "grad_norm": 0.862875507176608,
+      "learning_rate": 0.000315,
+      "loss": 8.4848,
+      "step": 105
+    },
+    {
+      "epoch": 0.00106,
+      "grad_norm": 0.858707935395144,
+      "learning_rate": 0.000318,
+      "loss": 8.4562,
+      "step": 106
+    },
+    {
+      "epoch": 0.00107,
+      "grad_norm": 0.8477892944676333,
+      "learning_rate": 0.000321,
+      "loss": 8.4313,
+      "step": 107
+    },
+    {
+      "epoch": 0.00108,
+      "grad_norm": 0.8523289020593828,
+      "learning_rate": 0.000324,
+      "loss": 8.4028,
+      "step": 108
+    },
+    {
+      "epoch": 0.00109,
+      "grad_norm": 0.847730490341604,
+      "learning_rate": 0.000327,
+      "loss": 8.3837,
+      "step": 109
+    },
+    {
+      "epoch": 0.0011,
+      "grad_norm": 0.8528691839316361,
+      "learning_rate": 0.00033,
+      "loss": 8.3605,
+      "step": 110
+    },
+    {
+      "epoch": 0.00111,
+      "grad_norm": 0.8490568938398722,
+      "learning_rate": 0.000333,
+      "loss": 8.3302,
+      "step": 111
+    },
+    {
+      "epoch": 0.00112,
+      "grad_norm": 0.8687159568709311,
+      "learning_rate": 0.00033600000000000004,
+      "loss": 8.3067,
+      "step": 112
+    },
+    {
+      "epoch": 0.00113,
+      "grad_norm": 0.9514329048402193,
+      "learning_rate": 0.000339,
+      "loss": 8.2682,
+      "step": 113
+    },
+    {
+      "epoch": 0.00114,
+      "grad_norm": 1.0232734531890686,
+      "learning_rate": 0.000342,
+      "loss": 8.2639,
+      "step": 114
+    },
+    {
+      "epoch": 0.00115,
+      "grad_norm": 0.88782361437861,
+      "learning_rate": 0.00034500000000000004,
+      "loss": 8.2339,
+      "step": 115
+    },
+    {
+      "epoch": 0.00116,
+      "grad_norm": 0.8477580889036649,
+      "learning_rate": 0.000348,
+      "loss": 8.2115,
+      "step": 116
+    },
+    {
+      "epoch": 0.00117,
+      "grad_norm": 0.9025353672223668,
+      "learning_rate": 0.000351,
+      "loss": 8.182,
+      "step": 117
+    },
+    {
+      "epoch": 0.00118,
+      "grad_norm": 0.8194804043714383,
+      "learning_rate": 0.000354,
+      "loss": 8.1632,
+      "step": 118
+    },
+    {
+      "epoch": 0.00119,
+      "grad_norm": 0.8747187640572928,
+      "learning_rate": 0.000357,
+      "loss": 8.1526,
+      "step": 119
+    },
+    {
+      "epoch": 0.0012,
+      "grad_norm": 0.819302051650556,
+      "learning_rate": 0.00035999999999999997,
+      "loss": 8.125,
+      "step": 120
+    },
+    {
+      "epoch": 0.00121,
+      "grad_norm": 0.8357670289489353,
+      "learning_rate": 0.000363,
+      "loss": 8.0977,
+      "step": 121
+    },
+    {
+      "epoch": 0.00122,
+      "grad_norm": 0.8269622916799205,
+      "learning_rate": 0.000366,
+      "loss": 8.0806,
+      "step": 122
+    },
+    {
+      "epoch": 0.00123,
+      "grad_norm": 0.7923909888868983,
+      "learning_rate": 0.000369,
+      "loss": 8.0558,
+      "step": 123
+    },
+    {
+      "epoch": 0.00124,
+      "grad_norm": 0.8405159164509748,
+      "learning_rate": 0.000372,
+      "loss": 8.0354,
+      "step": 124
+    },
+    {
+      "epoch": 0.00125,
+      "grad_norm": 0.7957014508382675,
+      "learning_rate": 0.000375,
+      "loss": 8.0077,
+      "step": 125
+    },
+    {
+      "epoch": 0.00126,
+      "grad_norm": 0.8032762193906265,
+      "learning_rate": 0.000378,
+      "loss": 7.9954,
+      "step": 126
+    },
+    {
+      "epoch": 0.00127,
+      "grad_norm": 0.8534337354315669,
+      "learning_rate": 0.000381,
+      "loss": 7.9712,
+      "step": 127
+    },
+    {
+      "epoch": 0.00128,
+      "grad_norm": 0.7889902179927859,
+      "learning_rate": 0.000384,
+      "loss": 7.9489,
+      "step": 128
+    },
+    {
+      "epoch": 0.00129,
+      "grad_norm": 0.7667804338310586,
+      "learning_rate": 0.00038700000000000003,
+      "loss": 7.936,
+      "step": 129
+    },
+    {
+      "epoch": 0.0013,
+      "grad_norm": 0.9090077886672961,
+      "learning_rate": 0.00039000000000000005,
+      "loss": 7.9113,
+      "step": 130
+    },
+    {
+      "epoch": 0.00131,
+      "grad_norm": 1.1017726636538738,
+      "learning_rate": 0.000393,
+      "loss": 7.8813,
+      "step": 131
+    },
+    {
+      "epoch": 0.00132,
+      "grad_norm": 0.9363739073599702,
+      "learning_rate": 0.00039600000000000003,
+      "loss": 7.8715,
+      "step": 132
+    },
+    {
+      "epoch": 0.00133,
+      "grad_norm": 0.7265974712286192,
+      "learning_rate": 0.00039900000000000005,
+      "loss": 7.8436,
+      "step": 133
+    },
+    {
+      "epoch": 0.00134,
+      "grad_norm": 1.085264808454749,
+      "learning_rate": 0.000402,
+      "loss": 7.8358,
+      "step": 134
+    },
+    {
+      "epoch": 0.00135,
+      "grad_norm": 1.1024504993278263,
+      "learning_rate": 0.00040500000000000003,
+      "loss": 7.8094,
+      "step": 135
+    },
+    {
+      "epoch": 0.00136,
+      "grad_norm": 0.7406778288782102,
+      "learning_rate": 0.00040800000000000005,
+      "loss": 7.7862,
+      "step": 136
+    },
+    {
+      "epoch": 0.00137,
+      "grad_norm": 0.948337541042023,
+      "learning_rate": 0.000411,
+      "loss": 7.7688,
+      "step": 137
+    },
+    {
+      "epoch": 0.00138,
+      "grad_norm": 0.9084991981726261,
+      "learning_rate": 0.00041400000000000003,
+      "loss": 7.7417,
+      "step": 138
+    },
+    {
+      "epoch": 0.00139,
+      "grad_norm": 0.7074494223870315,
+      "learning_rate": 0.00041700000000000005,
+      "loss": 7.7227,
+      "step": 139
+    },
+    {
+      "epoch": 0.0014,
+      "grad_norm": 0.7974190203181878,
+      "learning_rate": 0.00042000000000000007,
+      "loss": 7.7086,
+      "step": 140
+    },
+    {
+      "epoch": 0.00141,
+      "grad_norm": 0.6861992297830735,
+      "learning_rate": 0.000423,
+      "loss": 7.6913,
+      "step": 141
+    },
+    {
+      "epoch": 0.00142,
+      "grad_norm": 0.7349980301116962,
+      "learning_rate": 0.00042599999999999995,
+      "loss": 7.677,
+      "step": 142
+    },
+    {
+      "epoch": 0.00143,
+      "grad_norm": 0.7684994731116016,
+      "learning_rate": 0.00042899999999999997,
+      "loss": 7.6484,
+      "step": 143
+    },
+    {
+      "epoch": 0.00144,
+      "grad_norm": 0.6719380878952201,
+      "learning_rate": 0.000432,
+      "loss": 7.629,
+      "step": 144
+    },
+    {
+      "epoch": 0.00145,
+      "grad_norm": 0.7001362890446781,
+      "learning_rate": 0.000435,
+      "loss": 7.6232,
+      "step": 145
+    },
+    {
+      "epoch": 0.00146,
+      "grad_norm": 0.7384188504614574,
+      "learning_rate": 0.00043799999999999997,
+      "loss": 7.5849,
+      "step": 146
+    },
+    {
+      "epoch": 0.00147,
+      "grad_norm": 0.6536906507985312,
+      "learning_rate": 0.000441,
+      "loss": 7.5755,
+      "step": 147
+    },
+    {
+      "epoch": 0.00148,
+      "grad_norm": 0.6570387947682329,
+      "learning_rate": 0.000444,
+      "loss": 7.5632,
+      "step": 148
+    },
+    {
+      "epoch": 0.00149,
+      "grad_norm": 0.7127206041028632,
+      "learning_rate": 0.00044699999999999997,
+      "loss": 7.5571,
+      "step": 149
+    },
+    {
+      "epoch": 0.0015,
+      "grad_norm": 0.7308289900379658,
+      "learning_rate": 0.00045,
+      "loss": 7.5253,
+      "step": 150
+    },
+    {
+      "epoch": 0.00151,
+      "grad_norm": 0.8664894863114535,
+      "learning_rate": 0.000453,
+      "loss": 7.505,
+      "step": 151
+    },
+    {
+      "epoch": 0.00152,
+      "grad_norm": 0.8895841422117233,
+      "learning_rate": 0.000456,
+      "loss": 7.5034,
+      "step": 152
+    },
+    {
+      "epoch": 0.00153,
+      "grad_norm": 0.662621907628313,
+      "learning_rate": 0.000459,
+      "loss": 7.4855,
+      "step": 153
+    },
+    {
+      "epoch": 0.00154,
+      "grad_norm": 0.7176449004695503,
+      "learning_rate": 0.000462,
+      "loss": 7.4655,
+      "step": 154
+    },
+    {
+      "epoch": 0.00155,
+      "grad_norm": 0.7407325792933065,
+      "learning_rate": 0.000465,
+      "loss": 7.4508,
+      "step": 155
+    },
+    {
+      "epoch": 0.00156,
+      "grad_norm": 0.7037717820949844,
+      "learning_rate": 0.000468,
+      "loss": 7.4265,
+      "step": 156
+    },
+    {
+      "epoch": 0.00157,
+      "grad_norm": 0.5688508632010355,
+      "learning_rate": 0.000471,
+      "loss": 7.4159,
+      "step": 157
+    },
+    {
+      "epoch": 0.00158,
+      "grad_norm": 0.812094320370736,
+      "learning_rate": 0.00047400000000000003,
+      "loss": 7.4015,
+      "step": 158
+    },
+    {
+      "epoch": 0.00159,
+      "grad_norm": 0.6851465337071334,
+      "learning_rate": 0.000477,
+      "loss": 7.3879,
+      "step": 159
+    },
+    {
+      "epoch": 0.0016,
+      "grad_norm": 0.9553203809874583,
+      "learning_rate": 0.00048,
+      "loss": 7.3739,
+      "step": 160
+    },
+    {
+      "epoch": 0.00161,
+      "grad_norm": 0.7328141993841643,
+      "learning_rate": 0.00048300000000000003,
+      "loss": 7.3497,
+      "step": 161
+    },
+    {
+      "epoch": 0.00162,
+      "grad_norm": 0.6781537584586433,
+      "learning_rate": 0.00048600000000000005,
+      "loss": 7.3535,
+      "step": 162
+    },
+    {
+      "epoch": 0.00163,
+      "grad_norm": 0.5807738383940755,
+      "learning_rate": 0.0004890000000000001,
+      "loss": 7.3306,
+      "step": 163
+    },
+    {
+      "epoch": 0.00164,
+      "grad_norm": 0.6466438270814104,
+      "learning_rate": 0.000492,
+      "loss": 7.3371,
+      "step": 164
+    },
+    {
+      "epoch": 0.00165,
+      "grad_norm": 0.4365334647721832,
+      "learning_rate": 0.000495,
+      "loss": 7.3038,
+      "step": 165
+    },
+    {
+      "epoch": 0.00166,
+      "grad_norm": 0.7387692677149859,
+      "learning_rate": 0.0004980000000000001,
+      "loss": 7.2849,
+      "step": 166
+    },
+    {
+      "epoch": 0.00167,
+      "grad_norm": 0.6489238814400348,
+      "learning_rate": 0.000501,
+      "loss": 7.2765,
+      "step": 167
+    },
+    {
+      "epoch": 0.00168,
+      "grad_norm": 0.4943076416069544,
+      "learning_rate": 0.000504,
+      "loss": 7.2512,
+      "step": 168
+    },
+    {
+      "epoch": 0.00169,
+      "grad_norm": 0.6623047646201438,
+      "learning_rate": 0.0005070000000000001,
+      "loss": 7.2485,
+      "step": 169
+    },
+    {
+      "epoch": 0.0017,
+      "grad_norm": 0.5841690806336316,
+      "learning_rate": 0.00051,
+      "loss": 7.2331,
+      "step": 170
+    },
+    {
+      "epoch": 0.00171,
+      "grad_norm": 0.455313434659774,
+      "learning_rate": 0.000513,
+      "loss": 7.2132,
+      "step": 171
+    },
+    {
+      "epoch": 0.00172,
+      "grad_norm": 0.6012704724036779,
+      "learning_rate": 0.000516,
+      "loss": 7.214,
+      "step": 172
+    },
+    {
+      "epoch": 0.00173,
+      "grad_norm": 0.5433283029678282,
+      "learning_rate": 0.0005189999999999999,
+      "loss": 7.1993,
+      "step": 173
+    },
+    {
+      "epoch": 0.00174,
+      "grad_norm": 0.38251135313099477,
+      "learning_rate": 0.000522,
+      "loss": 7.2121,
+      "step": 174
+    },
+    {
+      "epoch": 0.00175,
+      "grad_norm": 0.6483519243845758,
+      "learning_rate": 0.000525,
+      "loss": 7.1796,
+      "step": 175
+    },
+    {
+      "epoch": 0.00176,
+      "grad_norm": 0.5106681215905619,
+      "learning_rate": 0.0005279999999999999,
+      "loss": 7.1531,
+      "step": 176
+    },
+    {
+      "epoch": 0.00177,
+      "grad_norm": 0.3867365673342131,
+      "learning_rate": 0.000531,
+      "loss": 7.1566,
+      "step": 177
+    },
+    {
+      "epoch": 0.00178,
+      "grad_norm": 0.6803005272577738,
+      "learning_rate": 0.000534,
+      "loss": 7.1525,
+      "step": 178
+    },
+    {
+      "epoch": 0.00179,
+      "grad_norm": 0.44918531151657043,
+      "learning_rate": 0.000537,
+      "loss": 7.1407,
+      "step": 179
+    },
+    {
+      "epoch": 0.0018,
+      "grad_norm": 0.39944612038572996,
+      "learning_rate": 0.00054,
+      "loss": 7.1121,
+      "step": 180
+    },
+    {
+      "epoch": 0.00181,
+      "grad_norm": 0.5937704312371144,
+      "learning_rate": 0.000543,
+      "loss": 7.1037,
+      "step": 181
+    },
+    {
+      "epoch": 0.00182,
+      "grad_norm": 0.3956001398507633,
+      "learning_rate": 0.000546,
+      "loss": 7.0976,
+      "step": 182
+    },
+    {
+      "epoch": 0.00183,
+      "grad_norm": 0.36202282333499824,
+      "learning_rate": 0.000549,
+      "loss": 7.0864,
+      "step": 183
+    },
+    {
+      "epoch": 0.00184,
+      "grad_norm": 0.564760920006622,
+      "learning_rate": 0.000552,
+      "loss": 7.0785,
+      "step": 184
+    },
+    {
+      "epoch": 0.00185,
+      "grad_norm": 0.36333201524042624,
+      "learning_rate": 0.000555,
+      "loss": 7.0515,
+      "step": 185
+    },
+    {
+      "epoch": 0.00186,
+      "grad_norm": 0.3517854453126823,
+      "learning_rate": 0.000558,
+      "loss": 7.0614,
+      "step": 186
+    },
+    {
+      "epoch": 0.00187,
+      "grad_norm": 0.5413049623390372,
+      "learning_rate": 0.000561,
+      "loss": 7.0394,
+      "step": 187
+    },
+    {
+      "epoch": 0.00188,
+      "grad_norm": 0.4452202297805183,
+      "learning_rate": 0.000564,
+      "loss": 7.0168,
+      "step": 188
+    },
+    {
+      "epoch": 0.00189,
+      "grad_norm": 0.2888404630990515,
+      "learning_rate": 0.000567,
+      "loss": 7.0108,
+      "step": 189
+    },
+    {
+      "epoch": 0.0019,
+      "grad_norm": 0.5069984229642092,
+      "learning_rate": 0.00057,
+      "loss": 6.9967,
+      "step": 190
+    },
+    {
+      "epoch": 0.00191,
+      "grad_norm": 0.5049159939798347,
+      "learning_rate": 0.000573,
+      "loss": 6.9897,
+      "step": 191
+    },
+    {
+      "epoch": 0.00192,
+      "grad_norm": 0.353459237780404,
+      "learning_rate": 0.000576,
+      "loss": 6.9784,
+      "step": 192
+    },
+    {
+      "epoch": 0.00193,
+      "grad_norm": 0.4583730098712965,
+      "learning_rate": 0.000579,
+      "loss": 6.9798,
+      "step": 193
+    },
+    {
+      "epoch": 0.00194,
+      "grad_norm": 0.5934016188318005,
+      "learning_rate": 0.000582,
+      "loss": 6.9757,
+      "step": 194
+    },
+    {
+      "epoch": 0.00195,
+      "grad_norm": 0.6611510755360627,
+      "learning_rate": 0.000585,
+      "loss": 6.9556,
+      "step": 195
+    },
+    {
+      "epoch": 0.00196,
+      "grad_norm": 0.6794801564658052,
+      "learning_rate": 0.000588,
+      "loss": 6.9455,
+      "step": 196
+    },
+    {
+      "epoch": 0.00197,
+      "grad_norm": 0.6965025324131076,
+      "learning_rate": 0.000591,
+      "loss": 6.9424,
+      "step": 197
+    },
+    {
+      "epoch": 0.00198,
+      "grad_norm": 1.0010424094015926,
+      "learning_rate": 0.000594,
+      "loss": 6.9428,
+      "step": 198
+    },
+    {
+      "epoch": 0.00199,
+      "grad_norm": 0.8473959375492113,
+      "learning_rate": 0.0005970000000000001,
+      "loss": 6.9257,
+      "step": 199
+    },
+    {
+      "epoch": 0.002,
+      "grad_norm": 0.4674164352666973,
+      "learning_rate": 0.0006000000000000001,
+      "loss": 6.9135,
+      "step": 200
+    },
+    {
+      "epoch": 0.00201,
+      "grad_norm": 0.6130101305251902,
+      "learning_rate": 0.000603,
+      "loss": 6.8907,
+      "step": 201
+    },
+    {
+      "epoch": 0.00202,
+      "grad_norm": 0.5766204915712545,
+      "learning_rate": 0.0006060000000000001,
+      "loss": 6.8786,
+      "step": 202
+    },
+    {
+      "epoch": 0.00203,
+      "grad_norm": 0.42726729176690026,
+      "learning_rate": 0.0006090000000000001,
+      "loss": 6.8819,
+      "step": 203
+    },
+    {
+      "epoch": 0.00204,
+      "grad_norm": 0.5161441207128264,
+      "learning_rate": 0.000612,
+      "loss": 6.8675,
+      "step": 204
+    },
+    {
+      "epoch": 0.00205,
+      "grad_norm": 0.44005998133884483,
+      "learning_rate": 0.000615,
+      "loss": 6.852,
+      "step": 205
+    },
+    {
+      "epoch": 0.00206,
+      "grad_norm": 0.4214710856814876,
+      "learning_rate": 0.000618,
+      "loss": 6.8455,
+      "step": 206
+    },
+    {
+      "epoch": 0.00207,
+      "grad_norm": 0.3835330802523119,
+      "learning_rate": 0.000621,
+      "loss": 6.8358,
+      "step": 207
+    },
+    {
+      "epoch": 0.00208,
+      "grad_norm": 0.4118282878888594,
+      "learning_rate": 0.000624,
+      "loss": 6.8325,
+      "step": 208
+    },
+    {
+      "epoch": 0.00209,
+      "grad_norm": 0.35207765117623135,
+      "learning_rate": 0.000627,
+      "loss": 6.8108,
+      "step": 209
+    },
+    {
+      "epoch": 0.0021,
+      "grad_norm": 0.35165038308098223,
+      "learning_rate": 0.00063,
+      "loss": 6.8001,
+      "step": 210
+    },
+    {
+      "epoch": 0.00211,
+      "grad_norm": 0.2907998960376576,
+      "learning_rate": 0.000633,
+      "loss": 6.794,
+      "step": 211
+    },
+    {
+      "epoch": 0.00212,
+      "grad_norm": 0.34183459291083,
+      "learning_rate": 0.000636,
+      "loss": 6.7864,
+      "step": 212
+    },
+    {
+      "epoch": 0.00213,
+      "grad_norm": 0.31625023606106545,
+      "learning_rate": 0.000639,
+      "loss": 6.7664,
+      "step": 213
+    },
+    {
+      "epoch": 0.00214,
+      "grad_norm": 0.3389167462098675,
+      "learning_rate": 0.000642,
+      "loss": 6.7645,
+      "step": 214
+    },
+    {
+      "epoch": 0.00215,
+      "grad_norm": 0.4689783367804683,
+      "learning_rate": 0.000645,
+      "loss": 6.7617,
+      "step": 215
+    },
+    {
+      "epoch": 0.00216,
+      "grad_norm": 0.7554849005001304,
+      "learning_rate": 0.000648,
+      "loss": 6.763,
+      "step": 216
+    },
+    {
+      "epoch": 0.00217,
+      "grad_norm": 1.3852957930548755,
+      "learning_rate": 0.000651,
+      "loss": 6.7541,
+      "step": 217
+    },
+    {
+      "epoch": 0.00218,
+      "grad_norm": 0.6607509792225358,
+      "learning_rate": 0.000654,
+      "loss": 6.7405,
+      "step": 218
+    },
+    {
+      "epoch": 0.00219,
+      "grad_norm": 0.6480761266032634,
+      "learning_rate": 0.000657,
+      "loss": 6.7122,
+      "step": 219
+    },
+    {
+      "epoch": 0.0022,
+      "grad_norm": 1.2227203355321279,
+      "learning_rate": 0.00066,
+      "loss": 6.7471,
+      "step": 220
+    },
+    {
+      "epoch": 0.00221,
+      "grad_norm": 1.1390371544106608,
+      "learning_rate": 0.0006630000000000001,
+      "loss": 6.7248,
+      "step": 221
+    },
+    {
+      "epoch": 0.00222,
+      "grad_norm": 0.7354052143767025,
+      "learning_rate": 0.000666,
+      "loss": 6.7013,
+      "step": 222
+    },
+    {
+      "epoch": 0.00223,
+      "grad_norm": 0.6134943537271429,
+      "learning_rate": 0.000669,
+      "loss": 6.6937,
+      "step": 223
+    },
+    {
+      "epoch": 0.00224,
+      "grad_norm": 0.5502679770385809,
+      "learning_rate": 0.0006720000000000001,
+      "loss": 6.6934,
+      "step": 224
+    },
+    {
+      "epoch": 0.00225,
+      "grad_norm": 0.47738132510350817,
+      "learning_rate": 0.000675,
+      "loss": 6.6778,
+      "step": 225
+    },
+    {
+      "epoch": 0.00226,
+      "grad_norm": 0.507114065421453,
+      "learning_rate": 0.000678,
+      "loss": 6.6712,
+      "step": 226
+    },
+    {
+      "epoch": 0.00227,
+      "grad_norm": 0.4403172481189424,
+      "learning_rate": 0.0006810000000000001,
+      "loss": 6.646,
+      "step": 227
+    },
+    {
+      "epoch": 0.00228,
+      "grad_norm": 0.34083753361811386,
+      "learning_rate": 0.000684,
+      "loss": 6.6465,
+      "step": 228
+    },
+    {
+      "epoch": 0.00229,
+      "grad_norm": 0.3502366193988861,
+      "learning_rate": 0.000687,
+      "loss": 6.6418,
+      "step": 229
+    },
+    {
+      "epoch": 0.0023,
+      "grad_norm": 0.40867670062411404,
+      "learning_rate": 0.0006900000000000001,
+      "loss": 6.6218,
+      "step": 230
+    },
+    {
+      "epoch": 0.00231,
+      "grad_norm": 0.3512396474179769,
+      "learning_rate": 0.000693,
+      "loss": 6.605,
+      "step": 231
+    },
+    {
+      "epoch": 0.00232,
+      "grad_norm": 0.37061772246720376,
+      "learning_rate": 0.000696,
+      "loss": 6.6164,
+      "step": 232
+    },
+    {
+      "epoch": 0.00233,
+      "grad_norm": 0.30975845835339183,
+      "learning_rate": 0.0006990000000000001,
+      "loss": 6.6026,
+      "step": 233
+    },
+    {
+      "epoch": 0.00234,
+      "grad_norm": 0.31056463949410484,
+      "learning_rate": 0.000702,
+      "loss": 6.5934,
+      "step": 234
+    },
+    {
+      "epoch": 0.00235,
+      "grad_norm": 0.3416829938754332,
+      "learning_rate": 0.000705,
+      "loss": 6.5798,
+      "step": 235
+    },
+    {
+      "epoch": 0.00236,
+      "grad_norm": 0.3580449551775802,
+      "learning_rate": 0.000708,
+      "loss": 6.5618,
+      "step": 236
+    },
+    {
+      "epoch": 0.00237,
+      "grad_norm": 0.31186166958910727,
+      "learning_rate": 0.0007109999999999999,
+      "loss": 6.5709,
+      "step": 237
+    },
+    {
+      "epoch": 0.00238,
+      "grad_norm": 0.35712039334009765,
+      "learning_rate": 0.000714,
+      "loss": 6.5601,
+      "step": 238
+    },
+    {
+      "epoch": 0.00239,
+      "grad_norm": 0.4689874670209286,
+      "learning_rate": 0.000717,
+      "loss": 6.5431,
+      "step": 239
+    },
+    {
+      "epoch": 0.0024,
+      "grad_norm": 0.8385556457838154,
+      "learning_rate": 0.0007199999999999999,
+      "loss": 6.5652,
+      "step": 240
+    },
+    {
+      "epoch": 0.00241,
+      "grad_norm": 1.2597764356570753,
+      "learning_rate": 0.000723,
+      "loss": 6.5721,
+      "step": 241
+    },
+    {
+      "epoch": 0.00242,
+      "grad_norm": 0.7018547016449656,
+      "learning_rate": 0.000726,
+      "loss": 6.5258,
+      "step": 242
+    },
+    {
+      "epoch": 0.00243,
+      "grad_norm": 0.9679840331331718,
+      "learning_rate": 0.000729,
+      "loss": 6.5235,
+      "step": 243
+    },
+    {
+      "epoch": 0.00244,
+      "grad_norm": 0.8151244219176279,
+      "learning_rate": 0.000732,
+      "loss": 6.5289,
+      "step": 244
+    },
+    {
+      "epoch": 0.00245,
+      "grad_norm": 0.7408239035858821,
+      "learning_rate": 0.000735,
+      "loss": 6.5159,
+      "step": 245
+    },
+    {
+      "epoch": 0.00246,
+      "grad_norm": 0.7269888029650857,
+      "learning_rate": 0.000738,
+      "loss": 6.5006,
+      "step": 246
+    },
+    {
+      "epoch": 0.00247,
+      "grad_norm": 0.6715573633422388,
+      "learning_rate": 0.000741,
+      "loss": 6.4916,
+      "step": 247
+    },
+    {
+      "epoch": 0.00248,
+      "grad_norm": 0.5355440416613977,
+      "learning_rate": 0.000744,
+      "loss": 6.4965,
+      "step": 248
+    },
+    {
+      "epoch": 0.00249,
+      "grad_norm": 0.47837466185087324,
+      "learning_rate": 0.000747,
+      "loss": 6.474,
+      "step": 249
+    },
+    {
+      "epoch": 0.0025,
+      "grad_norm": 0.5421674013984196,
+      "learning_rate": 0.00075,
+      "loss": 6.4781,
+      "step": 250
+    },
+    {
+      "epoch": 0.00251,
+      "grad_norm": 0.5046285559759315,
+      "learning_rate": 0.000753,
+      "loss": 6.4394,
+      "step": 251
+    },
+    {
+      "epoch": 0.00252,
+      "grad_norm": 0.4902399046559828,
+      "learning_rate": 0.000756,
+      "loss": 6.4478,
+      "step": 252
+    },
+    {
+      "epoch": 0.00253,
+      "grad_norm": 0.6026763597870729,
+      "learning_rate": 0.000759,
+      "loss": 6.4417,
+      "step": 253
+    },
+    {
+      "epoch": 0.00254,
+      "grad_norm": 0.6534629043130656,
+      "learning_rate": 0.000762,
+      "loss": 6.4357,
+      "step": 254
+    },
+    {
+      "epoch": 0.00255,
+      "grad_norm": 0.5879568865224861,
+      "learning_rate": 0.0007650000000000001,
+      "loss": 6.4331,
+      "step": 255
+    },
+    {
+      "epoch": 0.00256,
+      "grad_norm": 0.587549512889128,
+      "learning_rate": 0.000768,
+      "loss": 6.4121,
+      "step": 256
+    },
+    {
+      "epoch": 0.00257,
+      "grad_norm": 0.4948417519702818,
+      "learning_rate": 0.000771,
+      "loss": 6.4144,
+      "step": 257
+    },
+    {
+      "epoch": 0.00258,
+      "grad_norm": 0.500640645286767,
+      "learning_rate": 0.0007740000000000001,
+      "loss": 6.3915,
+      "step": 258
+    },
+    {
+      "epoch": 0.00259,
+      "grad_norm": 0.46096026590523803,
+      "learning_rate": 0.000777,
+      "loss": 6.3904,
+      "step": 259
+    },
+    {
+      "epoch": 0.0026,
+      "grad_norm": 0.36710644781296853,
+      "learning_rate": 0.0007800000000000001,
+      "loss": 6.3738,
+      "step": 260
+    },
+    {
+      "epoch": 0.00261,
+      "grad_norm": 0.3228962429333065,
+      "learning_rate": 0.0007830000000000001,
+      "loss": 6.3839,
+      "step": 261
+    },
+    {
+      "epoch": 0.00262,
+      "grad_norm": 0.33723485694460476,
+      "learning_rate": 0.000786,
+      "loss": 6.3601,
+      "step": 262
+    },
+    {
+      "epoch": 0.00263,
+      "grad_norm": 0.3792844357509868,
+      "learning_rate": 0.0007890000000000001,
+      "loss": 6.3664,
+      "step": 263
+    },
+    {
+      "epoch": 0.00264,
+      "grad_norm": 0.41999467297545995,
+      "learning_rate": 0.0007920000000000001,
+      "loss": 6.3505,
+      "step": 264
+    },
+    {
+      "epoch": 0.00265,
+      "grad_norm": 0.5058327981670172,
+      "learning_rate": 0.000795,
+      "loss": 6.3427,
+      "step": 265
+    },
+    {
+      "epoch": 0.00266,
+      "grad_norm": 0.728666013298328,
+      "learning_rate": 0.0007980000000000001,
+      "loss": 6.3445,
+      "step": 266
+    },
+    {
+      "epoch": 0.00267,
+      "grad_norm": 1.051597167429085,
+      "learning_rate": 0.0008010000000000001,
+      "loss": 6.3504,
+      "step": 267
+    },
+    {
+      "epoch": 0.00268,
+      "grad_norm": 0.8184600443662068,
+      "learning_rate": 0.000804,
+      "loss": 6.3425,
+      "step": 268
+    },
+    {
+      "epoch": 0.00269,
+      "grad_norm": 0.9044150906220578,
+      "learning_rate": 0.0008070000000000001,
+      "loss": 6.3249,
+      "step": 269
+    },
+    {
+      "epoch": 0.0027,
+      "grad_norm": 1.128589514160359,
+      "learning_rate": 0.0008100000000000001,
+      "loss": 6.3418,
+      "step": 270
+    },
+    {
+      "epoch": 0.00271,
+      "grad_norm": 0.8269681270889044,
+      "learning_rate": 0.000813,
+      "loss": 6.3324,
+      "step": 271
+    },
+    {
+      "epoch": 0.00272,
+      "grad_norm": 0.9605488749490486,
+      "learning_rate": 0.0008160000000000001,
+      "loss": 6.3229,
+      "step": 272
+    },
+    {
+      "epoch": 0.00273,
+      "grad_norm": 1.4256959145221717,
+      "learning_rate": 0.0008190000000000001,
+      "loss": 6.3332,
+      "step": 273
+    },
+    {
+      "epoch": 0.00274,
+      "grad_norm": 0.8746481700595169,
+      "learning_rate": 0.000822,
+      "loss": 6.301,
+      "step": 274
+    },
+    {
+      "epoch": 0.00275,
+      "grad_norm": 0.7483839590071554,
+      "learning_rate": 0.0008250000000000001,
+      "loss": 6.3047,
+      "step": 275
+    },
+    {
+      "epoch": 0.00276,
+      "grad_norm": 0.9683696395132253,
+      "learning_rate": 0.0008280000000000001,
+      "loss": 6.2965,
+      "step": 276
+    },
+    {
+      "epoch": 0.00277,
+      "grad_norm": 0.9445364353770574,
+      "learning_rate": 0.0008310000000000001,
+      "loss": 6.3021,
+      "step": 277
+    },
+    {
+      "epoch": 0.00278,
+      "grad_norm": 0.5637393663463164,
+      "learning_rate": 0.0008340000000000001,
+      "loss": 6.2711,
+      "step": 278
+    },
+    {
+      "epoch": 0.00279,
+      "grad_norm": 0.6474632709262491,
+      "learning_rate": 0.0008370000000000001,
+      "loss": 6.2722,
+      "step": 279
+    },
+    {
+      "epoch": 0.0028,
+      "grad_norm": 0.5477535984985658,
+      "learning_rate": 0.0008400000000000001,
+      "loss": 6.2548,
+      "step": 280
+    },
+    {
+      "epoch": 0.00281,
+      "grad_norm": 0.5235139271800053,
+      "learning_rate": 0.0008430000000000001,
+      "loss": 6.2428,
+      "step": 281
+    },
+    {
+      "epoch": 0.00282,
+      "grad_norm": 0.4309641347762167,
+      "learning_rate": 0.000846,
+      "loss": 6.2382,
+      "step": 282
+    },
+    {
+      "epoch": 0.00283,
+      "grad_norm": 0.47598469419389666,
+      "learning_rate": 0.0008489999999999999,
+      "loss": 6.2321,
+      "step": 283
+    },
+    {
+      "epoch": 0.00284,
+      "grad_norm": 0.41791767912211497,
+      "learning_rate": 0.0008519999999999999,
+      "loss": 6.2199,
+      "step": 284
+    },
+    {
+      "epoch": 0.00285,
+      "grad_norm": 0.4068034669968041,
+      "learning_rate": 0.000855,
+      "loss": 6.2226,
+      "step": 285
+    },
+    {
+      "epoch": 0.00286,
+      "grad_norm": 0.461449105404332,
+      "learning_rate": 0.0008579999999999999,
+      "loss": 6.2004,
+      "step": 286
+    },
+    {
+      "epoch": 0.00287,
+      "grad_norm": 0.3833700707849029,
+      "learning_rate": 0.000861,
+      "loss": 6.2013,
+      "step": 287
+    },
+    {
+      "epoch": 0.00288,
+      "grad_norm": 0.3960645710165419,
+      "learning_rate": 0.000864,
+      "loss": 6.1818,
+      "step": 288
+    },
+    {
+      "epoch": 0.00289,
+      "grad_norm": 0.44431212101949524,
+      "learning_rate": 0.0008669999999999999,
+      "loss": 6.1961,
+      "step": 289
+    },
+    {
+      "epoch": 0.0029,
+      "grad_norm": 0.5861155579490576,
+      "learning_rate": 0.00087,
+      "loss": 6.1786,
+      "step": 290
+    },
+    {
+      "epoch": 0.00291,
+      "grad_norm": 0.6639450218120423,
+      "learning_rate": 0.000873,
+      "loss": 6.1739,
+      "step": 291
+    },
+    {
+      "epoch": 0.00292,
+      "grad_norm": 0.8465664162859037,
+      "learning_rate": 0.0008759999999999999,
+      "loss": 6.1756,
+      "step": 292
+    },
+    {
+      "epoch": 0.00293,
+      "grad_norm": 0.9291009677838105,
+      "learning_rate": 0.000879,
+      "loss": 6.1618,
+      "step": 293
+    },
+    {
+      "epoch": 0.00294,
+      "grad_norm": 1.3082966000676708,
+      "learning_rate": 0.000882,
+      "loss": 6.1833,
+      "step": 294
+    },
+    {
+      "epoch": 0.00295,
+      "grad_norm": 1.1893342417992165,
+      "learning_rate": 0.0008849999999999999,
+      "loss": 6.1586,
+      "step": 295
+    },
+    {
+      "epoch": 0.00296,
+      "grad_norm": 0.6546841726146672,
+      "learning_rate": 0.000888,
+      "loss": 6.1551,
+      "step": 296
+    },
+    {
+      "epoch": 0.00297,
+      "grad_norm": 0.8416245966488266,
+      "learning_rate": 0.000891,
+      "loss": 6.1315,
+      "step": 297
+    },
+    {
+      "epoch": 0.00298,
+      "grad_norm": 0.806301942306892,
+      "learning_rate": 0.0008939999999999999,
+      "loss": 6.1452,
+      "step": 298
+    },
+    {
+      "epoch": 0.00299,
+      "grad_norm": 1.0994469943740992,
+      "learning_rate": 0.000897,
+      "loss": 6.1288,
+      "step": 299
+    },
+    {
+      "epoch": 0.003,
+      "grad_norm": 0.7425213774845364,
+      "learning_rate": 0.0009,
+      "loss": 6.123,
+      "step": 300
+    },
+    {
+      "epoch": 0.00301,
+      "grad_norm": 0.6433032250398036,
+      "learning_rate": 0.0009029999999999999,
+      "loss": 6.1149,
+      "step": 301
+    },
+    {
+      "epoch": 0.00302,
+      "grad_norm": 0.5449990789285414,
+      "learning_rate": 0.000906,
+      "loss": 6.0952,
+      "step": 302
+    },
+    {
+      "epoch": 0.00303,
+      "grad_norm": 0.5356674479383384,
+      "learning_rate": 0.000909,
+      "loss": 6.1067,
+      "step": 303
+    },
+    {
+      "epoch": 0.00304,
+      "grad_norm": 0.5049891146020773,
+      "learning_rate": 0.000912,
+      "loss": 6.1013,
+      "step": 304
+    },
+    {
+      "epoch": 0.00305,
+      "grad_norm": 0.48068374440179407,
+      "learning_rate": 0.000915,
+      "loss": 6.0985,
+      "step": 305
+    },
+    {
+      "epoch": 0.00306,
+      "grad_norm": 0.39305660615027715,
+      "learning_rate": 0.000918,
+      "loss": 6.0835,
+      "step": 306
+    },
+    {
+      "epoch": 0.00307,
+      "grad_norm": 0.4044411514506039,
+      "learning_rate": 0.000921,
+      "loss": 6.0774,
+      "step": 307
+    },
+    {
+      "epoch": 0.00308,
+      "grad_norm": 0.3317383414417003,
+      "learning_rate": 0.000924,
+      "loss": 6.0578,
+      "step": 308
+    },
+    {
+      "epoch": 0.00309,
+      "grad_norm": 0.4170081186564571,
+      "learning_rate": 0.000927,
+      "loss": 6.0741,
+      "step": 309
+    },
+    {
+      "epoch": 0.0031,
+      "grad_norm": 0.3537372632819937,
+      "learning_rate": 0.00093,
+      "loss": 6.0544,
+      "step": 310
+    },
+    {
+      "epoch": 0.00311,
+      "grad_norm": 0.3807795877969161,
+      "learning_rate": 0.000933,
+      "loss": 6.0455,
+      "step": 311
+    },
+    {
+      "epoch": 0.00312,
+      "grad_norm": 0.36301395919450175,
+      "learning_rate": 0.000936,
+      "loss": 6.025,
+      "step": 312
+    },
+    {
+      "epoch": 0.00313,
+      "grad_norm": 0.456154645430038,
+      "learning_rate": 0.0009390000000000001,
+      "loss": 6.0399,
+      "step": 313
+    },
+    {
+      "epoch": 0.00314,
+      "grad_norm": 0.6252307526263491,
+      "learning_rate": 0.000942,
+      "loss": 6.048,
+      "step": 314
+    },
+    {
+      "epoch": 0.00315,
+      "grad_norm": 1.0054536709858048,
+      "learning_rate": 0.000945,
+      "loss": 6.0332,
+      "step": 315
+    },
+    {
+      "epoch": 0.00316,
+      "grad_norm": 1.2671801420052267,
+      "learning_rate": 0.0009480000000000001,
+      "loss": 6.0322,
+      "step": 316
+    },
+    {
+      "epoch": 0.00317,
+      "grad_norm": 0.5267566088823273,
+      "learning_rate": 0.000951,
+      "loss": 6.0198,
+      "step": 317
+    },
+    {
+      "epoch": 0.00318,
+      "grad_norm": 1.2005145600955376,
+      "learning_rate": 0.000954,
+      "loss": 6.041,
+      "step": 318
+    },
+    {
+      "epoch": 0.00319,
+      "grad_norm": 1.004603435172909,
+      "learning_rate": 0.0009570000000000001,
+      "loss": 6.025,
+      "step": 319
+    },
+    {
+      "epoch": 0.0032,
+      "grad_norm": 1.1201777687880106,
+      "learning_rate": 0.00096,
+      "loss": 6.0356,
+      "step": 320
+    },
+    {
+      "epoch": 0.00321,
+      "grad_norm": 1.0167944206203925,
+      "learning_rate": 0.000963,
+      "loss": 6.012,
+      "step": 321
+    },
+    {
+      "epoch": 0.00322,
+      "grad_norm": 1.374478470085335,
+      "learning_rate": 0.0009660000000000001,
+      "loss": 6.025,
+      "step": 322
+    },
+    {
+      "epoch": 0.00323,
+      "grad_norm": 0.7765654772896999,
+      "learning_rate": 0.000969,
+      "loss": 6.0057,
+      "step": 323
+    },
+    {
+      "epoch": 0.00324,
+      "grad_norm": 0.6987857474960232,
+      "learning_rate": 0.0009720000000000001,
+      "loss": 6.0035,
+      "step": 324
+    },
+    {
+      "epoch": 0.00325,
+      "grad_norm": 0.7564069291902213,
+      "learning_rate": 0.0009750000000000001,
+      "loss": 5.9742,
+      "step": 325
+    },
+    {
+      "epoch": 0.00326,
+      "grad_norm": 0.7275905653375951,
+      "learning_rate": 0.0009780000000000001,
+      "loss": 5.9877,
+      "step": 326
+    },
+    {
+      "epoch": 0.00327,
+      "grad_norm": 0.6984989438722892,
+      "learning_rate": 0.000981,
+      "loss": 5.9857,
+      "step": 327
+    },
+    {
+      "epoch": 0.00328,
+      "grad_norm": 0.721382136086564,
+      "learning_rate": 0.000984,
+      "loss": 5.9627,
+      "step": 328
+    },
+    {
+      "epoch": 0.00329,
+      "grad_norm": 0.7841656285031653,
+      "learning_rate": 0.000987,
+      "loss": 5.9649,
+      "step": 329
+    },
+    {
+      "epoch": 0.0033,
+      "grad_norm": 0.8334397806028965,
+      "learning_rate": 0.00099,
+      "loss": 5.9568,
+      "step": 330
+    },
+    {
+      "epoch": 0.00331,
+      "grad_norm": 1.070828806370832,
+      "learning_rate": 0.0009930000000000002,
+      "loss": 5.9642,
+      "step": 331
+    },
+    {
+      "epoch": 0.00332,
+      "grad_norm": 1.0826703064607723,
+      "learning_rate": 0.0009960000000000001,
+      "loss": 5.9538,
+      "step": 332
+    },
+    {
+      "epoch": 0.00333,
+      "grad_norm": 0.9458285226462658,
+      "learning_rate": 0.000999,
+      "loss": 5.9361,
+      "step": 333
+    },
+    {
+      "epoch": 0.00334,
+      "grad_norm": 0.8783572499928525,
+      "learning_rate": 0.001002,
+      "loss": 5.9443,
+      "step": 334
+    },
+    {
+      "epoch": 0.00335,
+      "grad_norm": 0.5654549281403035,
+      "learning_rate": 0.001005,
+      "loss": 5.9354,
+      "step": 335
+    },
+    {
+      "epoch": 0.00336,
+      "grad_norm": 0.6516350129184051,
+      "learning_rate": 0.001008,
+      "loss": 5.9285,
+      "step": 336
+    },
+    {
+      "epoch": 0.00337,
+      "grad_norm": 0.5946351510133966,
+      "learning_rate": 0.0010110000000000002,
+      "loss": 5.9026,
+      "step": 337
+    },
+    {
+      "epoch": 0.00338,
+      "grad_norm": 0.46824568572142555,
+      "learning_rate": 0.0010140000000000001,
+      "loss": 5.9165,
+      "step": 338
+    },
+    {
+      "epoch": 0.00339,
+      "grad_norm": 0.5098867793571775,
+      "learning_rate": 0.0010170000000000001,
+      "loss": 5.9215,
+      "step": 339
+    },
+    {
+      "epoch": 0.0034,
+      "grad_norm": 0.5202026348670868,
+      "learning_rate": 0.00102,
+      "loss": 5.8881,
+      "step": 340
+    },
+    {
+      "epoch": 0.00341,
+      "grad_norm": 0.4748523902202147,
+      "learning_rate": 0.001023,
+      "loss": 5.8705,
+      "step": 341
+    },
+    {
+      "epoch": 0.00342,
+      "grad_norm": 0.42065780247182966,
+      "learning_rate": 0.001026,
+      "loss": 5.8732,
+      "step": 342
+    },
+    {
+      "epoch": 0.00343,
+      "grad_norm": 0.4385189981740614,
+      "learning_rate": 0.0010290000000000002,
+      "loss": 5.8762,
+      "step": 343
+    },
+    {
+      "epoch": 0.00344,
+      "grad_norm": 0.44131626809067537,
+      "learning_rate": 0.001032,
+      "loss": 5.8646,
+      "step": 344
+    },
+    {
+      "epoch": 0.00345,
+      "grad_norm": 0.5602404326473228,
+      "learning_rate": 0.001035,
+      "loss": 5.8576,
+      "step": 345
+    },
+    {
+      "epoch": 0.00346,
+      "grad_norm": 0.7338782884674983,
+      "learning_rate": 0.0010379999999999999,
+      "loss": 5.8684,
+      "step": 346
+    },
+    {
+      "epoch": 0.00347,
+      "grad_norm": 0.8871217127911775,
+      "learning_rate": 0.001041,
+      "loss": 5.8429,
+      "step": 347
+    },
+    {
+      "epoch": 0.00348,
+      "grad_norm": 0.7294005585257834,
+      "learning_rate": 0.001044,
+      "loss": 5.8498,
+      "step": 348
+    },
+    {
+      "epoch": 0.00349,
+      "grad_norm": 0.672488440407315,
+      "learning_rate": 0.001047,
+      "loss": 5.8268,
+      "step": 349
+    },
+    {
+      "epoch": 0.0035,
+      "grad_norm": 0.944274771030096,
+      "learning_rate": 0.00105,
+      "loss": 5.8574,
+      "step": 350
+    },
+    {
+      "epoch": 0.00351,
+      "grad_norm": 0.8431353927439175,
+      "learning_rate": 0.001053,
+      "loss": 5.8586,
+      "step": 351
+    },
+    {
+      "epoch": 0.00352,
+      "grad_norm": 0.8223595555288745,
+      "learning_rate": 0.0010559999999999999,
+      "loss": 5.8326,
+      "step": 352
+    },
+    {
+      "epoch": 0.00353,
+      "grad_norm": 0.7412573085023916,
+      "learning_rate": 0.001059,
+      "loss": 5.8336,
+      "step": 353
+    },
+    {
+      "epoch": 0.00354,
+      "grad_norm": 0.7219192917744488,
+      "learning_rate": 0.001062,
+      "loss": 5.8047,
+      "step": 354
+    },
+    {
+      "epoch": 0.00355,
+      "grad_norm": 0.7332482461779326,
+      "learning_rate": 0.001065,
+      "loss": 5.8213,
+      "step": 355
+    },
+    {
+      "epoch": 0.00356,
+      "grad_norm": 0.6300796792438244,
+      "learning_rate": 0.001068,
+      "loss": 5.8096,
+      "step": 356
+    },
+    {
+      "epoch": 0.00357,
+      "grad_norm": 0.7796926368866208,
+      "learning_rate": 0.001071,
+      "loss": 5.802,
+      "step": 357
+    },
+    {
+      "epoch": 0.00358,
+      "grad_norm": 1.048118676098695,
+      "learning_rate": 0.001074,
+      "loss": 5.8089,
+      "step": 358
+    },
+    {
+      "epoch": 0.00359,
+      "grad_norm": 0.8469283052653666,
+      "learning_rate": 0.001077,
+      "loss": 5.7908,
+      "step": 359
+    },
+    {
+      "epoch": 0.0036,
+      "grad_norm": 0.824969051981288,
+      "learning_rate": 0.00108,
+      "loss": 5.7948,
+      "step": 360
+    },
+    {
+      "epoch": 0.00361,
+      "grad_norm": 0.8007424119602606,
+      "learning_rate": 0.001083,
+      "loss": 5.7876,
+      "step": 361
+    },
+    {
+      "epoch": 0.00362,
+      "grad_norm": 0.7740442530973947,
+      "learning_rate": 0.001086,
+      "loss": 5.7713,
+      "step": 362
+    },
+    {
+      "epoch": 0.00363,
+      "grad_norm": 0.8217790100779416,
+      "learning_rate": 0.001089,
+      "loss": 5.7673,
+      "step": 363
+    },
+    {
+      "epoch": 0.00364,
+      "grad_norm": 0.9222078053619298,
+      "learning_rate": 0.001092,
+      "loss": 5.7813,
+      "step": 364
+    },
+    {
+      "epoch": 0.00365,
+      "grad_norm": 1.3725693861639678,
+      "learning_rate": 0.001095,
+      "loss": 5.801,
+      "step": 365
+    },
+    {
+      "epoch": 0.00366,
+      "grad_norm": 0.7809021294214551,
+      "learning_rate": 0.001098,
+      "loss": 5.7587,
+      "step": 366
+    },
+    {
+      "epoch": 0.00367,
+      "grad_norm": 1.1843193874441642,
+      "learning_rate": 0.001101,
+      "loss": 5.7731,
+      "step": 367
+    },
+    {
+      "epoch": 0.00368,
+      "grad_norm": 0.7875897064576857,
+      "learning_rate": 0.001104,
+      "loss": 5.7636,
+      "step": 368
+    },
+    {
+      "epoch": 0.00369,
+      "grad_norm": 0.6858717859751121,
+      "learning_rate": 0.001107,
+      "loss": 5.7637,
+      "step": 369
+    },
+    {
+      "epoch": 0.0037,
+      "grad_norm": 0.6252688126791845,
+      "learning_rate": 0.00111,
+      "loss": 5.7428,
+      "step": 370
+    },
+    {
+      "epoch": 0.00371,
+      "grad_norm": 0.6219765776345385,
+      "learning_rate": 0.001113,
+      "loss": 5.7116,
+      "step": 371
+    },
+    {
+      "epoch": 0.00372,
+      "grad_norm": 0.5998751494739524,
+      "learning_rate": 0.001116,
+      "loss": 5.7331,
+      "step": 372
+    },
+    {
+      "epoch": 0.00373,
+      "grad_norm": 0.5714053922949759,
+      "learning_rate": 0.001119,
+      "loss": 5.72,
+      "step": 373
+    },
+    {
+      "epoch": 0.00374,
+      "grad_norm": 0.628102864915935,
+      "learning_rate": 0.001122,
+      "loss": 5.7262,
+      "step": 374
+    },
+    {
+      "epoch": 0.00375,
+      "grad_norm": 0.5909991857975568,
+      "learning_rate": 0.0011250000000000001,
+      "loss": 5.7276,
+      "step": 375
+    },
+    {
+      "epoch": 0.00376,
+      "grad_norm": 0.5216592986380127,
+      "learning_rate": 0.001128,
+      "loss": 5.72,
+      "step": 376
+    },
+    {
+      "epoch": 0.00377,
+      "grad_norm": 0.5469626358369448,
+      "learning_rate": 0.001131,
+      "loss": 5.7145,
+      "step": 377
+    },
+    {
+      "epoch": 0.00378,
+      "grad_norm": 0.6291207985788981,
+      "learning_rate": 0.001134,
+      "loss": 5.7123,
+      "step": 378
+    },
+    {
+      "epoch": 0.00379,
+      "grad_norm": 0.6445779804368097,
+      "learning_rate": 0.001137,
+      "loss": 5.698,
+      "step": 379
+    },
+    {
+      "epoch": 0.0038,
+      "grad_norm": 0.7034810438425685,
+      "learning_rate": 0.00114,
+      "loss": 5.7177,
+      "step": 380
+    },
+    {
+      "epoch": 0.00381,
+      "grad_norm": 0.9330449756794958,
+      "learning_rate": 0.0011430000000000001,
+      "loss": 5.6924,
+      "step": 381
+    },
+    {
+      "epoch": 0.00382,
+      "grad_norm": 0.9449582673655755,
+      "learning_rate": 0.001146,
+      "loss": 5.6863,
+      "step": 382
+    },
+    {
+      "epoch": 0.00383,
+      "grad_norm": 0.6715143882341864,
+      "learning_rate": 0.001149,
+      "loss": 5.6846,
+      "step": 383
+    },
+    {
+      "epoch": 0.00384,
+      "grad_norm": 0.8662082921806001,
+      "learning_rate": 0.001152,
+      "loss": 5.699,
+      "step": 384
+    },
+    {
+      "epoch": 0.00385,
+      "grad_norm": 0.8482542645994808,
+      "learning_rate": 0.001155,
+      "loss": 5.6818,
+      "step": 385
+    },
+    {
+      "epoch": 0.00386,
+      "grad_norm": 0.8694120663581818,
+      "learning_rate": 0.001158,
+      "loss": 5.6914,
+      "step": 386
+    },
+    {
+      "epoch": 0.00387,
+      "grad_norm": 1.382803938865885,
+      "learning_rate": 0.0011610000000000001,
+      "loss": 5.6813,
+      "step": 387
+    },
+    {
+      "epoch": 0.00388,
+      "grad_norm": 0.9644228724095026,
+      "learning_rate": 0.001164,
+      "loss": 5.6689,
+      "step": 388
+    },
+    {
+      "epoch": 0.00389,
+      "grad_norm": 0.7396617808819147,
+      "learning_rate": 0.001167,
+      "loss": 5.6709,
+      "step": 389
+    },
+    {
+      "epoch": 0.0039,
+      "grad_norm": 0.5609185173455993,
+      "learning_rate": 0.00117,
+      "loss": 5.648,
+      "step": 390
+    },
+    {
+      "epoch": 0.00391,
+      "grad_norm": 0.735416794156057,
+      "learning_rate": 0.001173,
+      "loss": 5.6589,
+      "step": 391
+    },
+    {
+      "epoch": 0.00392,
+      "grad_norm": 0.7564046131732848,
+      "learning_rate": 0.001176,
+      "loss": 5.6662,
+      "step": 392
+    },
+    {
+      "epoch": 0.00393,
+      "grad_norm": 0.7233997204719024,
+      "learning_rate": 0.0011790000000000001,
+      "loss": 5.6377,
+      "step": 393
+    },
+    {
+      "epoch": 0.00394,
+      "grad_norm": 0.6288246845478384,
+      "learning_rate": 0.001182,
+      "loss": 5.6239,
+      "step": 394
+    },
+    {
+      "epoch": 0.00395,
+      "grad_norm": 0.6696319293361586,
+      "learning_rate": 0.001185,
+      "loss": 5.6102,
+      "step": 395
+    },
+    {
+      "epoch": 0.00396,
+      "grad_norm": 0.7214553632515296,
+      "learning_rate": 0.001188,
+      "loss": 5.6276,
+      "step": 396
+    },
+    {
+      "epoch": 0.00397,
+      "grad_norm": 0.7104651338358826,
+      "learning_rate": 0.001191,
+      "loss": 5.619,
+      "step": 397
+    },
+    {
+      "epoch": 0.00398,
+      "grad_norm": 0.5955487966253655,
+      "learning_rate": 0.0011940000000000002,
+      "loss": 5.6195,
+      "step": 398
+    },
+    {
+      "epoch": 0.00399,
+      "grad_norm": 0.6061151049974988,
+      "learning_rate": 0.0011970000000000001,
+      "loss": 5.6211,
+      "step": 399
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 0.747975845260549,
+      "learning_rate": 0.0012000000000000001,
+      "loss": 5.6119,
+      "step": 400
+    },
+    {
+      "epoch": 0.00401,
+      "grad_norm": 1.066378470212407,
+      "learning_rate": 0.001203,
+      "loss": 5.6055,
+      "step": 401
+    },
+    {
+      "epoch": 0.00402,
+      "grad_norm": 1.1521859370487337,
+      "learning_rate": 0.001206,
+      "loss": 5.6144,
+      "step": 402
+    },
+    {
+      "epoch": 0.00403,
+      "grad_norm": 1.0137206411467539,
+      "learning_rate": 0.001209,
+      "loss": 5.6121,
+      "step": 403
+    },
+    {
+      "epoch": 0.00404,
+      "grad_norm": 1.1663989264954397,
+      "learning_rate": 0.0012120000000000002,
+      "loss": 5.6177,
+      "step": 404
+    },
+    {
+      "epoch": 0.00405,
+      "grad_norm": 1.1363050593119097,
+      "learning_rate": 0.0012150000000000002,
+      "loss": 5.6098,
+      "step": 405
+    },
+    {
+      "epoch": 0.00406,
+      "grad_norm": 1.0087910495777368,
+      "learning_rate": 0.0012180000000000001,
+      "loss": 5.5952,
+      "step": 406
+    },
+    {
+      "epoch": 0.00407,
+      "grad_norm": 0.9520738669869687,
+      "learning_rate": 0.0012209999999999999,
+      "loss": 5.5915,
+      "step": 407
+    },
+    {
+      "epoch": 0.00408,
+      "grad_norm": 1.0377478362348644,
+      "learning_rate": 0.001224,
+      "loss": 5.6129,
+      "step": 408
+    },
+    {
+      "epoch": 0.00409,
+      "grad_norm": 1.1016437762309168,
+      "learning_rate": 0.001227,
+      "loss": 5.5927,
+      "step": 409
+    },
+    {
+      "epoch": 0.0041,
+      "grad_norm": 0.9709841842795465,
+      "learning_rate": 0.00123,
+      "loss": 5.5883,
+      "step": 410
+    },
+    {
+      "epoch": 0.00411,
+      "grad_norm": 0.8240534275873757,
+      "learning_rate": 0.001233,
+      "loss": 5.5977,
+      "step": 411
+    },
+    {
+      "epoch": 0.00412,
+      "grad_norm": 0.8077831494606339,
+      "learning_rate": 0.001236,
+      "loss": 5.5894,
+      "step": 412
+    },
+    {
+      "epoch": 0.00413,
+      "grad_norm": 0.7433003641205871,
+      "learning_rate": 0.0012389999999999999,
+      "loss": 5.5815,
+      "step": 413
+    },
+    {
+      "epoch": 0.00414,
+      "grad_norm": 0.615571577509929,
+      "learning_rate": 0.001242,
+      "loss": 5.5582,
+      "step": 414
+    },
+    {
+      "epoch": 0.00415,
+      "grad_norm": 0.656841702591939,
+      "learning_rate": 0.001245,
+      "loss": 5.5606,
+      "step": 415
+    },
+    {
+      "epoch": 0.00416,
+      "grad_norm": 0.613196865035469,
+      "learning_rate": 0.001248,
+      "loss": 5.5418,
+      "step": 416
+    },
+    {
+      "epoch": 0.00417,
+      "grad_norm": 0.48631699887544017,
+      "learning_rate": 0.001251,
+      "loss": 5.5683,
+      "step": 417
+    },
+    {
+      "epoch": 0.00418,
+      "grad_norm": 0.4611320175617472,
+      "learning_rate": 0.001254,
+      "loss": 5.5401,
+      "step": 418
+    },
+    {
+      "epoch": 0.00419,
+      "grad_norm": 0.5824396705507784,
+      "learning_rate": 0.0012569999999999999,
+      "loss": 5.5305,
+      "step": 419
+    },
+    {
+      "epoch": 0.0042,
+      "grad_norm": 0.8055087238093066,
+      "learning_rate": 0.00126,
+      "loss": 5.5467,
+      "step": 420
+    },
+    {
+      "epoch": 0.00421,
+      "grad_norm": 0.9831917562206494,
+      "learning_rate": 0.001263,
+      "loss": 5.5503,
+      "step": 421
+    },
+    {
+      "epoch": 0.00422,
+      "grad_norm": 0.8124592037579013,
+      "learning_rate": 0.001266,
+      "loss": 5.5357,
+      "step": 422
+    },
+    {
+      "epoch": 0.00423,
+      "grad_norm": 1.0116811247233115,
+      "learning_rate": 0.001269,
+      "loss": 5.5337,
+      "step": 423
+    },
+    {
+      "epoch": 0.00424,
+      "grad_norm": 1.368200107021336,
+      "learning_rate": 0.001272,
+      "loss": 5.5595,
+      "step": 424
+    },
+    {
+      "epoch": 0.00425,
+      "grad_norm": 0.9353292964204464,
+      "learning_rate": 0.001275,
+      "loss": 5.5445,
+      "step": 425
+    },
+    {
+      "epoch": 0.00426,
+      "grad_norm": 0.9604600896045714,
+      "learning_rate": 0.001278,
+      "loss": 5.5493,
+      "step": 426
+    },
+    {
+      "epoch": 0.00427,
+      "grad_norm": 0.9684632680550875,
+      "learning_rate": 0.001281,
+      "loss": 5.5328,
+      "step": 427
+    },
+    {
+      "epoch": 0.00428,
+      "grad_norm": 0.9533788870858234,
+      "learning_rate": 0.001284,
+      "loss": 5.5058,
+      "step": 428
+    },
+    {
+      "epoch": 0.00429,
+      "grad_norm": 1.113271451738206,
+      "learning_rate": 0.001287,
+      "loss": 5.5322,
+      "step": 429
+    },
+    {
+      "epoch": 0.0043,
+      "grad_norm": 0.9492520989573756,
+      "learning_rate": 0.00129,
+      "loss": 5.5226,
+      "step": 430
+    },
+    {
+      "epoch": 0.00431,
+      "grad_norm": 1.1309185966069144,
+      "learning_rate": 0.001293,
+      "loss": 5.5118,
+      "step": 431
+    },
+    {
+      "epoch": 0.00432,
+      "grad_norm": 1.2851571092183034,
+      "learning_rate": 0.001296,
+      "loss": 5.5367,
+      "step": 432
+    },
+    {
+      "epoch": 0.00433,
+      "grad_norm": 0.7219415163940603,
+      "learning_rate": 0.001299,
+      "loss": 5.5106,
+      "step": 433
+    },
+    {
+      "epoch": 0.00434,
+      "grad_norm": 0.7943786167833424,
+      "learning_rate": 0.001302,
+      "loss": 5.5109,
+      "step": 434
+    },
+    {
+      "epoch": 0.00435,
+      "grad_norm": 0.7514833526497924,
+      "learning_rate": 0.001305,
+      "loss": 5.5051,
+      "step": 435
+    },
+    {
+      "epoch": 0.00436,
+      "grad_norm": 0.5499552719714149,
+      "learning_rate": 0.001308,
+      "loss": 5.4758,
+      "step": 436
+    },
+    {
+      "epoch": 0.00437,
+      "grad_norm": 0.5803132978630802,
+      "learning_rate": 0.001311,
+      "loss": 5.494,
+      "step": 437
+    },
+    {
+      "epoch": 0.00438,
+      "grad_norm": 0.4343754614294673,
+      "learning_rate": 0.001314,
+      "loss": 5.4701,
+      "step": 438
+    },
+    {
+      "epoch": 0.00439,
+      "grad_norm": 0.49450539118744047,
+      "learning_rate": 0.001317,
+      "loss": 5.4997,
+      "step": 439
+    },
+    {
+      "epoch": 0.0044,
+      "grad_norm": 0.3670496641121159,
+      "learning_rate": 0.00132,
+      "loss": 5.4633,
+      "step": 440
+    },
+    {
+      "epoch": 0.00441,
+      "grad_norm": 0.44023069828614625,
+      "learning_rate": 0.001323,
+      "loss": 5.4624,
+      "step": 441
+    },
+    {
+      "epoch": 0.00442,
+      "grad_norm": 0.45124022563783683,
+      "learning_rate": 0.0013260000000000001,
+      "loss": 5.4541,
+      "step": 442
+    },
+    {
+      "epoch": 0.00443,
+      "grad_norm": 0.5170657075032803,
+      "learning_rate": 0.001329,
+      "loss": 5.4572,
+      "step": 443
+    },
+    {
+      "epoch": 0.00444,
+      "grad_norm": 0.6406930762525409,
+      "learning_rate": 0.001332,
+      "loss": 5.4608,
+      "step": 444
+    },
+    {
+      "epoch": 0.00445,
+      "grad_norm": 0.7431205120059805,
+      "learning_rate": 0.001335,
+      "loss": 5.4462,
+      "step": 445
+    },
+    {
+      "epoch": 0.00446,
+      "grad_norm": 0.9849435734941417,
+      "learning_rate": 0.001338,
+      "loss": 5.4578,
+      "step": 446
+    },
+    {
+      "epoch": 0.00447,
+      "grad_norm": 1.1031854356255555,
+      "learning_rate": 0.001341,
+      "loss": 5.469,
+      "step": 447
+    },
+    {
+      "epoch": 0.00448,
+      "grad_norm": 0.8550264287932019,
+      "learning_rate": 0.0013440000000000001,
+      "loss": 5.4411,
+      "step": 448
+    },
+    {
+      "epoch": 0.00449,
+      "grad_norm": 0.8569858280927948,
+      "learning_rate": 0.001347,
+      "loss": 5.4491,
+      "step": 449
+    },
+    {
+      "epoch": 0.0045,
+      "grad_norm": 0.766625402040655,
+      "learning_rate": 0.00135,
+      "loss": 5.4461,
+      "step": 450
+    },
+    {
+      "epoch": 0.00451,
+      "grad_norm": 0.8992860281272513,
+      "learning_rate": 0.001353,
+      "loss": 5.4411,
+      "step": 451
+    },
+    {
+      "epoch": 0.00452,
+      "grad_norm": 0.9587820940422671,
+      "learning_rate": 0.001356,
+      "loss": 5.4273,
+      "step": 452
+    },
+    {
+      "epoch": 0.00453,
+      "grad_norm": 0.9706881368414635,
+      "learning_rate": 0.001359,
+      "loss": 5.4204,
+      "step": 453
+    },
+    {
+      "epoch": 0.00454,
+      "grad_norm": 1.1942708782025127,
+      "learning_rate": 0.0013620000000000001,
+      "loss": 5.4481,
+      "step": 454
+    },
+    {
+      "epoch": 0.00455,
+      "grad_norm": 1.0651260215062406,
+      "learning_rate": 0.0013650000000000001,
+      "loss": 5.4483,
+      "step": 455
+    },
+    {
+      "epoch": 0.00456,
+      "grad_norm": 1.0894309440916736,
+      "learning_rate": 0.001368,
+      "loss": 5.4405,
+      "step": 456
+    },
+    {
+      "epoch": 0.00457,
+      "grad_norm": 1.3632436316792311,
+      "learning_rate": 0.001371,
+      "loss": 5.442,
+      "step": 457
+    },
+    {
+      "epoch": 0.00458,
+      "grad_norm": 0.9802910013598249,
+      "learning_rate": 0.001374,
+      "loss": 5.4423,
+      "step": 458
+    },
+    {
+      "epoch": 0.00459,
+      "grad_norm": 0.9681577835390196,
+      "learning_rate": 0.0013770000000000002,
+      "loss": 5.4314,
+      "step": 459
+    },
+    {
+      "epoch": 0.0046,
+      "grad_norm": 0.955881976726698,
+      "learning_rate": 0.0013800000000000002,
+      "loss": 5.4146,
+      "step": 460
+    },
+    {
+      "epoch": 0.00461,
+      "grad_norm": 1.072138092727722,
+      "learning_rate": 0.0013830000000000001,
+      "loss": 5.4186,
+      "step": 461
+    },
+    {
+      "epoch": 0.00462,
+      "grad_norm": 0.8672451154734739,
+      "learning_rate": 0.001386,
+      "loss": 5.4193,
+      "step": 462
+    },
+    {
+      "epoch": 0.00463,
+      "grad_norm": 0.9908998654710571,
+      "learning_rate": 0.001389,
+      "loss": 5.4011,
+      "step": 463
+    },
+    {
+      "epoch": 0.00464,
+      "grad_norm": 1.1599842863140275,
+      "learning_rate": 0.001392,
+      "loss": 5.4321,
+      "step": 464
+    },
+    {
+      "epoch": 0.00465,
+      "grad_norm": 0.698212811337685,
+      "learning_rate": 0.0013950000000000002,
+      "loss": 5.3861,
+      "step": 465
+    },
+    {
+      "epoch": 0.00466,
+      "grad_norm": 0.6209828526031017,
+      "learning_rate": 0.0013980000000000002,
+      "loss": 5.4004,
+      "step": 466
+    },
+    {
+      "epoch": 0.00467,
+      "grad_norm": 0.626908730655191,
+      "learning_rate": 0.0014010000000000001,
+      "loss": 5.3894,
+      "step": 467
+    },
+    {
+      "epoch": 0.00468,
+      "grad_norm": 0.6046317726003113,
+      "learning_rate": 0.001404,
+      "loss": 5.3787,
+      "step": 468
+    },
+    {
+      "epoch": 0.00469,
+      "grad_norm": 0.5283854127913247,
+      "learning_rate": 0.001407,
+      "loss": 5.3767,
+      "step": 469
+    },
+    {
+      "epoch": 0.0047,
+      "grad_norm": 0.5494342704278756,
+      "learning_rate": 0.00141,
+      "loss": 5.3859,
+      "step": 470
+    },
+    {
+      "epoch": 0.00471,
+      "grad_norm": 0.6548172150181142,
+      "learning_rate": 0.001413,
+      "loss": 5.3653,
+      "step": 471
+    },
+    {
+      "epoch": 0.00472,
+      "grad_norm": 0.8352552132751145,
+      "learning_rate": 0.001416,
+      "loss": 5.3799,
+      "step": 472
+    },
+    {
+      "epoch": 0.00473,
+      "grad_norm": 1.04775859687285,
+      "learning_rate": 0.001419,
+      "loss": 5.3985,
+      "step": 473
+    },
+    {
+      "epoch": 0.00474,
+      "grad_norm": 0.9808295146566943,
+      "learning_rate": 0.0014219999999999999,
+      "loss": 5.3787,
+      "step": 474
+    },
+    {
+      "epoch": 0.00475,
+      "grad_norm": 1.0065096729700158,
+      "learning_rate": 0.001425,
+      "loss": 5.3725,
+      "step": 475
+    },
+    {
+      "epoch": 0.00476,
+      "grad_norm": 1.0604383614131039,
+      "learning_rate": 0.001428,
+      "loss": 5.4172,
+      "step": 476
+    },
+    {
+      "epoch": 0.00477,
+      "grad_norm": 0.5551620314639825,
+      "learning_rate": 0.001431,
+      "loss": 5.3484,
+      "step": 477
+    },
+    {
+      "epoch": 0.00478,
+      "grad_norm": 0.6446758952809472,
+      "learning_rate": 0.001434,
+      "loss": 5.3819,
+      "step": 478
+    },
+    {
+      "epoch": 0.00479,
+      "grad_norm": 0.6811842486215005,
+      "learning_rate": 0.001437,
+      "loss": 5.3551,
+      "step": 479
+    },
+    {
+      "epoch": 0.0048,
+      "grad_norm": 0.6985415352827566,
+      "learning_rate": 0.0014399999999999999,
+      "loss": 5.3502,
+      "step": 480
+    },
+    {
+      "epoch": 0.00481,
+      "grad_norm": 0.7430626707063542,
+      "learning_rate": 0.001443,
+      "loss": 5.3537,
+      "step": 481
+    },
+    {
+      "epoch": 0.00482,
+      "grad_norm": 0.7883555304691023,
+      "learning_rate": 0.001446,
+      "loss": 5.3281,
+      "step": 482
+    },
+    {
+      "epoch": 0.00483,
+      "grad_norm": 0.9604272864715826,
+      "learning_rate": 0.001449,
+      "loss": 5.3455,
+      "step": 483
+    },
+    {
+      "epoch": 0.00484,
+      "grad_norm": 0.9435245256880889,
+      "learning_rate": 0.001452,
+      "loss": 5.3494,
+      "step": 484
+    },
+    {
+      "epoch": 0.00485,
+      "grad_norm": 0.8042883678590446,
+      "learning_rate": 0.001455,
+      "loss": 5.3511,
+      "step": 485
+    },
+    {
+      "epoch": 0.00486,
+      "grad_norm": 0.7958734942371152,
+      "learning_rate": 0.001458,
+      "loss": 5.3247,
+      "step": 486
+    },
+    {
+      "epoch": 0.00487,
+      "grad_norm": 0.7735191187819994,
+      "learning_rate": 0.001461,
+      "loss": 5.3397,
+      "step": 487
+    },
+    {
+      "epoch": 0.00488,
+      "grad_norm": 1.012081783792506,
+      "learning_rate": 0.001464,
+      "loss": 5.3265,
+      "step": 488
+    },
+    {
+      "epoch": 0.00489,
+      "grad_norm": 0.8905045147335564,
+      "learning_rate": 0.001467,
+      "loss": 5.3131,
+      "step": 489
+    },
+    {
+      "epoch": 0.0049,
+      "grad_norm": 0.8898357576674187,
+      "learning_rate": 0.00147,
+      "loss": 5.3259,
+      "step": 490
+    },
+    {
+      "epoch": 0.00491,
+      "grad_norm": 0.9415043482581119,
+      "learning_rate": 0.001473,
+      "loss": 5.3269,
+      "step": 491
+    },
+    {
+      "epoch": 0.00492,
+      "grad_norm": 0.8320755719970162,
+      "learning_rate": 0.001476,
+      "loss": 5.3196,
+      "step": 492
+    },
+    {
+      "epoch": 0.00493,
+      "grad_norm": 0.8445852574519659,
+      "learning_rate": 0.001479,
+      "loss": 5.3145,
+      "step": 493
+    },
+    {
+      "epoch": 0.00494,
+      "grad_norm": 1.055556747775556,
+      "learning_rate": 0.001482,
+      "loss": 5.3274,
+      "step": 494
+    },
+    {
+      "epoch": 0.00495,
+      "grad_norm": 0.8469107680360277,
+      "learning_rate": 0.001485,
+      "loss": 5.3268,
+      "step": 495
+    },
+    {
+      "epoch": 0.00496,
+      "grad_norm": 0.7635320378283988,
+      "learning_rate": 0.001488,
+      "loss": 5.3083,
+      "step": 496
+    },
+    {
+      "epoch": 0.00497,
+      "grad_norm": 0.8508130432624411,
+      "learning_rate": 0.001491,
+      "loss": 5.2927,
+      "step": 497
+    },
+    {
+      "epoch": 0.00498,
+      "grad_norm": 0.6847944056515368,
+      "learning_rate": 0.001494,
+      "loss": 5.3134,
+      "step": 498
+    },
+    {
+      "epoch": 0.00499,
+      "grad_norm": 0.73441383592788,
+      "learning_rate": 0.001497,
+      "loss": 5.2894,
+      "step": 499
+    },
+    {
+      "epoch": 0.005,
+      "grad_norm": 0.6599139353534871,
+      "learning_rate": 0.0015,
+      "loss": 5.2898,
+      "step": 500
+    },
+    {
+      "epoch": 0.00501,
+      "grad_norm": 0.5852764030687648,
+      "learning_rate": 0.001503,
+      "loss": 5.2881,
+      "step": 501
+    },
+    {
+      "epoch": 0.00502,
+      "grad_norm": 0.6853021422001514,
+      "learning_rate": 0.001506,
+      "loss": 5.2946,
+      "step": 502
+    },
+    {
+      "epoch": 0.00503,
+      "grad_norm": 0.720141729806561,
+      "learning_rate": 0.0015090000000000001,
+      "loss": 5.2988,
+      "step": 503
+    },
+    {
+      "epoch": 0.00504,
+      "grad_norm": 0.7571759373525393,
+      "learning_rate": 0.001512,
+      "loss": 5.2829,
+      "step": 504
+    },
+    {
+      "epoch": 0.00505,
+      "grad_norm": 0.7892633543044816,
+      "learning_rate": 0.001515,
+      "loss": 5.2795,
+      "step": 505
+    },
+    {
+      "epoch": 0.00506,
+      "grad_norm": 0.7628738296895545,
+      "learning_rate": 0.001518,
+      "loss": 5.2778,
+      "step": 506
+    },
+    {
+      "epoch": 0.00507,
+      "grad_norm": 0.7334519380566521,
+      "learning_rate": 0.001521,
+      "loss": 5.2722,
+      "step": 507
+    },
+    {
+      "epoch": 0.00508,
+      "grad_norm": 0.7350882061617702,
+      "learning_rate": 0.001524,
+      "loss": 5.2693,
+      "step": 508
+    },
+    {
+      "epoch": 0.00509,
+      "grad_norm": 0.7407389039658853,
+      "learning_rate": 0.0015270000000000001,
+      "loss": 5.2562,
+      "step": 509
+    },
+    {
+      "epoch": 0.0051,
+      "grad_norm": 0.7104400166661097,
+      "learning_rate": 0.0015300000000000001,
+      "loss": 5.2696,
+      "step": 510
+    },
+    {
+      "epoch": 0.00511,
+      "grad_norm": 0.8484400275698938,
+      "learning_rate": 0.001533,
+      "loss": 5.2541,
+      "step": 511
+    },
+    {
+      "epoch": 0.00512,
+      "grad_norm": 0.7909533536283434,
+      "learning_rate": 0.001536,
+      "loss": 5.2475,
+      "step": 512
+    },
+    {
+      "epoch": 0.00513,
+      "grad_norm": 0.9481517521047255,
+      "learning_rate": 0.001539,
+      "loss": 5.2684,
+      "step": 513
+    },
+    {
+      "epoch": 0.00514,
+      "grad_norm": 1.405665610624422,
+      "learning_rate": 0.001542,
+      "loss": 5.2742,
+      "step": 514
+    },
+    {
+      "epoch": 0.00515,
+      "grad_norm": 0.9237050984729354,
+      "learning_rate": 0.0015450000000000001,
+      "loss": 5.2636,
+      "step": 515
+    },
+    {
+      "epoch": 0.00516,
+      "grad_norm": 1.1634719669204763,
+      "learning_rate": 0.0015480000000000001,
+      "loss": 5.2755,
+      "step": 516
+    },
+    {
+      "epoch": 0.00517,
+      "grad_norm": 0.8585760471561034,
+      "learning_rate": 0.001551,
+      "loss": 5.2577,
+      "step": 517
+    },
+    {
+      "epoch": 0.00518,
+      "grad_norm": 0.728905665920263,
+      "learning_rate": 0.001554,
+      "loss": 5.2422,
+      "step": 518
+    },
+    {
+      "epoch": 0.00519,
+      "grad_norm": 0.841003643450708,
+      "learning_rate": 0.001557,
+      "loss": 5.2472,
+      "step": 519
+    },
+    {
+      "epoch": 0.0052,
+      "grad_norm": 0.8542329179478354,
+      "learning_rate": 0.0015600000000000002,
+      "loss": 5.2573,
+      "step": 520
+    },
+    {
+      "epoch": 0.00521,
+      "grad_norm": 0.9136344705581151,
+      "learning_rate": 0.0015630000000000002,
+      "loss": 5.2327,
+      "step": 521
+    },
+    {
+      "epoch": 0.00522,
+      "grad_norm": 1.1142541660381047,
+      "learning_rate": 0.0015660000000000001,
+      "loss": 5.2435,
+      "step": 522
+    },
+    {
+      "epoch": 0.00523,
+      "grad_norm": 1.2424782927197504,
+      "learning_rate": 0.001569,
+      "loss": 5.2642,
+      "step": 523
+    },
+    {
+      "epoch": 0.00524,
+      "grad_norm": 0.9564855927553343,
+      "learning_rate": 0.001572,
+      "loss": 5.264,
+      "step": 524
+    },
+    {
+      "epoch": 0.00525,
+      "grad_norm": 0.8662515649518988,
+      "learning_rate": 0.001575,
+      "loss": 5.2537,
+      "step": 525
+    },
+    {
+      "epoch": 0.00526,
+      "grad_norm": 0.72079723918141,
+      "learning_rate": 0.0015780000000000002,
+      "loss": 5.2224,
+      "step": 526
+    },
+    {
+      "epoch": 0.00527,
+      "grad_norm": 0.6822789926865414,
+      "learning_rate": 0.0015810000000000002,
+      "loss": 5.2353,
+      "step": 527
+    },
+    {
+      "epoch": 0.00528,
+      "grad_norm": 0.6986034229049856,
+      "learning_rate": 0.0015840000000000001,
+      "loss": 5.2128,
+      "step": 528
+    },
+    {
+      "epoch": 0.00529,
+      "grad_norm": 0.6935526974121139,
+      "learning_rate": 0.001587,
+      "loss": 5.2236,
+      "step": 529
+    },
+    {
+      "epoch": 0.0053,
+      "grad_norm": 0.6635846542220551,
+      "learning_rate": 0.00159,
+      "loss": 5.2302,
+      "step": 530
+    },
+    {
+      "epoch": 0.00531,
+      "grad_norm": 0.7197123783695031,
+      "learning_rate": 0.001593,
+      "loss": 5.2167,
+      "step": 531
+    },
+    {
+      "epoch": 0.00532,
+      "grad_norm": 0.6862921713046577,
+      "learning_rate": 0.0015960000000000002,
+      "loss": 5.2205,
+      "step": 532
+    },
+    {
+      "epoch": 0.00533,
+      "grad_norm": 0.672254585940336,
+      "learning_rate": 0.0015990000000000002,
+      "loss": 5.2165,
+      "step": 533
+    },
+    {
+      "epoch": 0.00534,
+      "grad_norm": 0.6180247915886188,
+      "learning_rate": 0.0016020000000000001,
+      "loss": 5.2018,
+      "step": 534
+    },
+    {
+      "epoch": 0.00535,
+      "grad_norm": 0.7076887136252737,
+      "learning_rate": 0.001605,
+      "loss": 5.2099,
+      "step": 535
+    },
+    {
+      "epoch": 0.00536,
+      "grad_norm": 0.8627381010586813,
+      "learning_rate": 0.001608,
+      "loss": 5.2158,
+      "step": 536
+    },
+    {
+      "epoch": 0.00537,
+      "grad_norm": 0.9890089503230703,
+      "learning_rate": 0.0016110000000000002,
+      "loss": 5.2125,
+      "step": 537
+    },
+    {
+      "epoch": 0.00538,
+      "grad_norm": 1.0890684861329838,
+      "learning_rate": 0.0016140000000000002,
+      "loss": 5.1997,
+      "step": 538
+    },
+    {
+      "epoch": 0.00539,
+      "grad_norm": 0.7898695514456295,
+      "learning_rate": 0.0016170000000000002,
+      "loss": 5.1885,
+      "step": 539
+    },
+    {
+      "epoch": 0.0054,
+      "grad_norm": 0.7877858658336557,
+      "learning_rate": 0.0016200000000000001,
+      "loss": 5.1979,
+      "step": 540
+    },
+    {
+      "epoch": 0.00541,
+      "grad_norm": 0.8613625325852945,
+      "learning_rate": 0.001623,
+      "loss": 5.1884,
+      "step": 541
+    },
+    {
+      "epoch": 0.00542,
+      "grad_norm": 0.9310959295325639,
+      "learning_rate": 0.001626,
+      "loss": 5.2135,
+      "step": 542
+    },
+    {
+      "epoch": 0.00543,
+      "grad_norm": 1.0650490028229627,
+      "learning_rate": 0.0016290000000000002,
+      "loss": 5.1821,
+      "step": 543
+    },
+    {
+      "epoch": 0.00544,
+      "grad_norm": 1.058155288535174,
+      "learning_rate": 0.0016320000000000002,
+      "loss": 5.2106,
+      "step": 544
+    },
+    {
+      "epoch": 0.00545,
+      "grad_norm": 0.7888748411290551,
+      "learning_rate": 0.0016350000000000002,
+      "loss": 5.2093,
+      "step": 545
+    },
+    {
+      "epoch": 0.00546,
+      "grad_norm": 0.8327623405474364,
+      "learning_rate": 0.0016380000000000001,
+      "loss": 5.1821,
+      "step": 546
+    },
+    {
+      "epoch": 0.00547,
+      "grad_norm": 0.9359889846839972,
+      "learning_rate": 0.001641,
+      "loss": 5.1867,
+      "step": 547
+    },
+    {
+      "epoch": 0.00548,
+      "grad_norm": 0.8250520922796094,
+      "learning_rate": 0.001644,
+      "loss": 5.1781,
+      "step": 548
+    },
+    {
+      "epoch": 0.00549,
+      "grad_norm": 0.8346002681162655,
+      "learning_rate": 0.0016470000000000002,
+      "loss": 5.1786,
+      "step": 549
+    },
+    {
+      "epoch": 0.0055,
+      "grad_norm": 0.8307775357404767,
+      "learning_rate": 0.0016500000000000002,
+      "loss": 5.1901,
+      "step": 550
+    },
+    {
+      "epoch": 0.00551,
+      "grad_norm": 0.8868587049067305,
+      "learning_rate": 0.0016530000000000002,
+      "loss": 5.1882,
+      "step": 551
+    },
+    {
+      "epoch": 0.00552,
+      "grad_norm": 1.0021721179116465,
+      "learning_rate": 0.0016560000000000001,
+      "loss": 5.1887,
+      "step": 552
+    },
+    {
+      "epoch": 0.00553,
+      "grad_norm": 1.0435640829012027,
+      "learning_rate": 0.001659,
+      "loss": 5.1828,
+      "step": 553
+    },
+    {
+      "epoch": 0.00554,
+      "grad_norm": 1.091952614784726,
+      "learning_rate": 0.0016620000000000003,
+      "loss": 5.1843,
+      "step": 554
+    },
+    {
+      "epoch": 0.00555,
+      "grad_norm": 0.7530351004511261,
+      "learning_rate": 0.0016650000000000002,
+      "loss": 5.1804,
+      "step": 555
+    },
+    {
+      "epoch": 0.00556,
+      "grad_norm": 0.7840106393058217,
+      "learning_rate": 0.0016680000000000002,
+      "loss": 5.1664,
+      "step": 556
+    },
+    {
+      "epoch": 0.00557,
+      "grad_norm": 0.6999816254444311,
+      "learning_rate": 0.0016710000000000002,
+      "loss": 5.1437,
+      "step": 557
+    },
+    {
+      "epoch": 0.00558,
+      "grad_norm": 0.7845980119871422,
+      "learning_rate": 0.0016740000000000001,
+      "loss": 5.1792,
+      "step": 558
+    },
+    {
+      "epoch": 0.00559,
+      "grad_norm": 1.002237738003299,
+      "learning_rate": 0.001677,
+      "loss": 5.165,
+      "step": 559
+    },
+    {
+      "epoch": 0.0056,
+      "grad_norm": 1.1590360602458978,
+      "learning_rate": 0.0016800000000000003,
+      "loss": 5.15,
+      "step": 560
+    },
+    {
+      "epoch": 0.00561,
+      "grad_norm": 0.9693350319936842,
+      "learning_rate": 0.0016830000000000003,
+      "loss": 5.1673,
+      "step": 561
+    },
+    {
+      "epoch": 0.00562,
+      "grad_norm": 0.9337806686381701,
+      "learning_rate": 0.0016860000000000002,
+      "loss": 5.1656,
+      "step": 562
+    },
+    {
+      "epoch": 0.00563,
+      "grad_norm": 0.7021371152666548,
+      "learning_rate": 0.001689,
+      "loss": 5.1554,
+      "step": 563
+    },
+    {
+      "epoch": 0.00564,
+      "grad_norm": 0.761762323285238,
+      "learning_rate": 0.001692,
+      "loss": 5.1364,
+      "step": 564
+    },
+    {
+      "epoch": 0.00565,
+      "grad_norm": 0.9962272803832337,
+      "learning_rate": 0.001695,
+      "loss": 5.1465,
+      "step": 565
+    },
+    {
+      "epoch": 0.00566,
+      "grad_norm": 0.9270663941212276,
+      "learning_rate": 0.0016979999999999999,
+      "loss": 5.1454,
+      "step": 566
+    },
+    {
+      "epoch": 0.00567,
+      "grad_norm": 1.0296751705367089,
+      "learning_rate": 0.0017009999999999998,
+      "loss": 5.1403,
+      "step": 567
+    },
+    {
+      "epoch": 0.00568,
+      "grad_norm": 1.0921308967550072,
+      "learning_rate": 0.0017039999999999998,
+      "loss": 5.1647,
+      "step": 568
+    },
+    {
+      "epoch": 0.00569,
+      "grad_norm": 0.9680118526613578,
+      "learning_rate": 0.001707,
+      "loss": 5.1376,
+      "step": 569
+    },
+    {
+      "epoch": 0.0057,
+      "grad_norm": 1.1373894554942883,
+      "learning_rate": 0.00171,
+      "loss": 5.1667,
+      "step": 570
+    },
+    {
+      "epoch": 0.00571,
+      "grad_norm": 0.9491793536820188,
+      "learning_rate": 0.001713,
+      "loss": 5.1618,
+      "step": 571
+    },
+    {
+      "epoch": 0.00572,
+      "grad_norm": 1.0938883440367575,
+      "learning_rate": 0.0017159999999999999,
+      "loss": 5.1499,
+      "step": 572
+    },
+    {
+      "epoch": 0.00573,
+      "grad_norm": 0.8953824806877001,
+      "learning_rate": 0.0017189999999999998,
+      "loss": 5.1459,
+      "step": 573
+    },
+    {
+      "epoch": 0.00574,
+      "grad_norm": 0.8777825611555061,
+      "learning_rate": 0.001722,
+      "loss": 5.1467,
+      "step": 574
+    },
+    {
+      "epoch": 0.00575,
+      "grad_norm": 0.9427806830230203,
+      "learning_rate": 0.001725,
+      "loss": 5.132,
+      "step": 575
+    },
+    {
+      "epoch": 0.00576,
+      "grad_norm": 0.8846520959631657,
+      "learning_rate": 0.001728,
+      "loss": 5.1269,
+      "step": 576
+    },
+    {
+      "epoch": 0.00577,
+      "grad_norm": 0.6344552018416748,
+      "learning_rate": 0.001731,
+      "loss": 5.1395,
+      "step": 577
+    },
+    {
+      "epoch": 0.00578,
+      "grad_norm": 0.6771922101340231,
+      "learning_rate": 0.0017339999999999999,
+      "loss": 5.1199,
+      "step": 578
+    },
+    {
+      "epoch": 0.00579,
+      "grad_norm": 0.5381676919488529,
+      "learning_rate": 0.0017369999999999998,
+      "loss": 5.1001,
+      "step": 579
+    },
+    {
+      "epoch": 0.0058,
+      "grad_norm": 0.5416857271346956,
+      "learning_rate": 0.00174,
+      "loss": 5.1179,
+      "step": 580
+    },
+    {
+      "epoch": 0.00581,
+      "grad_norm": 0.4886409554358658,
+      "learning_rate": 0.001743,
+      "loss": 5.1131,
+      "step": 581
+    },
+    {
+      "epoch": 0.00582,
+      "grad_norm": 0.3953796837190132,
+      "learning_rate": 0.001746,
+      "loss": 5.0842,
+      "step": 582
+    },
+    {
+      "epoch": 0.00583,
+      "grad_norm": 0.3816231764982968,
+      "learning_rate": 0.001749,
+      "loss": 5.1076,
+      "step": 583
+    },
+    {
+      "epoch": 0.00584,
+      "grad_norm": 0.39062356918599367,
+      "learning_rate": 0.0017519999999999999,
+      "loss": 5.1003,
+      "step": 584
+    },
+    {
+      "epoch": 0.00585,
+      "grad_norm": 0.3915858898605007,
+      "learning_rate": 0.0017549999999999998,
+      "loss": 5.0837,
+      "step": 585
+    },
+    {
+      "epoch": 0.00586,
+      "grad_norm": 0.4063811781565053,
+      "learning_rate": 0.001758,
+      "loss": 5.0866,
+      "step": 586
+    },
+    {
+      "epoch": 0.00587,
+      "grad_norm": 0.4274699383917251,
+      "learning_rate": 0.001761,
+      "loss": 5.0709,
+      "step": 587
+    },
+    {
+      "epoch": 0.00588,
+      "grad_norm": 0.5008029306674315,
+      "learning_rate": 0.001764,
+      "loss": 5.0767,
+      "step": 588
+    },
+    {
+      "epoch": 0.00589,
+      "grad_norm": 0.7229655060897183,
+      "learning_rate": 0.001767,
+      "loss": 5.084,
+      "step": 589
+    },
+    {
+      "epoch": 0.0059,
+      "grad_norm": 0.9281114393382421,
+      "learning_rate": 0.0017699999999999999,
+      "loss": 5.0779,
+      "step": 590
+    },
+    {
+      "epoch": 0.00591,
+      "grad_norm": 0.9689787780419555,
+      "learning_rate": 0.001773,
+      "loss": 5.0756,
+      "step": 591
+    },
+    {
+      "epoch": 0.00592,
+      "grad_norm": 0.8435580831880815,
+      "learning_rate": 0.001776,
+      "loss": 5.0975,
+      "step": 592
+    },
+    {
+      "epoch": 0.00593,
+      "grad_norm": 0.9986679613531879,
+      "learning_rate": 0.001779,
+      "loss": 5.098,
+      "step": 593
+    },
+    {
+      "epoch": 0.00594,
+      "grad_norm": 1.0996840470415932,
+      "learning_rate": 0.001782,
+      "loss": 5.1014,
+      "step": 594
+    },
+    {
+      "epoch": 0.00595,
+      "grad_norm": 0.9601369241773534,
+      "learning_rate": 0.001785,
+      "loss": 5.0957,
+      "step": 595
+    },
+    {
+      "epoch": 0.00596,
+      "grad_norm": 0.8606352953891906,
+      "learning_rate": 0.0017879999999999999,
+      "loss": 5.0932,
+      "step": 596
+    },
+    {
+      "epoch": 0.00597,
+      "grad_norm": 0.8716680293105395,
+      "learning_rate": 0.001791,
+      "loss": 5.066,
+      "step": 597
+    },
+    {
+      "epoch": 0.00598,
+      "grad_norm": 0.9954805607751368,
+      "learning_rate": 0.001794,
+      "loss": 5.0837,
+      "step": 598
+    },
+    {
+      "epoch": 0.00599,
+      "grad_norm": 1.0139742121604893,
+      "learning_rate": 0.001797,
+      "loss": 5.0892,
+      "step": 599
+    },
+    {
+      "epoch": 0.006,
+      "grad_norm": 0.8994983668637855,
+      "learning_rate": 0.0018,
+      "loss": 5.083,
+      "step": 600
+    },
+    {
+      "epoch": 0.00601,
+      "grad_norm": 1.012817382509115,
+      "learning_rate": 0.001803,
+      "loss": 5.0872,
+      "step": 601
+    },
+    {
+      "epoch": 0.00602,
+      "grad_norm": 0.829792766990954,
+      "learning_rate": 0.0018059999999999999,
+      "loss": 5.0655,
+      "step": 602
+    },
+    {
+      "epoch": 0.00603,
+      "grad_norm": 0.8978641621726422,
+      "learning_rate": 0.001809,
+      "loss": 5.0896,
+      "step": 603
+    },
+    {
+      "epoch": 0.00604,
+      "grad_norm": 1.006526980400353,
+      "learning_rate": 0.001812,
+      "loss": 5.0733,
+      "step": 604
+    },
+    {
+      "epoch": 0.00605,
+      "grad_norm": 0.9664384705752951,
+      "learning_rate": 0.001815,
+      "loss": 5.0828,
+      "step": 605
+    },
+    {
+      "epoch": 0.00606,
+      "grad_norm": 1.0962791606856168,
+      "learning_rate": 0.001818,
+      "loss": 5.0888,
+      "step": 606
+    },
+    {
+      "epoch": 0.00607,
+      "grad_norm": 0.9313257227556363,
+      "learning_rate": 0.001821,
+      "loss": 5.0876,
+      "step": 607
+    },
+    {
+      "epoch": 0.00608,
+      "grad_norm": 0.9207616431206566,
+      "learning_rate": 0.001824,
+      "loss": 5.068,
+      "step": 608
+    },
+    {
+      "epoch": 0.00609,
+      "grad_norm": 0.9552257534317795,
+      "learning_rate": 0.001827,
+      "loss": 5.0751,
+      "step": 609
+    },
+    {
+      "epoch": 0.0061,
+      "grad_norm": 1.1338881088580717,
+      "learning_rate": 0.00183,
+      "loss": 5.062,
+      "step": 610
+    },
+    {
+      "epoch": 0.00611,
+      "grad_norm": 0.907481169345242,
+      "learning_rate": 0.001833,
+      "loss": 5.0576,
+      "step": 611
+    },
+    {
+      "epoch": 0.00612,
+      "grad_norm": 1.04757168088542,
+      "learning_rate": 0.001836,
+      "loss": 5.0686,
+      "step": 612
+    },
+    {
+      "epoch": 0.00613,
+      "grad_norm": 0.9021638869008188,
+      "learning_rate": 0.001839,
+      "loss": 5.0559,
+      "step": 613
+    },
+    {
+      "epoch": 0.00614,
+      "grad_norm": 0.7877209937196055,
+      "learning_rate": 0.001842,
+      "loss": 5.0768,
+      "step": 614
+    },
+    {
+      "epoch": 0.00615,
+      "grad_norm": 0.7456491069500266,
+      "learning_rate": 0.001845,
+      "loss": 5.0572,
+      "step": 615
+    },
+    {
+      "epoch": 0.00616,
+      "grad_norm": 0.7643816652567319,
+      "learning_rate": 0.001848,
+      "loss": 5.043,
+      "step": 616
+    },
+    {
+      "epoch": 0.00617,
+      "grad_norm": 0.8920948602141958,
+      "learning_rate": 0.001851,
+      "loss": 5.0555,
+      "step": 617
+    },
+    {
+      "epoch": 0.00618,
+      "grad_norm": 1.0257178323795717,
+      "learning_rate": 0.001854,
+      "loss": 5.0507,
+      "step": 618
+    },
+    {
+      "epoch": 0.00619,
+      "grad_norm": 0.7092119630036949,
+      "learning_rate": 0.001857,
+      "loss": 5.0429,
+      "step": 619
+    },
+    {
+      "epoch": 0.0062,
+      "grad_norm": 0.6092469798525071,
+      "learning_rate": 0.00186,
+      "loss": 5.0262,
+      "step": 620
+    },
+    {
+      "epoch": 0.00621,
+      "grad_norm": 0.7552706683211996,
+      "learning_rate": 0.001863,
+      "loss": 5.032,
+      "step": 621
+    },
+    {
+      "epoch": 0.00622,
+      "grad_norm": 0.6988650940233329,
+      "learning_rate": 0.001866,
+      "loss": 5.0245,
+      "step": 622
+    },
+    {
+      "epoch": 0.00623,
+      "grad_norm": 0.6062593525325608,
+      "learning_rate": 0.001869,
+      "loss": 5.0091,
+      "step": 623
+    },
+    {
+      "epoch": 0.00624,
+      "grad_norm": 0.555479795278756,
+      "learning_rate": 0.001872,
+      "loss": 5.0161,
+      "step": 624
+    },
+    {
+      "epoch": 0.00625,
+      "grad_norm": 0.6339519418385815,
+      "learning_rate": 0.001875,
+      "loss": 5.0276,
+      "step": 625
+    },
+    {
+      "epoch": 0.00626,
+      "grad_norm": 0.7820017260467441,
+      "learning_rate": 0.0018780000000000001,
+      "loss": 5.0084,
+      "step": 626
+    },
+    {
+      "epoch": 0.00627,
+      "grad_norm": 0.8263278266088275,
+      "learning_rate": 0.001881,
+      "loss": 5.0003,
+      "step": 627
+    },
+    {
+      "epoch": 0.00628,
+      "grad_norm": 0.5717634931424201,
+      "learning_rate": 0.001884,
+      "loss": 5.0204,
+      "step": 628
+    },
+    {
+      "epoch": 0.00629,
+      "grad_norm": 0.5505525631139665,
+      "learning_rate": 0.001887,
+      "loss": 4.9928,
+      "step": 629
+    },
+    {
+      "epoch": 0.0063,
+      "grad_norm": 0.47331922258372455,
+      "learning_rate": 0.00189,
+      "loss": 4.9837,
+      "step": 630
+    },
+    {
+      "epoch": 0.00631,
+      "grad_norm": 0.4173073508747504,
+      "learning_rate": 0.0018930000000000002,
+      "loss": 4.9757,
+      "step": 631
+    },
+    {
+      "epoch": 0.00632,
+      "grad_norm": 0.42808966750972455,
+      "learning_rate": 0.0018960000000000001,
+      "loss": 4.9928,
+      "step": 632
+    },
+    {
+      "epoch": 0.00633,
+      "grad_norm": 0.46750467774391197,
+      "learning_rate": 0.001899,
+      "loss": 4.9879,
+      "step": 633
+    },
+    {
+      "epoch": 0.00634,
+      "grad_norm": 0.5696215613123979,
+      "learning_rate": 0.001902,
+      "loss": 4.9763,
+      "step": 634
+    },
+    {
+      "epoch": 0.00635,
+      "grad_norm": 0.7138673203190975,
+      "learning_rate": 0.001905,
+      "loss": 4.9686,
+      "step": 635
+    },
+    {
+      "epoch": 0.00636,
+      "grad_norm": 0.8836310721952346,
+      "learning_rate": 0.001908,
+      "loss": 4.9828,
+      "step": 636
+    },
+    {
+      "epoch": 0.00637,
+      "grad_norm": 0.9755328249694639,
+      "learning_rate": 0.0019110000000000002,
+      "loss": 5.002,
+      "step": 637
+    },
+    {
+      "epoch": 0.00638,
+      "grad_norm": 1.3272264449165134,
+      "learning_rate": 0.0019140000000000001,
+      "loss": 5.0051,
+      "step": 638
+    },
+    {
+      "epoch": 0.00639,
+      "grad_norm": 0.9140856571338883,
+      "learning_rate": 0.001917,
+      "loss": 4.9726,
+      "step": 639
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 1.0466736994412218,
+      "learning_rate": 0.00192,
+      "loss": 4.9869,
+      "step": 640
+    },
+    {
+      "epoch": 0.00641,
+      "grad_norm": 1.1161814084223103,
+      "learning_rate": 0.001923,
+      "loss": 5.0003,
+      "step": 641
+    },
+    {
+      "epoch": 0.00642,
+      "grad_norm": 1.1352153221992676,
+      "learning_rate": 0.001926,
+      "loss": 4.992,
+      "step": 642
+    },
+    {
+      "epoch": 0.00643,
+      "grad_norm": 1.2268384891507862,
+      "learning_rate": 0.0019290000000000002,
+      "loss": 5.008,
+      "step": 643
+    },
+    {
+      "epoch": 0.00644,
+      "grad_norm": 0.9564122465750423,
+      "learning_rate": 0.0019320000000000001,
+      "loss": 4.9857,
+      "step": 644
+    },
+    {
+      "epoch": 0.00645,
+      "grad_norm": 0.9066234784688915,
+      "learning_rate": 0.001935,
+      "loss": 4.9889,
+      "step": 645
+    },
+    {
+      "epoch": 0.00646,
+      "grad_norm": 0.894776309426942,
+      "learning_rate": 0.001938,
+      "loss": 4.995,
+      "step": 646
+    },
+    {
+      "epoch": 0.00647,
+      "grad_norm": 1.036514260058091,
+      "learning_rate": 0.001941,
+      "loss": 5.0081,
+      "step": 647
+    },
+    {
+      "epoch": 0.00648,
+      "grad_norm": 1.0669688261896326,
+      "learning_rate": 0.0019440000000000002,
+      "loss": 4.9991,
+      "step": 648
+    },
+    {
+      "epoch": 0.00649,
+      "grad_norm": 1.0027602192299327,
+      "learning_rate": 0.0019470000000000002,
+      "loss": 4.9834,
+      "step": 649
+    },
+    {
+      "epoch": 0.0065,
+      "grad_norm": 1.2453243505592877,
+      "learning_rate": 0.0019500000000000001,
+      "loss": 5.0183,
+      "step": 650
+    },
+    {
+      "epoch": 0.00651,
+      "grad_norm": 0.9810465710779535,
+      "learning_rate": 0.001953,
+      "loss": 4.9899,
+      "step": 651
+    },
+    {
+      "epoch": 0.00652,
+      "grad_norm": 1.0529607937914427,
+      "learning_rate": 0.0019560000000000003,
+      "loss": 4.9855,
+      "step": 652
+    },
+    {
+      "epoch": 0.00653,
+      "grad_norm": 0.8367844014470341,
+      "learning_rate": 0.0019590000000000002,
+      "loss": 4.9719,
+      "step": 653
+    },
+    {
+      "epoch": 0.00654,
+      "grad_norm": 0.9556352679930769,
+      "learning_rate": 0.001962,
+      "loss": 4.96,
+      "step": 654
+    },
+    {
+      "epoch": 0.00655,
+      "grad_norm": 1.0712859822115042,
+      "learning_rate": 0.001965,
+      "loss": 4.9811,
+      "step": 655
+    },
+    {
+      "epoch": 0.00656,
+      "grad_norm": 0.8381525542412449,
+      "learning_rate": 0.001968,
+      "loss": 4.9628,
+      "step": 656
+    },
+    {
+      "epoch": 0.00657,
+      "grad_norm": 0.853522104902103,
+      "learning_rate": 0.001971,
+      "loss": 4.9486,
+      "step": 657
+    },
+    {
+      "epoch": 0.00658,
+      "grad_norm": 0.9276507218825019,
+      "learning_rate": 0.001974,
+      "loss": 4.9746,
+      "step": 658
+    },
+    {
+      "epoch": 0.00659,
+      "grad_norm": 1.0041122671115763,
+      "learning_rate": 0.001977,
+      "loss": 4.9552,
+      "step": 659
+    },
+    {
+      "epoch": 0.0066,
+      "grad_norm": 0.925609681342879,
+      "learning_rate": 0.00198,
+      "loss": 4.9516,
+      "step": 660
+    },
+    {
+      "epoch": 0.00661,
+      "grad_norm": 0.9340897694356347,
+      "learning_rate": 0.001983,
+      "loss": 4.9301,
+      "step": 661
+    },
+    {
+      "epoch": 0.00662,
+      "grad_norm": 1.1039924911609775,
+      "learning_rate": 0.0019860000000000004,
+      "loss": 4.9353,
+      "step": 662
+    },
+    {
+      "epoch": 0.00663,
+      "grad_norm": 0.842271096957419,
+      "learning_rate": 0.0019890000000000003,
+      "loss": 4.942,
+      "step": 663
+    },
+    {
+      "epoch": 0.00664,
+      "grad_norm": 0.9255021185692508,
+      "learning_rate": 0.0019920000000000003,
+      "loss": 4.9294,
+      "step": 664
+    },
+    {
+      "epoch": 0.00665,
+      "grad_norm": 0.9624840652606003,
+      "learning_rate": 0.0019950000000000002,
+      "loss": 4.9033,
+      "step": 665
+    },
+    {
+      "epoch": 0.00666,
+      "grad_norm": 0.8706912476713129,
+      "learning_rate": 0.001998,
+      "loss": 4.9275,
+      "step": 666
+    },
+    {
+      "epoch": 0.00667,
+      "grad_norm": 0.9355886153293348,
+      "learning_rate": 0.002001,
+      "loss": 4.9186,
+      "step": 667
+    },
+    {
+      "epoch": 0.00668,
+      "grad_norm": 0.9525697321085718,
+      "learning_rate": 0.002004,
+      "loss": 4.9422,
+      "step": 668
+    },
+    {
+      "epoch": 0.00669,
+      "grad_norm": 0.9357519892866003,
+      "learning_rate": 0.002007,
+      "loss": 4.9423,
+      "step": 669
+    },
+    {
+      "epoch": 0.0067,
+      "grad_norm": 0.8227655724451596,
+      "learning_rate": 0.00201,
+      "loss": 4.9287,
+      "step": 670
+    },
+    {
+      "epoch": 0.00671,
+      "grad_norm": 0.6363703099565691,
+      "learning_rate": 0.002013,
+      "loss": 4.8975,
+      "step": 671
+    },
+    {
+      "epoch": 0.00672,
+      "grad_norm": 0.5774852286623988,
+      "learning_rate": 0.002016,
+      "loss": 4.894,
+      "step": 672
+    },
+    {
+      "epoch": 0.00673,
+      "grad_norm": 0.5472152449551199,
+      "learning_rate": 0.002019,
+      "loss": 4.8881,
+      "step": 673
+    },
+    {
+      "epoch": 0.00674,
+      "grad_norm": 0.5267919584843693,
+      "learning_rate": 0.0020220000000000004,
+      "loss": 4.8857,
+      "step": 674
+    },
+    {
+      "epoch": 0.00675,
+      "grad_norm": 0.4823533534454447,
+      "learning_rate": 0.0020250000000000003,
+      "loss": 4.8922,
+      "step": 675
+    },
+    {
+      "epoch": 0.00676,
+      "grad_norm": 0.5135326218761617,
+      "learning_rate": 0.0020280000000000003,
+      "loss": 4.8739,
+      "step": 676
+    },
+    {
+      "epoch": 0.00677,
+      "grad_norm": 0.5629070673817519,
+      "learning_rate": 0.0020310000000000003,
+      "loss": 4.8753,
+      "step": 677
+    },
+    {
+      "epoch": 0.00678,
+      "grad_norm": 0.6636518437686225,
+      "learning_rate": 0.0020340000000000002,
+      "loss": 4.8844,
+      "step": 678
+    },
+    {
+      "epoch": 0.00679,
+      "grad_norm": 0.7939150938071041,
+      "learning_rate": 0.002037,
+      "loss": 4.8694,
+      "step": 679
+    },
+    {
+      "epoch": 0.0068,
+      "grad_norm": 0.6961050241971515,
+      "learning_rate": 0.00204,
+      "loss": 4.8681,
+      "step": 680
+    },
+    {
+      "epoch": 0.00681,
+      "grad_norm": 0.55281231514776,
+      "learning_rate": 0.002043,
+      "loss": 4.854,
+      "step": 681
+    },
+    {
+      "epoch": 0.00682,
+      "grad_norm": 0.6966375314025457,
+      "learning_rate": 0.002046,
+      "loss": 4.8606,
+      "step": 682
+    },
+    {
+      "epoch": 0.00683,
+      "grad_norm": 0.7919918537287608,
+      "learning_rate": 0.002049,
+      "loss": 4.8437,
+      "step": 683
+    },
+    {
+      "epoch": 0.00684,
+      "grad_norm": 1.014505658611748,
+      "learning_rate": 0.002052,
+      "loss": 4.868,
+      "step": 684
+    },
+    {
+      "epoch": 0.00685,
+      "grad_norm": 1.1979460789898289,
+      "learning_rate": 0.0020550000000000004,
+      "loss": 4.8824,
+      "step": 685
+    },
+    {
+      "epoch": 0.00686,
+      "grad_norm": 1.0209197434565493,
+      "learning_rate": 0.0020580000000000004,
+      "loss": 4.8555,
+      "step": 686
+    },
+    {
+      "epoch": 0.00687,
+      "grad_norm": 0.8666785819902827,
+      "learning_rate": 0.0020610000000000003,
+      "loss": 4.8457,
+      "step": 687
+    },
+    {
+      "epoch": 0.00688,
+      "grad_norm": 1.066277682185938,
+      "learning_rate": 0.002064,
+      "loss": 4.8576,
+      "step": 688
+    },
+    {
+      "epoch": 0.00689,
+      "grad_norm": 0.9564373481813865,
+      "learning_rate": 0.002067,
+      "loss": 4.8511,
+      "step": 689
+    },
+    {
+      "epoch": 0.0069,
+      "grad_norm": 1.0921039434762347,
+      "learning_rate": 0.00207,
+      "loss": 4.854,
+      "step": 690
+    },
+    {
+      "epoch": 0.00691,
+      "grad_norm": 0.9682767243250711,
+      "learning_rate": 0.0020729999999999998,
+      "loss": 4.8265,
+      "step": 691
+    },
+    {
+      "epoch": 0.00692,
+      "grad_norm": 1.0182149348151321,
+      "learning_rate": 0.0020759999999999997,
+      "loss": 4.8523,
+      "step": 692
+    },
+    {
+      "epoch": 0.00693,
+      "grad_norm": 1.1559582715656997,
+      "learning_rate": 0.0020789999999999997,
+      "loss": 4.8626,
+      "step": 693
+    },
+    {
+      "epoch": 0.00694,
+      "grad_norm": 0.8731253854999043,
+      "learning_rate": 0.002082,
+      "loss": 4.8314,
+      "step": 694
+    },
+    {
+      "epoch": 0.00695,
+      "grad_norm": 1.02924772696324,
+      "learning_rate": 0.002085,
+      "loss": 4.8388,
+      "step": 695
+    },
+    {
+      "epoch": 0.00696,
+      "grad_norm": 1.1148665982097032,
+      "learning_rate": 0.002088,
+      "loss": 4.8532,
+      "step": 696
+    },
+    {
+      "epoch": 0.00697,
+      "grad_norm": 1.1306884538795905,
+      "learning_rate": 0.002091,
+      "loss": 4.8556,
+      "step": 697
+    },
+    {
+      "epoch": 0.00698,
+      "grad_norm": 1.267565149046305,
+      "learning_rate": 0.002094,
+      "loss": 4.8474,
+      "step": 698
+    },
+    {
+      "epoch": 0.00699,
+      "grad_norm": 0.9067921371830103,
+      "learning_rate": 0.002097,
+      "loss": 4.8312,
+      "step": 699
+    },
+    {
+      "epoch": 0.007,
+      "grad_norm": 0.8797095491461238,
+      "learning_rate": 0.0021,
+      "loss": 4.8378,
+      "step": 700
+    },
+    {
+      "epoch": 0.00701,
+      "grad_norm": 0.9764684557107473,
+      "learning_rate": 0.002103,
+      "loss": 4.8231,
+      "step": 701
+    },
+    {
+      "epoch": 0.00702,
+      "grad_norm": 0.9975561372265659,
+      "learning_rate": 0.002106,
+      "loss": 4.8269,
+      "step": 702
+    },
+    {
+      "epoch": 0.00703,
+      "grad_norm": 0.8409706105195134,
+      "learning_rate": 0.0021089999999999998,
+      "loss": 4.8046,
+      "step": 703
+    },
+    {
+      "epoch": 0.00704,
+      "grad_norm": 0.8602349489288845,
+      "learning_rate": 0.0021119999999999997,
+      "loss": 4.8112,
+      "step": 704
+    },
+    {
+      "epoch": 0.00705,
+      "grad_norm": 0.810283974291816,
+      "learning_rate": 0.002115,
+      "loss": 4.8189,
+      "step": 705
+    },
+    {
+      "epoch": 0.00706,
+      "grad_norm": 0.7904979262360543,
+      "learning_rate": 0.002118,
+      "loss": 4.8127,
+      "step": 706
+    },
+    {
+      "epoch": 0.00707,
+      "grad_norm": 0.8638007527609218,
+      "learning_rate": 0.002121,
+      "loss": 4.7902,
+      "step": 707
+    },
+    {
+      "epoch": 0.00708,
+      "grad_norm": 0.9274642295068019,
+      "learning_rate": 0.002124,
+      "loss": 4.7754,
+      "step": 708
+    },
+    {
+      "epoch": 0.00709,
+      "grad_norm": 0.9265048264631317,
+      "learning_rate": 0.002127,
+      "loss": 4.8051,
+      "step": 709
+    },
+    {
+      "epoch": 0.0071,
+      "grad_norm": 1.0021235226233682,
+      "learning_rate": 0.00213,
+      "loss": 4.8021,
+      "step": 710
+    },
+    {
+      "epoch": 0.00711,
+      "grad_norm": 0.8616225643918476,
+      "learning_rate": 0.002133,
+      "loss": 4.7687,
+      "step": 711
+    },
+    {
+      "epoch": 0.00712,
+      "grad_norm": 1.1509118831082872,
+      "learning_rate": 0.002136,
+      "loss": 4.8063,
+      "step": 712
+    },
+    {
+      "epoch": 0.00713,
+      "grad_norm": 0.8979386810595619,
+      "learning_rate": 0.002139,
+      "loss": 4.8085,
+      "step": 713
+    },
+    {
+      "epoch": 0.00714,
+      "grad_norm": 0.9672478380991173,
+      "learning_rate": 0.002142,
+      "loss": 4.7999,
+      "step": 714
+    },
+    {
+      "epoch": 0.00715,
+      "grad_norm": 1.0283238664767786,
+      "learning_rate": 0.0021449999999999998,
+      "loss": 4.7746,
+      "step": 715
+    },
+    {
+      "epoch": 0.00716,
+      "grad_norm": 0.8031836486660412,
+      "learning_rate": 0.002148,
+      "loss": 4.7611,
+      "step": 716
+    },
+    {
+      "epoch": 0.00717,
+      "grad_norm": 0.8079202930068127,
+      "learning_rate": 0.002151,
+      "loss": 4.7802,
+      "step": 717
+    },
+    {
+      "epoch": 0.00718,
+      "grad_norm": 0.7625878769693025,
+      "learning_rate": 0.002154,
+      "loss": 4.7607,
+      "step": 718
+    },
+    {
+      "epoch": 0.00719,
+      "grad_norm": 0.8833410128202536,
+      "learning_rate": 0.002157,
+      "loss": 4.7767,
+      "step": 719
+    },
+    {
+      "epoch": 0.0072,
+      "grad_norm": 1.0099718860880083,
+      "learning_rate": 0.00216,
+      "loss": 4.7733,
+      "step": 720
+    },
+    {
+      "epoch": 0.00721,
+      "grad_norm": 1.0037155368349988,
+      "learning_rate": 0.002163,
+      "loss": 4.78,
+      "step": 721
+    },
+    {
+      "epoch": 0.00722,
+      "grad_norm": 0.8678480066188063,
+      "learning_rate": 0.002166,
+      "loss": 4.7474,
+      "step": 722
+    },
+    {
+      "epoch": 0.00723,
+      "grad_norm": 0.9399210002572385,
+      "learning_rate": 0.002169,
+      "loss": 4.7657,
+      "step": 723
+    },
+    {
+      "epoch": 0.00724,
+      "grad_norm": 0.8184829699796181,
+      "learning_rate": 0.002172,
+      "loss": 4.7533,
+      "step": 724
+    },
+    {
+      "epoch": 0.00725,
+      "grad_norm": 0.8676839571587074,
+      "learning_rate": 0.002175,
+      "loss": 4.7513,
+      "step": 725
+    },
+    {
+      "epoch": 0.00726,
+      "grad_norm": 0.9799992988904748,
+      "learning_rate": 0.002178,
+      "loss": 4.7626,
+      "step": 726
+    },
+    {
+      "epoch": 0.00727,
+      "grad_norm": 1.2281779129682024,
+      "learning_rate": 0.0021809999999999998,
+      "loss": 4.7581,
+      "step": 727
+    },
+    {
+      "epoch": 0.00728,
+      "grad_norm": 1.082945747060172,
+      "learning_rate": 0.002184,
+      "loss": 4.7657,
+      "step": 728
+    },
+    {
+      "epoch": 0.00729,
+      "grad_norm": 1.0915510364818644,
+      "learning_rate": 0.002187,
+      "loss": 4.7617,
+      "step": 729
+    },
+    {
+      "epoch": 0.0073,
+      "grad_norm": 1.0738468909531949,
+      "learning_rate": 0.00219,
+      "loss": 4.7676,
+      "step": 730
+    },
+    {
+      "epoch": 0.00731,
+      "grad_norm": 1.0774407965183543,
+      "learning_rate": 0.002193,
+      "loss": 4.7572,
+      "step": 731
+    },
+    {
+      "epoch": 0.00732,
+      "grad_norm": 0.9732910355796593,
+      "learning_rate": 0.002196,
+      "loss": 4.7638,
+      "step": 732
+    },
+    {
+      "epoch": 0.00733,
+      "grad_norm": 1.02567545008427,
+      "learning_rate": 0.002199,
+      "loss": 4.7467,
+      "step": 733
+    },
+    {
+      "epoch": 0.00734,
+      "grad_norm": 0.9058765241181546,
+      "learning_rate": 0.002202,
+      "loss": 4.7686,
+      "step": 734
+    },
+    {
+      "epoch": 0.00735,
+      "grad_norm": 0.9682392169542167,
+      "learning_rate": 0.002205,
+      "loss": 4.7759,
+      "step": 735
+    },
+    {
+      "epoch": 0.00736,
+      "grad_norm": 0.9938811207200824,
+      "learning_rate": 0.002208,
+      "loss": 4.7615,
+      "step": 736
+    },
+    {
+      "epoch": 0.00737,
+      "grad_norm": 1.0964297344539389,
+      "learning_rate": 0.002211,
+      "loss": 4.7347,
+      "step": 737
+    },
+    {
+      "epoch": 0.00738,
+      "grad_norm": 0.8707349769325928,
+      "learning_rate": 0.002214,
+      "loss": 4.7342,
+      "step": 738
+    },
+    {
+      "epoch": 0.00739,
+      "grad_norm": 0.7399818233744658,
+      "learning_rate": 0.0022170000000000002,
+      "loss": 4.717,
+      "step": 739
+    },
+    {
+      "epoch": 0.0074,
+      "grad_norm": 0.9377539124718652,
+      "learning_rate": 0.00222,
+      "loss": 4.7301,
+      "step": 740
+    },
+    {
+      "epoch": 0.00741,
+      "grad_norm": 0.8596400455739317,
+      "learning_rate": 0.002223,
+      "loss": 4.699,
+      "step": 741
+    },
+    {
+      "epoch": 0.00742,
+      "grad_norm": 0.6578432901740889,
+      "learning_rate": 0.002226,
+      "loss": 4.7249,
+      "step": 742
+    },
+    {
+      "epoch": 0.00743,
+      "grad_norm": 0.6175884285032084,
+      "learning_rate": 0.002229,
+      "loss": 4.6843,
+      "step": 743
+    },
+    {
+      "epoch": 0.00744,
+      "grad_norm": 0.6615203369086972,
+      "learning_rate": 0.002232,
+      "loss": 4.6918,
+      "step": 744
+    },
+    {
+      "epoch": 0.00745,
+      "grad_norm": 0.6999197355703424,
+      "learning_rate": 0.002235,
+      "loss": 4.7005,
+      "step": 745
+    },
+    {
+      "epoch": 0.00746,
+      "grad_norm": 0.7056349857734648,
+      "learning_rate": 0.002238,
+      "loss": 4.6964,
+      "step": 746
+    },
+    {
+      "epoch": 0.00747,
+      "grad_norm": 0.6678161149510893,
+      "learning_rate": 0.002241,
+      "loss": 4.6817,
+      "step": 747
+    },
+    {
+      "epoch": 0.00748,
+      "grad_norm": 0.6725287147155753,
+      "learning_rate": 0.002244,
+      "loss": 4.6915,
+      "step": 748
+    },
+    {
+      "epoch": 0.00749,
+      "grad_norm": 0.7368138262221237,
+      "learning_rate": 0.002247,
+      "loss": 4.6725,
+      "step": 749
+    },
+    {
+      "epoch": 0.0075,
+      "grad_norm": 0.7550252977049275,
+      "learning_rate": 0.0022500000000000003,
+      "loss": 4.7124,
+      "step": 750
+    },
+    {
+      "epoch": 0.00751,
+      "grad_norm": 0.6461697616177359,
+      "learning_rate": 0.0022530000000000002,
+      "loss": 4.6948,
+      "step": 751
+    },
+    {
+      "epoch": 0.00752,
+      "grad_norm": 0.6473605328229959,
+      "learning_rate": 0.002256,
+      "loss": 4.648,
+      "step": 752
+    },
+    {
+      "epoch": 0.00753,
+      "grad_norm": 0.8172272904356894,
+      "learning_rate": 0.002259,
+      "loss": 4.6929,
+      "step": 753
+    },
+    {
+      "epoch": 0.00754,
+      "grad_norm": 0.8400684755887758,
+      "learning_rate": 0.002262,
+      "loss": 4.7068,
+      "step": 754
+    },
+    {
+      "epoch": 0.00755,
+      "grad_norm": 0.7078832518794317,
+      "learning_rate": 0.002265,
+      "loss": 4.6656,
+      "step": 755
+    },
+    {
+      "epoch": 0.00756,
+      "grad_norm": 0.6076060117836831,
+      "learning_rate": 0.002268,
+      "loss": 4.6484,
+      "step": 756
+    },
+    {
+      "epoch": 0.00757,
+      "grad_norm": 0.7133093934008413,
+      "learning_rate": 0.002271,
+      "loss": 4.6658,
+      "step": 757
+    },
+    {
+      "epoch": 0.00758,
+      "grad_norm": 0.7661771348142844,
+      "learning_rate": 0.002274,
+      "loss": 4.6521,
+      "step": 758
+    },
+    {
+      "epoch": 0.00759,
+      "grad_norm": 0.9250288948777622,
+      "learning_rate": 0.002277,
+      "loss": 4.6753,
+      "step": 759
+    },
+    {
+      "epoch": 0.0076,
+      "grad_norm": 1.037174236565274,
+      "learning_rate": 0.00228,
+      "loss": 4.669,
+      "step": 760
+    },
+    {
+      "epoch": 0.00761,
+      "grad_norm": 0.9678315157211191,
+      "learning_rate": 0.002283,
+      "loss": 4.6392,
+      "step": 761
+    },
+    {
+      "epoch": 0.00762,
+      "grad_norm": 1.3728001530688312,
+      "learning_rate": 0.0022860000000000003,
+      "loss": 4.6453,
+      "step": 762
+    },
+    {
+      "epoch": 0.00763,
+      "grad_norm": 1.0284727877786697,
+      "learning_rate": 0.0022890000000000002,
+      "loss": 4.6793,
+      "step": 763
+    },
+    {
+      "epoch": 0.00764,
+      "grad_norm": 0.9914794664489192,
+      "learning_rate": 0.002292,
+      "loss": 4.6942,
+      "step": 764
+    },
+    {
+      "epoch": 0.00765,
+      "grad_norm": 0.984322504117537,
+      "learning_rate": 0.002295,
+      "loss": 4.6765,
+      "step": 765
+    },
+    {
+      "epoch": 0.00766,
+      "grad_norm": 0.9320893698991433,
+      "learning_rate": 0.002298,
+      "loss": 4.6792,
+      "step": 766
+    },
+    {
+      "epoch": 0.00767,
+      "grad_norm": 1.0651442494276249,
+      "learning_rate": 0.002301,
+      "loss": 4.6823,
+      "step": 767
+    },
+    {
+      "epoch": 0.00768,
+      "grad_norm": 0.9179111624711317,
+      "learning_rate": 0.002304,
+      "loss": 4.6817,
+      "step": 768
+    },
+    {
+      "epoch": 0.00769,
+      "grad_norm": 1.026483766110404,
+      "learning_rate": 0.002307,
+      "loss": 4.6987,
+      "step": 769
+    },
+    {
+      "epoch": 0.0077,
+      "grad_norm": 1.1653119571960542,
+      "learning_rate": 0.00231,
+      "loss": 4.6621,
+      "step": 770
+    },
+    {
+      "epoch": 0.00771,
+      "grad_norm": 0.8477238808348645,
+      "learning_rate": 0.002313,
+      "loss": 4.689,
+      "step": 771
+    },
+    {
+      "epoch": 0.00772,
+      "grad_norm": 0.6519421622488206,
+      "learning_rate": 0.002316,
+      "loss": 4.6631,
+      "step": 772
+    },
+    {
+      "epoch": 0.00773,
+      "grad_norm": 0.6177861857364649,
+      "learning_rate": 0.0023190000000000003,
+      "loss": 4.6627,
+      "step": 773
+    },
+    {
+      "epoch": 0.00774,
+      "grad_norm": 0.6901665734497584,
+      "learning_rate": 0.0023220000000000003,
+      "loss": 4.6775,
+      "step": 774
+    },
+    {
+      "epoch": 0.00775,
+      "grad_norm": 0.7356087164350898,
+      "learning_rate": 0.0023250000000000002,
+      "loss": 4.6725,
+      "step": 775
+    },
+    {
+      "epoch": 0.00776,
+      "grad_norm": 0.8693432194982287,
+      "learning_rate": 0.002328,
+      "loss": 4.6722,
+      "step": 776
+    },
+    {
+      "epoch": 0.00777,
+      "grad_norm": 1.099570573598906,
+      "learning_rate": 0.002331,
+      "loss": 4.6596,
+      "step": 777
+    },
+    {
+      "epoch": 0.00778,
+      "grad_norm": 1.0694357671416344,
+      "learning_rate": 0.002334,
+      "loss": 4.6725,
+      "step": 778
+    },
+    {
+      "epoch": 0.00779,
+      "grad_norm": 1.0144407661707453,
+      "learning_rate": 0.002337,
+      "loss": 4.6345,
+      "step": 779
+    },
+    {
+      "epoch": 0.0078,
+      "grad_norm": 0.9392788681741788,
+      "learning_rate": 0.00234,
+      "loss": 4.6579,
+      "step": 780
+    },
+    {
+      "epoch": 0.00781,
+      "grad_norm": 0.9039044158767507,
+      "learning_rate": 0.002343,
+      "loss": 4.6528,
+      "step": 781
+    },
+    {
+      "epoch": 0.00782,
+      "grad_norm": 0.9671545635863801,
+      "learning_rate": 0.002346,
+      "loss": 4.6552,
+      "step": 782
+    },
+    {
+      "epoch": 0.00783,
+      "grad_norm": 0.9789758043675277,
+      "learning_rate": 0.002349,
+      "loss": 4.6404,
+      "step": 783
+    },
+    {
+      "epoch": 0.00784,
+      "grad_norm": 0.7674708275760124,
+      "learning_rate": 0.002352,
+      "loss": 4.6449,
+      "step": 784
+    },
+    {
+      "epoch": 0.00785,
+      "grad_norm": 0.6509064102772842,
+      "learning_rate": 0.0023550000000000003,
+      "loss": 4.6391,
+      "step": 785
+    },
+    {
+      "epoch": 0.00786,
+      "grad_norm": 0.6282839162170526,
+      "learning_rate": 0.0023580000000000003,
+      "loss": 4.6356,
+      "step": 786
+    },
+    {
+      "epoch": 0.00787,
+      "grad_norm": 0.6520823986249177,
+      "learning_rate": 0.0023610000000000003,
+      "loss": 4.6384,
+      "step": 787
+    },
+    {
+      "epoch": 0.00788,
+      "grad_norm": 0.6693545084259133,
+      "learning_rate": 0.002364,
+      "loss": 4.6342,
+      "step": 788
+    },
+    {
+      "epoch": 0.00789,
+      "grad_norm": 0.7061838607079715,
+      "learning_rate": 0.002367,
+      "loss": 4.6597,
+      "step": 789
+    },
+    {
+      "epoch": 0.0079,
+      "grad_norm": 0.7001074671969121,
+      "learning_rate": 0.00237,
+      "loss": 4.6333,
+      "step": 790
+    },
+    {
+      "epoch": 0.00791,
+      "grad_norm": 0.6934841557036142,
+      "learning_rate": 0.002373,
+      "loss": 4.5873,
+      "step": 791
+    },
+    {
+      "epoch": 0.00792,
+      "grad_norm": 0.6555126375785874,
+      "learning_rate": 0.002376,
+      "loss": 4.6158,
+      "step": 792
+    },
+    {
+      "epoch": 0.00793,
+      "grad_norm": 0.6198619755446345,
+      "learning_rate": 0.002379,
+      "loss": 4.6082,
+      "step": 793
+    },
+    {
+      "epoch": 0.00794,
+      "grad_norm": 0.5540734864838481,
+      "learning_rate": 0.002382,
+      "loss": 4.615,
+      "step": 794
+    },
+    {
+      "epoch": 0.00795,
+      "grad_norm": 0.5756469670820633,
+      "learning_rate": 0.002385,
+      "loss": 4.5927,
+      "step": 795
+    },
+    {
+      "epoch": 0.00796,
+      "grad_norm": 0.5912602735893169,
+      "learning_rate": 0.0023880000000000004,
+      "loss": 4.5839,
+      "step": 796
+    },
+    {
+      "epoch": 0.00797,
+      "grad_norm": 0.6477479187436139,
+      "learning_rate": 0.0023910000000000003,
+      "loss": 4.628,
+      "step": 797
+    },
+    {
+      "epoch": 0.00798,
+      "grad_norm": 0.6568610896012951,
+      "learning_rate": 0.0023940000000000003,
+      "loss": 4.5975,
+      "step": 798
+    },
+    {
+      "epoch": 0.00799,
+      "grad_norm": 0.8964867071559416,
+      "learning_rate": 0.0023970000000000003,
+      "loss": 4.6327,
+      "step": 799
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 1.1130272400690795,
+      "learning_rate": 0.0024000000000000002,
+      "loss": 4.6127,
+      "step": 800
+    },
+    {
+      "epoch": 0.00801,
+      "grad_norm": 0.8936330318178004,
+      "learning_rate": 0.002403,
+      "loss": 4.624,
+      "step": 801
+    },
+    {
+      "epoch": 0.00802,
+      "grad_norm": 1.279259611288336,
+      "learning_rate": 0.002406,
+      "loss": 4.6431,
+      "step": 802
+    },
+    {
+      "epoch": 0.00803,
+      "grad_norm": 0.7588568023085343,
+      "learning_rate": 0.002409,
+      "loss": 4.5967,
+      "step": 803
+    },
+    {
+      "epoch": 0.00804,
+      "grad_norm": 0.9911172738466991,
+      "learning_rate": 0.002412,
+      "loss": 4.6083,
+      "step": 804
+    },
+    {
+      "epoch": 0.00805,
+      "grad_norm": 1.0297855025384943,
+      "learning_rate": 0.002415,
+      "loss": 4.6095,
+      "step": 805
+    },
+    {
+      "epoch": 0.00806,
+      "grad_norm": 0.8656546102408385,
+      "learning_rate": 0.002418,
+      "loss": 4.6231,
+      "step": 806
+    },
+    {
+      "epoch": 0.00807,
+      "grad_norm": 0.7922998530369505,
+      "learning_rate": 0.0024210000000000004,
+      "loss": 4.6144,
+      "step": 807
+    },
+    {
+      "epoch": 0.00808,
+      "grad_norm": 0.7119238411669042,
+      "learning_rate": 0.0024240000000000004,
+      "loss": 4.5933,
+      "step": 808
+    },
+    {
+      "epoch": 0.00809,
+      "grad_norm": 0.6923611180056076,
+      "learning_rate": 0.0024270000000000003,
+      "loss": 4.5855,
+      "step": 809
+    },
+    {
+      "epoch": 0.0081,
+      "grad_norm": 0.759382946117684,
+      "learning_rate": 0.0024300000000000003,
+      "loss": 4.6149,
+      "step": 810
+    },
+    {
+      "epoch": 0.00811,
+      "grad_norm": 0.9629068291697402,
+      "learning_rate": 0.0024330000000000003,
+      "loss": 4.5955,
+      "step": 811
+    },
+    {
+      "epoch": 0.00812,
+      "grad_norm": 0.9897740547487952,
+      "learning_rate": 0.0024360000000000002,
+      "loss": 4.5793,
+      "step": 812
+    },
+    {
+      "epoch": 0.00813,
+      "grad_norm": 1.0343029805479964,
+      "learning_rate": 0.0024389999999999998,
+      "loss": 4.6192,
+      "step": 813
+    },
+    {
+      "epoch": 0.00814,
+      "grad_norm": 1.188458156645309,
+      "learning_rate": 0.0024419999999999997,
+      "loss": 4.6409,
+      "step": 814
+    },
+    {
+      "epoch": 0.00815,
+      "grad_norm": 0.960728927074141,
+      "learning_rate": 0.0024449999999999997,
+      "loss": 4.6319,
+      "step": 815
+    },
+    {
+      "epoch": 0.00816,
+      "grad_norm": 1.0099133377105225,
+      "learning_rate": 0.002448,
+      "loss": 4.6186,
+      "step": 816
+    },
+    {
+      "epoch": 0.00817,
+      "grad_norm": 0.9311887346236664,
+      "learning_rate": 0.002451,
+      "loss": 4.6246,
+      "step": 817
+    },
+    {
+      "epoch": 0.00818,
+      "grad_norm": 1.2609797944179002,
+      "learning_rate": 0.002454,
+      "loss": 4.625,
+      "step": 818
+    },
+    {
+      "epoch": 0.00819,
+      "grad_norm": 1.0447566978787928,
+      "learning_rate": 0.002457,
+      "loss": 4.6401,
+      "step": 819
+    },
+    {
+      "epoch": 0.0082,
+      "grad_norm": 1.1000322233283122,
+      "learning_rate": 0.00246,
+      "loss": 4.6239,
+      "step": 820
+    },
+    {
+      "epoch": 0.00821,
+      "grad_norm": 0.8928816997822231,
+      "learning_rate": 0.002463,
+      "loss": 4.5932,
+      "step": 821
+    },
+    {
+      "epoch": 0.00822,
+      "grad_norm": 0.9105337296182261,
+      "learning_rate": 0.002466,
+      "loss": 4.5996,
+      "step": 822
+    },
+    {
+      "epoch": 0.00823,
+      "grad_norm": 0.6537296194510775,
+      "learning_rate": 0.002469,
+      "loss": 4.5905,
+      "step": 823
+    },
+    {
+      "epoch": 0.00824,
+      "grad_norm": 0.6615102644633621,
+      "learning_rate": 0.002472,
+      "loss": 4.61,
+      "step": 824
+    },
+    {
+      "epoch": 0.00825,
+      "grad_norm": 0.7605760557018463,
+      "learning_rate": 0.0024749999999999998,
+      "loss": 4.6148,
+      "step": 825
+    },
+    {
+      "epoch": 0.00826,
+      "grad_norm": 0.9288808111624368,
+      "learning_rate": 0.0024779999999999997,
+      "loss": 4.5806,
+      "step": 826
+    },
+    {
+      "epoch": 0.00827,
+      "grad_norm": 0.8099205132023622,
+      "learning_rate": 0.002481,
+      "loss": 4.6084,
+      "step": 827
+    },
+    {
+      "epoch": 0.00828,
+      "grad_norm": 0.7444084275861881,
+      "learning_rate": 0.002484,
+      "loss": 4.5965,
+      "step": 828
+    },
+    {
+      "epoch": 0.00829,
+      "grad_norm": 0.6438605280155706,
+      "learning_rate": 0.002487,
+      "loss": 4.5891,
+      "step": 829
+    },
+    {
+      "epoch": 0.0083,
+      "grad_norm": 0.6242869251575957,
+      "learning_rate": 0.00249,
+      "loss": 4.5955,
+      "step": 830
+    },
+    {
+      "epoch": 0.00831,
+      "grad_norm": 0.6705073637500226,
+      "learning_rate": 0.002493,
+      "loss": 4.576,
+      "step": 831
+    },
+    {
+      "epoch": 0.00832,
+      "grad_norm": 0.7603459804613621,
+      "learning_rate": 0.002496,
+      "loss": 4.5759,
+      "step": 832
+    },
+    {
+      "epoch": 0.00833,
+      "grad_norm": 0.7662808437283888,
+      "learning_rate": 0.002499,
+      "loss": 4.5696,
+      "step": 833
+    },
+    {
+      "epoch": 0.00834,
+      "grad_norm": 0.8012380275176963,
+      "learning_rate": 0.002502,
+      "loss": 4.5743,
+      "step": 834
+    },
+    {
+      "epoch": 0.00835,
+      "grad_norm": 1.0075156101089233,
+      "learning_rate": 0.002505,
+      "loss": 4.5783,
+      "step": 835
+    },
+    {
+      "epoch": 0.00836,
+      "grad_norm": 1.0847073328294785,
+      "learning_rate": 0.002508,
+      "loss": 4.6155,
+      "step": 836
+    },
+    {
+      "epoch": 0.00837,
+      "grad_norm": 0.8849400794535106,
+      "learning_rate": 0.0025109999999999998,
+      "loss": 4.5603,
+      "step": 837
+    },
+    {
+      "epoch": 0.00838,
+      "grad_norm": 0.8832989831172253,
+      "learning_rate": 0.0025139999999999997,
+      "loss": 4.5816,
+      "step": 838
+    },
+    {
+      "epoch": 0.00839,
+      "grad_norm": 0.9511730781538094,
+      "learning_rate": 0.002517,
+      "loss": 4.5843,
+      "step": 839
+    },
+    {
+      "epoch": 0.0084,
+      "grad_norm": 1.037817208386755,
+      "learning_rate": 0.00252,
+      "loss": 4.5831,
+      "step": 840
+    },
+    {
+      "epoch": 0.00841,
+      "grad_norm": 0.8472466492390959,
+      "learning_rate": 0.002523,
+      "loss": 4.5624,
+      "step": 841
+    },
+    {
+      "epoch": 0.00842,
+      "grad_norm": 0.9269217528832409,
+      "learning_rate": 0.002526,
+      "loss": 4.5661,
+      "step": 842
+    },
+    {
+      "epoch": 0.00843,
+      "grad_norm": 0.8716247137682855,
+      "learning_rate": 0.002529,
+      "loss": 4.5688,
+      "step": 843
+    },
+    {
+      "epoch": 0.00844,
+      "grad_norm": 0.7416972066179122,
+      "learning_rate": 0.002532,
+      "loss": 4.585,
+      "step": 844
+    },
+    {
+      "epoch": 0.00845,
+      "grad_norm": 0.6177824387251759,
+      "learning_rate": 0.002535,
+      "loss": 4.5442,
+      "step": 845
+    },
+    {
+      "epoch": 0.00846,
+      "grad_norm": 0.5865348849804463,
+      "learning_rate": 0.002538,
+      "loss": 4.5656,
+      "step": 846
+    },
+    {
+      "epoch": 0.00847,
+      "grad_norm": 0.5324615561823309,
+      "learning_rate": 0.002541,
+      "loss": 4.5442,
+      "step": 847
+    },
+    {
+      "epoch": 0.00848,
+      "grad_norm": 0.5568583953268653,
+      "learning_rate": 0.002544,
+      "loss": 4.5407,
+      "step": 848
+    },
+    {
+      "epoch": 0.00849,
+      "grad_norm": 0.600307108588631,
+      "learning_rate": 0.002547,
+      "loss": 4.5371,
+      "step": 849
+    },
+    {
+      "epoch": 0.0085,
+      "grad_norm": 0.5343909127282478,
+      "learning_rate": 0.00255,
+      "loss": 4.5435,
+      "step": 850
+    },
+    {
+      "epoch": 0.00851,
+      "grad_norm": 0.5790732432599173,
+      "learning_rate": 0.002553,
+      "loss": 4.5588,
+      "step": 851
+    },
+    {
+      "epoch": 0.00852,
+      "grad_norm": 0.6406298666409433,
+      "learning_rate": 0.002556,
+      "loss": 4.5503,
+      "step": 852
+    },
+    {
+      "epoch": 0.00853,
+      "grad_norm": 0.6711876835719848,
+      "learning_rate": 0.002559,
+      "loss": 4.5298,
+      "step": 853
+    },
+    {
+      "epoch": 0.00854,
+      "grad_norm": 0.643558144142947,
+      "learning_rate": 0.002562,
+      "loss": 4.5219,
+      "step": 854
+    },
+    {
+      "epoch": 0.00855,
+      "grad_norm": 0.621905866188772,
+      "learning_rate": 0.002565,
+      "loss": 4.5026,
+      "step": 855
+    },
+    {
+      "epoch": 0.00856,
+      "grad_norm": 0.7503391003054042,
+      "learning_rate": 0.002568,
+      "loss": 4.5375,
+      "step": 856
+    },
+    {
+      "epoch": 0.00857,
+      "grad_norm": 0.90802719077466,
+      "learning_rate": 0.002571,
+      "loss": 4.5344,
+      "step": 857
+    },
+    {
+      "epoch": 0.00858,
+      "grad_norm": 0.9157789056738207,
+      "learning_rate": 0.002574,
+      "loss": 4.5308,
+      "step": 858
+    },
+    {
+      "epoch": 0.00859,
+      "grad_norm": 0.8455467899517649,
+      "learning_rate": 0.002577,
+      "loss": 4.5208,
+      "step": 859
+    },
+    {
+      "epoch": 0.0086,
+      "grad_norm": 0.929986585155377,
+      "learning_rate": 0.00258,
+      "loss": 4.5557,
+      "step": 860
+    },
+    {
+      "epoch": 0.00861,
+      "grad_norm": 1.062443351155347,
+      "learning_rate": 0.0025830000000000002,
+      "loss": 4.5815,
+      "step": 861
+    },
+    {
+      "epoch": 0.00862,
+      "grad_norm": 1.0476479971551458,
+      "learning_rate": 0.002586,
+      "loss": 4.5382,
+      "step": 862
+    },
+    {
+      "epoch": 0.00863,
+      "grad_norm": 0.9516272804079478,
+      "learning_rate": 0.002589,
+      "loss": 4.5686,
+      "step": 863
+    },
+    {
+      "epoch": 0.00864,
+      "grad_norm": 1.0775847362135182,
+      "learning_rate": 0.002592,
+      "loss": 4.5222,
+      "step": 864
+    },
+    {
+      "epoch": 0.00865,
+      "grad_norm": 0.902924351244226,
+      "learning_rate": 0.002595,
+      "loss": 4.5357,
+      "step": 865
+    },
+    {
+      "epoch": 0.00866,
+      "grad_norm": 0.889409265471235,
+      "learning_rate": 0.002598,
+      "loss": 4.5623,
+      "step": 866
+    },
+    {
+      "epoch": 0.00867,
+      "grad_norm": 1.1312843257696636,
+      "learning_rate": 0.002601,
+      "loss": 4.5565,
+      "step": 867
+    },
+    {
+      "epoch": 0.00868,
+      "grad_norm": 1.0195609890673947,
+      "learning_rate": 0.002604,
+      "loss": 4.5479,
+      "step": 868
+    },
+    {
+      "epoch": 0.00869,
+      "grad_norm": 0.8378311515658349,
+      "learning_rate": 0.002607,
+      "loss": 4.5492,
+      "step": 869
+    },
+    {
+      "epoch": 0.0087,
+      "grad_norm": 0.93569973985326,
+      "learning_rate": 0.00261,
+      "loss": 4.5413,
+      "step": 870
+    },
+    {
+      "epoch": 0.00871,
+      "grad_norm": 0.9947247250751194,
+      "learning_rate": 0.002613,
+      "loss": 4.5608,
+      "step": 871
+    },
+    {
+      "epoch": 0.00872,
+      "grad_norm": 0.9456334146879876,
+      "learning_rate": 0.002616,
+      "loss": 4.5489,
+      "step": 872
+    },
+    {
+      "epoch": 0.00873,
+      "grad_norm": 0.9088952463307589,
+      "learning_rate": 0.0026190000000000002,
+      "loss": 4.5587,
+      "step": 873
+    },
+    {
+      "epoch": 0.00874,
+      "grad_norm": 0.7636008668853458,
+      "learning_rate": 0.002622,
+      "loss": 4.5702,
+      "step": 874
+    },
+    {
+      "epoch": 0.00875,
+      "grad_norm": 0.8446171091890929,
+      "learning_rate": 0.002625,
+      "loss": 4.5191,
+      "step": 875
+    },
+    {
+      "epoch": 0.00876,
+      "grad_norm": 0.8731048122579586,
+      "learning_rate": 0.002628,
+      "loss": 4.551,
+      "step": 876
+    },
+    {
+      "epoch": 0.00877,
+      "grad_norm": 0.9444127454444267,
+      "learning_rate": 0.002631,
+      "loss": 4.5701,
+      "step": 877
+    },
+    {
+      "epoch": 0.00878,
+      "grad_norm": 0.9863621265162925,
+      "learning_rate": 0.002634,
+      "loss": 4.5108,
+      "step": 878
+    },
+    {
+      "epoch": 0.00879,
+      "grad_norm": 0.7746522972030878,
+      "learning_rate": 0.002637,
+      "loss": 4.5293,
+      "step": 879
+    },
+    {
+      "epoch": 0.0088,
+      "grad_norm": 0.7150869014843059,
+      "learning_rate": 0.00264,
+      "loss": 4.5368,
+      "step": 880
+    },
+    {
+      "epoch": 0.00881,
+      "grad_norm": 0.866142425879982,
+      "learning_rate": 0.002643,
+      "loss": 4.5448,
+      "step": 881
+    },
+    {
+      "epoch": 0.00882,
+      "grad_norm": 1.0565004714930601,
+      "learning_rate": 0.002646,
+      "loss": 4.5522,
+      "step": 882
+    },
+    {
+      "epoch": 0.00883,
+      "grad_norm": 0.88323512650563,
+      "learning_rate": 0.002649,
+      "loss": 4.5382,
+      "step": 883
+    },
+    {
+      "epoch": 0.00884,
+      "grad_norm": 0.8377440854137777,
+      "learning_rate": 0.0026520000000000003,
+      "loss": 4.5227,
+      "step": 884
+    },
+    {
+      "epoch": 0.00885,
+      "grad_norm": 0.8587569363150891,
+      "learning_rate": 0.0026550000000000002,
+      "loss": 4.5189,
+      "step": 885
+    },
+    {
+      "epoch": 0.00886,
+      "grad_norm": 0.9455034420832737,
+      "learning_rate": 0.002658,
+      "loss": 4.5539,
+      "step": 886
+    },
+    {
+      "epoch": 0.00887,
+      "grad_norm": 0.8258368480900744,
+      "learning_rate": 0.002661,
+      "loss": 4.536,
+      "step": 887
+    },
+    {
+      "epoch": 0.00888,
+      "grad_norm": 0.927023832062946,
+      "learning_rate": 0.002664,
+      "loss": 4.5392,
+      "step": 888
+    },
+    {
+      "epoch": 0.00889,
+      "grad_norm": 0.9905400773231482,
+      "learning_rate": 0.002667,
+      "loss": 4.5077,
+      "step": 889
+    },
+    {
+      "epoch": 0.0089,
+      "grad_norm": 1.0181625806478707,
+      "learning_rate": 0.00267,
+      "loss": 4.551,
+      "step": 890
+    },
+    {
+      "epoch": 0.00891,
+      "grad_norm": 1.0618776306697646,
+      "learning_rate": 0.002673,
+      "loss": 4.5446,
+      "step": 891
+    },
+    {
+      "epoch": 0.00892,
+      "grad_norm": 0.9464629097549706,
+      "learning_rate": 0.002676,
+      "loss": 4.5493,
+      "step": 892
+    },
+    {
+      "epoch": 0.00893,
+      "grad_norm": 1.02959831042168,
+      "learning_rate": 0.002679,
+      "loss": 4.5321,
+      "step": 893
+    },
+    {
+      "epoch": 0.00894,
+      "grad_norm": 0.8717588229222071,
+      "learning_rate": 0.002682,
+      "loss": 4.5126,
+      "step": 894
+    },
+    {
+      "epoch": 0.00895,
+      "grad_norm": 0.7766302283006379,
+      "learning_rate": 0.0026850000000000003,
+      "loss": 4.5191,
+      "step": 895
+    },
+    {
+      "epoch": 0.00896,
+      "grad_norm": 0.7210349653989065,
+      "learning_rate": 0.0026880000000000003,
+      "loss": 4.5061,
+      "step": 896
+    },
+    {
+      "epoch": 0.00897,
+      "grad_norm": 0.7840425119747009,
+      "learning_rate": 0.0026910000000000002,
+      "loss": 4.5249,
+      "step": 897
+    },
+    {
+      "epoch": 0.00898,
+      "grad_norm": 0.7907811575022647,
+      "learning_rate": 0.002694,
+      "loss": 4.518,
+      "step": 898
+    },
+    {
+      "epoch": 0.00899,
+      "grad_norm": 0.7866169386193732,
+      "learning_rate": 0.002697,
+      "loss": 4.5195,
+      "step": 899
+    },
+    {
+      "epoch": 0.009,
+      "grad_norm": 0.7303478296484299,
+      "learning_rate": 0.0027,
+      "loss": 4.5449,
+      "step": 900
+    },
+    {
+      "epoch": 0.00901,
+      "grad_norm": 0.6444049592088867,
+      "learning_rate": 0.002703,
+      "loss": 4.4943,
+      "step": 901
+    },
+    {
+      "epoch": 0.00902,
+      "grad_norm": 0.5222361304775145,
+      "learning_rate": 0.002706,
+      "loss": 4.498,
+      "step": 902
+    },
+    {
+      "epoch": 0.00903,
+      "grad_norm": 0.6152635474063327,
+      "learning_rate": 0.002709,
+      "loss": 4.4967,
+      "step": 903
+    },
+    {
+      "epoch": 0.00904,
+      "grad_norm": 0.6931791624762859,
+      "learning_rate": 0.002712,
+      "loss": 4.4818,
+      "step": 904
+    },
+    {
+      "epoch": 0.00905,
+      "grad_norm": 0.6467753244335184,
+      "learning_rate": 0.002715,
+      "loss": 4.4929,
+      "step": 905
+    },
+    {
+      "epoch": 0.00906,
+      "grad_norm": 0.6638188976320862,
+      "learning_rate": 0.002718,
+      "loss": 4.5049,
+      "step": 906
+    },
+    {
+      "epoch": 0.00907,
+      "grad_norm": 0.503631484987646,
+      "learning_rate": 0.0027210000000000003,
+      "loss": 4.5121,
+      "step": 907
+    },
+    {
+      "epoch": 0.00908,
+      "grad_norm": 0.4745734069368428,
+      "learning_rate": 0.0027240000000000003,
+      "loss": 4.481,
+      "step": 908
+    },
+    {
+      "epoch": 0.00909,
+      "grad_norm": 0.5648420216556126,
+      "learning_rate": 0.0027270000000000003,
+      "loss": 4.5013,
+      "step": 909
+    },
+    {
+      "epoch": 0.0091,
+      "grad_norm": 0.5471966544704259,
+      "learning_rate": 0.0027300000000000002,
+      "loss": 4.5027,
+      "step": 910
+    },
+    {
+      "epoch": 0.00911,
+      "grad_norm": 0.6580432739675335,
+      "learning_rate": 0.002733,
+      "loss": 4.4619,
+      "step": 911
+    },
+    {
+      "epoch": 0.00912,
+      "grad_norm": 0.652804172410743,
+      "learning_rate": 0.002736,
+      "loss": 4.4678,
+      "step": 912
+    },
+    {
+      "epoch": 0.00913,
+      "grad_norm": 0.6108886726833864,
+      "learning_rate": 0.002739,
+      "loss": 4.4601,
+      "step": 913
+    },
+    {
+      "epoch": 0.00914,
+      "grad_norm": 0.5246452629450868,
+      "learning_rate": 0.002742,
+      "loss": 4.4865,
+      "step": 914
+    },
+    {
+      "epoch": 0.00915,
+      "grad_norm": 0.6046562570851716,
+      "learning_rate": 0.002745,
+      "loss": 4.4501,
+      "step": 915
+    },
+    {
+      "epoch": 0.00916,
+      "grad_norm": 0.6470672223067736,
+      "learning_rate": 0.002748,
+      "loss": 4.4304,
+      "step": 916
+    },
+    {
+      "epoch": 0.00917,
+      "grad_norm": 0.6458845430146066,
+      "learning_rate": 0.002751,
+      "loss": 4.4431,
+      "step": 917
+    },
+    {
+      "epoch": 0.00918,
+      "grad_norm": 0.6107710101855935,
+      "learning_rate": 0.0027540000000000004,
+      "loss": 4.4622,
+      "step": 918
+    },
+    {
+      "epoch": 0.00919,
+      "grad_norm": 0.6056157888916294,
+      "learning_rate": 0.0027570000000000003,
+      "loss": 4.4436,
+      "step": 919
+    },
+    {
+      "epoch": 0.0092,
+      "grad_norm": 0.6858065732895877,
+      "learning_rate": 0.0027600000000000003,
+      "loss": 4.462,
+      "step": 920
+    },
+    {
+      "epoch": 0.00921,
+      "grad_norm": 0.9391926745722488,
+      "learning_rate": 0.0027630000000000003,
+      "loss": 4.4556,
+      "step": 921
+    },
+    {
+      "epoch": 0.00922,
+      "grad_norm": 1.1348542218598812,
+      "learning_rate": 0.0027660000000000002,
+      "loss": 4.5063,
+      "step": 922
+    },
+    {
+      "epoch": 0.00923,
+      "grad_norm": 1.0421503736233508,
+      "learning_rate": 0.002769,
+      "loss": 4.4713,
+      "step": 923
+    },
+    {
+      "epoch": 0.00924,
+      "grad_norm": 1.0588478925879097,
+      "learning_rate": 0.002772,
+      "loss": 4.4959,
+      "step": 924
+    },
+    {
+      "epoch": 0.00925,
+      "grad_norm": 0.9781832929515508,
+      "learning_rate": 0.002775,
+      "loss": 4.4806,
+      "step": 925
+    },
+    {
+      "epoch": 0.00926,
+      "grad_norm": 0.7776497655560727,
+      "learning_rate": 0.002778,
+      "loss": 4.488,
+      "step": 926
+    },
+    {
+      "epoch": 0.00927,
+      "grad_norm": 0.7034519817521436,
+      "learning_rate": 0.002781,
+      "loss": 4.4448,
+      "step": 927
+    },
+    {
+      "epoch": 0.00928,
+      "grad_norm": 0.9270286412247504,
+      "learning_rate": 0.002784,
+      "loss": 4.4965,
+      "step": 928
+    },
+    {
+      "epoch": 0.00929,
+      "grad_norm": 1.0728170784278697,
+      "learning_rate": 0.0027870000000000004,
+      "loss": 4.4869,
+      "step": 929
+    },
+    {
+      "epoch": 0.0093,
+      "grad_norm": 0.9492127377122095,
+      "learning_rate": 0.0027900000000000004,
+      "loss": 4.4732,
+      "step": 930
+    },
+    {
+      "epoch": 0.00931,
+      "grad_norm": 0.9513900088751025,
+      "learning_rate": 0.0027930000000000003,
+      "loss": 4.4833,
+      "step": 931
+    },
+    {
+      "epoch": 0.00932,
+      "grad_norm": 1.4536440294804005,
+      "learning_rate": 0.0027960000000000003,
+      "loss": 4.5168,
+      "step": 932
+    },
+    {
+      "epoch": 0.00933,
+      "grad_norm": 1.132862147568844,
+      "learning_rate": 0.0027990000000000003,
+      "loss": 4.4769,
+      "step": 933
+    },
+    {
+      "epoch": 0.00934,
+      "grad_norm": 0.8513409648274727,
+      "learning_rate": 0.0028020000000000002,
+      "loss": 4.4941,
+      "step": 934
+    },
+    {
+      "epoch": 0.00935,
+      "grad_norm": 0.8677938211616196,
+      "learning_rate": 0.002805,
+      "loss": 4.5057,
+      "step": 935
+    },
+    {
+      "epoch": 0.00936,
+      "grad_norm": 0.8298971693271944,
+      "learning_rate": 0.002808,
+      "loss": 4.5081,
+      "step": 936
+    },
+    {
+      "epoch": 0.00937,
+      "grad_norm": 0.8363829119527492,
+      "learning_rate": 0.002811,
+      "loss": 4.521,
+      "step": 937
+    },
+    {
+      "epoch": 0.00938,
+      "grad_norm": 1.0036953395826609,
+      "learning_rate": 0.002814,
+      "loss": 4.5006,
+      "step": 938
+    },
+    {
+      "epoch": 0.00939,
+      "grad_norm": 1.0054078613176451,
+      "learning_rate": 0.002817,
+      "loss": 4.5073,
+      "step": 939
+    },
+    {
+      "epoch": 0.0094,
+      "grad_norm": 0.8667374621688471,
+      "learning_rate": 0.00282,
+      "loss": 4.4958,
+      "step": 940
+    },
+    {
+      "epoch": 0.00941,
+      "grad_norm": 0.9205808599892458,
+      "learning_rate": 0.002823,
+      "loss": 4.4733,
+      "step": 941
+    },
+    {
+      "epoch": 0.00942,
+      "grad_norm": 0.9778408651584425,
+      "learning_rate": 0.002826,
+      "loss": 4.5008,
+      "step": 942
+    },
+    {
+      "epoch": 0.00943,
+      "grad_norm": 1.082550194860624,
+      "learning_rate": 0.002829,
+      "loss": 4.4958,
+      "step": 943
+    },
+    {
+      "epoch": 0.00944,
+      "grad_norm": 1.355725245571646,
+      "learning_rate": 0.002832,
+      "loss": 4.5179,
+      "step": 944
+    },
+    {
+      "epoch": 0.00945,
+      "grad_norm": 0.8345488861943434,
+      "learning_rate": 0.002835,
+      "loss": 4.5015,
+      "step": 945
+    },
+    {
+      "epoch": 0.00946,
+      "grad_norm": 1.0009931674850154,
+      "learning_rate": 0.002838,
+      "loss": 4.5193,
+      "step": 946
+    },
+    {
+      "epoch": 0.00947,
+      "grad_norm": 1.0176650405493206,
+      "learning_rate": 0.0028409999999999998,
+      "loss": 4.4912,
+      "step": 947
+    },
+    {
+      "epoch": 0.00948,
+      "grad_norm": 0.836896216552642,
+      "learning_rate": 0.0028439999999999997,
+      "loss": 4.472,
+      "step": 948
+    },
+    {
+      "epoch": 0.00949,
+      "grad_norm": 0.6881971993105954,
+      "learning_rate": 0.002847,
+      "loss": 4.5057,
+      "step": 949
+    },
+    {
+      "epoch": 0.0095,
+      "grad_norm": 0.8926787293989068,
+      "learning_rate": 0.00285,
+      "loss": 4.5027,
+      "step": 950
+    },
+    {
+      "epoch": 0.00951,
+      "grad_norm": 1.0437772730376889,
+      "learning_rate": 0.002853,
+      "loss": 4.4861,
+      "step": 951
+    },
+    {
+      "epoch": 0.00952,
+      "grad_norm": 0.8745743913439339,
+      "learning_rate": 0.002856,
+      "loss": 4.5041,
+      "step": 952
+    },
+    {
+      "epoch": 0.00953,
+      "grad_norm": 0.7847706213592531,
+      "learning_rate": 0.002859,
+      "loss": 4.4446,
+      "step": 953
+    },
+    {
+      "epoch": 0.00954,
+      "grad_norm": 0.6236105730880978,
+      "learning_rate": 0.002862,
+      "loss": 4.4945,
+      "step": 954
+    },
+    {
+      "epoch": 0.00955,
+      "grad_norm": 0.5696186051972435,
+      "learning_rate": 0.002865,
+      "loss": 4.475,
+      "step": 955
+    },
+    {
+      "epoch": 0.00956,
+      "grad_norm": 0.5459272314199634,
+      "learning_rate": 0.002868,
+      "loss": 4.4518,
+      "step": 956
+    },
+    {
+      "epoch": 0.00957,
+      "grad_norm": 0.5100325019322003,
+      "learning_rate": 0.002871,
+      "loss": 4.4721,
+      "step": 957
+    },
+    {
+      "epoch": 0.00958,
+      "grad_norm": 0.6617022302690957,
+      "learning_rate": 0.002874,
+      "loss": 4.4567,
+      "step": 958
+    },
+    {
+      "epoch": 0.00959,
+      "grad_norm": 0.7948420381771908,
+      "learning_rate": 0.002877,
+      "loss": 4.4693,
+      "step": 959
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 1.0277128972108451,
+      "learning_rate": 0.0028799999999999997,
+      "loss": 4.4747,
+      "step": 960
+    },
+    {
+      "epoch": 0.00961,
+      "grad_norm": 0.9925275083373442,
+      "learning_rate": 0.002883,
+      "loss": 4.4507,
+      "step": 961
+    },
+    {
+      "epoch": 0.00962,
+      "grad_norm": 1.0177847800658486,
+      "learning_rate": 0.002886,
+      "loss": 4.4861,
+      "step": 962
+    },
+    {
+      "epoch": 0.00963,
+      "grad_norm": 1.0118802628275685,
+      "learning_rate": 0.002889,
+      "loss": 4.4865,
+      "step": 963
+    },
+    {
+      "epoch": 0.00964,
+      "grad_norm": 1.1856958623453784,
+      "learning_rate": 0.002892,
+      "loss": 4.4868,
+      "step": 964
+    },
+    {
+      "epoch": 0.00965,
+      "grad_norm": 0.8341279969334199,
+      "learning_rate": 0.002895,
+      "loss": 4.4524,
+      "step": 965
+    },
+    {
+      "epoch": 0.00966,
+      "grad_norm": 0.6756655743310646,
+      "learning_rate": 0.002898,
+      "loss": 4.4552,
+      "step": 966
+    },
+    {
+      "epoch": 0.00967,
+      "grad_norm": 0.6634929266596646,
+      "learning_rate": 0.002901,
+      "loss": 4.4933,
+      "step": 967
+    },
+    {
+      "epoch": 0.00968,
+      "grad_norm": 0.6850072615364151,
+      "learning_rate": 0.002904,
+      "loss": 4.4564,
+      "step": 968
+    },
+    {
+      "epoch": 0.00969,
+      "grad_norm": 0.6166935294692494,
+      "learning_rate": 0.002907,
+      "loss": 4.4862,
+      "step": 969
+    },
+    {
+      "epoch": 0.0097,
+      "grad_norm": 0.6616720297110108,
+      "learning_rate": 0.00291,
+      "loss": 4.4401,
+      "step": 970
+    },
+    {
+      "epoch": 0.00971,
+      "grad_norm": 0.5463010292381552,
+      "learning_rate": 0.002913,
+      "loss": 4.4415,
+      "step": 971
+    },
+    {
+      "epoch": 0.00972,
+      "grad_norm": 0.5115285821904162,
+      "learning_rate": 0.002916,
+      "loss": 4.4454,
+      "step": 972
+    },
+    {
+      "epoch": 0.00973,
+      "grad_norm": 0.543781975527911,
+      "learning_rate": 0.002919,
+      "loss": 4.4402,
+      "step": 973
+    },
+    {
+      "epoch": 0.00974,
+      "grad_norm": 0.5167263203162235,
+      "learning_rate": 0.002922,
+      "loss": 4.4246,
+      "step": 974
+    },
+    {
+      "epoch": 0.00975,
+      "grad_norm": 0.4676730441900605,
+      "learning_rate": 0.002925,
+      "loss": 4.4434,
+      "step": 975
+    },
+    {
+      "epoch": 0.00976,
+      "grad_norm": 0.4483140513716565,
+      "learning_rate": 0.002928,
+      "loss": 4.4259,
+      "step": 976
+    },
+    {
+      "epoch": 0.00977,
+      "grad_norm": 0.45300675248114197,
+      "learning_rate": 0.002931,
+      "loss": 4.4582,
+      "step": 977
+    },
+    {
+      "epoch": 0.00978,
+      "grad_norm": 0.524704391185419,
+      "learning_rate": 0.002934,
+      "loss": 4.4335,
+      "step": 978
+    },
+    {
+      "epoch": 0.00979,
+      "grad_norm": 0.6643809565846889,
+      "learning_rate": 0.002937,
+      "loss": 4.4266,
+      "step": 979
+    },
+    {
+      "epoch": 0.0098,
+      "grad_norm": 0.9159629342753403,
+      "learning_rate": 0.00294,
+      "loss": 4.433,
+      "step": 980
+    },
+    {
+      "epoch": 0.00981,
+      "grad_norm": 1.0134318026936866,
+      "learning_rate": 0.002943,
+      "loss": 4.4516,
+      "step": 981
+    },
+    {
+      "epoch": 0.00982,
+      "grad_norm": 0.7532241089289973,
+      "learning_rate": 0.002946,
+      "loss": 4.4387,
+      "step": 982
+    },
+    {
+      "epoch": 0.00983,
+      "grad_norm": 0.7539130672753217,
+      "learning_rate": 0.0029490000000000002,
+      "loss": 4.4372,
+      "step": 983
+    },
+    {
+      "epoch": 0.00984,
+      "grad_norm": 0.7045823796689694,
+      "learning_rate": 0.002952,
+      "loss": 4.4213,
+      "step": 984
+    },
+    {
+      "epoch": 0.00985,
+      "grad_norm": 0.6478785171714704,
+      "learning_rate": 0.002955,
+      "loss": 4.444,
+      "step": 985
+    },
+    {
+      "epoch": 0.00986,
+      "grad_norm": 0.5640139479823427,
+      "learning_rate": 0.002958,
+      "loss": 4.3997,
+      "step": 986
+    },
+    {
+      "epoch": 0.00987,
+      "grad_norm": 0.6145958247621988,
+      "learning_rate": 0.002961,
+      "loss": 4.4141,
+      "step": 987
+    },
+    {
+      "epoch": 0.00988,
+      "grad_norm": 0.7121366087401472,
+      "learning_rate": 0.002964,
+      "loss": 4.4512,
+      "step": 988
+    },
+    {
+      "epoch": 0.00989,
+      "grad_norm": 0.7474063595618389,
+      "learning_rate": 0.002967,
+      "loss": 4.3897,
+      "step": 989
+    },
+    {
+      "epoch": 0.0099,
+      "grad_norm": 0.7608901784540225,
+      "learning_rate": 0.00297,
+      "loss": 4.4296,
+      "step": 990
+    },
+    {
+      "epoch": 0.00991,
+      "grad_norm": 0.8136541225519112,
+      "learning_rate": 0.002973,
+      "loss": 4.4314,
+      "step": 991
+    },
+    {
+      "epoch": 0.00992,
+      "grad_norm": 0.7865701920195308,
+      "learning_rate": 0.002976,
+      "loss": 4.4266,
+      "step": 992
+    },
+    {
+      "epoch": 0.00993,
+      "grad_norm": 0.8315737176917932,
+      "learning_rate": 0.002979,
+      "loss": 4.4267,
+      "step": 993
+    },
+    {
+      "epoch": 0.00994,
+      "grad_norm": 0.7821802322270756,
+      "learning_rate": 0.002982,
+      "loss": 4.4281,
+      "step": 994
+    },
+    {
+      "epoch": 0.00995,
+      "grad_norm": 0.8705272973695986,
+      "learning_rate": 0.0029850000000000002,
+      "loss": 4.4331,
+      "step": 995
+    },
+    {
+      "epoch": 0.00996,
+      "grad_norm": 0.9954881536889274,
+      "learning_rate": 0.002988,
+      "loss": 4.4408,
+      "step": 996
+    },
+    {
+      "epoch": 0.00997,
+      "grad_norm": 0.9720370341600497,
+      "learning_rate": 0.002991,
+      "loss": 4.4354,
+      "step": 997
+    },
+    {
+      "epoch": 0.00998,
+      "grad_norm": 0.7522042168889891,
+      "learning_rate": 0.002994,
+      "loss": 4.4331,
+      "step": 998
+    },
+    {
+      "epoch": 0.00999,
+      "grad_norm": 0.9425882614336212,
+      "learning_rate": 0.002997,
+      "loss": 4.4261,
+      "step": 999
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9689827453070083,
+      "learning_rate": 0.003,
+      "loss": 4.458,
+      "step": 1000
+    },
+    {
+      "epoch": 0.01001,
+      "grad_norm": 0.9802085236972514,
+      "learning_rate": 0.003,
+      "loss": 4.4569,
+      "step": 1001
+    },
+    {
+      "epoch": 0.01002,
+      "grad_norm": 0.9742392022619593,
+      "learning_rate": 0.003,
+      "loss": 4.4262,
+      "step": 1002
+    },
+    {
+      "epoch": 0.01003,
+      "grad_norm": 0.9115330327806416,
+      "learning_rate": 0.003,
+      "loss": 4.4513,
+      "step": 1003
+    },
+    {
+      "epoch": 0.01004,
+      "grad_norm": 0.8562148232052564,
+      "learning_rate": 0.003,
+      "loss": 4.4397,
+      "step": 1004
+    },
+    {
+      "epoch": 0.01005,
+      "grad_norm": 0.9105663755998641,
+      "learning_rate": 0.003,
+      "loss": 4.4154,
+      "step": 1005
+    },
+    {
+      "epoch": 0.01006,
+      "grad_norm": 1.018045955439956,
+      "learning_rate": 0.003,
+      "loss": 4.433,
+      "step": 1006
+    },
+    {
+      "epoch": 0.01007,
+      "grad_norm": 0.9357911521230117,
+      "learning_rate": 0.003,
+      "loss": 4.4713,
+      "step": 1007
+    },
+    {
+      "epoch": 0.01008,
+      "grad_norm": 0.9681416399703533,
+      "learning_rate": 0.003,
+      "loss": 4.4846,
+      "step": 1008
+    },
+    {
+      "epoch": 0.01009,
+      "grad_norm": 1.030244827646589,
+      "learning_rate": 0.003,
+      "loss": 4.4619,
+      "step": 1009
+    },
+    {
+      "epoch": 0.0101,
+      "grad_norm": 1.0010362528613534,
+      "learning_rate": 0.003,
+      "loss": 4.4688,
+      "step": 1010
+    },
+    {
+      "epoch": 0.01011,
+      "grad_norm": 1.0284858383133122,
+      "learning_rate": 0.003,
+      "loss": 4.4881,
+      "step": 1011
+    },
+    {
+      "epoch": 0.01012,
+      "grad_norm": 0.9778345429910184,
+      "learning_rate": 0.003,
+      "loss": 4.4674,
+      "step": 1012
+    },
+    {
+      "epoch": 0.01013,
+      "grad_norm": 0.8705611016855861,
+      "learning_rate": 0.003,
+      "loss": 4.4919,
+      "step": 1013
+    },
+    {
+      "epoch": 0.01014,
+      "grad_norm": 0.8083144078788229,
+      "learning_rate": 0.003,
+      "loss": 4.4268,
+      "step": 1014
+    },
+    {
+      "epoch": 0.01015,
+      "grad_norm": 0.7155205086193928,
+      "learning_rate": 0.003,
+      "loss": 4.4657,
+      "step": 1015
+    },
+    {
+      "epoch": 0.01016,
+      "grad_norm": 0.5844948538982189,
+      "learning_rate": 0.003,
+      "loss": 4.466,
+      "step": 1016
+    },
+    {
+      "epoch": 0.01017,
+      "grad_norm": 0.50419422786756,
+      "learning_rate": 0.003,
+      "loss": 4.4019,
+      "step": 1017
+    },
+    {
+      "epoch": 0.01018,
+      "grad_norm": 0.5273528303530755,
+      "learning_rate": 0.003,
+      "loss": 4.4545,
+      "step": 1018
+    },
+    {
+      "epoch": 0.01019,
+      "grad_norm": 0.5329688106940915,
+      "learning_rate": 0.003,
+      "loss": 4.4315,
+      "step": 1019
+    },
+    {
+      "epoch": 0.0102,
+      "grad_norm": 0.6285459843623249,
+      "learning_rate": 0.003,
+      "loss": 4.4292,
+      "step": 1020
+    },
+    {
+      "epoch": 0.01021,
+      "grad_norm": 0.6444525489483212,
+      "learning_rate": 0.003,
+      "loss": 4.412,
+      "step": 1021
+    },
+    {
+      "epoch": 0.01022,
+      "grad_norm": 0.6609826768689684,
+      "learning_rate": 0.003,
+      "loss": 4.4194,
+      "step": 1022
+    },
+    {
+      "epoch": 0.01023,
+      "grad_norm": 0.6479610287689606,
+      "learning_rate": 0.003,
+      "loss": 4.4151,
+      "step": 1023
+    },
+    {
+      "epoch": 0.01024,
+      "grad_norm": 0.7091931447524652,
+      "learning_rate": 0.003,
+      "loss": 4.3798,
+      "step": 1024
+    },
+    {
+      "epoch": 0.01025,
+      "grad_norm": 0.7026391214213478,
+      "learning_rate": 0.003,
+      "loss": 4.4116,
+      "step": 1025
+    },
+    {
+      "epoch": 0.01026,
+      "grad_norm": 0.6731332289892269,
+      "learning_rate": 0.003,
+      "loss": 4.3991,
+      "step": 1026
+    },
+    {
+      "epoch": 0.01027,
+      "grad_norm": 0.6590644472165706,
+      "learning_rate": 0.003,
+      "loss": 4.4177,
+      "step": 1027
+    },
+    {
+      "epoch": 0.01028,
+      "grad_norm": 0.7485101036485022,
+      "learning_rate": 0.003,
+      "loss": 4.4014,
+      "step": 1028
+    },
+    {
+      "epoch": 0.01029,
+      "grad_norm": 0.7198477689690366,
+      "learning_rate": 0.003,
+      "loss": 4.3803,
+      "step": 1029
+    },
+    {
+      "epoch": 0.0103,
+      "grad_norm": 0.5542753635749327,
+      "learning_rate": 0.003,
+      "loss": 4.4023,
+      "step": 1030
+    },
+    {
+      "epoch": 0.01031,
+      "grad_norm": 0.5292390629019561,
+      "learning_rate": 0.003,
+      "loss": 4.4144,
+      "step": 1031
+    },
+    {
+      "epoch": 0.01032,
+      "grad_norm": 0.500926566427153,
+      "learning_rate": 0.003,
+      "loss": 4.3837,
+      "step": 1032
+    },
+    {
+      "epoch": 0.01033,
+      "grad_norm": 0.49056326531128164,
+      "learning_rate": 0.003,
+      "loss": 4.4201,
+      "step": 1033
+    },
+    {
+      "epoch": 0.01034,
+      "grad_norm": 0.5309292795236984,
+      "learning_rate": 0.003,
+      "loss": 4.3865,
+      "step": 1034
+    },
+    {
+      "epoch": 0.01035,
+      "grad_norm": 0.6084296376545847,
+      "learning_rate": 0.003,
+      "loss": 4.3892,
+      "step": 1035
+    },
+    {
+      "epoch": 0.01036,
+      "grad_norm": 0.8615895093148164,
+      "learning_rate": 0.003,
+      "loss": 4.3883,
+      "step": 1036
+    },
+    {
+      "epoch": 0.01037,
+      "grad_norm": 0.8936988458454226,
+      "learning_rate": 0.003,
+      "loss": 4.4336,
+      "step": 1037
+    },
+    {
+      "epoch": 0.01038,
+      "grad_norm": 0.6948017038229403,
+      "learning_rate": 0.003,
+      "loss": 4.4116,
+      "step": 1038
+    },
+    {
+      "epoch": 0.01039,
+      "grad_norm": 0.9114790426144561,
+      "learning_rate": 0.003,
+      "loss": 4.4152,
+      "step": 1039
+    },
+    {
+      "epoch": 0.0104,
+      "grad_norm": 0.8448966818619524,
+      "learning_rate": 0.003,
+      "loss": 4.4288,
+      "step": 1040
+    },
+    {
+      "epoch": 0.01041,
+      "grad_norm": 0.7995140521375168,
+      "learning_rate": 0.003,
+      "loss": 4.404,
+      "step": 1041
+    },
+    {
+      "epoch": 0.01042,
+      "grad_norm": 0.8979223927667839,
+      "learning_rate": 0.003,
+      "loss": 4.4357,
+      "step": 1042
+    },
+    {
+      "epoch": 0.01043,
+      "grad_norm": 0.7233892787514891,
+      "learning_rate": 0.003,
+      "loss": 4.3775,
+      "step": 1043
+    },
+    {
+      "epoch": 0.01044,
+      "grad_norm": 0.782866039824708,
+      "learning_rate": 0.003,
+      "loss": 4.4144,
+      "step": 1044
+    },
+    {
+      "epoch": 0.01045,
+      "grad_norm": 0.841755171402396,
+      "learning_rate": 0.003,
+      "loss": 4.4353,
+      "step": 1045
+    },
+    {
+      "epoch": 0.01046,
+      "grad_norm": 0.780348729633882,
+      "learning_rate": 0.003,
+      "loss": 4.4043,
+      "step": 1046
+    },
+    {
+      "epoch": 0.01047,
+      "grad_norm": 0.8623775519511051,
+      "learning_rate": 0.003,
+      "loss": 4.4135,
+      "step": 1047
+    },
+    {
+      "epoch": 0.01048,
+      "grad_norm": 0.9009461265034386,
+      "learning_rate": 0.003,
+      "loss": 4.4049,
+      "step": 1048
+    },
+    {
+      "epoch": 0.01049,
+      "grad_norm": 0.739794637514069,
+      "learning_rate": 0.003,
+      "loss": 4.3786,
+      "step": 1049
+    },
+    {
+      "epoch": 0.0105,
+      "grad_norm": 0.6259101207085414,
+      "learning_rate": 0.003,
+      "loss": 4.4154,
+      "step": 1050
+    },
+    {
+      "epoch": 0.01051,
+      "grad_norm": 0.6694575857098493,
+      "learning_rate": 0.003,
+      "loss": 4.393,
+      "step": 1051
+    },
+    {
+      "epoch": 0.01052,
+      "grad_norm": 0.7141380180592181,
+      "learning_rate": 0.003,
+      "loss": 4.4052,
+      "step": 1052
+    },
+    {
+      "epoch": 0.01053,
+      "grad_norm": 0.7943106393078887,
+      "learning_rate": 0.003,
+      "loss": 4.4143,
+      "step": 1053
+    },
+    {
+      "epoch": 0.01054,
+      "grad_norm": 0.7478655479661611,
+      "learning_rate": 0.003,
+      "loss": 4.4077,
+      "step": 1054
+    },
+    {
+      "epoch": 0.01055,
+      "grad_norm": 0.707966249364238,
+      "learning_rate": 0.003,
+      "loss": 4.3861,
+      "step": 1055
+    },
+    {
+      "epoch": 0.01056,
+      "grad_norm": 0.6358438963919717,
+      "learning_rate": 0.003,
+      "loss": 4.3823,
+      "step": 1056
+    },
+    {
+      "epoch": 0.01057,
+      "grad_norm": 0.5691063929797074,
+      "learning_rate": 0.003,
+      "loss": 4.3849,
+      "step": 1057
+    },
+    {
+      "epoch": 0.01058,
+      "grad_norm": 0.5109096197008457,
+      "learning_rate": 0.003,
+      "loss": 4.3737,
+      "step": 1058
+    },
+    {
+      "epoch": 0.01059,
+      "grad_norm": 0.419033279294021,
+      "learning_rate": 0.003,
+      "loss": 4.3749,
+      "step": 1059
+    },
+    {
+      "epoch": 0.0106,
+      "grad_norm": 0.4506837031966603,
+      "learning_rate": 0.003,
+      "loss": 4.3654,
+      "step": 1060
+    },
+    {
+      "epoch": 0.01061,
+      "grad_norm": 0.42662194113596513,
+      "learning_rate": 0.003,
+      "loss": 4.361,
+      "step": 1061
+    },
+    {
+      "epoch": 0.01062,
+      "grad_norm": 0.5018523046776621,
+      "learning_rate": 0.003,
+      "loss": 4.3568,
+      "step": 1062
+    },
+    {
+      "epoch": 0.01063,
+      "grad_norm": 0.6580287488917459,
+      "learning_rate": 0.003,
+      "loss": 4.3552,
+      "step": 1063
+    },
+    {
+      "epoch": 0.01064,
+      "grad_norm": 0.831793270479749,
+      "learning_rate": 0.003,
+      "loss": 4.3844,
+      "step": 1064
+    },
+    {
+      "epoch": 0.01065,
+      "grad_norm": 1.0701105543701324,
+      "learning_rate": 0.003,
+      "loss": 4.3903,
+      "step": 1065
+    },
+    {
+      "epoch": 0.01066,
+      "grad_norm": 0.943731953453328,
+      "learning_rate": 0.003,
+      "loss": 4.4118,
+      "step": 1066
+    },
+    {
+      "epoch": 0.01067,
+      "grad_norm": 0.9236025360095688,
+      "learning_rate": 0.003,
+      "loss": 4.4136,
+      "step": 1067
+    },
+    {
+      "epoch": 0.01068,
+      "grad_norm": 1.1282330525862354,
+      "learning_rate": 0.003,
+      "loss": 4.4118,
+      "step": 1068
+    },
+    {
+      "epoch": 0.01069,
+      "grad_norm": 0.840410887973158,
+      "learning_rate": 0.003,
+      "loss": 4.4054,
+      "step": 1069
+    },
+    {
+      "epoch": 0.0107,
+      "grad_norm": 0.733631994312363,
+      "learning_rate": 0.003,
+      "loss": 4.4566,
+      "step": 1070
+    },
+    {
+      "epoch": 0.01071,
+      "grad_norm": 0.7799052277350242,
+      "learning_rate": 0.003,
+      "loss": 4.3774,
+      "step": 1071
+    },
+    {
+      "epoch": 0.01072,
+      "grad_norm": 0.7960071687741883,
+      "learning_rate": 0.003,
+      "loss": 4.4136,
+      "step": 1072
+    },
+    {
+      "epoch": 0.01073,
+      "grad_norm": 0.6975430946737453,
+      "learning_rate": 0.003,
+      "loss": 4.4079,
+      "step": 1073
+    },
+    {
+      "epoch": 0.01074,
+      "grad_norm": 0.6813769264433432,
+      "learning_rate": 0.003,
+      "loss": 4.37,
+      "step": 1074
+    },
+    {
+      "epoch": 0.01075,
+      "grad_norm": 0.5971783233576602,
+      "learning_rate": 0.003,
+      "loss": 4.4063,
+      "step": 1075
+    },
+    {
+      "epoch": 0.01076,
+      "grad_norm": 0.5510713740534053,
+      "learning_rate": 0.003,
+      "loss": 4.3967,
+      "step": 1076
+    },
+    {
+      "epoch": 0.01077,
+      "grad_norm": 0.5595097233894012,
+      "learning_rate": 0.003,
+      "loss": 4.3917,
+      "step": 1077
+    },
+    {
+      "epoch": 0.01078,
+      "grad_norm": 0.6230934184673731,
+      "learning_rate": 0.003,
+      "loss": 4.3491,
+      "step": 1078
+    },
+    {
+      "epoch": 0.01079,
+      "grad_norm": 0.7178707958532615,
+      "learning_rate": 0.003,
+      "loss": 4.3825,
+      "step": 1079
+    },
+    {
+      "epoch": 0.0108,
+      "grad_norm": 0.7484447643520958,
+      "learning_rate": 0.003,
+      "loss": 4.3665,
+      "step": 1080
+    },
+    {
+      "epoch": 0.01081,
+      "grad_norm": 0.780034137982597,
+      "learning_rate": 0.003,
+      "loss": 4.3727,
+      "step": 1081
+    },
+    {
+      "epoch": 0.01082,
+      "grad_norm": 0.7403701975238451,
+      "learning_rate": 0.003,
+      "loss": 4.3894,
+      "step": 1082
+    },
+    {
+      "epoch": 0.01083,
+      "grad_norm": 0.6305537452051372,
+      "learning_rate": 0.003,
+      "loss": 4.3733,
+      "step": 1083
+    },
+    {
+      "epoch": 0.01084,
+      "grad_norm": 0.7435200213630843,
+      "learning_rate": 0.003,
+      "loss": 4.3825,
+      "step": 1084
+    },
+    {
+      "epoch": 0.01085,
+      "grad_norm": 0.7405129376016774,
+      "learning_rate": 0.003,
+      "loss": 4.3735,
+      "step": 1085
+    },
+    {
+      "epoch": 0.01086,
+      "grad_norm": 0.6037526750507048,
+      "learning_rate": 0.003,
+      "loss": 4.3449,
+      "step": 1086
+    },
+    {
+      "epoch": 0.01087,
+      "grad_norm": 0.5349574840042758,
+      "learning_rate": 0.003,
+      "loss": 4.3549,
+      "step": 1087
+    },
+    {
+      "epoch": 0.01088,
+      "grad_norm": 0.5555341529780033,
+      "learning_rate": 0.003,
+      "loss": 4.3917,
+      "step": 1088
+    },
+    {
+      "epoch": 0.01089,
+      "grad_norm": 0.6254527345250118,
+      "learning_rate": 0.003,
+      "loss": 4.3599,
+      "step": 1089
+    },
+    {
+      "epoch": 0.0109,
+      "grad_norm": 0.6828156926551208,
+      "learning_rate": 0.003,
+      "loss": 4.3884,
+      "step": 1090
+    },
+    {
+      "epoch": 0.01091,
+      "grad_norm": 0.5753021180468998,
+      "learning_rate": 0.003,
+      "loss": 4.3549,
+      "step": 1091
+    },
+    {
+      "epoch": 0.01092,
+      "grad_norm": 0.4077341561870379,
+      "learning_rate": 0.003,
+      "loss": 4.3436,
+      "step": 1092
+    },
+    {
+      "epoch": 0.01093,
+      "grad_norm": 0.47615819716416974,
+      "learning_rate": 0.003,
+      "loss": 4.3664,
+      "step": 1093
+    },
+    {
+      "epoch": 0.01094,
+      "grad_norm": 0.5845002574061751,
+      "learning_rate": 0.003,
+      "loss": 4.3553,
+      "step": 1094
+    },
+    {
+      "epoch": 0.01095,
+      "grad_norm": 0.8583788887394415,
+      "learning_rate": 0.003,
+      "loss": 4.3944,
+      "step": 1095
+    },
+    {
+      "epoch": 0.01096,
+      "grad_norm": 1.0479503783625854,
+      "learning_rate": 0.003,
+      "loss": 4.3824,
+      "step": 1096
+    },
+    {
+      "epoch": 0.01097,
+      "grad_norm": 0.9584408455564232,
+      "learning_rate": 0.003,
+      "loss": 4.3695,
+      "step": 1097
+    },
+    {
+      "epoch": 0.01098,
+      "grad_norm": 1.2545731298718157,
+      "learning_rate": 0.003,
+      "loss": 4.4068,
+      "step": 1098
+    },
+    {
+      "epoch": 0.01099,
+      "grad_norm": 0.8305012272413557,
+      "learning_rate": 0.003,
+      "loss": 4.3706,
+      "step": 1099
+    },
+    {
+      "epoch": 0.011,
+      "grad_norm": 0.7311868902259275,
+      "learning_rate": 0.003,
+      "loss": 4.3677,
+      "step": 1100
+    },
+    {
+      "epoch": 0.01101,
+      "grad_norm": 0.6742597336847759,
+      "learning_rate": 0.003,
+      "loss": 4.3752,
+      "step": 1101
+    },
+    {
+      "epoch": 0.01102,
+      "grad_norm": 0.6773795133108548,
+      "learning_rate": 0.003,
+      "loss": 4.3802,
+      "step": 1102
+    },
+    {
+      "epoch": 0.01103,
+      "grad_norm": 0.6414092553075785,
+      "learning_rate": 0.003,
+      "loss": 4.3701,
+      "step": 1103
+    },
+    {
+      "epoch": 0.01104,
+      "grad_norm": 0.6378763268399013,
+      "learning_rate": 0.003,
+      "loss": 4.3589,
+      "step": 1104
+    },
+    {
+      "epoch": 0.01105,
+      "grad_norm": 0.6837675260878959,
+      "learning_rate": 0.003,
+      "loss": 4.3856,
+      "step": 1105
+    },
+    {
+      "epoch": 0.01106,
+      "grad_norm": 0.667950958790211,
+      "learning_rate": 0.003,
+      "loss": 4.3574,
+      "step": 1106
+    },
+    {
+      "epoch": 0.01107,
+      "grad_norm": 0.6082482122734244,
+      "learning_rate": 0.003,
+      "loss": 4.3623,
+      "step": 1107
+    },
+    {
+      "epoch": 0.01108,
+      "grad_norm": 0.5983800302348057,
+      "learning_rate": 0.003,
+      "loss": 4.3313,
+      "step": 1108
+    },
+    {
+      "epoch": 0.01109,
+      "grad_norm": 0.6046679697278923,
+      "learning_rate": 0.003,
+      "loss": 4.3476,
+      "step": 1109
+    },
+    {
+      "epoch": 0.0111,
+      "grad_norm": 0.6353880163110637,
+      "learning_rate": 0.003,
+      "loss": 4.3675,
+      "step": 1110
+    },
+    {
+      "epoch": 0.01111,
+      "grad_norm": 0.7049628214149826,
+      "learning_rate": 0.003,
+      "loss": 4.3473,
+      "step": 1111
+    },
+    {
+      "epoch": 0.01112,
+      "grad_norm": 0.758960503794264,
+      "learning_rate": 0.003,
+      "loss": 4.369,
+      "step": 1112
+    },
+    {
+      "epoch": 0.01113,
+      "grad_norm": 0.6091686772357772,
+      "learning_rate": 0.003,
+      "loss": 4.3335,
+      "step": 1113
+    },
+    {
+      "epoch": 0.01114,
+      "grad_norm": 0.628169460357088,
+      "learning_rate": 0.003,
+      "loss": 4.3462,
+      "step": 1114
+    },
+    {
+      "epoch": 0.01115,
+      "grad_norm": 0.7453432538813878,
+      "learning_rate": 0.003,
+      "loss": 4.3779,
+      "step": 1115
+    },
+    {
+      "epoch": 0.01116,
+      "grad_norm": 0.9959163456029771,
+      "learning_rate": 0.003,
+      "loss": 4.3771,
+      "step": 1116
+    },
+    {
+      "epoch": 0.01117,
+      "grad_norm": 0.9566018042437344,
+      "learning_rate": 0.003,
+      "loss": 4.3915,
+      "step": 1117
+    },
+    {
+      "epoch": 0.01118,
+      "grad_norm": 0.6794219206934775,
+      "learning_rate": 0.003,
+      "loss": 4.372,
+      "step": 1118
+    },
+    {
+      "epoch": 0.01119,
+      "grad_norm": 0.6027310157842346,
+      "learning_rate": 0.003,
+      "loss": 4.375,
+      "step": 1119
+    },
+    {
+      "epoch": 0.0112,
+      "grad_norm": 0.6064414406921254,
+      "learning_rate": 0.003,
+      "loss": 4.3588,
+      "step": 1120
+    },
+    {
+      "epoch": 0.01121,
+      "grad_norm": 0.643465354938861,
+      "learning_rate": 0.003,
+      "loss": 4.3691,
+      "step": 1121
+    },
+    {
+      "epoch": 0.01122,
+      "grad_norm": 0.8439491151148678,
+      "learning_rate": 0.003,
+      "loss": 4.3539,
+      "step": 1122
+    },
+    {
+      "epoch": 0.01123,
+      "grad_norm": 0.857123487934385,
+      "learning_rate": 0.003,
+      "loss": 4.345,
+      "step": 1123
+    },
+    {
+      "epoch": 0.01124,
+      "grad_norm": 0.7412296015988188,
+      "learning_rate": 0.003,
+      "loss": 4.3865,
+      "step": 1124
+    },
+    {
+      "epoch": 0.01125,
+      "grad_norm": 0.8143761816040683,
+      "learning_rate": 0.003,
+      "loss": 4.3583,
+      "step": 1125
+    },
+    {
+      "epoch": 0.01126,
+      "grad_norm": 0.741767094028342,
+      "learning_rate": 0.003,
+      "loss": 4.3618,
+      "step": 1126
+    },
+    {
+      "epoch": 0.01127,
+      "grad_norm": 0.737420855232079,
+      "learning_rate": 0.003,
+      "loss": 4.3948,
+      "step": 1127
+    },
+    {
+      "epoch": 0.01128,
+      "grad_norm": 0.7656500975813709,
+      "learning_rate": 0.003,
+      "loss": 4.372,
+      "step": 1128
+    },
+    {
+      "epoch": 0.01129,
+      "grad_norm": 0.8248659664813461,
+      "learning_rate": 0.003,
+      "loss": 4.3685,
+      "step": 1129
+    },
+    {
+      "epoch": 0.0113,
+      "grad_norm": 0.7756542452356111,
+      "learning_rate": 0.003,
+      "loss": 4.3447,
+      "step": 1130
+    },
+    {
+      "epoch": 0.01131,
+      "grad_norm": 0.7965849423705411,
+      "learning_rate": 0.003,
+      "loss": 4.3615,
+      "step": 1131
+    },
+    {
+      "epoch": 0.01132,
+      "grad_norm": 0.718692737781693,
+      "learning_rate": 0.003,
+      "loss": 4.3824,
+      "step": 1132
+    },
+    {
+      "epoch": 0.01133,
+      "grad_norm": 0.6500905111846983,
+      "learning_rate": 0.003,
+      "loss": 4.34,
+      "step": 1133
+    },
+    {
+      "epoch": 0.01134,
+      "grad_norm": 0.6052834019155894,
+      "learning_rate": 0.003,
+      "loss": 4.3525,
+      "step": 1134
+    },
+    {
+      "epoch": 0.01135,
+      "grad_norm": 0.57728723025885,
+      "learning_rate": 0.003,
+      "loss": 4.3767,
+      "step": 1135
+    },
+    {
+      "epoch": 0.01136,
+      "grad_norm": 0.6622614298653483,
+      "learning_rate": 0.003,
+      "loss": 4.3514,
+      "step": 1136
+    },
+    {
+      "epoch": 0.01137,
+      "grad_norm": 0.6945355696134266,
+      "learning_rate": 0.003,
+      "loss": 4.3784,
+      "step": 1137
+    },
+    {
+      "epoch": 0.01138,
+      "grad_norm": 0.7423434054596434,
+      "learning_rate": 0.003,
+      "loss": 4.3361,
+      "step": 1138
+    },
+    {
+      "epoch": 0.01139,
+      "grad_norm": 0.7678127530736037,
+      "learning_rate": 0.003,
+      "loss": 4.3463,
+      "step": 1139
+    },
+    {
+      "epoch": 0.0114,
+      "grad_norm": 0.7470822118949633,
+      "learning_rate": 0.003,
+      "loss": 4.3322,
+      "step": 1140
+    },
+    {
+      "epoch": 0.01141,
+      "grad_norm": 0.903198095826591,
+      "learning_rate": 0.003,
+      "loss": 4.3608,
+      "step": 1141
+    },
+    {
+      "epoch": 0.01142,
+      "grad_norm": 1.0797547624524322,
+      "learning_rate": 0.003,
+      "loss": 4.3612,
+      "step": 1142
+    },
+    {
+      "epoch": 0.01143,
+      "grad_norm": 0.7877548466868329,
+      "learning_rate": 0.003,
+      "loss": 4.3542,
+      "step": 1143
+    },
+    {
+      "epoch": 0.01144,
+      "grad_norm": 0.8147882944795127,
+      "learning_rate": 0.003,
+      "loss": 4.3253,
+      "step": 1144
+    },
+    {
+      "epoch": 0.01145,
+      "grad_norm": 0.7148112106554612,
+      "learning_rate": 0.003,
+      "loss": 4.3655,
+      "step": 1145
+    },
+    {
+      "epoch": 0.01146,
+      "grad_norm": 0.6580426381015894,
+      "learning_rate": 0.003,
+      "loss": 4.3445,
+      "step": 1146
+    },
+    {
+      "epoch": 0.01147,
+      "grad_norm": 0.7448256416265617,
+      "learning_rate": 0.003,
+      "loss": 4.3367,
+      "step": 1147
+    },
+    {
+      "epoch": 0.01148,
+      "grad_norm": 0.6968832253815664,
+      "learning_rate": 0.003,
+      "loss": 4.3386,
+      "step": 1148
+    },
+    {
+      "epoch": 0.01149,
+      "grad_norm": 0.627190656091196,
+      "learning_rate": 0.003,
+      "loss": 4.3527,
+      "step": 1149
+    },
+    {
+      "epoch": 0.0115,
+      "grad_norm": 0.6739971655272956,
+      "learning_rate": 0.003,
+      "loss": 4.3445,
+      "step": 1150
+    },
+    {
+      "epoch": 0.01151,
+      "grad_norm": 0.7518825421443525,
+      "learning_rate": 0.003,
+      "loss": 4.3445,
+      "step": 1151
+    },
+    {
+      "epoch": 0.01152,
+      "grad_norm": 0.7628403894732813,
+      "learning_rate": 0.003,
+      "loss": 4.3529,
+      "step": 1152
+    },
+    {
+      "epoch": 0.01153,
+      "grad_norm": 0.726109999889872,
+      "learning_rate": 0.003,
+      "loss": 4.3668,
+      "step": 1153
+    },
+    {
+      "epoch": 0.01154,
+      "grad_norm": 0.7747132516080261,
+      "learning_rate": 0.003,
+      "loss": 4.3438,
+      "step": 1154
+    },
+    {
+      "epoch": 0.01155,
+      "grad_norm": 0.8180276502719757,
+      "learning_rate": 0.003,
+      "loss": 4.3369,
+      "step": 1155
+    },
+    {
+      "epoch": 0.01156,
+      "grad_norm": 0.7757994684934312,
+      "learning_rate": 0.003,
+      "loss": 4.3405,
+      "step": 1156
+    },
+    {
+      "epoch": 0.01157,
+      "grad_norm": 0.7233155240278132,
+      "learning_rate": 0.003,
+      "loss": 4.3659,
+      "step": 1157
+    },
+    {
+      "epoch": 0.01158,
+      "grad_norm": 0.6552894912204377,
+      "learning_rate": 0.003,
+      "loss": 4.332,
+      "step": 1158
+    },
+    {
+      "epoch": 0.01159,
+      "grad_norm": 0.7654422010962233,
+      "learning_rate": 0.003,
+      "loss": 4.3316,
+      "step": 1159
+    },
+    {
+      "epoch": 0.0116,
+      "grad_norm": 0.7129248873282286,
+      "learning_rate": 0.003,
+      "loss": 4.3462,
+      "step": 1160
+    },
+    {
+      "epoch": 0.01161,
+      "grad_norm": 0.728340877307682,
+      "learning_rate": 0.003,
+      "loss": 4.3213,
+      "step": 1161
+    },
+    {
+      "epoch": 0.01162,
+      "grad_norm": 0.6672912532139038,
+      "learning_rate": 0.003,
+      "loss": 4.3344,
+      "step": 1162
+    },
+    {
+      "epoch": 0.01163,
+      "grad_norm": 0.6087688372398936,
+      "learning_rate": 0.003,
+      "loss": 4.3512,
+      "step": 1163
+    },
+    {
+      "epoch": 0.01164,
+      "grad_norm": 0.5012921787632111,
+      "learning_rate": 0.003,
+      "loss": 4.3329,
+      "step": 1164
+    },
+    {
+      "epoch": 0.01165,
+      "grad_norm": 0.4972192156782153,
+      "learning_rate": 0.003,
+      "loss": 4.3277,
+      "step": 1165
+    },
+    {
+      "epoch": 0.01166,
+      "grad_norm": 0.48804976467805194,
+      "learning_rate": 0.003,
+      "loss": 4.3262,
+      "step": 1166
+    },
+    {
+      "epoch": 0.01167,
+      "grad_norm": 0.5481538887564317,
+      "learning_rate": 0.003,
+      "loss": 4.2952,
+      "step": 1167
+    },
+    {
+      "epoch": 0.01168,
+      "grad_norm": 0.6090792597366076,
+      "learning_rate": 0.003,
+      "loss": 4.3026,
+      "step": 1168
+    },
+    {
+      "epoch": 0.01169,
+      "grad_norm": 0.653085333922602,
+      "learning_rate": 0.003,
+      "loss": 4.347,
+      "step": 1169
+    },
+    {
+      "epoch": 0.0117,
+      "grad_norm": 0.62436544437129,
+      "learning_rate": 0.003,
+      "loss": 4.3309,
+      "step": 1170
+    },
+    {
+      "epoch": 0.01171,
+      "grad_norm": 0.654802486268078,
+      "learning_rate": 0.003,
+      "loss": 4.3143,
+      "step": 1171
+    },
+    {
+      "epoch": 0.01172,
+      "grad_norm": 0.6246624927886145,
+      "learning_rate": 0.003,
+      "loss": 4.3038,
+      "step": 1172
+    },
+    {
+      "epoch": 0.01173,
+      "grad_norm": 0.6813303457409844,
+      "learning_rate": 0.003,
+      "loss": 4.3382,
+      "step": 1173
+    },
+    {
+      "epoch": 0.01174,
+      "grad_norm": 0.7721136771646261,
+      "learning_rate": 0.003,
+      "loss": 4.3203,
+      "step": 1174
+    },
+    {
+      "epoch": 0.01175,
+      "grad_norm": 0.823782632958531,
+      "learning_rate": 0.003,
+      "loss": 4.3152,
+      "step": 1175
+    },
+    {
+      "epoch": 0.01176,
+      "grad_norm": 0.9763749462159719,
+      "learning_rate": 0.003,
+      "loss": 4.3509,
+      "step": 1176
+    },
+    {
+      "epoch": 0.01177,
+      "grad_norm": 1.0894015492716702,
+      "learning_rate": 0.003,
+      "loss": 4.3444,
+      "step": 1177
+    },
+    {
+      "epoch": 0.01178,
+      "grad_norm": 0.8156537670820067,
+      "learning_rate": 0.003,
+      "loss": 4.3473,
+      "step": 1178
+    },
+    {
+      "epoch": 0.01179,
+      "grad_norm": 0.8529362760715462,
+      "learning_rate": 0.003,
+      "loss": 4.3276,
+      "step": 1179
+    },
+    {
+      "epoch": 0.0118,
+      "grad_norm": 0.857392382910556,
+      "learning_rate": 0.003,
+      "loss": 4.395,
+      "step": 1180
+    },
+    {
+      "epoch": 0.01181,
+      "grad_norm": 0.8128436446210868,
+      "learning_rate": 0.003,
+      "loss": 4.3809,
+      "step": 1181
+    },
+    {
+      "epoch": 0.01182,
+      "grad_norm": 0.911584010746237,
+      "learning_rate": 0.003,
+      "loss": 4.3433,
+      "step": 1182
+    },
+    {
+      "epoch": 0.01183,
+      "grad_norm": 0.8460613119287651,
+      "learning_rate": 0.003,
+      "loss": 4.337,
+      "step": 1183
+    },
+    {
+      "epoch": 0.01184,
+      "grad_norm": 0.7730717809982315,
+      "learning_rate": 0.003,
+      "loss": 4.3753,
+      "step": 1184
+    },
+    {
+      "epoch": 0.01185,
+      "grad_norm": 0.7718355908910234,
+      "learning_rate": 0.003,
+      "loss": 4.3449,
+      "step": 1185
+    },
+    {
+      "epoch": 0.01186,
+      "grad_norm": 0.749856576054878,
+      "learning_rate": 0.003,
+      "loss": 4.3669,
+      "step": 1186
+    },
+    {
+      "epoch": 0.01187,
+      "grad_norm": 0.6703031442863907,
+      "learning_rate": 0.003,
+      "loss": 4.33,
+      "step": 1187
+    },
+    {
+      "epoch": 0.01188,
+      "grad_norm": 0.6597409636847196,
+      "learning_rate": 0.003,
+      "loss": 4.3387,
+      "step": 1188
+    },
+    {
+      "epoch": 0.01189,
+      "grad_norm": 0.7528540145175824,
+      "learning_rate": 0.003,
+      "loss": 4.3431,
+      "step": 1189
+    },
+    {
+      "epoch": 0.0119,
+      "grad_norm": 0.7990197433249405,
+      "learning_rate": 0.003,
+      "loss": 4.3273,
+      "step": 1190
+    },
+    {
+      "epoch": 0.01191,
+      "grad_norm": 0.8012185403960035,
+      "learning_rate": 0.003,
+      "loss": 4.3081,
+      "step": 1191
+    },
+    {
+      "epoch": 0.01192,
+      "grad_norm": 0.7571323847445259,
+      "learning_rate": 0.003,
+      "loss": 4.3381,
+      "step": 1192
+    },
+    {
+      "epoch": 0.01193,
+      "grad_norm": 0.6624809456477357,
+      "learning_rate": 0.003,
+      "loss": 4.3371,
+      "step": 1193
+    },
+    {
+      "epoch": 0.01194,
+      "grad_norm": 0.7303795208901569,
+      "learning_rate": 0.003,
+      "loss": 4.3395,
+      "step": 1194
+    },
+    {
+      "epoch": 0.01195,
+      "grad_norm": 0.7841704264970735,
+      "learning_rate": 0.003,
+      "loss": 4.3211,
+      "step": 1195
+    },
+    {
+      "epoch": 0.01196,
+      "grad_norm": 0.8218420512118195,
+      "learning_rate": 0.003,
+      "loss": 4.3286,
+      "step": 1196
+    },
+    {
+      "epoch": 0.01197,
+      "grad_norm": 0.6346054129298183,
+      "learning_rate": 0.003,
+      "loss": 4.2874,
+      "step": 1197
+    },
+    {
+      "epoch": 0.01198,
+      "grad_norm": 0.648198494061273,
+      "learning_rate": 0.003,
+      "loss": 4.3575,
+      "step": 1198
+    },
+    {
+      "epoch": 0.01199,
+      "grad_norm": 0.719551428234481,
+      "learning_rate": 0.003,
+      "loss": 4.3378,
+      "step": 1199
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 0.7355036739818444,
+      "learning_rate": 0.003,
+      "loss": 4.3295,
+      "step": 1200
+    },
+    {
+      "epoch": 0.01201,
+      "grad_norm": 0.9984238489578499,
+      "learning_rate": 0.003,
+      "loss": 4.3657,
+      "step": 1201
+    },
+    {
+      "epoch": 0.01202,
+      "grad_norm": 1.2422658156142912,
+      "learning_rate": 0.003,
+      "loss": 4.3719,
+      "step": 1202
+    },
+    {
+      "epoch": 0.01203,
+      "grad_norm": 0.7461879644958674,
+      "learning_rate": 0.003,
+      "loss": 4.3511,
+      "step": 1203
+    },
+    {
+      "epoch": 0.01204,
+      "grad_norm": 0.8846740490522026,
+      "learning_rate": 0.003,
+      "loss": 4.3624,
+      "step": 1204
+    },
+    {
+      "epoch": 0.01205,
+      "grad_norm": 0.7423867106895455,
+      "learning_rate": 0.003,
+      "loss": 4.3403,
+      "step": 1205
+    },
+    {
+      "epoch": 0.01206,
+      "grad_norm": 0.7502877844645354,
+      "learning_rate": 0.003,
+      "loss": 4.3858,
+      "step": 1206
+    },
+    {
+      "epoch": 0.01207,
+      "grad_norm": 0.8010490453321524,
+      "learning_rate": 0.003,
+      "loss": 4.3053,
+      "step": 1207
+    },
+    {
+      "epoch": 0.01208,
+      "grad_norm": 0.7863992142209323,
+      "learning_rate": 0.003,
+      "loss": 4.3665,
+      "step": 1208
+    },
+    {
+      "epoch": 0.01209,
+      "grad_norm": 0.6936314129825434,
+      "learning_rate": 0.003,
+      "loss": 4.3419,
+      "step": 1209
+    },
+    {
+      "epoch": 0.0121,
+      "grad_norm": 0.5834058626530795,
+      "learning_rate": 0.003,
+      "loss": 4.3425,
+      "step": 1210
+    },
+    {
+      "epoch": 0.01211,
+      "grad_norm": 0.5442454390371272,
+      "learning_rate": 0.003,
+      "loss": 4.3248,
+      "step": 1211
+    },
+    {
+      "epoch": 0.01212,
+      "grad_norm": 0.5640531404426855,
+      "learning_rate": 0.003,
+      "loss": 4.3485,
+      "step": 1212
+    },
+    {
+      "epoch": 0.01213,
+      "grad_norm": 0.4686052961934471,
+      "learning_rate": 0.003,
+      "loss": 4.3204,
+      "step": 1213
+    },
+    {
+      "epoch": 0.01214,
+      "grad_norm": 0.4623933928179833,
+      "learning_rate": 0.003,
+      "loss": 4.3023,
+      "step": 1214
+    },
+    {
+      "epoch": 0.01215,
+      "grad_norm": 0.390314877629531,
+      "learning_rate": 0.003,
+      "loss": 4.3075,
+      "step": 1215
+    },
+    {
+      "epoch": 0.01216,
+      "grad_norm": 0.3573034139707242,
+      "learning_rate": 0.003,
+      "loss": 4.2995,
+      "step": 1216
+    },
+    {
+      "epoch": 0.01217,
+      "grad_norm": 0.38593293619021,
+      "learning_rate": 0.003,
+      "loss": 4.3063,
+      "step": 1217
+    },
+    {
+      "epoch": 0.01218,
+      "grad_norm": 0.40705674810377235,
+      "learning_rate": 0.003,
+      "loss": 4.3006,
+      "step": 1218
+    },
+    {
+      "epoch": 0.01219,
+      "grad_norm": 0.45251062517773044,
+      "learning_rate": 0.003,
+      "loss": 4.3,
+      "step": 1219
+    },
+    {
+      "epoch": 0.0122,
+      "grad_norm": 0.5627372695280709,
+      "learning_rate": 0.003,
+      "loss": 4.2943,
+      "step": 1220
+    },
+    {
+      "epoch": 0.01221,
+      "grad_norm": 0.6302297504421508,
+      "learning_rate": 0.003,
+      "loss": 4.3122,
+      "step": 1221
+    },
+    {
+      "epoch": 0.01222,
+      "grad_norm": 0.695146633094906,
+      "learning_rate": 0.003,
+      "loss": 4.2975,
+      "step": 1222
+    },
+    {
+      "epoch": 0.01223,
+      "grad_norm": 0.6745684022343855,
+      "learning_rate": 0.003,
+      "loss": 4.2965,
+      "step": 1223
+    },
+    {
+      "epoch": 0.01224,
+      "grad_norm": 0.6275537412165774,
+      "learning_rate": 0.003,
+      "loss": 4.2794,
+      "step": 1224
+    },
+    {
+      "epoch": 0.01225,
+      "grad_norm": 0.732873035624088,
+      "learning_rate": 0.003,
+      "loss": 4.3084,
+      "step": 1225
+    },
+    {
+      "epoch": 0.01226,
+      "grad_norm": 0.8536265776058696,
+      "learning_rate": 0.003,
+      "loss": 4.3061,
+      "step": 1226
+    },
+    {
+      "epoch": 0.01227,
+      "grad_norm": 0.7710687106511338,
+      "learning_rate": 0.003,
+      "loss": 4.3371,
+      "step": 1227
+    },
+    {
+      "epoch": 0.01228,
+      "grad_norm": 0.6158196915186589,
+      "learning_rate": 0.003,
+      "loss": 4.3122,
+      "step": 1228
+    },
+    {
+      "epoch": 0.01229,
+      "grad_norm": 0.5647355795097493,
+      "learning_rate": 0.003,
+      "loss": 4.2952,
+      "step": 1229
+    },
+    {
+      "epoch": 0.0123,
+      "grad_norm": 0.5989687141441427,
+      "learning_rate": 0.003,
+      "loss": 4.3304,
+      "step": 1230
+    },
+    {
+      "epoch": 0.01231,
+      "grad_norm": 0.6070784312189734,
+      "learning_rate": 0.003,
+      "loss": 4.2993,
+      "step": 1231
+    },
+    {
+      "epoch": 0.01232,
+      "grad_norm": 0.6098189672837228,
+      "learning_rate": 0.003,
+      "loss": 4.276,
+      "step": 1232
+    },
+    {
+      "epoch": 0.01233,
+      "grad_norm": 0.6187452049200426,
+      "learning_rate": 0.003,
+      "loss": 4.2843,
+      "step": 1233
+    },
+    {
+      "epoch": 0.01234,
+      "grad_norm": 0.6847326322735632,
+      "learning_rate": 0.003,
+      "loss": 4.3016,
+      "step": 1234
+    },
+    {
+      "epoch": 0.01235,
+      "grad_norm": 0.7347950056629514,
+      "learning_rate": 0.003,
+      "loss": 4.2975,
+      "step": 1235
+    },
+    {
+      "epoch": 0.01236,
+      "grad_norm": 0.8404243381213906,
+      "learning_rate": 0.003,
+      "loss": 4.3037,
+      "step": 1236
+    },
+    {
+      "epoch": 0.01237,
+      "grad_norm": 1.1666025530795177,
+      "learning_rate": 0.003,
+      "loss": 4.3338,
+      "step": 1237
+    },
+    {
+      "epoch": 0.01238,
+      "grad_norm": 0.7706517401085584,
+      "learning_rate": 0.003,
+      "loss": 4.3184,
+      "step": 1238
+    },
+    {
+      "epoch": 0.01239,
+      "grad_norm": 0.8892529662707559,
+      "learning_rate": 0.003,
+      "loss": 4.3055,
+      "step": 1239
+    },
+    {
+      "epoch": 0.0124,
+      "grad_norm": 0.9665932126023419,
+      "learning_rate": 0.003,
+      "loss": 4.3004,
+      "step": 1240
+    },
+    {
+      "epoch": 0.01241,
+      "grad_norm": 1.0935280340663984,
+      "learning_rate": 0.003,
+      "loss": 4.3233,
+      "step": 1241
+    },
+    {
+      "epoch": 0.01242,
+      "grad_norm": 1.117726648141726,
+      "learning_rate": 0.003,
+      "loss": 4.3312,
+      "step": 1242
+    },
+    {
+      "epoch": 0.01243,
+      "grad_norm": 1.0553809973617114,
+      "learning_rate": 0.003,
+      "loss": 4.3181,
+      "step": 1243
+    },
+    {
+      "epoch": 0.01244,
+      "grad_norm": 0.9192132297518948,
+      "learning_rate": 0.003,
+      "loss": 4.3267,
+      "step": 1244
+    },
+    {
+      "epoch": 0.01245,
+      "grad_norm": 0.9382654947912115,
+      "learning_rate": 0.003,
+      "loss": 4.3187,
+      "step": 1245
+    },
+    {
+      "epoch": 0.01246,
+      "grad_norm": 1.0892504099540619,
+      "learning_rate": 0.003,
+      "loss": 4.3062,
+      "step": 1246
+    },
+    {
+      "epoch": 0.01247,
+      "grad_norm": 0.8993491326759976,
+      "learning_rate": 0.003,
+      "loss": 4.3511,
+      "step": 1247
+    },
+    {
+      "epoch": 0.01248,
+      "grad_norm": 0.908268119830267,
+      "learning_rate": 0.003,
+      "loss": 4.3379,
+      "step": 1248
+    },
+    {
+      "epoch": 0.01249,
+      "grad_norm": 0.8800131758337482,
+      "learning_rate": 0.003,
+      "loss": 4.3278,
+      "step": 1249
+    },
+    {
+      "epoch": 0.0125,
+      "grad_norm": 0.8337196500160362,
+      "learning_rate": 0.003,
+      "loss": 4.3052,
+      "step": 1250
+    },
+    {
+      "epoch": 0.01251,
+      "grad_norm": 0.9249131658343839,
+      "learning_rate": 0.003,
+      "loss": 4.3424,
+      "step": 1251
+    },
+    {
+      "epoch": 0.01252,
+      "grad_norm": 0.8524650338611843,
+      "learning_rate": 0.003,
+      "loss": 4.3172,
+      "step": 1252
+    },
+    {
+      "epoch": 0.01253,
+      "grad_norm": 0.7163285364378025,
+      "learning_rate": 0.003,
+      "loss": 4.2931,
+      "step": 1253
+    },
+    {
+      "epoch": 0.01254,
+      "grad_norm": 0.6991567907525832,
+      "learning_rate": 0.003,
+      "loss": 4.3341,
+      "step": 1254
+    },
+    {
+      "epoch": 0.01255,
+      "grad_norm": 0.6949992430981787,
+      "learning_rate": 0.003,
+      "loss": 4.2856,
+      "step": 1255
+    },
+    {
+      "epoch": 0.01256,
+      "grad_norm": 0.7100922258125638,
+      "learning_rate": 0.003,
+      "loss": 4.3118,
+      "step": 1256
+    },
+    {
+      "epoch": 0.01257,
+      "grad_norm": 0.6612217096643498,
+      "learning_rate": 0.003,
+      "loss": 4.2941,
+      "step": 1257
+    },
+    {
+      "epoch": 0.01258,
+      "grad_norm": 0.7747579760551054,
+      "learning_rate": 0.003,
+      "loss": 4.3107,
+      "step": 1258
+    },
+    {
+      "epoch": 0.01259,
+      "grad_norm": 1.124107981669201,
+      "learning_rate": 0.003,
+      "loss": 4.3222,
+      "step": 1259
+    },
+    {
+      "epoch": 0.0126,
+      "grad_norm": 0.9528743820284102,
+      "learning_rate": 0.003,
+      "loss": 4.3627,
+      "step": 1260
+    },
+    {
+      "epoch": 0.01261,
+      "grad_norm": 0.842570464638209,
+      "learning_rate": 0.003,
+      "loss": 4.3486,
+      "step": 1261
+    },
+    {
+      "epoch": 0.01262,
+      "grad_norm": 0.8393239407149024,
+      "learning_rate": 0.003,
+      "loss": 4.3397,
+      "step": 1262
+    },
+    {
+      "epoch": 0.01263,
+      "grad_norm": 0.8299261180456607,
+      "learning_rate": 0.003,
+      "loss": 4.3362,
+      "step": 1263
+    },
+    {
+      "epoch": 0.01264,
+      "grad_norm": 0.8141230470068299,
+      "learning_rate": 0.003,
+      "loss": 4.3187,
+      "step": 1264
+    },
+    {
+      "epoch": 0.01265,
+      "grad_norm": 1.0141550421015684,
+      "learning_rate": 0.003,
+      "loss": 4.3395,
+      "step": 1265
+    },
+    {
+      "epoch": 0.01266,
+      "grad_norm": 0.7612600647582426,
+      "learning_rate": 0.003,
+      "loss": 4.3297,
+      "step": 1266
+    },
+    {
+      "epoch": 0.01267,
+      "grad_norm": 0.6187340841665547,
+      "learning_rate": 0.003,
+      "loss": 4.3447,
+      "step": 1267
+    },
+    {
+      "epoch": 0.01268,
+      "grad_norm": 0.5822458314315527,
+      "learning_rate": 0.003,
+      "loss": 4.3275,
+      "step": 1268
+    },
+    {
+      "epoch": 0.01269,
+      "grad_norm": 0.4492593229301203,
+      "learning_rate": 0.003,
+      "loss": 4.2855,
+      "step": 1269
+    },
+    {
+      "epoch": 0.0127,
+      "grad_norm": 0.5016783253235925,
+      "learning_rate": 0.003,
+      "loss": 4.3213,
+      "step": 1270
+    },
+    {
+      "epoch": 0.01271,
+      "grad_norm": 0.44977006545763915,
+      "learning_rate": 0.003,
+      "loss": 4.297,
+      "step": 1271
+    },
+    {
+      "epoch": 0.01272,
+      "grad_norm": 0.43028280089396154,
+      "learning_rate": 0.003,
+      "loss": 4.291,
+      "step": 1272
+    },
+    {
+      "epoch": 0.01273,
+      "grad_norm": 0.4520162301705406,
+      "learning_rate": 0.003,
+      "loss": 4.3158,
+      "step": 1273
+    },
+    {
+      "epoch": 0.01274,
+      "grad_norm": 0.451409910695501,
+      "learning_rate": 0.003,
+      "loss": 4.3063,
+      "step": 1274
+    },
+    {
+      "epoch": 0.01275,
+      "grad_norm": 0.4633201534438662,
+      "learning_rate": 0.003,
+      "loss": 4.2756,
+      "step": 1275
+    },
+    {
+      "epoch": 0.01276,
+      "grad_norm": 0.4818283038515168,
+      "learning_rate": 0.003,
+      "loss": 4.2804,
+      "step": 1276
+    },
+    {
+      "epoch": 0.01277,
+      "grad_norm": 0.5612989347677225,
+      "learning_rate": 0.003,
+      "loss": 4.2895,
+      "step": 1277
+    },
+    {
+      "epoch": 0.01278,
+      "grad_norm": 0.6134565833773333,
+      "learning_rate": 0.003,
+      "loss": 4.2966,
+      "step": 1278
+    },
+    {
+      "epoch": 0.01279,
+      "grad_norm": 0.548830310802857,
+      "learning_rate": 0.003,
+      "loss": 4.2778,
+      "step": 1279
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 0.5192954436192417,
+      "learning_rate": 0.003,
+      "loss": 4.2736,
+      "step": 1280
+    },
+    {
+      "epoch": 0.01281,
+      "grad_norm": 0.7079203023923951,
+      "learning_rate": 0.003,
+      "loss": 4.2645,
+      "step": 1281
+    },
+    {
+      "epoch": 0.01282,
+      "grad_norm": 0.8640412732426497,
+      "learning_rate": 0.003,
+      "loss": 4.3216,
+      "step": 1282
+    },
+    {
+      "epoch": 0.01283,
+      "grad_norm": 0.7685939231737142,
+      "learning_rate": 0.003,
+      "loss": 4.2859,
+      "step": 1283
+    },
+    {
+      "epoch": 0.01284,
+      "grad_norm": 0.6000861053702143,
+      "learning_rate": 0.003,
+      "loss": 4.2792,
+      "step": 1284
+    },
+    {
+      "epoch": 0.01285,
+      "grad_norm": 0.7254306454469097,
+      "learning_rate": 0.003,
+      "loss": 4.2881,
+      "step": 1285
+    },
+    {
+      "epoch": 0.01286,
+      "grad_norm": 0.6733987067909273,
+      "learning_rate": 0.003,
+      "loss": 4.3069,
+      "step": 1286
+    },
+    {
+      "epoch": 0.01287,
+      "grad_norm": 0.6347430576087306,
+      "learning_rate": 0.003,
+      "loss": 4.2706,
+      "step": 1287
+    },
+    {
+      "epoch": 0.01288,
+      "grad_norm": 0.6742039696219508,
+      "learning_rate": 0.003,
+      "loss": 4.2877,
+      "step": 1288
+    },
+    {
+      "epoch": 0.01289,
+      "grad_norm": 0.7964754563506287,
+      "learning_rate": 0.003,
+      "loss": 4.2851,
+      "step": 1289
+    },
+    {
+      "epoch": 0.0129,
+      "grad_norm": 0.7466700913788555,
+      "learning_rate": 0.003,
+      "loss": 4.3108,
+      "step": 1290
+    },
+    {
+      "epoch": 0.01291,
+      "grad_norm": 0.7257835946382555,
+      "learning_rate": 0.003,
+      "loss": 4.2878,
+      "step": 1291
+    },
+    {
+      "epoch": 0.01292,
+      "grad_norm": 0.7356576793031216,
+      "learning_rate": 0.003,
+      "loss": 4.3222,
+      "step": 1292
+    },
+    {
+      "epoch": 0.01293,
+      "grad_norm": 0.7652258994712906,
+      "learning_rate": 0.003,
+      "loss": 4.312,
+      "step": 1293
+    },
+    {
+      "epoch": 0.01294,
+      "grad_norm": 0.8074663392977137,
+      "learning_rate": 0.003,
+      "loss": 4.2851,
+      "step": 1294
+    },
+    {
+      "epoch": 0.01295,
+      "grad_norm": 0.8419246282427761,
+      "learning_rate": 0.003,
+      "loss": 4.3171,
+      "step": 1295
+    },
+    {
+      "epoch": 0.01296,
+      "grad_norm": 0.9067886743130651,
+      "learning_rate": 0.003,
+      "loss": 4.2936,
+      "step": 1296
+    },
+    {
+      "epoch": 0.01297,
+      "grad_norm": 0.9867944428505203,
+      "learning_rate": 0.003,
+      "loss": 4.3168,
+      "step": 1297
+    },
+    {
+      "epoch": 0.01298,
+      "grad_norm": 0.9008581643124973,
+      "learning_rate": 0.003,
+      "loss": 4.2927,
+      "step": 1298
+    },
+    {
+      "epoch": 0.01299,
+      "grad_norm": 0.9659427448976995,
+      "learning_rate": 0.003,
+      "loss": 4.3085,
+      "step": 1299
+    },
+    {
+      "epoch": 0.013,
+      "grad_norm": 1.004138764919463,
+      "learning_rate": 0.003,
+      "loss": 4.3011,
+      "step": 1300
+    },
+    {
+      "epoch": 0.01301,
+      "grad_norm": 0.9518499756671687,
+      "learning_rate": 0.003,
+      "loss": 4.3191,
+      "step": 1301
+    },
+    {
+      "epoch": 0.01302,
+      "grad_norm": 0.8711824338075493,
+      "learning_rate": 0.003,
+      "loss": 4.361,
+      "step": 1302
+    },
+    {
+      "epoch": 0.01303,
+      "grad_norm": 0.8676764239234932,
+      "learning_rate": 0.003,
+      "loss": 4.3063,
+      "step": 1303
+    },
+    {
+      "epoch": 0.01304,
+      "grad_norm": 0.7125360490503003,
+      "learning_rate": 0.003,
+      "loss": 4.3284,
+      "step": 1304
+    },
+    {
+      "epoch": 0.01305,
+      "grad_norm": 0.6046564741110017,
+      "learning_rate": 0.003,
+      "loss": 4.3089,
+      "step": 1305
+    },
+    {
+      "epoch": 0.01306,
+      "grad_norm": 0.5647927538938984,
+      "learning_rate": 0.003,
+      "loss": 4.2992,
+      "step": 1306
+    },
+    {
+      "epoch": 0.01307,
+      "grad_norm": 0.5913576309809552,
+      "learning_rate": 0.003,
+      "loss": 4.3149,
+      "step": 1307
+    },
+    {
+      "epoch": 0.01308,
+      "grad_norm": 0.5924251650199668,
+      "learning_rate": 0.003,
+      "loss": 4.2932,
+      "step": 1308
+    },
+    {
+      "epoch": 0.01309,
+      "grad_norm": 0.507504674715556,
+      "learning_rate": 0.003,
+      "loss": 4.326,
+      "step": 1309
+    },
+    {
+      "epoch": 0.0131,
+      "grad_norm": 0.6387323300673942,
+      "learning_rate": 0.003,
+      "loss": 4.2756,
+      "step": 1310
+    },
+    {
+      "epoch": 0.01311,
+      "grad_norm": 0.7281996988052075,
+      "learning_rate": 0.003,
+      "loss": 4.3202,
+      "step": 1311
+    },
+    {
+      "epoch": 0.01312,
+      "grad_norm": 0.8834351506579509,
+      "learning_rate": 0.003,
+      "loss": 4.2908,
+      "step": 1312
+    },
+    {
+      "epoch": 0.01313,
+      "grad_norm": 0.840185011015699,
+      "learning_rate": 0.003,
+      "loss": 4.2898,
+      "step": 1313
+    },
+    {
+      "epoch": 0.01314,
+      "grad_norm": 0.6830490768340751,
+      "learning_rate": 0.003,
+      "loss": 4.2745,
+      "step": 1314
+    },
+    {
+      "epoch": 0.01315,
+      "grad_norm": 0.6358981723627178,
+      "learning_rate": 0.003,
+      "loss": 4.2992,
+      "step": 1315
+    },
+    {
+      "epoch": 0.01316,
+      "grad_norm": 0.6007791887068558,
+      "learning_rate": 0.003,
+      "loss": 4.2829,
+      "step": 1316
+    },
+    {
+      "epoch": 0.01317,
+      "grad_norm": 0.5048125779943076,
+      "learning_rate": 0.003,
+      "loss": 4.3051,
+      "step": 1317
+    },
+    {
+      "epoch": 0.01318,
+      "grad_norm": 0.49450675942854344,
+      "learning_rate": 0.003,
+      "loss": 4.2837,
+      "step": 1318
+    },
+    {
+      "epoch": 0.01319,
+      "grad_norm": 0.5258905508186564,
+      "learning_rate": 0.003,
+      "loss": 4.292,
+      "step": 1319
+    },
+    {
+      "epoch": 0.0132,
+      "grad_norm": 0.5585710036479011,
+      "learning_rate": 0.003,
+      "loss": 4.2888,
+      "step": 1320
+    },
+    {
+      "epoch": 0.01321,
+      "grad_norm": 0.5365220915283813,
+      "learning_rate": 0.003,
+      "loss": 4.2608,
+      "step": 1321
+    },
+    {
+      "epoch": 0.01322,
+      "grad_norm": 0.5557887511106342,
+      "learning_rate": 0.003,
+      "loss": 4.292,
+      "step": 1322
+    },
+    {
+      "epoch": 0.01323,
+      "grad_norm": 0.5531377716759448,
+      "learning_rate": 0.003,
+      "loss": 4.2837,
+      "step": 1323
+    },
+    {
+      "epoch": 0.01324,
+      "grad_norm": 0.48156733948461256,
+      "learning_rate": 0.003,
+      "loss": 4.2672,
+      "step": 1324
+    },
+    {
+      "epoch": 0.01325,
+      "grad_norm": 0.4219074369900267,
+      "learning_rate": 0.003,
+      "loss": 4.2693,
+      "step": 1325
+    },
+    {
+      "epoch": 0.01326,
+      "grad_norm": 0.43890778305013683,
+      "learning_rate": 0.003,
+      "loss": 4.2559,
+      "step": 1326
+    },
+    {
+      "epoch": 0.01327,
+      "grad_norm": 0.4164892745664214,
+      "learning_rate": 0.003,
+      "loss": 4.2724,
+      "step": 1327
+    },
+    {
+      "epoch": 0.01328,
+      "grad_norm": 0.49229273243603294,
+      "learning_rate": 0.003,
+      "loss": 4.2646,
+      "step": 1328
+    },
+    {
+      "epoch": 0.01329,
+      "grad_norm": 0.6121730970661665,
+      "learning_rate": 0.003,
+      "loss": 4.2329,
+      "step": 1329
+    },
+    {
+      "epoch": 0.0133,
+      "grad_norm": 0.8848354653817555,
+      "learning_rate": 0.003,
+      "loss": 4.282,
+      "step": 1330
+    },
+    {
+      "epoch": 0.01331,
+      "grad_norm": 1.0334105334221095,
+      "learning_rate": 0.003,
+      "loss": 4.2945,
+      "step": 1331
+    },
+    {
+      "epoch": 0.01332,
+      "grad_norm": 0.8234894915328277,
+      "learning_rate": 0.003,
+      "loss": 4.2994,
+      "step": 1332
+    },
+    {
+      "epoch": 0.01333,
+      "grad_norm": 0.7856986790322864,
+      "learning_rate": 0.003,
+      "loss": 4.2716,
+      "step": 1333
+    },
+    {
+      "epoch": 0.01334,
+      "grad_norm": 0.7455452553988269,
+      "learning_rate": 0.003,
+      "loss": 4.2662,
+      "step": 1334
+    },
+    {
+      "epoch": 0.01335,
+      "grad_norm": 0.623751194689146,
+      "learning_rate": 0.003,
+      "loss": 4.2787,
+      "step": 1335
+    },
+    {
+      "epoch": 0.01336,
+      "grad_norm": 0.8238963440270202,
+      "learning_rate": 0.003,
+      "loss": 4.2839,
+      "step": 1336
+    },
+    {
+      "epoch": 0.01337,
+      "grad_norm": 0.8159528784128957,
+      "learning_rate": 0.003,
+      "loss": 4.3035,
+      "step": 1337
+    },
+    {
+      "epoch": 0.01338,
+      "grad_norm": 0.691075290250966,
+      "learning_rate": 0.003,
+      "loss": 4.2694,
+      "step": 1338
+    },
+    {
+      "epoch": 0.01339,
+      "grad_norm": 0.6872888519447673,
+      "learning_rate": 0.003,
+      "loss": 4.2893,
+      "step": 1339
+    },
+    {
+      "epoch": 0.0134,
+      "grad_norm": 0.6938872016813308,
+      "learning_rate": 0.003,
+      "loss": 4.2701,
+      "step": 1340
+    },
+    {
+      "epoch": 0.01341,
+      "grad_norm": 0.7051901323106005,
+      "learning_rate": 0.003,
+      "loss": 4.2883,
+      "step": 1341
+    },
+    {
+      "epoch": 0.01342,
+      "grad_norm": 0.7365511131614688,
+      "learning_rate": 0.003,
+      "loss": 4.257,
+      "step": 1342
+    },
+    {
+      "epoch": 0.01343,
+      "grad_norm": 0.7564141855011662,
+      "learning_rate": 0.003,
+      "loss": 4.2701,
+      "step": 1343
+    },
+    {
+      "epoch": 0.01344,
+      "grad_norm": 0.695387503279461,
+      "learning_rate": 0.003,
+      "loss": 4.2649,
+      "step": 1344
+    },
+    {
+      "epoch": 0.01345,
+      "grad_norm": 0.6914588150610467,
+      "learning_rate": 0.003,
+      "loss": 4.3033,
+      "step": 1345
+    },
+    {
+      "epoch": 0.01346,
+      "grad_norm": 0.6768362060970508,
+      "learning_rate": 0.003,
+      "loss": 4.277,
+      "step": 1346
+    },
+    {
+      "epoch": 0.01347,
+      "grad_norm": 0.5492073753536512,
+      "learning_rate": 0.003,
+      "loss": 4.2633,
+      "step": 1347
+    },
+    {
+      "epoch": 0.01348,
+      "grad_norm": 0.5626283903224933,
+      "learning_rate": 0.003,
+      "loss": 4.2635,
+      "step": 1348
+    },
+    {
+      "epoch": 0.01349,
+      "grad_norm": 0.6438865489267382,
+      "learning_rate": 0.003,
+      "loss": 4.2856,
+      "step": 1349
+    },
+    {
+      "epoch": 0.0135,
+      "grad_norm": 0.7199005579630849,
+      "learning_rate": 0.003,
+      "loss": 4.2796,
+      "step": 1350
+    },
+    {
+      "epoch": 0.01351,
+      "grad_norm": 0.5428932527008233,
+      "learning_rate": 0.003,
+      "loss": 4.3038,
+      "step": 1351
+    },
+    {
+      "epoch": 0.01352,
+      "grad_norm": 0.5284901745728212,
+      "learning_rate": 0.003,
+      "loss": 4.2779,
+      "step": 1352
+    },
+    {
+      "epoch": 0.01353,
+      "grad_norm": 0.5600878945602594,
+      "learning_rate": 0.003,
+      "loss": 4.2635,
+      "step": 1353
+    },
+    {
+      "epoch": 0.01354,
+      "grad_norm": 0.6391444612725596,
+      "learning_rate": 0.003,
+      "loss": 4.2525,
+      "step": 1354
+    },
+    {
+      "epoch": 0.01355,
+      "grad_norm": 0.5641606375474643,
+      "learning_rate": 0.003,
+      "loss": 4.2634,
+      "step": 1355
+    },
+    {
+      "epoch": 0.01356,
+      "grad_norm": 0.5080744885475224,
+      "learning_rate": 0.003,
+      "loss": 4.2545,
+      "step": 1356
+    },
+    {
+      "epoch": 0.01357,
+      "grad_norm": 0.5093639445431166,
+      "learning_rate": 0.003,
+      "loss": 4.2441,
+      "step": 1357
+    },
+    {
+      "epoch": 0.01358,
+      "grad_norm": 0.5049437394862782,
+      "learning_rate": 0.003,
+      "loss": 4.2383,
+      "step": 1358
+    },
+    {
+      "epoch": 0.01359,
+      "grad_norm": 0.534762392832106,
+      "learning_rate": 0.003,
+      "loss": 4.2789,
+      "step": 1359
+    },
+    {
+      "epoch": 0.0136,
+      "grad_norm": 0.5654562399234901,
+      "learning_rate": 0.003,
+      "loss": 4.2447,
+      "step": 1360
+    },
+    {
+      "epoch": 0.01361,
+      "grad_norm": 0.7911685516807686,
+      "learning_rate": 0.003,
+      "loss": 4.2666,
+      "step": 1361
+    },
+    {
+      "epoch": 0.01362,
+      "grad_norm": 1.133172158468294,
+      "learning_rate": 0.003,
+      "loss": 4.3073,
+      "step": 1362
+    },
+    {
+      "epoch": 0.01363,
+      "grad_norm": 0.9631930238400119,
+      "learning_rate": 0.003,
+      "loss": 4.2987,
+      "step": 1363
+    },
+    {
+      "epoch": 0.01364,
+      "grad_norm": 0.7443319461425727,
+      "learning_rate": 0.003,
+      "loss": 4.2839,
+      "step": 1364
+    },
+    {
+      "epoch": 0.01365,
+      "grad_norm": 0.7368707977522249,
+      "learning_rate": 0.003,
+      "loss": 4.3041,
+      "step": 1365
+    },
+    {
+      "epoch": 0.01366,
+      "grad_norm": 0.6506933778775009,
+      "learning_rate": 0.003,
+      "loss": 4.255,
+      "step": 1366
+    },
+    {
+      "epoch": 0.01367,
+      "grad_norm": 0.7425630695378723,
+      "learning_rate": 0.003,
+      "loss": 4.3072,
+      "step": 1367
+    },
+    {
+      "epoch": 0.01368,
+      "grad_norm": 0.7495522896004446,
+      "learning_rate": 0.003,
+      "loss": 4.2833,
+      "step": 1368
+    },
+    {
+      "epoch": 0.01369,
+      "grad_norm": 0.7542187487500857,
+      "learning_rate": 0.003,
+      "loss": 4.2721,
+      "step": 1369
+    },
+    {
+      "epoch": 0.0137,
+      "grad_norm": 0.9330809300322568,
+      "learning_rate": 0.003,
+      "loss": 4.3029,
+      "step": 1370
+    },
+    {
+      "epoch": 0.01371,
+      "grad_norm": 1.2015218016331504,
+      "learning_rate": 0.003,
+      "loss": 4.3156,
+      "step": 1371
+    },
+    {
+      "epoch": 0.01372,
+      "grad_norm": 1.0453049109497266,
+      "learning_rate": 0.003,
+      "loss": 4.2577,
+      "step": 1372
+    },
+    {
+      "epoch": 0.01373,
+      "grad_norm": 1.086666562724284,
+      "learning_rate": 0.003,
+      "loss": 4.308,
+      "step": 1373
+    },
+    {
+      "epoch": 0.01374,
+      "grad_norm": 1.0320592526365153,
+      "learning_rate": 0.003,
+      "loss": 4.3227,
+      "step": 1374
+    },
+    {
+      "epoch": 0.01375,
+      "grad_norm": 1.1967086046120219,
+      "learning_rate": 0.003,
+      "loss": 4.2884,
+      "step": 1375
+    },
+    {
+      "epoch": 0.01376,
+      "grad_norm": 0.8548157633990001,
+      "learning_rate": 0.003,
+      "loss": 4.307,
+      "step": 1376
+    },
+    {
+      "epoch": 0.01377,
+      "grad_norm": 0.9102072373918407,
+      "learning_rate": 0.003,
+      "loss": 4.2985,
+      "step": 1377
+    },
+    {
+      "epoch": 0.01378,
+      "grad_norm": 0.9843975361934545,
+      "learning_rate": 0.003,
+      "loss": 4.3358,
+      "step": 1378
+    },
+    {
+      "epoch": 0.01379,
+      "grad_norm": 1.0283723711953892,
+      "learning_rate": 0.003,
+      "loss": 4.3102,
+      "step": 1379
+    },
+    {
+      "epoch": 0.0138,
+      "grad_norm": 0.9006954176701093,
+      "learning_rate": 0.003,
+      "loss": 4.3045,
+      "step": 1380
+    },
+    {
+      "epoch": 0.01381,
+      "grad_norm": 1.123717795950056,
+      "learning_rate": 0.003,
+      "loss": 4.3217,
+      "step": 1381
+    },
+    {
+      "epoch": 0.01382,
+      "grad_norm": 1.0986902300743402,
+      "learning_rate": 0.003,
+      "loss": 4.3218,
+      "step": 1382
+    },
+    {
+      "epoch": 0.01383,
+      "grad_norm": 0.9410643097534779,
+      "learning_rate": 0.003,
+      "loss": 4.3113,
+      "step": 1383
+    },
+    {
+      "epoch": 0.01384,
+      "grad_norm": 0.8436673398281763,
+      "learning_rate": 0.003,
+      "loss": 4.3327,
+      "step": 1384
+    },
+    {
+      "epoch": 0.01385,
+      "grad_norm": 0.8422095588989059,
+      "learning_rate": 0.003,
+      "loss": 4.3233,
+      "step": 1385
+    },
+    {
+      "epoch": 0.01386,
+      "grad_norm": 0.9633211233231406,
+      "learning_rate": 0.003,
+      "loss": 4.3006,
+      "step": 1386
+    },
+    {
+      "epoch": 0.01387,
+      "grad_norm": 1.2716300561274656,
+      "learning_rate": 0.003,
+      "loss": 4.3052,
+      "step": 1387
+    },
+    {
+      "epoch": 0.01388,
+      "grad_norm": 0.786746450686237,
+      "learning_rate": 0.003,
+      "loss": 4.33,
+      "step": 1388
+    },
+    {
+      "epoch": 0.01389,
+      "grad_norm": 0.763957561776965,
+      "learning_rate": 0.003,
+      "loss": 4.3069,
+      "step": 1389
+    },
+    {
+      "epoch": 0.0139,
+      "grad_norm": 0.7874363277644975,
+      "learning_rate": 0.003,
+      "loss": 4.3041,
+      "step": 1390
+    },
+    {
+      "epoch": 0.01391,
+      "grad_norm": 0.7493071050038139,
+      "learning_rate": 0.003,
+      "loss": 4.3036,
+      "step": 1391
+    },
+    {
+      "epoch": 0.01392,
+      "grad_norm": 0.5930612778435612,
+      "learning_rate": 0.003,
+      "loss": 4.3076,
+      "step": 1392
+    },
+    {
+      "epoch": 0.01393,
+      "grad_norm": 0.5012167316836135,
+      "learning_rate": 0.003,
+      "loss": 4.2738,
+      "step": 1393
+    },
+    {
+      "epoch": 0.01394,
+      "grad_norm": 0.4948661937091383,
+      "learning_rate": 0.003,
+      "loss": 4.2568,
+      "step": 1394
+    },
+    {
+      "epoch": 0.01395,
+      "grad_norm": 0.4622888611133768,
+      "learning_rate": 0.003,
+      "loss": 4.2868,
+      "step": 1395
+    },
+    {
+      "epoch": 0.01396,
+      "grad_norm": 0.4205886993889684,
+      "learning_rate": 0.003,
+      "loss": 4.275,
+      "step": 1396
+    },
+    {
+      "epoch": 0.01397,
+      "grad_norm": 0.47767051412590744,
+      "learning_rate": 0.003,
+      "loss": 4.2862,
+      "step": 1397
+    },
+    {
+      "epoch": 0.01398,
+      "grad_norm": 0.417341308437963,
+      "learning_rate": 0.003,
+      "loss": 4.2647,
+      "step": 1398
+    },
+    {
+      "epoch": 0.01399,
+      "grad_norm": 0.38332601666132154,
+      "learning_rate": 0.003,
+      "loss": 4.2835,
+      "step": 1399
+    },
+    {
+      "epoch": 0.014,
+      "grad_norm": 0.49255586643940574,
+      "learning_rate": 0.003,
+      "loss": 4.2444,
+      "step": 1400
+    },
+    {
+      "epoch": 0.01401,
+      "grad_norm": 0.8450629843994065,
+      "learning_rate": 0.003,
+      "loss": 4.2739,
+      "step": 1401
+    },
+    {
+      "epoch": 0.01402,
+      "grad_norm": 1.0913061976723808,
+      "learning_rate": 0.003,
+      "loss": 4.3172,
+      "step": 1402
+    },
+    {
+      "epoch": 0.01403,
+      "grad_norm": 0.6222948640126749,
+      "learning_rate": 0.003,
+      "loss": 4.2634,
+      "step": 1403
+    },
+    {
+      "epoch": 0.01404,
+      "grad_norm": 0.6692390535634434,
+      "learning_rate": 0.003,
+      "loss": 4.291,
+      "step": 1404
+    },
+    {
+      "epoch": 0.01405,
+      "grad_norm": 0.531565118139576,
+      "learning_rate": 0.003,
+      "loss": 4.2997,
+      "step": 1405
+    },
+    {
+      "epoch": 0.01406,
+      "grad_norm": 0.4824477961295427,
+      "learning_rate": 0.003,
+      "loss": 4.2779,
+      "step": 1406
+    },
+    {
+      "epoch": 0.01407,
+      "grad_norm": 0.5759838031356613,
+      "learning_rate": 0.003,
+      "loss": 4.2942,
+      "step": 1407
+    },
+    {
+      "epoch": 0.01408,
+      "grad_norm": 0.4476558981995588,
+      "learning_rate": 0.003,
+      "loss": 4.245,
+      "step": 1408
+    },
+    {
+      "epoch": 0.01409,
+      "grad_norm": 0.43291671999910936,
+      "learning_rate": 0.003,
+      "loss": 4.2753,
+      "step": 1409
+    },
+    {
+      "epoch": 0.0141,
+      "grad_norm": 0.48354199581441126,
+      "learning_rate": 0.003,
+      "loss": 4.2565,
+      "step": 1410
+    },
+    {
+      "epoch": 0.01411,
+      "grad_norm": 0.4777579756136276,
+      "learning_rate": 0.003,
+      "loss": 4.2302,
+      "step": 1411
+    },
+    {
+      "epoch": 0.01412,
+      "grad_norm": 0.5224654831649898,
+      "learning_rate": 0.003,
+      "loss": 4.2693,
+      "step": 1412
+    },
+    {
+      "epoch": 0.01413,
+      "grad_norm": 0.5669597734894755,
+      "learning_rate": 0.003,
+      "loss": 4.2639,
+      "step": 1413
+    },
+    {
+      "epoch": 0.01414,
+      "grad_norm": 0.5645597477136786,
+      "learning_rate": 0.003,
+      "loss": 4.2453,
+      "step": 1414
+    },
+    {
+      "epoch": 0.01415,
+      "grad_norm": 0.5887759339552011,
+      "learning_rate": 0.003,
+      "loss": 4.238,
+      "step": 1415
+    },
+    {
+      "epoch": 0.01416,
+      "grad_norm": 0.6104838140849604,
+      "learning_rate": 0.003,
+      "loss": 4.2641,
+      "step": 1416
+    },
+    {
+      "epoch": 0.01417,
+      "grad_norm": 0.6908167320390044,
+      "learning_rate": 0.003,
+      "loss": 4.2559,
+      "step": 1417
+    },
+    {
+      "epoch": 0.01418,
+      "grad_norm": 0.8241127076876378,
+      "learning_rate": 0.003,
+      "loss": 4.28,
+      "step": 1418
+    },
+    {
+      "epoch": 0.01419,
+      "grad_norm": 0.8554302831681504,
+      "learning_rate": 0.003,
+      "loss": 4.2733,
+      "step": 1419
+    },
+    {
+      "epoch": 0.0142,
+      "grad_norm": 0.8153359619668534,
+      "learning_rate": 0.003,
+      "loss": 4.28,
+      "step": 1420
+    },
+    {
+      "epoch": 0.01421,
+      "grad_norm": 0.808004884443839,
+      "learning_rate": 0.003,
+      "loss": 4.2527,
+      "step": 1421
+    },
+    {
+      "epoch": 0.01422,
+      "grad_norm": 0.6302573298023686,
+      "learning_rate": 0.003,
+      "loss": 4.2516,
+      "step": 1422
+    },
+    {
+      "epoch": 0.01423,
+      "grad_norm": 0.6834582033971294,
+      "learning_rate": 0.003,
+      "loss": 4.2585,
+      "step": 1423
+    },
+    {
+      "epoch": 0.01424,
+      "grad_norm": 0.7665806417295189,
+      "learning_rate": 0.003,
+      "loss": 4.2761,
+      "step": 1424
+    },
+    {
+      "epoch": 0.01425,
+      "grad_norm": 0.8479056405573007,
+      "learning_rate": 0.003,
+      "loss": 4.2857,
+      "step": 1425
+    },
+    {
+      "epoch": 0.01426,
+      "grad_norm": 0.8307259541179206,
+      "learning_rate": 0.003,
+      "loss": 4.2636,
+      "step": 1426
+    },
+    {
+      "epoch": 0.01427,
+      "grad_norm": 0.7328418116871758,
+      "learning_rate": 0.003,
+      "loss": 4.257,
+      "step": 1427
+    },
+    {
+      "epoch": 0.01428,
+      "grad_norm": 0.834478644245489,
+      "learning_rate": 0.003,
+      "loss": 4.2596,
+      "step": 1428
+    },
+    {
+      "epoch": 0.01429,
+      "grad_norm": 0.8679311866430979,
+      "learning_rate": 0.003,
+      "loss": 4.2742,
+      "step": 1429
+    },
+    {
+      "epoch": 0.0143,
+      "grad_norm": 0.9567780417694972,
+      "learning_rate": 0.003,
+      "loss": 4.2902,
+      "step": 1430
+    },
+    {
+      "epoch": 0.01431,
+      "grad_norm": 0.99274156172833,
+      "learning_rate": 0.003,
+      "loss": 4.2901,
+      "step": 1431
+    },
+    {
+      "epoch": 0.01432,
+      "grad_norm": 0.9575505507003149,
+      "learning_rate": 0.003,
+      "loss": 4.2782,
+      "step": 1432
+    },
+    {
+      "epoch": 0.01433,
+      "grad_norm": 1.0120649680762432,
+      "learning_rate": 0.003,
+      "loss": 4.2802,
+      "step": 1433
+    },
+    {
+      "epoch": 0.01434,
+      "grad_norm": 0.9698542777423387,
+      "learning_rate": 0.003,
+      "loss": 4.2694,
+      "step": 1434
+    },
+    {
+      "epoch": 0.01435,
+      "grad_norm": 0.8592446572371273,
+      "learning_rate": 0.003,
+      "loss": 4.2745,
+      "step": 1435
+    },
+    {
+      "epoch": 0.01436,
+      "grad_norm": 0.7299491282894184,
+      "learning_rate": 0.003,
+      "loss": 4.2863,
+      "step": 1436
+    },
+    {
+      "epoch": 0.01437,
+      "grad_norm": 0.6523703034324407,
+      "learning_rate": 0.003,
+      "loss": 4.296,
+      "step": 1437
+    },
+    {
+      "epoch": 0.01438,
+      "grad_norm": 0.6201165515106615,
+      "learning_rate": 0.003,
+      "loss": 4.2864,
+      "step": 1438
+    },
+    {
+      "epoch": 0.01439,
+      "grad_norm": 0.6536973612494624,
+      "learning_rate": 0.003,
+      "loss": 4.25,
+      "step": 1439
+    },
+    {
+      "epoch": 0.0144,
+      "grad_norm": 0.729776581664813,
+      "learning_rate": 0.003,
+      "loss": 4.2648,
+      "step": 1440
+    },
+    {
+      "epoch": 0.01441,
+      "grad_norm": 0.7206829643898741,
+      "learning_rate": 0.003,
+      "loss": 4.2702,
+      "step": 1441
+    },
+    {
+      "epoch": 0.01442,
+      "grad_norm": 0.7419558696417659,
+      "learning_rate": 0.003,
+      "loss": 4.2561,
+      "step": 1442
+    },
+    {
+      "epoch": 0.01443,
+      "grad_norm": 0.8093095130841327,
+      "learning_rate": 0.003,
+      "loss": 4.3153,
+      "step": 1443
+    },
+    {
+      "epoch": 0.01444,
+      "grad_norm": 0.7933277975079275,
+      "learning_rate": 0.003,
+      "loss": 4.2825,
+      "step": 1444
+    },
+    {
+      "epoch": 0.01445,
+      "grad_norm": 0.7264821716906565,
+      "learning_rate": 0.003,
+      "loss": 4.2399,
+      "step": 1445
+    },
+    {
+      "epoch": 0.01446,
+      "grad_norm": 0.6240641235999153,
+      "learning_rate": 0.003,
+      "loss": 4.2506,
+      "step": 1446
+    },
+    {
+      "epoch": 0.01447,
+      "grad_norm": 0.5288921284639453,
+      "learning_rate": 0.003,
+      "loss": 4.256,
+      "step": 1447
+    },
+    {
+      "epoch": 0.01448,
+      "grad_norm": 0.534171003657093,
+      "learning_rate": 0.003,
+      "loss": 4.2664,
+      "step": 1448
+    },
+    {
+      "epoch": 0.01449,
+      "grad_norm": 0.504252276028137,
+      "learning_rate": 0.003,
+      "loss": 4.2565,
+      "step": 1449
+    },
+    {
+      "epoch": 0.0145,
+      "grad_norm": 0.44556324681710546,
+      "learning_rate": 0.003,
+      "loss": 4.27,
+      "step": 1450
+    },
+    {
+      "epoch": 0.01451,
+      "grad_norm": 0.4269835752600227,
+      "learning_rate": 0.003,
+      "loss": 4.2459,
+      "step": 1451
+    },
+    {
+      "epoch": 0.01452,
+      "grad_norm": 0.43297277393213274,
+      "learning_rate": 0.003,
+      "loss": 4.2457,
+      "step": 1452
+    },
+    {
+      "epoch": 0.01453,
+      "grad_norm": 0.48066055951097997,
+      "learning_rate": 0.003,
+      "loss": 4.2578,
+      "step": 1453
+    },
+    {
+      "epoch": 0.01454,
+      "grad_norm": 0.7307010489462921,
+      "learning_rate": 0.003,
+      "loss": 4.2308,
+      "step": 1454
+    },
+    {
+      "epoch": 0.01455,
+      "grad_norm": 1.061032754030789,
+      "learning_rate": 0.003,
+      "loss": 4.2921,
+      "step": 1455
+    },
+    {
+      "epoch": 0.01456,
+      "grad_norm": 0.8307341818819957,
+      "learning_rate": 0.003,
+      "loss": 4.2586,
+      "step": 1456
+    },
+    {
+      "epoch": 0.01457,
+      "grad_norm": 0.6035503218291812,
+      "learning_rate": 0.003,
+      "loss": 4.2382,
+      "step": 1457
+    },
+    {
+      "epoch": 0.01458,
+      "grad_norm": 0.6933692267109066,
+      "learning_rate": 0.003,
+      "loss": 4.242,
+      "step": 1458
+    },
+    {
+      "epoch": 0.01459,
+      "grad_norm": 0.6220443333798514,
+      "learning_rate": 0.003,
+      "loss": 4.2607,
+      "step": 1459
+    },
+    {
+      "epoch": 0.0146,
+      "grad_norm": 0.8143223428105976,
+      "learning_rate": 0.003,
+      "loss": 4.2715,
+      "step": 1460
+    },
+    {
+      "epoch": 0.01461,
+      "grad_norm": 0.7987144915177533,
+      "learning_rate": 0.003,
+      "loss": 4.2513,
+      "step": 1461
+    },
+    {
+      "epoch": 0.01462,
+      "grad_norm": 0.6408744129720689,
+      "learning_rate": 0.003,
+      "loss": 4.2512,
+      "step": 1462
+    },
+    {
+      "epoch": 0.01463,
+      "grad_norm": 0.7757864438293093,
+      "learning_rate": 0.003,
+      "loss": 4.2548,
+      "step": 1463
+    },
+    {
+      "epoch": 0.01464,
+      "grad_norm": 0.7699033474111219,
+      "learning_rate": 0.003,
+      "loss": 4.2388,
+      "step": 1464
+    },
+    {
+      "epoch": 0.01465,
+      "grad_norm": 0.5797229129062035,
+      "learning_rate": 0.003,
+      "loss": 4.2736,
+      "step": 1465
+    },
+    {
+      "epoch": 0.01466,
+      "grad_norm": 0.611021916593894,
+      "learning_rate": 0.003,
+      "loss": 4.238,
+      "step": 1466
+    },
+    {
+      "epoch": 0.01467,
+      "grad_norm": 0.7716393804810123,
+      "learning_rate": 0.003,
+      "loss": 4.2809,
+      "step": 1467
+    },
+    {
+      "epoch": 0.01468,
+      "grad_norm": 0.684598948135191,
+      "learning_rate": 0.003,
+      "loss": 4.2576,
+      "step": 1468
+    },
+    {
+      "epoch": 0.01469,
+      "grad_norm": 0.5983303592308452,
+      "learning_rate": 0.003,
+      "loss": 4.2537,
+      "step": 1469
+    },
+    {
+      "epoch": 0.0147,
+      "grad_norm": 0.5951720244159906,
+      "learning_rate": 0.003,
+      "loss": 4.246,
+      "step": 1470
+    },
+    {
+      "epoch": 0.01471,
+      "grad_norm": 0.595230494931454,
+      "learning_rate": 0.003,
+      "loss": 4.2579,
+      "step": 1471
+    },
+    {
+      "epoch": 0.01472,
+      "grad_norm": 0.5179540783713997,
+      "learning_rate": 0.003,
+      "loss": 4.2385,
+      "step": 1472
+    },
+    {
+      "epoch": 0.01473,
+      "grad_norm": 0.513255678190744,
+      "learning_rate": 0.003,
+      "loss": 4.2524,
+      "step": 1473
+    },
+    {
+      "epoch": 0.01474,
+      "grad_norm": 0.5312567488001211,
+      "learning_rate": 0.003,
+      "loss": 4.2368,
+      "step": 1474
+    },
+    {
+      "epoch": 0.01475,
+      "grad_norm": 0.5231277112786824,
+      "learning_rate": 0.003,
+      "loss": 4.2547,
+      "step": 1475
+    },
+    {
+      "epoch": 0.01476,
+      "grad_norm": 0.524915592276574,
+      "learning_rate": 0.003,
+      "loss": 4.2412,
+      "step": 1476
+    },
+    {
+      "epoch": 0.01477,
+      "grad_norm": 0.4868094804842935,
+      "learning_rate": 0.003,
+      "loss": 4.2453,
+      "step": 1477
+    },
+    {
+      "epoch": 0.01478,
+      "grad_norm": 0.46072622581307954,
+      "learning_rate": 0.003,
+      "loss": 4.2277,
+      "step": 1478
+    },
+    {
+      "epoch": 0.01479,
+      "grad_norm": 0.5672195126213446,
+      "learning_rate": 0.003,
+      "loss": 4.2625,
+      "step": 1479
+    },
+    {
+      "epoch": 0.0148,
+      "grad_norm": 0.7550630041249586,
+      "learning_rate": 0.003,
+      "loss": 4.2383,
+      "step": 1480
+    },
+    {
+      "epoch": 0.01481,
+      "grad_norm": 0.9561266542090932,
+      "learning_rate": 0.003,
+      "loss": 4.2523,
+      "step": 1481
+    },
+    {
+      "epoch": 0.01482,
+      "grad_norm": 1.4635954400671285,
+      "learning_rate": 0.003,
+      "loss": 4.2602,
+      "step": 1482
+    },
+    {
+      "epoch": 0.01483,
+      "grad_norm": 0.7350989176290839,
+      "learning_rate": 0.003,
+      "loss": 4.2401,
+      "step": 1483
+    },
+    {
+      "epoch": 0.01484,
+      "grad_norm": 0.687260607047077,
+      "learning_rate": 0.003,
+      "loss": 4.2906,
+      "step": 1484
+    },
+    {
+      "epoch": 0.01485,
+      "grad_norm": 0.6759306581993629,
+      "learning_rate": 0.003,
+      "loss": 4.2492,
+      "step": 1485
+    },
+    {
+      "epoch": 0.01486,
+      "grad_norm": 0.5738231891215967,
+      "learning_rate": 0.003,
+      "loss": 4.2438,
+      "step": 1486
+    },
+    {
+      "epoch": 0.01487,
+      "grad_norm": 0.67662740183017,
+      "learning_rate": 0.003,
+      "loss": 4.2486,
+      "step": 1487
+    },
+    {
+      "epoch": 0.01488,
+      "grad_norm": 0.6465587043004464,
+      "learning_rate": 0.003,
+      "loss": 4.2584,
+      "step": 1488
+    },
+    {
+      "epoch": 0.01489,
+      "grad_norm": 0.6278262825245318,
+      "learning_rate": 0.003,
+      "loss": 4.251,
+      "step": 1489
+    },
+    {
+      "epoch": 0.0149,
+      "grad_norm": 0.7125204242602325,
+      "learning_rate": 0.003,
+      "loss": 4.2472,
+      "step": 1490
+    },
+    {
+      "epoch": 0.01491,
+      "grad_norm": 0.739446677383364,
+      "learning_rate": 0.003,
+      "loss": 4.263,
+      "step": 1491
+    },
+    {
+      "epoch": 0.01492,
+      "grad_norm": 0.7351492446611294,
+      "learning_rate": 0.003,
+      "loss": 4.2298,
+      "step": 1492
+    },
+    {
+      "epoch": 0.01493,
+      "grad_norm": 0.7206112830630914,
+      "learning_rate": 0.003,
+      "loss": 4.2585,
+      "step": 1493
+    },
+    {
+      "epoch": 0.01494,
+      "grad_norm": 0.7389347083990496,
+      "learning_rate": 0.003,
+      "loss": 4.239,
+      "step": 1494
+    },
+    {
+      "epoch": 0.01495,
+      "grad_norm": 0.6924446817172788,
+      "learning_rate": 0.003,
+      "loss": 4.2476,
+      "step": 1495
+    },
+    {
+      "epoch": 0.01496,
+      "grad_norm": 0.6954181917083473,
+      "learning_rate": 0.003,
+      "loss": 4.2266,
+      "step": 1496
+    },
+    {
+      "epoch": 0.01497,
+      "grad_norm": 0.6994096236505176,
+      "learning_rate": 0.003,
+      "loss": 4.1995,
+      "step": 1497
+    },
+    {
+      "epoch": 0.01498,
+      "grad_norm": 0.737027067879948,
+      "learning_rate": 0.003,
+      "loss": 4.2577,
+      "step": 1498
+    },
+    {
+      "epoch": 0.01499,
+      "grad_norm": 0.8018965005148498,
+      "learning_rate": 0.003,
+      "loss": 4.2828,
+      "step": 1499
+    },
+    {
+      "epoch": 0.015,
+      "grad_norm": 0.673322646768884,
+      "learning_rate": 0.003,
+      "loss": 4.2302,
+      "step": 1500
+    },
+    {
+      "epoch": 0.01501,
+      "grad_norm": 0.6899086847228731,
+      "learning_rate": 0.003,
+      "loss": 4.23,
+      "step": 1501
+    },
+    {
+      "epoch": 0.01502,
+      "grad_norm": 0.6591283488970741,
+      "learning_rate": 0.003,
+      "loss": 4.2708,
+      "step": 1502
+    },
+    {
+      "epoch": 0.01503,
+      "grad_norm": 0.6852371530334846,
+      "learning_rate": 0.003,
+      "loss": 4.2501,
+      "step": 1503
+    },
+    {
+      "epoch": 0.01504,
+      "grad_norm": 0.6823538206406574,
+      "learning_rate": 0.003,
+      "loss": 4.2505,
+      "step": 1504
+    },
+    {
+      "epoch": 0.01505,
+      "grad_norm": 0.7899988696399359,
+      "learning_rate": 0.003,
+      "loss": 4.2395,
+      "step": 1505
+    },
+    {
+      "epoch": 0.01506,
+      "grad_norm": 0.8794700649213404,
+      "learning_rate": 0.003,
+      "loss": 4.2637,
+      "step": 1506
+    },
+    {
+      "epoch": 0.01507,
+      "grad_norm": 0.8573682870616134,
+      "learning_rate": 0.003,
+      "loss": 4.2788,
+      "step": 1507
+    },
+    {
+      "epoch": 0.01508,
+      "grad_norm": 0.9379452075670608,
+      "learning_rate": 0.003,
+      "loss": 4.257,
+      "step": 1508
+    },
+    {
+      "epoch": 0.01509,
+      "grad_norm": 1.155120366213554,
+      "learning_rate": 0.003,
+      "loss": 4.2868,
+      "step": 1509
+    },
+    {
+      "epoch": 0.0151,
+      "grad_norm": 0.8564988533087747,
+      "learning_rate": 0.003,
+      "loss": 4.2523,
+      "step": 1510
+    },
+    {
+      "epoch": 0.01511,
+      "grad_norm": 0.7943851233402885,
+      "learning_rate": 0.003,
+      "loss": 4.242,
+      "step": 1511
+    },
+    {
+      "epoch": 0.01512,
+      "grad_norm": 0.7290925664752087,
+      "learning_rate": 0.003,
+      "loss": 4.2483,
+      "step": 1512
+    },
+    {
+      "epoch": 0.01513,
+      "grad_norm": 0.8302189630029514,
+      "learning_rate": 0.003,
+      "loss": 4.2538,
+      "step": 1513
+    },
+    {
+      "epoch": 0.01514,
+      "grad_norm": 0.9180439778149861,
+      "learning_rate": 0.003,
+      "loss": 4.2658,
+      "step": 1514
+    },
+    {
+      "epoch": 0.01515,
+      "grad_norm": 0.8252164561072897,
+      "learning_rate": 0.003,
+      "loss": 4.2572,
+      "step": 1515
+    },
+    {
+      "epoch": 0.01516,
+      "grad_norm": 0.7235677802672953,
+      "learning_rate": 0.003,
+      "loss": 4.2628,
+      "step": 1516
+    },
+    {
+      "epoch": 0.01517,
+      "grad_norm": 0.669771960126658,
+      "learning_rate": 0.003,
+      "loss": 4.2114,
+      "step": 1517
+    },
+    {
+      "epoch": 0.01518,
+      "grad_norm": 0.6555410571298043,
+      "learning_rate": 0.003,
+      "loss": 4.248,
+      "step": 1518
+    },
+    {
+      "epoch": 0.01519,
+      "grad_norm": 0.7596573642242733,
+      "learning_rate": 0.003,
+      "loss": 4.2822,
+      "step": 1519
+    },
+    {
+      "epoch": 0.0152,
+      "grad_norm": 0.8963421668718663,
+      "learning_rate": 0.003,
+      "loss": 4.2455,
+      "step": 1520
+    },
+    {
+      "epoch": 0.01521,
+      "grad_norm": 1.008660081317184,
+      "learning_rate": 0.003,
+      "loss": 4.2626,
+      "step": 1521
+    },
+    {
+      "epoch": 0.01522,
+      "grad_norm": 1.0941342292979577,
+      "learning_rate": 0.003,
+      "loss": 4.2769,
+      "step": 1522
+    },
+    {
+      "epoch": 0.01523,
+      "grad_norm": 0.8501402046319685,
+      "learning_rate": 0.003,
+      "loss": 4.2571,
+      "step": 1523
+    },
+    {
+      "epoch": 0.01524,
+      "grad_norm": 0.9113983493960105,
+      "learning_rate": 0.003,
+      "loss": 4.2967,
+      "step": 1524
+    },
+    {
+      "epoch": 0.01525,
+      "grad_norm": 0.7627206625135139,
+      "learning_rate": 0.003,
+      "loss": 4.2846,
+      "step": 1525
+    },
+    {
+      "epoch": 0.01526,
+      "grad_norm": 0.7118386576342678,
+      "learning_rate": 0.003,
+      "loss": 4.2733,
+      "step": 1526
+    },
+    {
+      "epoch": 0.01527,
+      "grad_norm": 0.6832379742298202,
+      "learning_rate": 0.003,
+      "loss": 4.2396,
+      "step": 1527
+    },
+    {
+      "epoch": 0.01528,
+      "grad_norm": 0.6558129881498832,
+      "learning_rate": 0.003,
+      "loss": 4.2558,
+      "step": 1528
+    },
+    {
+      "epoch": 0.01529,
+      "grad_norm": 0.653598932900876,
+      "learning_rate": 0.003,
+      "loss": 4.2506,
+      "step": 1529
+    },
+    {
+      "epoch": 0.0153,
+      "grad_norm": 0.6481863699832326,
+      "learning_rate": 0.003,
+      "loss": 4.2569,
+      "step": 1530
+    },
+    {
+      "epoch": 0.01531,
+      "grad_norm": 0.7137971422876086,
+      "learning_rate": 0.003,
+      "loss": 4.2516,
+      "step": 1531
+    },
+    {
+      "epoch": 0.01532,
+      "grad_norm": 0.7682370963055534,
+      "learning_rate": 0.003,
+      "loss": 4.232,
+      "step": 1532
+    },
+    {
+      "epoch": 0.01533,
+      "grad_norm": 0.6869067765486859,
+      "learning_rate": 0.003,
+      "loss": 4.2552,
+      "step": 1533
+    },
+    {
+      "epoch": 0.01534,
+      "grad_norm": 0.6763479705568691,
+      "learning_rate": 0.003,
+      "loss": 4.251,
+      "step": 1534
+    },
+    {
+      "epoch": 0.01535,
+      "grad_norm": 0.6278011204652878,
+      "learning_rate": 0.003,
+      "loss": 4.2624,
+      "step": 1535
+    },
+    {
+      "epoch": 0.01536,
+      "grad_norm": 0.5449726408675756,
+      "learning_rate": 0.003,
+      "loss": 4.2371,
+      "step": 1536
+    },
+    {
+      "epoch": 0.01537,
+      "grad_norm": 0.5702070777355899,
+      "learning_rate": 0.003,
+      "loss": 4.2348,
+      "step": 1537
+    },
+    {
+      "epoch": 0.01538,
+      "grad_norm": 0.5568148406111664,
+      "learning_rate": 0.003,
+      "loss": 4.2106,
+      "step": 1538
+    },
+    {
+      "epoch": 0.01539,
+      "grad_norm": 0.5990162223338897,
+      "learning_rate": 0.003,
+      "loss": 4.2411,
+      "step": 1539
+    },
+    {
+      "epoch": 0.0154,
+      "grad_norm": 0.6709932169117396,
+      "learning_rate": 0.003,
+      "loss": 4.2471,
+      "step": 1540
+    },
+    {
+      "epoch": 0.01541,
+      "grad_norm": 0.839479307713138,
+      "learning_rate": 0.003,
+      "loss": 4.2333,
+      "step": 1541
+    },
+    {
+      "epoch": 0.01542,
+      "grad_norm": 0.998609935659052,
+      "learning_rate": 0.003,
+      "loss": 4.2723,
+      "step": 1542
+    },
+    {
+      "epoch": 0.01543,
+      "grad_norm": 0.9175226186557726,
+      "learning_rate": 0.003,
+      "loss": 4.2597,
+      "step": 1543
+    },
+    {
+      "epoch": 0.01544,
+      "grad_norm": 0.9698336862541426,
+      "learning_rate": 0.003,
+      "loss": 4.2493,
+      "step": 1544
+    },
+    {
+      "epoch": 0.01545,
+      "grad_norm": 0.8646283897782253,
+      "learning_rate": 0.003,
+      "loss": 4.3133,
+      "step": 1545
+    },
+    {
+      "epoch": 0.01546,
+      "grad_norm": 0.7489694669979721,
+      "learning_rate": 0.003,
+      "loss": 4.2786,
+      "step": 1546
+    },
+    {
+      "epoch": 0.01547,
+      "grad_norm": 0.7346787237600204,
+      "learning_rate": 0.003,
+      "loss": 4.2793,
+      "step": 1547
+    },
+    {
+      "epoch": 0.01548,
+      "grad_norm": 0.7843295792698941,
+      "learning_rate": 0.003,
+      "loss": 4.2451,
+      "step": 1548
+    },
+    {
+      "epoch": 0.01549,
+      "grad_norm": 0.7547938162635378,
+      "learning_rate": 0.003,
+      "loss": 4.2323,
+      "step": 1549
+    },
+    {
+      "epoch": 0.0155,
+      "grad_norm": 0.8033101017591833,
+      "learning_rate": 0.003,
+      "loss": 4.2543,
+      "step": 1550
+    },
+    {
+      "epoch": 0.01551,
+      "grad_norm": 0.784324716245673,
+      "learning_rate": 0.003,
+      "loss": 4.2319,
+      "step": 1551
+    },
+    {
+      "epoch": 0.01552,
+      "grad_norm": 0.6675659235069369,
+      "learning_rate": 0.003,
+      "loss": 4.2485,
+      "step": 1552
+    },
+    {
+      "epoch": 0.01553,
+      "grad_norm": 0.7200584927165973,
+      "learning_rate": 0.003,
+      "loss": 4.2735,
+      "step": 1553
+    },
+    {
+      "epoch": 0.01554,
+      "grad_norm": 0.833889265994982,
+      "learning_rate": 0.003,
+      "loss": 4.259,
+      "step": 1554
+    },
+    {
+      "epoch": 0.01555,
+      "grad_norm": 0.7292713050790204,
+      "learning_rate": 0.003,
+      "loss": 4.2302,
+      "step": 1555
+    },
+    {
+      "epoch": 0.01556,
+      "grad_norm": 0.6334535318603711,
+      "learning_rate": 0.003,
+      "loss": 4.2647,
+      "step": 1556
+    },
+    {
+      "epoch": 0.01557,
+      "grad_norm": 0.6665206815174863,
+      "learning_rate": 0.003,
+      "loss": 4.2396,
+      "step": 1557
+    },
+    {
+      "epoch": 0.01558,
+      "grad_norm": 0.5467983416304218,
+      "learning_rate": 0.003,
+      "loss": 4.2264,
+      "step": 1558
+    },
+    {
+      "epoch": 0.01559,
+      "grad_norm": 0.5303338136727427,
+      "learning_rate": 0.003,
+      "loss": 4.2625,
+      "step": 1559
+    },
+    {
+      "epoch": 0.0156,
+      "grad_norm": 0.5964427896199266,
+      "learning_rate": 0.003,
+      "loss": 4.2526,
+      "step": 1560
+    },
+    {
+      "epoch": 0.01561,
+      "grad_norm": 0.7487394215839274,
+      "learning_rate": 0.003,
+      "loss": 4.2358,
+      "step": 1561
+    },
+    {
+      "epoch": 0.01562,
+      "grad_norm": 0.8937121416966828,
+      "learning_rate": 0.003,
+      "loss": 4.2394,
+      "step": 1562
+    },
+    {
+      "epoch": 0.01563,
+      "grad_norm": 1.2085214855734254,
+      "learning_rate": 0.003,
+      "loss": 4.2602,
+      "step": 1563
+    },
+    {
+      "epoch": 0.01564,
+      "grad_norm": 0.8696563758847599,
+      "learning_rate": 0.003,
+      "loss": 4.284,
+      "step": 1564
+    },
+    {
+      "epoch": 0.01565,
+      "grad_norm": 0.739740990652482,
+      "learning_rate": 0.003,
+      "loss": 4.2566,
+      "step": 1565
+    },
+    {
+      "epoch": 0.01566,
+      "grad_norm": 0.7600676271323985,
+      "learning_rate": 0.003,
+      "loss": 4.2738,
+      "step": 1566
+    },
+    {
+      "epoch": 0.01567,
+      "grad_norm": 0.7222900608043384,
+      "learning_rate": 0.003,
+      "loss": 4.2646,
+      "step": 1567
+    },
+    {
+      "epoch": 0.01568,
+      "grad_norm": 0.7569225436342659,
+      "learning_rate": 0.003,
+      "loss": 4.2466,
+      "step": 1568
+    },
+    {
+      "epoch": 0.01569,
+      "grad_norm": 0.7441561581750152,
+      "learning_rate": 0.003,
+      "loss": 4.252,
+      "step": 1569
+    },
+    {
+      "epoch": 0.0157,
+      "grad_norm": 0.7580218902280629,
+      "learning_rate": 0.003,
+      "loss": 4.2727,
+      "step": 1570
+    },
+    {
+      "epoch": 0.01571,
+      "grad_norm": 0.771634247408187,
+      "learning_rate": 0.003,
+      "loss": 4.2581,
+      "step": 1571
+    },
+    {
+      "epoch": 0.01572,
+      "grad_norm": 0.95958549182467,
+      "learning_rate": 0.003,
+      "loss": 4.2713,
+      "step": 1572
+    },
+    {
+      "epoch": 0.01573,
+      "grad_norm": 0.9029057415025205,
+      "learning_rate": 0.003,
+      "loss": 4.261,
+      "step": 1573
+    },
+    {
+      "epoch": 0.01574,
+      "grad_norm": 0.8044062509450788,
+      "learning_rate": 0.003,
+      "loss": 4.2692,
+      "step": 1574
+    },
+    {
+      "epoch": 0.01575,
+      "grad_norm": 0.7286639057606195,
+      "learning_rate": 0.003,
+      "loss": 4.2305,
+      "step": 1575
+    },
+    {
+      "epoch": 0.01576,
+      "grad_norm": 0.6821195872107877,
+      "learning_rate": 0.003,
+      "loss": 4.2492,
+      "step": 1576
+    },
+    {
+      "epoch": 0.01577,
+      "grad_norm": 0.6033878670810847,
+      "learning_rate": 0.003,
+      "loss": 4.2068,
+      "step": 1577
+    },
+    {
+      "epoch": 0.01578,
+      "grad_norm": 0.519693589168086,
+      "learning_rate": 0.003,
+      "loss": 4.2343,
+      "step": 1578
+    },
+    {
+      "epoch": 0.01579,
+      "grad_norm": 0.46078437812161954,
+      "learning_rate": 0.003,
+      "loss": 4.2271,
+      "step": 1579
+    },
+    {
+      "epoch": 0.0158,
+      "grad_norm": 0.3811995998581025,
+      "learning_rate": 0.003,
+      "loss": 4.2344,
+      "step": 1580
+    },
+    {
+      "epoch": 0.01581,
+      "grad_norm": 0.40258363386782764,
+      "learning_rate": 0.003,
+      "loss": 4.2406,
+      "step": 1581
+    },
+    {
+      "epoch": 0.01582,
+      "grad_norm": 0.4070479328001984,
+      "learning_rate": 0.003,
+      "loss": 4.2222,
+      "step": 1582
+    },
+    {
+      "epoch": 0.01583,
+      "grad_norm": 0.39481584768858025,
+      "learning_rate": 0.003,
+      "loss": 4.2093,
+      "step": 1583
+    },
+    {
+      "epoch": 0.01584,
+      "grad_norm": 0.4212949759278112,
+      "learning_rate": 0.003,
+      "loss": 4.24,
+      "step": 1584
+    },
+    {
+      "epoch": 0.01585,
+      "grad_norm": 0.6375222350508811,
+      "learning_rate": 0.003,
+      "loss": 4.2351,
+      "step": 1585
+    },
+    {
+      "epoch": 0.01586,
+      "grad_norm": 1.1087267261978884,
+      "learning_rate": 0.003,
+      "loss": 4.2389,
+      "step": 1586
+    },
+    {
+      "epoch": 0.01587,
+      "grad_norm": 1.095427209738268,
+      "learning_rate": 0.003,
+      "loss": 4.2576,
+      "step": 1587
+    },
+    {
+      "epoch": 0.01588,
+      "grad_norm": 0.5974585636629688,
+      "learning_rate": 0.003,
+      "loss": 4.2673,
+      "step": 1588
+    },
+    {
+      "epoch": 0.01589,
+      "grad_norm": 0.6434065905263348,
+      "learning_rate": 0.003,
+      "loss": 4.2657,
+      "step": 1589
+    },
+    {
+      "epoch": 0.0159,
+      "grad_norm": 0.5922765275307262,
+      "learning_rate": 0.003,
+      "loss": 4.2507,
+      "step": 1590
+    },
+    {
+      "epoch": 0.01591,
+      "grad_norm": 0.6559622239667369,
+      "learning_rate": 0.003,
+      "loss": 4.2576,
+      "step": 1591
+    },
+    {
+      "epoch": 0.01592,
+      "grad_norm": 0.5876956364208827,
+      "learning_rate": 0.003,
+      "loss": 4.2446,
+      "step": 1592
+    },
+    {
+      "epoch": 0.01593,
+      "grad_norm": 0.4847713630300587,
+      "learning_rate": 0.003,
+      "loss": 4.2428,
+      "step": 1593
+    },
+    {
+      "epoch": 0.01594,
+      "grad_norm": 0.5605434129471829,
+      "learning_rate": 0.003,
+      "loss": 4.2313,
+      "step": 1594
+    },
+    {
+      "epoch": 0.01595,
+      "grad_norm": 0.5718187801259297,
+      "learning_rate": 0.003,
+      "loss": 4.2581,
+      "step": 1595
+    },
+    {
+      "epoch": 0.01596,
+      "grad_norm": 0.620870673073496,
+      "learning_rate": 0.003,
+      "loss": 4.2509,
+      "step": 1596
+    },
+    {
+      "epoch": 0.01597,
+      "grad_norm": 0.6969552121884817,
+      "learning_rate": 0.003,
+      "loss": 4.2154,
+      "step": 1597
+    },
+    {
+      "epoch": 0.01598,
+      "grad_norm": 0.8094238052393485,
+      "learning_rate": 0.003,
+      "loss": 4.2172,
+      "step": 1598
+    },
+    {
+      "epoch": 0.01599,
+      "grad_norm": 0.7713054312041273,
+      "learning_rate": 0.003,
+      "loss": 4.2333,
+      "step": 1599
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 0.7787996781195005,
+      "learning_rate": 0.003,
+      "loss": 4.2062,
+      "step": 1600
+    },
+    {
+      "epoch": 0.01601,
+      "grad_norm": 0.620719741277366,
+      "learning_rate": 0.003,
+      "loss": 4.2478,
+      "step": 1601
+    },
+    {
+      "epoch": 0.01602,
+      "grad_norm": 0.5965590372558249,
+      "learning_rate": 0.003,
+      "loss": 4.223,
+      "step": 1602
+    },
+    {
+      "epoch": 0.01603,
+      "grad_norm": 0.6057948188618382,
+      "learning_rate": 0.003,
+      "loss": 4.2386,
+      "step": 1603
+    },
+    {
+      "epoch": 0.01604,
+      "grad_norm": 0.5859806802733505,
+      "learning_rate": 0.003,
+      "loss": 4.2313,
+      "step": 1604
+    },
+    {
+      "epoch": 0.01605,
+      "grad_norm": 0.5985785202970987,
+      "learning_rate": 0.003,
+      "loss": 4.2093,
+      "step": 1605
+    },
+    {
+      "epoch": 0.01606,
+      "grad_norm": 0.6329923375985832,
+      "learning_rate": 0.003,
+      "loss": 4.2131,
+      "step": 1606
+    },
+    {
+      "epoch": 0.01607,
+      "grad_norm": 0.6777814987430973,
+      "learning_rate": 0.003,
+      "loss": 4.2418,
+      "step": 1607
+    },
+    {
+      "epoch": 0.01608,
+      "grad_norm": 0.6454825679602344,
+      "learning_rate": 0.003,
+      "loss": 4.2152,
+      "step": 1608
+    },
+    {
+      "epoch": 0.01609,
+      "grad_norm": 0.6234331303306465,
+      "learning_rate": 0.003,
+      "loss": 4.222,
+      "step": 1609
+    },
+    {
+      "epoch": 0.0161,
+      "grad_norm": 0.717661852780552,
+      "learning_rate": 0.003,
+      "loss": 4.1917,
+      "step": 1610
+    },
+    {
+      "epoch": 0.01611,
+      "grad_norm": 0.814514408732785,
+      "learning_rate": 0.003,
+      "loss": 4.2133,
+      "step": 1611
+    },
+    {
+      "epoch": 0.01612,
+      "grad_norm": 0.968378243767811,
+      "learning_rate": 0.003,
+      "loss": 4.2184,
+      "step": 1612
+    },
+    {
+      "epoch": 0.01613,
+      "grad_norm": 1.1889382213143287,
+      "learning_rate": 0.003,
+      "loss": 4.2747,
+      "step": 1613
+    },
+    {
+      "epoch": 0.01614,
+      "grad_norm": 0.902755449051569,
+      "learning_rate": 0.003,
+      "loss": 4.2894,
+      "step": 1614
+    },
+    {
+      "epoch": 0.01615,
+      "grad_norm": 0.8743919537144361,
+      "learning_rate": 0.003,
+      "loss": 4.2622,
+      "step": 1615
+    },
+    {
+      "epoch": 0.01616,
+      "grad_norm": 0.8601657492637806,
+      "learning_rate": 0.003,
+      "loss": 4.2396,
+      "step": 1616
+    },
+    {
+      "epoch": 0.01617,
+      "grad_norm": 0.8301577875616364,
+      "learning_rate": 0.003,
+      "loss": 4.2333,
+      "step": 1617
+    },
+    {
+      "epoch": 0.01618,
+      "grad_norm": 0.829332785699801,
+      "learning_rate": 0.003,
+      "loss": 4.2387,
+      "step": 1618
+    },
+    {
+      "epoch": 0.01619,
+      "grad_norm": 0.7047548182139096,
+      "learning_rate": 0.003,
+      "loss": 4.2209,
+      "step": 1619
+    },
+    {
+      "epoch": 0.0162,
+      "grad_norm": 0.7270289420377364,
+      "learning_rate": 0.003,
+      "loss": 4.2317,
+      "step": 1620
+    },
+    {
+      "epoch": 0.01621,
+      "grad_norm": 0.6929887184456018,
+      "learning_rate": 0.003,
+      "loss": 4.2257,
+      "step": 1621
+    },
+    {
+      "epoch": 0.01622,
+      "grad_norm": 0.7471865137769718,
+      "learning_rate": 0.003,
+      "loss": 4.2589,
+      "step": 1622
+    },
+    {
+      "epoch": 0.01623,
+      "grad_norm": 0.8282860007733814,
+      "learning_rate": 0.003,
+      "loss": 4.2399,
+      "step": 1623
+    },
+    {
+      "epoch": 0.01624,
+      "grad_norm": 0.8871911840665138,
+      "learning_rate": 0.003,
+      "loss": 4.2581,
+      "step": 1624
+    },
+    {
+      "epoch": 0.01625,
+      "grad_norm": 1.0259944995678087,
+      "learning_rate": 0.003,
+      "loss": 4.2469,
+      "step": 1625
+    },
+    {
+      "epoch": 0.01626,
+      "grad_norm": 0.937815191301713,
+      "learning_rate": 0.003,
+      "loss": 4.2602,
+      "step": 1626
+    },
+    {
+      "epoch": 0.01627,
+      "grad_norm": 1.064022068457424,
+      "learning_rate": 0.003,
+      "loss": 4.2368,
+      "step": 1627
+    },
+    {
+      "epoch": 0.01628,
+      "grad_norm": 1.0710330182336567,
+      "learning_rate": 0.003,
+      "loss": 4.2567,
+      "step": 1628
+    },
+    {
+      "epoch": 0.01629,
+      "grad_norm": 0.7877325862421115,
+      "learning_rate": 0.003,
+      "loss": 4.2414,
+      "step": 1629
+    },
+    {
+      "epoch": 0.0163,
+      "grad_norm": 0.6743454531988056,
+      "learning_rate": 0.003,
+      "loss": 4.2288,
+      "step": 1630
+    },
+    {
+      "epoch": 0.01631,
+      "grad_norm": 0.701073161255015,
+      "learning_rate": 0.003,
+      "loss": 4.2499,
+      "step": 1631
+    },
+    {
+      "epoch": 0.01632,
+      "grad_norm": 0.6004677978717051,
+      "learning_rate": 0.003,
+      "loss": 4.252,
+      "step": 1632
+    },
+    {
+      "epoch": 0.01633,
+      "grad_norm": 0.5772521307266697,
+      "learning_rate": 0.003,
+      "loss": 4.2255,
+      "step": 1633
+    },
+    {
+      "epoch": 0.01634,
+      "grad_norm": 0.5454153935480286,
+      "learning_rate": 0.003,
+      "loss": 4.2406,
+      "step": 1634
+    },
+    {
+      "epoch": 0.01635,
+      "grad_norm": 0.5154455969479952,
+      "learning_rate": 0.003,
+      "loss": 4.218,
+      "step": 1635
+    },
+    {
+      "epoch": 0.01636,
+      "grad_norm": 0.5267549110628605,
+      "learning_rate": 0.003,
+      "loss": 4.2348,
+      "step": 1636
+    },
+    {
+      "epoch": 0.01637,
+      "grad_norm": 0.5971697216641335,
+      "learning_rate": 0.003,
+      "loss": 4.2205,
+      "step": 1637
+    },
+    {
+      "epoch": 0.01638,
+      "grad_norm": 0.7211767188224439,
+      "learning_rate": 0.003,
+      "loss": 4.229,
+      "step": 1638
+    },
+    {
+      "epoch": 0.01639,
+      "grad_norm": 0.7816125211534669,
+      "learning_rate": 0.003,
+      "loss": 4.2548,
+      "step": 1639
+    },
+    {
+      "epoch": 0.0164,
+      "grad_norm": 0.7389027343982134,
+      "learning_rate": 0.003,
+      "loss": 4.2362,
+      "step": 1640
+    },
+    {
+      "epoch": 0.01641,
+      "grad_norm": 0.5975265072148448,
+      "learning_rate": 0.003,
+      "loss": 4.2026,
+      "step": 1641
+    },
+    {
+      "epoch": 0.01642,
+      "grad_norm": 0.7083476744478381,
+      "learning_rate": 0.003,
+      "loss": 4.2507,
+      "step": 1642
+    },
+    {
+      "epoch": 0.01643,
+      "grad_norm": 0.7791000270261257,
+      "learning_rate": 0.003,
+      "loss": 4.2378,
+      "step": 1643
+    },
+    {
+      "epoch": 0.01644,
+      "grad_norm": 0.9006394984546569,
+      "learning_rate": 0.003,
+      "loss": 4.2625,
+      "step": 1644
+    },
+    {
+      "epoch": 0.01645,
+      "grad_norm": 1.2334686487745912,
+      "learning_rate": 0.003,
+      "loss": 4.2236,
+      "step": 1645
+    },
+    {
+      "epoch": 0.01646,
+      "grad_norm": 0.9932721101220016,
+      "learning_rate": 0.003,
+      "loss": 4.263,
+      "step": 1646
+    },
+    {
+      "epoch": 0.01647,
+      "grad_norm": 0.989511986212608,
+      "learning_rate": 0.003,
+      "loss": 4.2576,
+      "step": 1647
+    },
+    {
+      "epoch": 0.01648,
+      "grad_norm": 0.8142750385583939,
+      "learning_rate": 0.003,
+      "loss": 4.23,
+      "step": 1648
+    },
+    {
+      "epoch": 0.01649,
+      "grad_norm": 0.6967651791258922,
+      "learning_rate": 0.003,
+      "loss": 4.2583,
+      "step": 1649
+    },
+    {
+      "epoch": 0.0165,
+      "grad_norm": 0.7348292229935104,
+      "learning_rate": 0.003,
+      "loss": 4.2361,
+      "step": 1650
+    },
+    {
+      "epoch": 0.01651,
+      "grad_norm": 0.6565673319405235,
+      "learning_rate": 0.003,
+      "loss": 4.2402,
+      "step": 1651
+    },
+    {
+      "epoch": 0.01652,
+      "grad_norm": 0.5674400404701213,
+      "learning_rate": 0.003,
+      "loss": 4.1936,
+      "step": 1652
+    },
+    {
+      "epoch": 0.01653,
+      "grad_norm": 0.5571261023539824,
+      "learning_rate": 0.003,
+      "loss": 4.2378,
+      "step": 1653
+    },
+    {
+      "epoch": 0.01654,
+      "grad_norm": 0.5083863828402352,
+      "learning_rate": 0.003,
+      "loss": 4.2229,
+      "step": 1654
+    },
+    {
+      "epoch": 0.01655,
+      "grad_norm": 0.49820131551804164,
+      "learning_rate": 0.003,
+      "loss": 4.23,
+      "step": 1655
+    },
+    {
+      "epoch": 0.01656,
+      "grad_norm": 0.5201001189396881,
+      "learning_rate": 0.003,
+      "loss": 4.2012,
+      "step": 1656
+    },
+    {
+      "epoch": 0.01657,
+      "grad_norm": 0.6652455003065759,
+      "learning_rate": 0.003,
+      "loss": 4.2237,
+      "step": 1657
+    },
+    {
+      "epoch": 0.01658,
+      "grad_norm": 0.8870121051704852,
+      "learning_rate": 0.003,
+      "loss": 4.219,
+      "step": 1658
+    },
+    {
+      "epoch": 0.01659,
+      "grad_norm": 1.1634698909054708,
+      "learning_rate": 0.003,
+      "loss": 4.2361,
+      "step": 1659
+    },
+    {
+      "epoch": 0.0166,
+      "grad_norm": 0.6579478278546654,
+      "learning_rate": 0.003,
+      "loss": 4.2096,
+      "step": 1660
+    },
+    {
+      "epoch": 0.01661,
+      "grad_norm": 0.5917680182889076,
+      "learning_rate": 0.003,
+      "loss": 4.2092,
+      "step": 1661
+    },
+    {
+      "epoch": 0.01662,
+      "grad_norm": 0.7821000738049482,
+      "learning_rate": 0.003,
+      "loss": 4.2171,
+      "step": 1662
+    },
+    {
+      "epoch": 0.01663,
+      "grad_norm": 0.812312164693218,
+      "learning_rate": 0.003,
+      "loss": 4.2422,
+      "step": 1663
+    },
+    {
+      "epoch": 0.01664,
+      "grad_norm": 0.7053733462479915,
+      "learning_rate": 0.003,
+      "loss": 4.2567,
+      "step": 1664
+    },
+    {
+      "epoch": 0.01665,
+      "grad_norm": 0.713897124959619,
+      "learning_rate": 0.003,
+      "loss": 4.2198,
+      "step": 1665
+    },
+    {
+      "epoch": 0.01666,
+      "grad_norm": 0.6600299709203347,
+      "learning_rate": 0.003,
+      "loss": 4.2205,
+      "step": 1666
+    },
+    {
+      "epoch": 0.01667,
+      "grad_norm": 0.5832085680202569,
+      "learning_rate": 0.003,
+      "loss": 4.2321,
+      "step": 1667
+    },
+    {
+      "epoch": 0.01668,
+      "grad_norm": 0.6589799435966834,
+      "learning_rate": 0.003,
+      "loss": 4.2397,
+      "step": 1668
+    },
+    {
+      "epoch": 0.01669,
+      "grad_norm": 0.7719970182316414,
+      "learning_rate": 0.003,
+      "loss": 4.2273,
+      "step": 1669
+    },
+    {
+      "epoch": 0.0167,
+      "grad_norm": 0.7846205843605817,
+      "learning_rate": 0.003,
+      "loss": 4.2184,
+      "step": 1670
+    },
+    {
+      "epoch": 0.01671,
+      "grad_norm": 0.7635694428992538,
+      "learning_rate": 0.003,
+      "loss": 4.2087,
+      "step": 1671
+    },
+    {
+      "epoch": 0.01672,
+      "grad_norm": 0.7758449881926132,
+      "learning_rate": 0.003,
+      "loss": 4.2199,
+      "step": 1672
+    },
+    {
+      "epoch": 0.01673,
+      "grad_norm": 0.6618290355489082,
+      "learning_rate": 0.003,
+      "loss": 4.2134,
+      "step": 1673
+    },
+    {
+      "epoch": 0.01674,
+      "grad_norm": 0.6727466502726199,
+      "learning_rate": 0.003,
+      "loss": 4.232,
+      "step": 1674
+    },
+    {
+      "epoch": 0.01675,
+      "grad_norm": 0.6690360284678838,
+      "learning_rate": 0.003,
+      "loss": 4.2158,
+      "step": 1675
+    },
+    {
+      "epoch": 0.01676,
+      "grad_norm": 0.6672382539021496,
+      "learning_rate": 0.003,
+      "loss": 4.2242,
+      "step": 1676
+    },
+    {
+      "epoch": 0.01677,
+      "grad_norm": 0.6311361325143124,
+      "learning_rate": 0.003,
+      "loss": 4.2132,
+      "step": 1677
+    },
+    {
+      "epoch": 0.01678,
+      "grad_norm": 0.727198962494881,
+      "learning_rate": 0.003,
+      "loss": 4.2299,
+      "step": 1678
+    },
+    {
+      "epoch": 0.01679,
+      "grad_norm": 0.8433802333808034,
+      "learning_rate": 0.003,
+      "loss": 4.2177,
+      "step": 1679
+    },
+    {
+      "epoch": 0.0168,
+      "grad_norm": 0.8402151525100402,
+      "learning_rate": 0.003,
+      "loss": 4.2352,
+      "step": 1680
+    },
+    {
+      "epoch": 0.01681,
+      "grad_norm": 0.9608253642789906,
+      "learning_rate": 0.003,
+      "loss": 4.2433,
+      "step": 1681
+    },
+    {
+      "epoch": 0.01682,
+      "grad_norm": 1.160024891342376,
+      "learning_rate": 0.003,
+      "loss": 4.246,
+      "step": 1682
+    },
+    {
+      "epoch": 0.01683,
+      "grad_norm": 0.8201452744681862,
+      "learning_rate": 0.003,
+      "loss": 4.2539,
+      "step": 1683
+    },
+    {
+      "epoch": 0.01684,
+      "grad_norm": 0.7434576192749681,
+      "learning_rate": 0.003,
+      "loss": 4.2242,
+      "step": 1684
+    },
+    {
+      "epoch": 0.01685,
+      "grad_norm": 0.8733167238677911,
+      "learning_rate": 0.003,
+      "loss": 4.2448,
+      "step": 1685
+    },
+    {
+      "epoch": 0.01686,
+      "grad_norm": 0.7251849705005492,
+      "learning_rate": 0.003,
+      "loss": 4.2342,
+      "step": 1686
+    },
+    {
+      "epoch": 0.01687,
+      "grad_norm": 0.623355393019925,
+      "learning_rate": 0.003,
+      "loss": 4.2258,
+      "step": 1687
+    },
+    {
+      "epoch": 0.01688,
+      "grad_norm": 0.6485004834811148,
+      "learning_rate": 0.003,
+      "loss": 4.2113,
+      "step": 1688
+    },
+    {
+      "epoch": 0.01689,
+      "grad_norm": 0.6622097332232696,
+      "learning_rate": 0.003,
+      "loss": 4.2246,
+      "step": 1689
+    },
+    {
+      "epoch": 0.0169,
+      "grad_norm": 0.5609920955895745,
+      "learning_rate": 0.003,
+      "loss": 4.2095,
+      "step": 1690
+    },
+    {
+      "epoch": 0.01691,
+      "grad_norm": 0.5554950872814335,
+      "learning_rate": 0.003,
+      "loss": 4.2208,
+      "step": 1691
+    },
+    {
+      "epoch": 0.01692,
+      "grad_norm": 0.578754263387868,
+      "learning_rate": 0.003,
+      "loss": 4.2223,
+      "step": 1692
+    },
+    {
+      "epoch": 0.01693,
+      "grad_norm": 0.54763155586202,
+      "learning_rate": 0.003,
+      "loss": 4.2453,
+      "step": 1693
+    },
+    {
+      "epoch": 0.01694,
+      "grad_norm": 0.5925337152955293,
+      "learning_rate": 0.003,
+      "loss": 4.1963,
+      "step": 1694
+    },
+    {
+      "epoch": 0.01695,
+      "grad_norm": 0.5883008115307695,
+      "learning_rate": 0.003,
+      "loss": 4.2285,
+      "step": 1695
+    },
+    {
+      "epoch": 0.01696,
+      "grad_norm": 0.5777962347479031,
+      "learning_rate": 0.003,
+      "loss": 4.2174,
+      "step": 1696
+    },
+    {
+      "epoch": 0.01697,
+      "grad_norm": 0.6137264663560853,
+      "learning_rate": 0.003,
+      "loss": 4.1953,
+      "step": 1697
+    },
+    {
+      "epoch": 0.01698,
+      "grad_norm": 0.5567779485618047,
+      "learning_rate": 0.003,
+      "loss": 4.1887,
+      "step": 1698
+    },
+    {
+      "epoch": 0.01699,
+      "grad_norm": 0.563692648440794,
+      "learning_rate": 0.003,
+      "loss": 4.1995,
+      "step": 1699
+    },
+    {
+      "epoch": 0.017,
+      "grad_norm": 0.5378862525995165,
+      "learning_rate": 0.003,
+      "loss": 4.2165,
+      "step": 1700
+    },
+    {
+      "epoch": 0.01701,
+      "grad_norm": 0.5131924880718308,
+      "learning_rate": 0.003,
+      "loss": 4.1973,
+      "step": 1701
+    },
+    {
+      "epoch": 0.01702,
+      "grad_norm": 0.5473378733884345,
+      "learning_rate": 0.003,
+      "loss": 4.2207,
+      "step": 1702
+    },
+    {
+      "epoch": 0.01703,
+      "grad_norm": 0.659714976195776,
+      "learning_rate": 0.003,
+      "loss": 4.2099,
+      "step": 1703
+    },
+    {
+      "epoch": 0.01704,
+      "grad_norm": 0.7407833108135735,
+      "learning_rate": 0.003,
+      "loss": 4.2001,
+      "step": 1704
+    },
+    {
+      "epoch": 0.01705,
+      "grad_norm": 0.7567928276503215,
+      "learning_rate": 0.003,
+      "loss": 4.2073,
+      "step": 1705
+    },
+    {
+      "epoch": 0.01706,
+      "grad_norm": 0.7694420146818104,
+      "learning_rate": 0.003,
+      "loss": 4.2055,
+      "step": 1706
+    },
+    {
+      "epoch": 0.01707,
+      "grad_norm": 1.0045884849963176,
+      "learning_rate": 0.003,
+      "loss": 4.2374,
+      "step": 1707
+    },
+    {
+      "epoch": 0.01708,
+      "grad_norm": 1.2575213437630939,
+      "learning_rate": 0.003,
+      "loss": 4.2465,
+      "step": 1708
+    },
+    {
+      "epoch": 0.01709,
+      "grad_norm": 0.848642979076845,
+      "learning_rate": 0.003,
+      "loss": 4.2418,
+      "step": 1709
+    },
+    {
+      "epoch": 0.0171,
+      "grad_norm": 0.7781603648787689,
+      "learning_rate": 0.003,
+      "loss": 4.1936,
+      "step": 1710
+    },
+    {
+      "epoch": 0.01711,
+      "grad_norm": 0.7996923845242195,
+      "learning_rate": 0.003,
+      "loss": 4.2245,
+      "step": 1711
+    },
+    {
+      "epoch": 0.01712,
+      "grad_norm": 0.7302785582736923,
+      "learning_rate": 0.003,
+      "loss": 4.2298,
+      "step": 1712
+    },
+    {
+      "epoch": 0.01713,
+      "grad_norm": 0.7495189068505729,
+      "learning_rate": 0.003,
+      "loss": 4.204,
+      "step": 1713
+    },
+    {
+      "epoch": 0.01714,
+      "grad_norm": 0.68320470368433,
+      "learning_rate": 0.003,
+      "loss": 4.2407,
+      "step": 1714
+    },
+    {
+      "epoch": 0.01715,
+      "grad_norm": 0.744728870764733,
+      "learning_rate": 0.003,
+      "loss": 4.2481,
+      "step": 1715
+    },
+    {
+      "epoch": 0.01716,
+      "grad_norm": 0.7728817520220844,
+      "learning_rate": 0.003,
+      "loss": 4.2317,
+      "step": 1716
+    },
+    {
+      "epoch": 0.01717,
+      "grad_norm": 0.8671487727140617,
+      "learning_rate": 0.003,
+      "loss": 4.2664,
+      "step": 1717
+    },
+    {
+      "epoch": 0.01718,
+      "grad_norm": 0.8286537250448458,
+      "learning_rate": 0.003,
+      "loss": 4.237,
+      "step": 1718
+    },
+    {
+      "epoch": 0.01719,
+      "grad_norm": 0.7719901017400573,
+      "learning_rate": 0.003,
+      "loss": 4.2535,
+      "step": 1719
+    },
+    {
+      "epoch": 0.0172,
+      "grad_norm": 0.7363513590865703,
+      "learning_rate": 0.003,
+      "loss": 4.2105,
+      "step": 1720
+    },
+    {
+      "epoch": 0.01721,
+      "grad_norm": 0.6255496683743654,
+      "learning_rate": 0.003,
+      "loss": 4.2258,
+      "step": 1721
+    },
+    {
+      "epoch": 0.01722,
+      "grad_norm": 0.5928702135655549,
+      "learning_rate": 0.003,
+      "loss": 4.2129,
+      "step": 1722
+    },
+    {
+      "epoch": 0.01723,
+      "grad_norm": 0.6570038276937221,
+      "learning_rate": 0.003,
+      "loss": 4.2248,
+      "step": 1723
+    },
+    {
+      "epoch": 0.01724,
+      "grad_norm": 0.6380856228410711,
+      "learning_rate": 0.003,
+      "loss": 4.2241,
+      "step": 1724
+    },
+    {
+      "epoch": 0.01725,
+      "grad_norm": 0.7477214539729845,
+      "learning_rate": 0.003,
+      "loss": 4.2088,
+      "step": 1725
+    },
+    {
+      "epoch": 0.01726,
+      "grad_norm": 0.7064643502935442,
+      "learning_rate": 0.003,
+      "loss": 4.2027,
+      "step": 1726
+    },
+    {
+      "epoch": 0.01727,
+      "grad_norm": 0.7018895227983273,
+      "learning_rate": 0.003,
+      "loss": 4.2082,
+      "step": 1727
+    },
+    {
+      "epoch": 0.01728,
+      "grad_norm": 0.7268534843228023,
+      "learning_rate": 0.003,
+      "loss": 4.2116,
+      "step": 1728
+    },
+    {
+      "epoch": 0.01729,
+      "grad_norm": 0.8265185231394759,
+      "learning_rate": 0.003,
+      "loss": 4.2226,
+      "step": 1729
+    },
+    {
+      "epoch": 0.0173,
+      "grad_norm": 0.9216458529069881,
+      "learning_rate": 0.003,
+      "loss": 4.2023,
+      "step": 1730
+    },
+    {
+      "epoch": 0.01731,
+      "grad_norm": 1.0984877424905655,
+      "learning_rate": 0.003,
+      "loss": 4.2294,
+      "step": 1731
+    },
+    {
+      "epoch": 0.01732,
+      "grad_norm": 0.9457932448227158,
+      "learning_rate": 0.003,
+      "loss": 4.2417,
+      "step": 1732
+    },
+    {
+      "epoch": 0.01733,
+      "grad_norm": 0.8744510824253744,
+      "learning_rate": 0.003,
+      "loss": 4.2124,
+      "step": 1733
+    },
+    {
+      "epoch": 0.01734,
+      "grad_norm": 0.8464357668562363,
+      "learning_rate": 0.003,
+      "loss": 4.2277,
+      "step": 1734
+    },
+    {
+      "epoch": 0.01735,
+      "grad_norm": 0.7933851274075868,
+      "learning_rate": 0.003,
+      "loss": 4.1993,
+      "step": 1735
+    },
+    {
+      "epoch": 0.01736,
+      "grad_norm": 0.7800431566899245,
+      "learning_rate": 0.003,
+      "loss": 4.2002,
+      "step": 1736
+    },
+    {
+      "epoch": 0.01737,
+      "grad_norm": 0.7049958430065779,
+      "learning_rate": 0.003,
+      "loss": 4.2254,
+      "step": 1737
+    },
+    {
+      "epoch": 0.01738,
+      "grad_norm": 0.7322441940210889,
+      "learning_rate": 0.003,
+      "loss": 4.2135,
+      "step": 1738
+    },
+    {
+      "epoch": 0.01739,
+      "grad_norm": 0.7327690195119888,
+      "learning_rate": 0.003,
+      "loss": 4.2297,
+      "step": 1739
+    },
+    {
+      "epoch": 0.0174,
+      "grad_norm": 0.7530152956649067,
+      "learning_rate": 0.003,
+      "loss": 4.231,
+      "step": 1740
+    },
+    {
+      "epoch": 0.01741,
+      "grad_norm": 0.7681419574064243,
+      "learning_rate": 0.003,
+      "loss": 4.2369,
+      "step": 1741
+    },
+    {
+      "epoch": 0.01742,
+      "grad_norm": 0.8179621872439043,
+      "learning_rate": 0.003,
+      "loss": 4.2278,
+      "step": 1742
+    },
+    {
+      "epoch": 0.01743,
+      "grad_norm": 0.9578047731817773,
+      "learning_rate": 0.003,
+      "loss": 4.2432,
+      "step": 1743
+    },
+    {
+      "epoch": 0.01744,
+      "grad_norm": 1.0853544694502002,
+      "learning_rate": 0.003,
+      "loss": 4.2354,
+      "step": 1744
+    },
+    {
+      "epoch": 0.01745,
+      "grad_norm": 0.901300254011202,
+      "learning_rate": 0.003,
+      "loss": 4.1985,
+      "step": 1745
+    },
+    {
+      "epoch": 0.01746,
+      "grad_norm": 0.7977925606337166,
+      "learning_rate": 0.003,
+      "loss": 4.2177,
+      "step": 1746
+    },
+    {
+      "epoch": 0.01747,
+      "grad_norm": 0.7805003864858971,
+      "learning_rate": 0.003,
+      "loss": 4.2153,
+      "step": 1747
+    },
+    {
+      "epoch": 0.01748,
+      "grad_norm": 0.5958068748685724,
+      "learning_rate": 0.003,
+      "loss": 4.2004,
+      "step": 1748
+    },
+    {
+      "epoch": 0.01749,
+      "grad_norm": 0.5609955740586448,
+      "learning_rate": 0.003,
+      "loss": 4.1969,
+      "step": 1749
+    },
+    {
+      "epoch": 0.0175,
+      "grad_norm": 0.5677742109277946,
+      "learning_rate": 0.003,
+      "loss": 4.2121,
+      "step": 1750
+    },
+    {
+      "epoch": 0.01751,
+      "grad_norm": 0.5067010918467684,
+      "learning_rate": 0.003,
+      "loss": 4.2009,
+      "step": 1751
+    },
+    {
+      "epoch": 0.01752,
+      "grad_norm": 0.4469462107071404,
+      "learning_rate": 0.003,
+      "loss": 4.239,
+      "step": 1752
+    },
+    {
+      "epoch": 0.01753,
+      "grad_norm": 0.4299529922000094,
+      "learning_rate": 0.003,
+      "loss": 4.1859,
+      "step": 1753
+    },
+    {
+      "epoch": 0.01754,
+      "grad_norm": 0.39327472210537173,
+      "learning_rate": 0.003,
+      "loss": 4.2141,
+      "step": 1754
+    },
+    {
+      "epoch": 0.01755,
+      "grad_norm": 0.4833428525390586,
+      "learning_rate": 0.003,
+      "loss": 4.1582,
+      "step": 1755
+    },
+    {
+      "epoch": 0.01756,
+      "grad_norm": 0.5692339112673285,
+      "learning_rate": 0.003,
+      "loss": 4.2351,
+      "step": 1756
+    },
+    {
+      "epoch": 0.01757,
+      "grad_norm": 0.765825485079122,
+      "learning_rate": 0.003,
+      "loss": 4.2204,
+      "step": 1757
+    },
+    {
+      "epoch": 0.01758,
+      "grad_norm": 0.9042836101782444,
+      "learning_rate": 0.003,
+      "loss": 4.2384,
+      "step": 1758
+    },
+    {
+      "epoch": 0.01759,
+      "grad_norm": 0.8742540565152447,
+      "learning_rate": 0.003,
+      "loss": 4.2064,
+      "step": 1759
+    },
+    {
+      "epoch": 0.0176,
+      "grad_norm": 0.7243758331092184,
+      "learning_rate": 0.003,
+      "loss": 4.2193,
+      "step": 1760
+    },
+    {
+      "epoch": 0.01761,
+      "grad_norm": 0.6848453720193391,
+      "learning_rate": 0.003,
+      "loss": 4.2074,
+      "step": 1761
+    },
+    {
+      "epoch": 0.01762,
+      "grad_norm": 0.8704455967545811,
+      "learning_rate": 0.003,
+      "loss": 4.2328,
+      "step": 1762
+    },
+    {
+      "epoch": 0.01763,
+      "grad_norm": 0.8471557315960334,
+      "learning_rate": 0.003,
+      "loss": 4.2246,
+      "step": 1763
+    },
+    {
+      "epoch": 0.01764,
+      "grad_norm": 0.7827279567320375,
+      "learning_rate": 0.003,
+      "loss": 4.1921,
+      "step": 1764
+    },
+    {
+      "epoch": 0.01765,
+      "grad_norm": 0.8790545608575575,
+      "learning_rate": 0.003,
+      "loss": 4.2237,
+      "step": 1765
+    },
+    {
+      "epoch": 0.01766,
+      "grad_norm": 0.763900084595892,
+      "learning_rate": 0.003,
+      "loss": 4.2322,
+      "step": 1766
+    },
+    {
+      "epoch": 0.01767,
+      "grad_norm": 0.7449808930716393,
+      "learning_rate": 0.003,
+      "loss": 4.1962,
+      "step": 1767
+    },
+    {
+      "epoch": 0.01768,
+      "grad_norm": 0.6931774004435453,
+      "learning_rate": 0.003,
+      "loss": 4.2017,
+      "step": 1768
+    },
+    {
+      "epoch": 0.01769,
+      "grad_norm": 0.7426550774463054,
+      "learning_rate": 0.003,
+      "loss": 4.2066,
+      "step": 1769
+    },
+    {
+      "epoch": 0.0177,
+      "grad_norm": 0.7834040480889046,
+      "learning_rate": 0.003,
+      "loss": 4.1901,
+      "step": 1770
+    },
+    {
+      "epoch": 0.01771,
+      "grad_norm": 0.6750175361724853,
+      "learning_rate": 0.003,
+      "loss": 4.2195,
+      "step": 1771
+    },
+    {
+      "epoch": 0.01772,
+      "grad_norm": 0.6376494933849937,
+      "learning_rate": 0.003,
+      "loss": 4.1888,
+      "step": 1772
+    },
+    {
+      "epoch": 0.01773,
+      "grad_norm": 0.6192068250431033,
+      "learning_rate": 0.003,
+      "loss": 4.1807,
+      "step": 1773
+    },
+    {
+      "epoch": 0.01774,
+      "grad_norm": 0.6106095483716042,
+      "learning_rate": 0.003,
+      "loss": 4.2034,
+      "step": 1774
+    },
+    {
+      "epoch": 0.01775,
+      "grad_norm": 0.6196510426340663,
+      "learning_rate": 0.003,
+      "loss": 4.1895,
+      "step": 1775
+    },
+    {
+      "epoch": 0.01776,
+      "grad_norm": 0.7097137120766858,
+      "learning_rate": 0.003,
+      "loss": 4.1828,
+      "step": 1776
+    },
+    {
+      "epoch": 0.01777,
+      "grad_norm": 0.7063319061887237,
+      "learning_rate": 0.003,
+      "loss": 4.2044,
+      "step": 1777
+    },
+    {
+      "epoch": 0.01778,
+      "grad_norm": 0.629621456935323,
+      "learning_rate": 0.003,
+      "loss": 4.194,
+      "step": 1778
+    },
+    {
+      "epoch": 0.01779,
+      "grad_norm": 0.7088930396533265,
+      "learning_rate": 0.003,
+      "loss": 4.215,
+      "step": 1779
+    },
+    {
+      "epoch": 0.0178,
+      "grad_norm": 0.7253919374521124,
+      "learning_rate": 0.003,
+      "loss": 4.191,
+      "step": 1780
+    },
+    {
+      "epoch": 0.01781,
+      "grad_norm": 0.7852702442407208,
+      "learning_rate": 0.003,
+      "loss": 4.2044,
+      "step": 1781
+    },
+    {
+      "epoch": 0.01782,
+      "grad_norm": 0.8394755648372281,
+      "learning_rate": 0.003,
+      "loss": 4.2328,
+      "step": 1782
+    },
+    {
+      "epoch": 0.01783,
+      "grad_norm": 0.6910931858922343,
+      "learning_rate": 0.003,
+      "loss": 4.2068,
+      "step": 1783
+    },
+    {
+      "epoch": 0.01784,
+      "grad_norm": 0.7674520462715927,
+      "learning_rate": 0.003,
+      "loss": 4.2107,
+      "step": 1784
+    },
+    {
+      "epoch": 0.01785,
+      "grad_norm": 0.8199387977488617,
+      "learning_rate": 0.003,
+      "loss": 4.1741,
+      "step": 1785
+    },
+    {
+      "epoch": 0.01786,
+      "grad_norm": 0.9173162484249777,
+      "learning_rate": 0.003,
+      "loss": 4.2074,
+      "step": 1786
+    },
+    {
+      "epoch": 0.01787,
+      "grad_norm": 1.334253988039247,
+      "learning_rate": 0.003,
+      "loss": 4.2336,
+      "step": 1787
+    },
+    {
+      "epoch": 0.01788,
+      "grad_norm": 1.0880162732024794,
+      "learning_rate": 0.003,
+      "loss": 4.2464,
+      "step": 1788
+    },
+    {
+      "epoch": 0.01789,
+      "grad_norm": 0.8768448988019708,
+      "learning_rate": 0.003,
+      "loss": 4.2114,
+      "step": 1789
+    },
+    {
+      "epoch": 0.0179,
+      "grad_norm": 0.9478883510444388,
+      "learning_rate": 0.003,
+      "loss": 4.2362,
+      "step": 1790
+    },
+    {
+      "epoch": 0.01791,
+      "grad_norm": 0.9789788338963192,
+      "learning_rate": 0.003,
+      "loss": 4.2475,
+      "step": 1791
+    },
+    {
+      "epoch": 0.01792,
+      "grad_norm": 0.9379061880413183,
+      "learning_rate": 0.003,
+      "loss": 4.2484,
+      "step": 1792
+    },
+    {
+      "epoch": 0.01793,
+      "grad_norm": 0.9327271640213346,
+      "learning_rate": 0.003,
+      "loss": 4.215,
+      "step": 1793
+    },
+    {
+      "epoch": 0.01794,
+      "grad_norm": 0.7690204695866035,
+      "learning_rate": 0.003,
+      "loss": 4.2222,
+      "step": 1794
+    },
+    {
+      "epoch": 0.01795,
+      "grad_norm": 0.6971902449310058,
+      "learning_rate": 0.003,
+      "loss": 4.2172,
+      "step": 1795
+    },
+    {
+      "epoch": 0.01796,
+      "grad_norm": 0.762901849453012,
+      "learning_rate": 0.003,
+      "loss": 4.2347,
+      "step": 1796
+    },
+    {
+      "epoch": 0.01797,
+      "grad_norm": 0.8440627560282756,
+      "learning_rate": 0.003,
+      "loss": 4.2037,
+      "step": 1797
+    },
+    {
+      "epoch": 0.01798,
+      "grad_norm": 0.8368125917294373,
+      "learning_rate": 0.003,
+      "loss": 4.2148,
+      "step": 1798
+    },
+    {
+      "epoch": 0.01799,
+      "grad_norm": 0.7348763481882639,
+      "learning_rate": 0.003,
+      "loss": 4.2247,
+      "step": 1799
+    },
+    {
+      "epoch": 0.018,
+      "grad_norm": 0.7680359832381647,
+      "learning_rate": 0.003,
+      "loss": 4.2184,
+      "step": 1800
+    },
+    {
+      "epoch": 0.01801,
+      "grad_norm": 0.7404783794030995,
+      "learning_rate": 0.003,
+      "loss": 4.2101,
+      "step": 1801
+    },
+    {
+      "epoch": 0.01802,
+      "grad_norm": 0.6580953290368566,
+      "learning_rate": 0.003,
+      "loss": 4.2238,
+      "step": 1802
+    },
+    {
+      "epoch": 0.01803,
+      "grad_norm": 0.6630631878191348,
+      "learning_rate": 0.003,
+      "loss": 4.2284,
+      "step": 1803
+    },
+    {
+      "epoch": 0.01804,
+      "grad_norm": 0.6396012711111247,
+      "learning_rate": 0.003,
+      "loss": 4.1992,
+      "step": 1804
+    },
+    {
+      "epoch": 0.01805,
+      "grad_norm": 0.5559219636409923,
+      "learning_rate": 0.003,
+      "loss": 4.2154,
+      "step": 1805
+    },
+    {
+      "epoch": 0.01806,
+      "grad_norm": 0.6144421410960899,
+      "learning_rate": 0.003,
+      "loss": 4.1763,
+      "step": 1806
+    },
+    {
+      "epoch": 0.01807,
+      "grad_norm": 0.6146777760728932,
+      "learning_rate": 0.003,
+      "loss": 4.2121,
+      "step": 1807
+    },
+    {
+      "epoch": 0.01808,
+      "grad_norm": 0.6116485891270191,
+      "learning_rate": 0.003,
+      "loss": 4.2001,
+      "step": 1808
+    },
+    {
+      "epoch": 0.01809,
+      "grad_norm": 0.8096274979504547,
+      "learning_rate": 0.003,
+      "loss": 4.2275,
+      "step": 1809
+    },
+    {
+      "epoch": 0.0181,
+      "grad_norm": 1.1102312559563035,
+      "learning_rate": 0.003,
+      "loss": 4.2346,
+      "step": 1810
+    },
+    {
+      "epoch": 0.01811,
+      "grad_norm": 0.8754620418282368,
+      "learning_rate": 0.003,
+      "loss": 4.2216,
+      "step": 1811
+    },
+    {
+      "epoch": 0.01812,
+      "grad_norm": 0.6567434234036816,
+      "learning_rate": 0.003,
+      "loss": 4.1947,
+      "step": 1812
+    },
+    {
+      "epoch": 0.01813,
+      "grad_norm": 0.7375490051000831,
+      "learning_rate": 0.003,
+      "loss": 4.2409,
+      "step": 1813
+    },
+    {
+      "epoch": 0.01814,
+      "grad_norm": 0.7842570681491089,
+      "learning_rate": 0.003,
+      "loss": 4.2075,
+      "step": 1814
+    },
+    {
+      "epoch": 0.01815,
+      "grad_norm": 0.6332278380543743,
+      "learning_rate": 0.003,
+      "loss": 4.1891,
+      "step": 1815
+    },
+    {
+      "epoch": 0.01816,
+      "grad_norm": 0.6578340036445237,
+      "learning_rate": 0.003,
+      "loss": 4.2166,
+      "step": 1816
+    },
+    {
+      "epoch": 0.01817,
+      "grad_norm": 0.6624785828428507,
+      "learning_rate": 0.003,
+      "loss": 4.2068,
+      "step": 1817
+    },
+    {
+      "epoch": 0.01818,
+      "grad_norm": 0.5897502755122268,
+      "learning_rate": 0.003,
+      "loss": 4.2034,
+      "step": 1818
+    },
+    {
+      "epoch": 0.01819,
+      "grad_norm": 0.49357607734419184,
+      "learning_rate": 0.003,
+      "loss": 4.1821,
+      "step": 1819
+    },
+    {
+      "epoch": 0.0182,
+      "grad_norm": 0.5025205629657381,
+      "learning_rate": 0.003,
+      "loss": 4.2022,
+      "step": 1820
+    },
+    {
+      "epoch": 0.01821,
+      "grad_norm": 0.5231764315153676,
+      "learning_rate": 0.003,
+      "loss": 4.188,
+      "step": 1821
+    },
+    {
+      "epoch": 0.01822,
+      "grad_norm": 0.589855961825142,
+      "learning_rate": 0.003,
+      "loss": 4.2017,
+      "step": 1822
+    },
+    {
+      "epoch": 0.01823,
+      "grad_norm": 0.6454531707936206,
+      "learning_rate": 0.003,
+      "loss": 4.2159,
+      "step": 1823
+    },
+    {
+      "epoch": 0.01824,
+      "grad_norm": 0.616450651672451,
+      "learning_rate": 0.003,
+      "loss": 4.1987,
+      "step": 1824
+    },
+    {
+      "epoch": 0.01825,
+      "grad_norm": 0.5393299556773667,
+      "learning_rate": 0.003,
+      "loss": 4.199,
+      "step": 1825
+    },
+    {
+      "epoch": 0.01826,
+      "grad_norm": 0.6162633558651508,
+      "learning_rate": 0.003,
+      "loss": 4.2135,
+      "step": 1826
+    },
+    {
+      "epoch": 0.01827,
+      "grad_norm": 0.6366286833338192,
+      "learning_rate": 0.003,
+      "loss": 4.1847,
+      "step": 1827
+    },
+    {
+      "epoch": 0.01828,
+      "grad_norm": 0.5519308619846083,
+      "learning_rate": 0.003,
+      "loss": 4.1935,
+      "step": 1828
+    },
+    {
+      "epoch": 0.01829,
+      "grad_norm": 0.556700588713125,
+      "learning_rate": 0.003,
+      "loss": 4.2083,
+      "step": 1829
+    },
+    {
+      "epoch": 0.0183,
+      "grad_norm": 0.6225243144725614,
+      "learning_rate": 0.003,
+      "loss": 4.2039,
+      "step": 1830
+    },
+    {
+      "epoch": 0.01831,
+      "grad_norm": 0.8092938095153425,
+      "learning_rate": 0.003,
+      "loss": 4.2017,
+      "step": 1831
+    },
+    {
+      "epoch": 0.01832,
+      "grad_norm": 1.156880999962459,
+      "learning_rate": 0.003,
+      "loss": 4.214,
+      "step": 1832
+    },
+    {
+      "epoch": 0.01833,
+      "grad_norm": 0.8498167035958548,
+      "learning_rate": 0.003,
+      "loss": 4.2422,
+      "step": 1833
+    },
+    {
+      "epoch": 0.01834,
+      "grad_norm": 0.6651811196480657,
+      "learning_rate": 0.003,
+      "loss": 4.2008,
+      "step": 1834
+    },
+    {
+      "epoch": 0.01835,
+      "grad_norm": 0.7438128363368247,
+      "learning_rate": 0.003,
+      "loss": 4.205,
+      "step": 1835
+    },
+    {
+      "epoch": 0.01836,
+      "grad_norm": 0.7712205326044778,
+      "learning_rate": 0.003,
+      "loss": 4.2014,
+      "step": 1836
+    },
+    {
+      "epoch": 0.01837,
+      "grad_norm": 0.7094945295905764,
+      "learning_rate": 0.003,
+      "loss": 4.1968,
+      "step": 1837
+    },
+    {
+      "epoch": 0.01838,
+      "grad_norm": 0.8255996162995448,
+      "learning_rate": 0.003,
+      "loss": 4.2049,
+      "step": 1838
+    },
+    {
+      "epoch": 0.01839,
+      "grad_norm": 0.7001056258854736,
+      "learning_rate": 0.003,
+      "loss": 4.2234,
+      "step": 1839
+    },
+    {
+      "epoch": 0.0184,
+      "grad_norm": 0.7156089343491636,
+      "learning_rate": 0.003,
+      "loss": 4.2248,
+      "step": 1840
+    },
+    {
+      "epoch": 0.01841,
+      "grad_norm": 0.7514284093505675,
+      "learning_rate": 0.003,
+      "loss": 4.164,
+      "step": 1841
+    },
+    {
+      "epoch": 0.01842,
+      "grad_norm": 0.7804042052375482,
+      "learning_rate": 0.003,
+      "loss": 4.1995,
+      "step": 1842
+    },
+    {
+      "epoch": 0.01843,
+      "grad_norm": 0.7741073572881331,
+      "learning_rate": 0.003,
+      "loss": 4.1826,
+      "step": 1843
+    },
+    {
+      "epoch": 0.01844,
+      "grad_norm": 0.8011416312871611,
+      "learning_rate": 0.003,
+      "loss": 4.1842,
+      "step": 1844
+    },
+    {
+      "epoch": 0.01845,
+      "grad_norm": 0.8386720056960224,
+      "learning_rate": 0.003,
+      "loss": 4.187,
+      "step": 1845
+    },
+    {
+      "epoch": 0.01846,
+      "grad_norm": 0.9781723483143008,
+      "learning_rate": 0.003,
+      "loss": 4.2242,
+      "step": 1846
+    },
+    {
+      "epoch": 0.01847,
+      "grad_norm": 0.9173016227544075,
+      "learning_rate": 0.003,
+      "loss": 4.2189,
+      "step": 1847
+    },
+    {
+      "epoch": 0.01848,
+      "grad_norm": 0.961352235410328,
+      "learning_rate": 0.003,
+      "loss": 4.2359,
+      "step": 1848
+    },
+    {
+      "epoch": 0.01849,
+      "grad_norm": 0.9115360913743059,
+      "learning_rate": 0.003,
+      "loss": 4.241,
+      "step": 1849
+    },
+    {
+      "epoch": 0.0185,
+      "grad_norm": 0.949227003474929,
+      "learning_rate": 0.003,
+      "loss": 4.2298,
+      "step": 1850
+    },
+    {
+      "epoch": 0.01851,
+      "grad_norm": 0.756851342238193,
+      "learning_rate": 0.003,
+      "loss": 4.2063,
+      "step": 1851
+    },
+    {
+      "epoch": 0.01852,
+      "grad_norm": 0.7062657388831308,
+      "learning_rate": 0.003,
+      "loss": 4.2058,
+      "step": 1852
+    },
+    {
+      "epoch": 0.01853,
+      "grad_norm": 0.804960952863702,
+      "learning_rate": 0.003,
+      "loss": 4.206,
+      "step": 1853
+    },
+    {
+      "epoch": 0.01854,
+      "grad_norm": 0.8048402157100555,
+      "learning_rate": 0.003,
+      "loss": 4.2379,
+      "step": 1854
+    },
+    {
+      "epoch": 0.01855,
+      "grad_norm": 0.8514933979131342,
+      "learning_rate": 0.003,
+      "loss": 4.2236,
+      "step": 1855
+    },
+    {
+      "epoch": 0.01856,
+      "grad_norm": 0.9293561457561126,
+      "learning_rate": 0.003,
+      "loss": 4.2144,
+      "step": 1856
+    },
+    {
+      "epoch": 0.01857,
+      "grad_norm": 0.8452785281639297,
+      "learning_rate": 0.003,
+      "loss": 4.1937,
+      "step": 1857
+    },
+    {
+      "epoch": 0.01858,
+      "grad_norm": 0.805524803230092,
+      "learning_rate": 0.003,
+      "loss": 4.2015,
+      "step": 1858
+    },
+    {
+      "epoch": 0.01859,
+      "grad_norm": 0.826226040676062,
+      "learning_rate": 0.003,
+      "loss": 4.2023,
+      "step": 1859
+    },
+    {
+      "epoch": 0.0186,
+      "grad_norm": 0.7646190876843869,
+      "learning_rate": 0.003,
+      "loss": 4.1808,
+      "step": 1860
+    },
+    {
+      "epoch": 0.01861,
+      "grad_norm": 0.9398444707205434,
+      "learning_rate": 0.003,
+      "loss": 4.229,
+      "step": 1861
+    },
+    {
+      "epoch": 0.01862,
+      "grad_norm": 1.0799822817783482,
+      "learning_rate": 0.003,
+      "loss": 4.208,
+      "step": 1862
+    },
+    {
+      "epoch": 0.01863,
+      "grad_norm": 1.0663393244093409,
+      "learning_rate": 0.003,
+      "loss": 4.2176,
+      "step": 1863
+    },
+    {
+      "epoch": 0.01864,
+      "grad_norm": 0.9063119158507146,
+      "learning_rate": 0.003,
+      "loss": 4.243,
+      "step": 1864
+    },
+    {
+      "epoch": 0.01865,
+      "grad_norm": 0.8878085186485055,
+      "learning_rate": 0.003,
+      "loss": 4.222,
+      "step": 1865
+    },
+    {
+      "epoch": 0.01866,
+      "grad_norm": 0.8278859311900081,
+      "learning_rate": 0.003,
+      "loss": 4.2161,
+      "step": 1866
+    },
+    {
+      "epoch": 0.01867,
+      "grad_norm": 0.9175811468419192,
+      "learning_rate": 0.003,
+      "loss": 4.1983,
+      "step": 1867
+    },
+    {
+      "epoch": 0.01868,
+      "grad_norm": 0.8195895794655929,
+      "learning_rate": 0.003,
+      "loss": 4.2097,
+      "step": 1868
+    },
+    {
+      "epoch": 0.01869,
+      "grad_norm": 0.6958671652018995,
+      "learning_rate": 0.003,
+      "loss": 4.2233,
+      "step": 1869
+    },
+    {
+      "epoch": 0.0187,
+      "grad_norm": 0.7276701584872959,
+      "learning_rate": 0.003,
+      "loss": 4.2132,
+      "step": 1870
+    },
+    {
+      "epoch": 0.01871,
+      "grad_norm": 0.700252662578374,
+      "learning_rate": 0.003,
+      "loss": 4.1886,
+      "step": 1871
+    },
+    {
+      "epoch": 0.01872,
+      "grad_norm": 0.7936164889803413,
+      "learning_rate": 0.003,
+      "loss": 4.2323,
+      "step": 1872
+    },
+    {
+      "epoch": 0.01873,
+      "grad_norm": 0.8026794723116766,
+      "learning_rate": 0.003,
+      "loss": 4.1938,
+      "step": 1873
+    },
+    {
+      "epoch": 0.01874,
+      "grad_norm": 0.6813841695856067,
+      "learning_rate": 0.003,
+      "loss": 4.1939,
+      "step": 1874
+    },
+    {
+      "epoch": 0.01875,
+      "grad_norm": 0.7703759265763755,
+      "learning_rate": 0.003,
+      "loss": 4.2154,
+      "step": 1875
+    },
+    {
+      "epoch": 0.01876,
+      "grad_norm": 0.7048464208203703,
+      "learning_rate": 0.003,
+      "loss": 4.2195,
+      "step": 1876
+    },
+    {
+      "epoch": 0.01877,
+      "grad_norm": 0.6552231659596246,
+      "learning_rate": 0.003,
+      "loss": 4.2204,
+      "step": 1877
+    },
+    {
+      "epoch": 0.01878,
+      "grad_norm": 0.6594564394551935,
+      "learning_rate": 0.003,
+      "loss": 4.2061,
+      "step": 1878
+    },
+    {
+      "epoch": 0.01879,
+      "grad_norm": 0.6775220505077814,
+      "learning_rate": 0.003,
+      "loss": 4.1841,
+      "step": 1879
+    },
+    {
+      "epoch": 0.0188,
+      "grad_norm": 0.7469198030113778,
+      "learning_rate": 0.003,
+      "loss": 4.1948,
+      "step": 1880
+    },
+    {
+      "epoch": 0.01881,
+      "grad_norm": 0.8052292593043413,
+      "learning_rate": 0.003,
+      "loss": 4.1959,
+      "step": 1881
+    },
+    {
+      "epoch": 0.01882,
+      "grad_norm": 0.9310051602523918,
+      "learning_rate": 0.003,
+      "loss": 4.2256,
+      "step": 1882
+    },
+    {
+      "epoch": 0.01883,
+      "grad_norm": 0.8470067207894152,
+      "learning_rate": 0.003,
+      "loss": 4.2057,
+      "step": 1883
+    },
+    {
+      "epoch": 0.01884,
+      "grad_norm": 0.7163623431526733,
+      "learning_rate": 0.003,
+      "loss": 4.1948,
+      "step": 1884
+    },
+    {
+      "epoch": 0.01885,
+      "grad_norm": 0.7917195791610646,
+      "learning_rate": 0.003,
+      "loss": 4.1745,
+      "step": 1885
+    },
+    {
+      "epoch": 0.01886,
+      "grad_norm": 0.7950048534183088,
+      "learning_rate": 0.003,
+      "loss": 4.2171,
+      "step": 1886
+    },
+    {
+      "epoch": 0.01887,
+      "grad_norm": 0.7209518927340428,
+      "learning_rate": 0.003,
+      "loss": 4.2025,
+      "step": 1887
+    },
+    {
+      "epoch": 0.01888,
+      "grad_norm": 0.633107358042472,
+      "learning_rate": 0.003,
+      "loss": 4.1883,
+      "step": 1888
+    },
+    {
+      "epoch": 0.01889,
+      "grad_norm": 0.6143038001546575,
+      "learning_rate": 0.003,
+      "loss": 4.1996,
+      "step": 1889
+    },
+    {
+      "epoch": 0.0189,
+      "grad_norm": 0.5158052453680656,
+      "learning_rate": 0.003,
+      "loss": 4.1757,
+      "step": 1890
+    },
+    {
+      "epoch": 0.01891,
+      "grad_norm": 0.4633760186396196,
+      "learning_rate": 0.003,
+      "loss": 4.2224,
+      "step": 1891
+    },
+    {
+      "epoch": 0.01892,
+      "grad_norm": 0.4927650862046063,
+      "learning_rate": 0.003,
+      "loss": 4.1956,
+      "step": 1892
+    },
+    {
+      "epoch": 0.01893,
+      "grad_norm": 0.4373024700349087,
+      "learning_rate": 0.003,
+      "loss": 4.1773,
+      "step": 1893
+    },
+    {
+      "epoch": 0.01894,
+      "grad_norm": 0.40335105556509215,
+      "learning_rate": 0.003,
+      "loss": 4.1784,
+      "step": 1894
+    },
+    {
+      "epoch": 0.01895,
+      "grad_norm": 0.3931015972388579,
+      "learning_rate": 0.003,
+      "loss": 4.1649,
+      "step": 1895
+    },
+    {
+      "epoch": 0.01896,
+      "grad_norm": 0.369330374322213,
+      "learning_rate": 0.003,
+      "loss": 4.1861,
+      "step": 1896
+    },
+    {
+      "epoch": 0.01897,
+      "grad_norm": 0.38394762475942135,
+      "learning_rate": 0.003,
+      "loss": 4.1841,
+      "step": 1897
+    },
+    {
+      "epoch": 0.01898,
+      "grad_norm": 0.4616420599151427,
+      "learning_rate": 0.003,
+      "loss": 4.1533,
+      "step": 1898
+    },
+    {
+      "epoch": 0.01899,
+      "grad_norm": 0.6786086644128849,
+      "learning_rate": 0.003,
+      "loss": 4.1744,
+      "step": 1899
+    },
+    {
+      "epoch": 0.019,
+      "grad_norm": 1.1028955308001205,
+      "learning_rate": 0.003,
+      "loss": 4.1821,
+      "step": 1900
+    },
+    {
+      "epoch": 0.01901,
+      "grad_norm": 1.0381658886424407,
+      "learning_rate": 0.003,
+      "loss": 4.2416,
+      "step": 1901
+    },
+    {
+      "epoch": 0.01902,
+      "grad_norm": 0.6228046274507568,
+      "learning_rate": 0.003,
+      "loss": 4.202,
+      "step": 1902
+    },
+    {
+      "epoch": 0.01903,
+      "grad_norm": 0.552687373313921,
+      "learning_rate": 0.003,
+      "loss": 4.1547,
+      "step": 1903
+    },
+    {
+      "epoch": 0.01904,
+      "grad_norm": 0.5642583939104693,
+      "learning_rate": 0.003,
+      "loss": 4.2113,
+      "step": 1904
+    },
+    {
+      "epoch": 0.01905,
+      "grad_norm": 0.4866940659015072,
+      "learning_rate": 0.003,
+      "loss": 4.1815,
+      "step": 1905
+    },
+    {
+      "epoch": 0.01906,
+      "grad_norm": 0.6770303387740269,
+      "learning_rate": 0.003,
+      "loss": 4.2038,
+      "step": 1906
+    },
+    {
+      "epoch": 0.01907,
+      "grad_norm": 0.6799220004643979,
+      "learning_rate": 0.003,
+      "loss": 4.199,
+      "step": 1907
+    },
+    {
+      "epoch": 0.01908,
+      "grad_norm": 0.5738027965979297,
+      "learning_rate": 0.003,
+      "loss": 4.2084,
+      "step": 1908
+    },
+    {
+      "epoch": 0.01909,
+      "grad_norm": 0.5646602374989558,
+      "learning_rate": 0.003,
+      "loss": 4.1743,
+      "step": 1909
+    },
+    {
+      "epoch": 0.0191,
+      "grad_norm": 0.6702075278822099,
+      "learning_rate": 0.003,
+      "loss": 4.1866,
+      "step": 1910
+    },
+    {
+      "epoch": 0.01911,
+      "grad_norm": 0.7731308719042311,
+      "learning_rate": 0.003,
+      "loss": 4.1672,
+      "step": 1911
+    },
+    {
+      "epoch": 0.01912,
+      "grad_norm": 0.6258475347121828,
+      "learning_rate": 0.003,
+      "loss": 4.1744,
+      "step": 1912
+    },
+    {
+      "epoch": 0.01913,
+      "grad_norm": 0.6373334035841824,
+      "learning_rate": 0.003,
+      "loss": 4.1923,
+      "step": 1913
+    },
+    {
+      "epoch": 0.01914,
+      "grad_norm": 0.7497697739080498,
+      "learning_rate": 0.003,
+      "loss": 4.212,
+      "step": 1914
+    },
+    {
+      "epoch": 0.01915,
+      "grad_norm": 0.8443649873420552,
+      "learning_rate": 0.003,
+      "loss": 4.204,
+      "step": 1915
+    },
+    {
+      "epoch": 0.01916,
+      "grad_norm": 0.8327185244516057,
+      "learning_rate": 0.003,
+      "loss": 4.1829,
+      "step": 1916
+    },
+    {
+      "epoch": 0.01917,
+      "grad_norm": 0.8831134466171928,
+      "learning_rate": 0.003,
+      "loss": 4.2052,
+      "step": 1917
+    },
+    {
+      "epoch": 0.01918,
+      "grad_norm": 0.9318795049142128,
+      "learning_rate": 0.003,
+      "loss": 4.1944,
+      "step": 1918
+    },
+    {
+      "epoch": 0.01919,
+      "grad_norm": 0.944334593915859,
+      "learning_rate": 0.003,
+      "loss": 4.2049,
+      "step": 1919
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 0.9365986219494583,
+      "learning_rate": 0.003,
+      "loss": 4.2076,
+      "step": 1920
+    },
+    {
+      "epoch": 0.01921,
+      "grad_norm": 0.9617762005888578,
+      "learning_rate": 0.003,
+      "loss": 4.2016,
+      "step": 1921
+    },
+    {
+      "epoch": 0.01922,
+      "grad_norm": 0.9805293516298391,
+      "learning_rate": 0.003,
+      "loss": 4.2181,
+      "step": 1922
+    },
+    {
+      "epoch": 0.01923,
+      "grad_norm": 0.8498471851851511,
+      "learning_rate": 0.003,
+      "loss": 4.2204,
+      "step": 1923
+    },
+    {
+      "epoch": 0.01924,
+      "grad_norm": 0.9257747078341632,
+      "learning_rate": 0.003,
+      "loss": 4.2284,
+      "step": 1924
+    },
+    {
+      "epoch": 0.01925,
+      "grad_norm": 1.0324220402854314,
+      "learning_rate": 0.003,
+      "loss": 4.2182,
+      "step": 1925
+    },
+    {
+      "epoch": 0.01926,
+      "grad_norm": 0.9745230962825202,
+      "learning_rate": 0.003,
+      "loss": 4.2282,
+      "step": 1926
+    },
+    {
+      "epoch": 0.01927,
+      "grad_norm": 0.7705843813355484,
+      "learning_rate": 0.003,
+      "loss": 4.2129,
+      "step": 1927
+    },
+    {
+      "epoch": 0.01928,
+      "grad_norm": 0.5911790602013839,
+      "learning_rate": 0.003,
+      "loss": 4.1789,
+      "step": 1928
+    },
+    {
+      "epoch": 0.01929,
+      "grad_norm": 0.5782217294801848,
+      "learning_rate": 0.003,
+      "loss": 4.1986,
+      "step": 1929
+    },
+    {
+      "epoch": 0.0193,
+      "grad_norm": 0.5556254348196169,
+      "learning_rate": 0.003,
+      "loss": 4.1871,
+      "step": 1930
+    },
+    {
+      "epoch": 0.01931,
+      "grad_norm": 0.6337897693351197,
+      "learning_rate": 0.003,
+      "loss": 4.188,
+      "step": 1931
+    },
+    {
+      "epoch": 0.01932,
+      "grad_norm": 0.6851141809276902,
+      "learning_rate": 0.003,
+      "loss": 4.1982,
+      "step": 1932
+    },
+    {
+      "epoch": 0.01933,
+      "grad_norm": 0.6532471380653952,
+      "learning_rate": 0.003,
+      "loss": 4.1838,
+      "step": 1933
+    },
+    {
+      "epoch": 0.01934,
+      "grad_norm": 0.660311918680962,
+      "learning_rate": 0.003,
+      "loss": 4.1881,
+      "step": 1934
+    },
+    {
+      "epoch": 0.01935,
+      "grad_norm": 0.6202434802996027,
+      "learning_rate": 0.003,
+      "loss": 4.2019,
+      "step": 1935
+    },
+    {
+      "epoch": 0.01936,
+      "grad_norm": 0.6456697693032586,
+      "learning_rate": 0.003,
+      "loss": 4.1829,
+      "step": 1936
+    },
+    {
+      "epoch": 0.01937,
+      "grad_norm": 0.6945407817649866,
+      "learning_rate": 0.003,
+      "loss": 4.1968,
+      "step": 1937
+    },
+    {
+      "epoch": 0.01938,
+      "grad_norm": 0.7393543695473613,
+      "learning_rate": 0.003,
+      "loss": 4.1957,
+      "step": 1938
+    },
+    {
+      "epoch": 0.01939,
+      "grad_norm": 0.7713579649815137,
+      "learning_rate": 0.003,
+      "loss": 4.223,
+      "step": 1939
+    },
+    {
+      "epoch": 0.0194,
+      "grad_norm": 0.6228971544711813,
+      "learning_rate": 0.003,
+      "loss": 4.1928,
+      "step": 1940
+    },
+    {
+      "epoch": 0.01941,
+      "grad_norm": 0.5726639902946573,
+      "learning_rate": 0.003,
+      "loss": 4.1739,
+      "step": 1941
+    },
+    {
+      "epoch": 0.01942,
+      "grad_norm": 0.6064973298236461,
+      "learning_rate": 0.003,
+      "loss": 4.1989,
+      "step": 1942
+    },
+    {
+      "epoch": 0.01943,
+      "grad_norm": 0.6888737571597997,
+      "learning_rate": 0.003,
+      "loss": 4.1711,
+      "step": 1943
+    },
+    {
+      "epoch": 0.01944,
+      "grad_norm": 0.6708431670479557,
+      "learning_rate": 0.003,
+      "loss": 4.2158,
+      "step": 1944
+    },
+    {
+      "epoch": 0.01945,
+      "grad_norm": 0.5681415129494921,
+      "learning_rate": 0.003,
+      "loss": 4.1787,
+      "step": 1945
+    },
+    {
+      "epoch": 0.01946,
+      "grad_norm": 0.47730421654011235,
+      "learning_rate": 0.003,
+      "loss": 4.1716,
+      "step": 1946
+    },
+    {
+      "epoch": 0.01947,
+      "grad_norm": 0.47116667407771645,
+      "learning_rate": 0.003,
+      "loss": 4.1791,
+      "step": 1947
+    },
+    {
+      "epoch": 0.01948,
+      "grad_norm": 0.6002169491076368,
+      "learning_rate": 0.003,
+      "loss": 4.2004,
+      "step": 1948
+    },
+    {
+      "epoch": 0.01949,
+      "grad_norm": 0.6804506442238769,
+      "learning_rate": 0.003,
+      "loss": 4.1961,
+      "step": 1949
+    },
+    {
+      "epoch": 0.0195,
+      "grad_norm": 0.9031150085350885,
+      "learning_rate": 0.003,
+      "loss": 4.1909,
+      "step": 1950
+    },
+    {
+      "epoch": 0.01951,
+      "grad_norm": 1.055108763514552,
+      "learning_rate": 0.003,
+      "loss": 4.1972,
+      "step": 1951
+    },
+    {
+      "epoch": 0.01952,
+      "grad_norm": 0.7942759742680753,
+      "learning_rate": 0.003,
+      "loss": 4.1947,
+      "step": 1952
+    },
+    {
+      "epoch": 0.01953,
+      "grad_norm": 0.6995419703208098,
+      "learning_rate": 0.003,
+      "loss": 4.2007,
+      "step": 1953
+    },
+    {
+      "epoch": 0.01954,
+      "grad_norm": 0.7323475440015588,
+      "learning_rate": 0.003,
+      "loss": 4.2017,
+      "step": 1954
+    },
+    {
+      "epoch": 0.01955,
+      "grad_norm": 0.654602209622962,
+      "learning_rate": 0.003,
+      "loss": 4.1897,
+      "step": 1955
+    },
+    {
+      "epoch": 0.01956,
+      "grad_norm": 0.6795706239523924,
+      "learning_rate": 0.003,
+      "loss": 4.188,
+      "step": 1956
+    },
+    {
+      "epoch": 0.01957,
+      "grad_norm": 0.822104190978127,
+      "learning_rate": 0.003,
+      "loss": 4.1611,
+      "step": 1957
+    },
+    {
+      "epoch": 0.01958,
+      "grad_norm": 0.7829306267633842,
+      "learning_rate": 0.003,
+      "loss": 4.1555,
+      "step": 1958
+    },
+    {
+      "epoch": 0.01959,
+      "grad_norm": 0.7521187388682534,
+      "learning_rate": 0.003,
+      "loss": 4.1801,
+      "step": 1959
+    },
+    {
+      "epoch": 0.0196,
+      "grad_norm": 0.8699203949276841,
+      "learning_rate": 0.003,
+      "loss": 4.1777,
+      "step": 1960
+    },
+    {
+      "epoch": 0.01961,
+      "grad_norm": 0.9397547951451798,
+      "learning_rate": 0.003,
+      "loss": 4.1846,
+      "step": 1961
+    },
+    {
+      "epoch": 0.01962,
+      "grad_norm": 0.8367517905224471,
+      "learning_rate": 0.003,
+      "loss": 4.1848,
+      "step": 1962
+    },
+    {
+      "epoch": 0.01963,
+      "grad_norm": 0.8914011026895058,
+      "learning_rate": 0.003,
+      "loss": 4.1735,
+      "step": 1963
+    },
+    {
+      "epoch": 0.01964,
+      "grad_norm": 1.069474116819015,
+      "learning_rate": 0.003,
+      "loss": 4.2192,
+      "step": 1964
+    },
+    {
+      "epoch": 0.01965,
+      "grad_norm": 0.9846418582261246,
+      "learning_rate": 0.003,
+      "loss": 4.2097,
+      "step": 1965
+    },
+    {
+      "epoch": 0.01966,
+      "grad_norm": 1.190189959584559,
+      "learning_rate": 0.003,
+      "loss": 4.2246,
+      "step": 1966
+    },
+    {
+      "epoch": 0.01967,
+      "grad_norm": 0.9331071177303962,
+      "learning_rate": 0.003,
+      "loss": 4.1941,
+      "step": 1967
+    },
+    {
+      "epoch": 0.01968,
+      "grad_norm": 0.8236928876485803,
+      "learning_rate": 0.003,
+      "loss": 4.2092,
+      "step": 1968
+    },
+    {
+      "epoch": 0.01969,
+      "grad_norm": 0.9035337012628801,
+      "learning_rate": 0.003,
+      "loss": 4.2128,
+      "step": 1969
+    },
+    {
+      "epoch": 0.0197,
+      "grad_norm": 0.9584839773684666,
+      "learning_rate": 0.003,
+      "loss": 4.2139,
+      "step": 1970
+    },
+    {
+      "epoch": 0.01971,
+      "grad_norm": 1.1976612578146555,
+      "learning_rate": 0.003,
+      "loss": 4.2078,
+      "step": 1971
+    },
+    {
+      "epoch": 0.01972,
+      "grad_norm": 1.03766882598544,
+      "learning_rate": 0.003,
+      "loss": 4.216,
+      "step": 1972
+    },
+    {
+      "epoch": 0.01973,
+      "grad_norm": 1.157490820023187,
+      "learning_rate": 0.003,
+      "loss": 4.204,
+      "step": 1973
+    },
+    {
+      "epoch": 0.01974,
+      "grad_norm": 1.0168705909824025,
+      "learning_rate": 0.003,
+      "loss": 4.2365,
+      "step": 1974
+    },
+    {
+      "epoch": 0.01975,
+      "grad_norm": 0.9807823986551194,
+      "learning_rate": 0.003,
+      "loss": 4.2359,
+      "step": 1975
+    },
+    {
+      "epoch": 0.01976,
+      "grad_norm": 0.7975927738530828,
+      "learning_rate": 0.003,
+      "loss": 4.2115,
+      "step": 1976
+    },
+    {
+      "epoch": 0.01977,
+      "grad_norm": 0.9111601004721311,
+      "learning_rate": 0.003,
+      "loss": 4.2503,
+      "step": 1977
+    },
+    {
+      "epoch": 0.01978,
+      "grad_norm": 0.95585271722792,
+      "learning_rate": 0.003,
+      "loss": 4.2456,
+      "step": 1978
+    },
+    {
+      "epoch": 0.01979,
+      "grad_norm": 0.8967358768483624,
+      "learning_rate": 0.003,
+      "loss": 4.2367,
+      "step": 1979
+    },
+    {
+      "epoch": 0.0198,
+      "grad_norm": 0.9098539062714608,
+      "learning_rate": 0.003,
+      "loss": 4.2135,
+      "step": 1980
+    },
+    {
+      "epoch": 0.01981,
+      "grad_norm": 0.9619207371647327,
+      "learning_rate": 0.003,
+      "loss": 4.2036,
+      "step": 1981
+    },
+    {
+      "epoch": 0.01982,
+      "grad_norm": 1.004618203421798,
+      "learning_rate": 0.003,
+      "loss": 4.2311,
+      "step": 1982
+    },
+    {
+      "epoch": 0.01983,
+      "grad_norm": 0.8778331022367669,
+      "learning_rate": 0.003,
+      "loss": 4.2203,
+      "step": 1983
+    },
+    {
+      "epoch": 0.01984,
+      "grad_norm": 0.7409261134848839,
+      "learning_rate": 0.003,
+      "loss": 4.2234,
+      "step": 1984
+    },
+    {
+      "epoch": 0.01985,
+      "grad_norm": 0.7987127824711159,
+      "learning_rate": 0.003,
+      "loss": 4.2126,
+      "step": 1985
+    },
+    {
+      "epoch": 0.01986,
+      "grad_norm": 0.8488590004969258,
+      "learning_rate": 0.003,
+      "loss": 4.2379,
+      "step": 1986
+    },
+    {
+      "epoch": 0.01987,
+      "grad_norm": 0.856993962490694,
+      "learning_rate": 0.003,
+      "loss": 4.1987,
+      "step": 1987
+    },
+    {
+      "epoch": 0.01988,
+      "grad_norm": 0.6814479531298734,
+      "learning_rate": 0.003,
+      "loss": 4.2021,
+      "step": 1988
+    },
+    {
+      "epoch": 0.01989,
+      "grad_norm": 0.5447842315255877,
+      "learning_rate": 0.003,
+      "loss": 4.1871,
+      "step": 1989
+    },
+    {
+      "epoch": 0.0199,
+      "grad_norm": 0.5760022915320738,
+      "learning_rate": 0.003,
+      "loss": 4.1964,
+      "step": 1990
+    },
+    {
+      "epoch": 0.01991,
+      "grad_norm": 0.5132919071482156,
+      "learning_rate": 0.003,
+      "loss": 4.1738,
+      "step": 1991
+    },
+    {
+      "epoch": 0.01992,
+      "grad_norm": 0.5851245859941666,
+      "learning_rate": 0.003,
+      "loss": 4.2128,
+      "step": 1992
+    },
+    {
+      "epoch": 0.01993,
+      "grad_norm": 0.6267354593691943,
+      "learning_rate": 0.003,
+      "loss": 4.1956,
+      "step": 1993
+    },
+    {
+      "epoch": 0.01994,
+      "grad_norm": 0.7099792076486661,
+      "learning_rate": 0.003,
+      "loss": 4.2188,
+      "step": 1994
+    },
+    {
+      "epoch": 0.01995,
+      "grad_norm": 0.7678935327418717,
+      "learning_rate": 0.003,
+      "loss": 4.2111,
+      "step": 1995
+    },
+    {
+      "epoch": 0.01996,
+      "grad_norm": 0.6394486685805509,
+      "learning_rate": 0.003,
+      "loss": 4.1755,
+      "step": 1996
+    },
+    {
+      "epoch": 0.01997,
+      "grad_norm": 0.4226153497592854,
+      "learning_rate": 0.003,
+      "loss": 4.1972,
+      "step": 1997
+    },
+    {
+      "epoch": 0.01998,
+      "grad_norm": 0.5031225205891672,
+      "learning_rate": 0.003,
+      "loss": 4.1844,
+      "step": 1998
+    },
+    {
+      "epoch": 0.01999,
+      "grad_norm": 0.5395430302746123,
+      "learning_rate": 0.003,
+      "loss": 4.1747,
+      "step": 1999
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.46234156015204186,
+      "learning_rate": 0.003,
+      "loss": 4.1586,
+      "step": 2000
+    },
+    {
+      "epoch": 0.02001,
+      "grad_norm": 0.4395345274961725,
+      "learning_rate": 0.003,
+      "loss": 4.1934,
+      "step": 2001
+    },
+    {
+      "epoch": 0.02002,
+      "grad_norm": 0.3957661324841875,
+      "learning_rate": 0.003,
+      "loss": 4.1863,
+      "step": 2002
+    },
+    {
+      "epoch": 0.02003,
+      "grad_norm": 0.4711152480437331,
+      "learning_rate": 0.003,
+      "loss": 4.1873,
+      "step": 2003
+    },
+    {
+      "epoch": 0.02004,
+      "grad_norm": 0.546136823152871,
+      "learning_rate": 0.003,
+      "loss": 4.1727,
+      "step": 2004
+    },
+    {
+      "epoch": 0.02005,
+      "grad_norm": 0.654074266015456,
+      "learning_rate": 0.003,
+      "loss": 4.1803,
+      "step": 2005
+    },
+    {
+      "epoch": 0.02006,
+      "grad_norm": 0.710370558494441,
+      "learning_rate": 0.003,
+      "loss": 4.1784,
+      "step": 2006
+    },
+    {
+      "epoch": 0.02007,
+      "grad_norm": 0.7542070025547006,
+      "learning_rate": 0.003,
+      "loss": 4.1945,
+      "step": 2007
+    },
+    {
+      "epoch": 0.02008,
+      "grad_norm": 0.7773827468349818,
+      "learning_rate": 0.003,
+      "loss": 4.191,
+      "step": 2008
+    },
+    {
+      "epoch": 0.02009,
+      "grad_norm": 0.6899352222366664,
+      "learning_rate": 0.003,
+      "loss": 4.1857,
+      "step": 2009
+    },
+    {
+      "epoch": 0.0201,
+      "grad_norm": 0.5614651061397582,
+      "learning_rate": 0.003,
+      "loss": 4.1935,
+      "step": 2010
+    },
+    {
+      "epoch": 0.02011,
+      "grad_norm": 0.6625002839338998,
+      "learning_rate": 0.003,
+      "loss": 4.183,
+      "step": 2011
+    },
+    {
+      "epoch": 0.02012,
+      "grad_norm": 0.6997403670690155,
+      "learning_rate": 0.003,
+      "loss": 4.1822,
+      "step": 2012
+    },
+    {
+      "epoch": 0.02013,
+      "grad_norm": 0.6697515151480535,
+      "learning_rate": 0.003,
+      "loss": 4.1896,
+      "step": 2013
+    },
+    {
+      "epoch": 0.02014,
+      "grad_norm": 0.6386903403073065,
+      "learning_rate": 0.003,
+      "loss": 4.1729,
+      "step": 2014
+    },
+    {
+      "epoch": 0.02015,
+      "grad_norm": 0.5907166723123874,
+      "learning_rate": 0.003,
+      "loss": 4.1657,
+      "step": 2015
+    },
+    {
+      "epoch": 0.02016,
+      "grad_norm": 0.605106268984101,
+      "learning_rate": 0.003,
+      "loss": 4.1769,
+      "step": 2016
+    },
+    {
+      "epoch": 0.02017,
+      "grad_norm": 0.6631806062284518,
+      "learning_rate": 0.003,
+      "loss": 4.1711,
+      "step": 2017
+    },
+    {
+      "epoch": 0.02018,
+      "grad_norm": 0.6347297750042842,
+      "learning_rate": 0.003,
+      "loss": 4.1746,
+      "step": 2018
+    },
+    {
+      "epoch": 0.02019,
+      "grad_norm": 0.8248022848153107,
+      "learning_rate": 0.003,
+      "loss": 4.1834,
+      "step": 2019
+    },
+    {
+      "epoch": 0.0202,
+      "grad_norm": 0.9904970732072179,
+      "learning_rate": 0.003,
+      "loss": 4.2031,
+      "step": 2020
+    },
+    {
+      "epoch": 0.02021,
+      "grad_norm": 0.9086090012812373,
+      "learning_rate": 0.003,
+      "loss": 4.2148,
+      "step": 2021
+    },
+    {
+      "epoch": 0.02022,
+      "grad_norm": 0.9232445860299345,
+      "learning_rate": 0.003,
+      "loss": 4.175,
+      "step": 2022
+    },
+    {
+      "epoch": 0.02023,
+      "grad_norm": 0.8050131783240017,
+      "learning_rate": 0.003,
+      "loss": 4.1493,
+      "step": 2023
+    },
+    {
+      "epoch": 0.02024,
+      "grad_norm": 0.7442373542055678,
+      "learning_rate": 0.003,
+      "loss": 4.1717,
+      "step": 2024
+    },
+    {
+      "epoch": 0.02025,
+      "grad_norm": 0.9832310341451624,
+      "learning_rate": 0.003,
+      "loss": 4.2324,
+      "step": 2025
+    },
+    {
+      "epoch": 0.02026,
+      "grad_norm": 0.9556185585548506,
+      "learning_rate": 0.003,
+      "loss": 4.2147,
+      "step": 2026
+    },
+    {
+      "epoch": 0.02027,
+      "grad_norm": 1.0633653823279028,
+      "learning_rate": 0.003,
+      "loss": 4.218,
+      "step": 2027
+    },
+    {
+      "epoch": 0.02028,
+      "grad_norm": 0.8361708667618777,
+      "learning_rate": 0.003,
+      "loss": 4.1941,
+      "step": 2028
+    },
+    {
+      "epoch": 0.02029,
+      "grad_norm": 0.8621114948206428,
+      "learning_rate": 0.003,
+      "loss": 4.1995,
+      "step": 2029
+    },
+    {
+      "epoch": 0.0203,
+      "grad_norm": 0.7505639644342775,
+      "learning_rate": 0.003,
+      "loss": 4.1873,
+      "step": 2030
+    },
+    {
+      "epoch": 0.02031,
+      "grad_norm": 0.7688539298918926,
+      "learning_rate": 0.003,
+      "loss": 4.1919,
+      "step": 2031
+    },
+    {
+      "epoch": 0.02032,
+      "grad_norm": 0.8008832349814213,
+      "learning_rate": 0.003,
+      "loss": 4.1682,
+      "step": 2032
+    },
+    {
+      "epoch": 0.02033,
+      "grad_norm": 0.7744220160391694,
+      "learning_rate": 0.003,
+      "loss": 4.2069,
+      "step": 2033
+    },
+    {
+      "epoch": 0.02034,
+      "grad_norm": 0.9334621419858093,
+      "learning_rate": 0.003,
+      "loss": 4.2063,
+      "step": 2034
+    },
+    {
+      "epoch": 0.02035,
+      "grad_norm": 1.0003714475706196,
+      "learning_rate": 0.003,
+      "loss": 4.1981,
+      "step": 2035
+    },
+    {
+      "epoch": 0.02036,
+      "grad_norm": 1.0165755074456069,
+      "learning_rate": 0.003,
+      "loss": 4.2039,
+      "step": 2036
+    },
+    {
+      "epoch": 0.02037,
+      "grad_norm": 0.8752656941969562,
+      "learning_rate": 0.003,
+      "loss": 4.2133,
+      "step": 2037
+    },
+    {
+      "epoch": 0.02038,
+      "grad_norm": 0.8030828089902575,
+      "learning_rate": 0.003,
+      "loss": 4.1951,
+      "step": 2038
+    },
+    {
+      "epoch": 0.02039,
+      "grad_norm": 0.6849651287466818,
+      "learning_rate": 0.003,
+      "loss": 4.1984,
+      "step": 2039
+    },
+    {
+      "epoch": 0.0204,
+      "grad_norm": 0.7048992687138306,
+      "learning_rate": 0.003,
+      "loss": 4.2015,
+      "step": 2040
+    },
+    {
+      "epoch": 0.02041,
+      "grad_norm": 0.5846622816322785,
+      "learning_rate": 0.003,
+      "loss": 4.1601,
+      "step": 2041
+    },
+    {
+      "epoch": 0.02042,
+      "grad_norm": 0.579059647971908,
+      "learning_rate": 0.003,
+      "loss": 4.159,
+      "step": 2042
+    },
+    {
+      "epoch": 0.02043,
+      "grad_norm": 0.5320216814522557,
+      "learning_rate": 0.003,
+      "loss": 4.2018,
+      "step": 2043
+    },
+    {
+      "epoch": 0.02044,
+      "grad_norm": 0.5749364817997652,
+      "learning_rate": 0.003,
+      "loss": 4.174,
+      "step": 2044
+    },
+    {
+      "epoch": 0.02045,
+      "grad_norm": 0.5872195224989512,
+      "learning_rate": 0.003,
+      "loss": 4.1881,
+      "step": 2045
+    },
+    {
+      "epoch": 0.02046,
+      "grad_norm": 0.6166345493647961,
+      "learning_rate": 0.003,
+      "loss": 4.1634,
+      "step": 2046
+    },
+    {
+      "epoch": 0.02047,
+      "grad_norm": 0.7223919170905914,
+      "learning_rate": 0.003,
+      "loss": 4.2037,
+      "step": 2047
+    },
+    {
+      "epoch": 0.02048,
+      "grad_norm": 0.7799560828695167,
+      "learning_rate": 0.003,
+      "loss": 4.1877,
+      "step": 2048
+    },
+    {
+      "epoch": 0.02049,
+      "grad_norm": 0.8702209657581298,
+      "learning_rate": 0.003,
+      "loss": 4.2131,
+      "step": 2049
+    },
+    {
+      "epoch": 0.0205,
+      "grad_norm": 0.9375771238974213,
+      "learning_rate": 0.003,
+      "loss": 4.1631,
+      "step": 2050
+    },
+    {
+      "epoch": 0.02051,
+      "grad_norm": 1.1873073578231947,
+      "learning_rate": 0.003,
+      "loss": 4.178,
+      "step": 2051
+    },
+    {
+      "epoch": 0.02052,
+      "grad_norm": 1.03209456773,
+      "learning_rate": 0.003,
+      "loss": 4.1777,
+      "step": 2052
+    },
+    {
+      "epoch": 0.02053,
+      "grad_norm": 1.0555848829046368,
+      "learning_rate": 0.003,
+      "loss": 4.1832,
+      "step": 2053
+    },
+    {
+      "epoch": 0.02054,
+      "grad_norm": 0.9050417954028227,
+      "learning_rate": 0.003,
+      "loss": 4.2168,
+      "step": 2054
+    },
+    {
+      "epoch": 0.02055,
+      "grad_norm": 0.8496945287219897,
+      "learning_rate": 0.003,
+      "loss": 4.1765,
+      "step": 2055
+    },
+    {
+      "epoch": 0.02056,
+      "grad_norm": 0.7540734931162184,
+      "learning_rate": 0.003,
+      "loss": 4.17,
+      "step": 2056
+    },
+    {
+      "epoch": 0.02057,
+      "grad_norm": 0.79729190390271,
+      "learning_rate": 0.003,
+      "loss": 4.2076,
+      "step": 2057
+    },
+    {
+      "epoch": 0.02058,
+      "grad_norm": 0.8092766990194165,
+      "learning_rate": 0.003,
+      "loss": 4.1879,
+      "step": 2058
+    },
+    {
+      "epoch": 0.02059,
+      "grad_norm": 0.787520651985971,
+      "learning_rate": 0.003,
+      "loss": 4.1662,
+      "step": 2059
+    },
+    {
+      "epoch": 0.0206,
+      "grad_norm": 0.8179849805821454,
+      "learning_rate": 0.003,
+      "loss": 4.2123,
+      "step": 2060
+    },
+    {
+      "epoch": 0.02061,
+      "grad_norm": 0.765277725532741,
+      "learning_rate": 0.003,
+      "loss": 4.1867,
+      "step": 2061
+    },
+    {
+      "epoch": 0.02062,
+      "grad_norm": 0.6833167046153735,
+      "learning_rate": 0.003,
+      "loss": 4.1822,
+      "step": 2062
+    },
+    {
+      "epoch": 0.02063,
+      "grad_norm": 0.7852297427280791,
+      "learning_rate": 0.003,
+      "loss": 4.1788,
+      "step": 2063
+    },
+    {
+      "epoch": 0.02064,
+      "grad_norm": 0.7558602646875056,
+      "learning_rate": 0.003,
+      "loss": 4.1894,
+      "step": 2064
+    },
+    {
+      "epoch": 0.02065,
+      "grad_norm": 0.7461513201926369,
+      "learning_rate": 0.003,
+      "loss": 4.1686,
+      "step": 2065
+    },
+    {
+      "epoch": 0.02066,
+      "grad_norm": 0.8279588238364882,
+      "learning_rate": 0.003,
+      "loss": 4.1898,
+      "step": 2066
+    },
+    {
+      "epoch": 0.02067,
+      "grad_norm": 0.7670898564018478,
+      "learning_rate": 0.003,
+      "loss": 4.1869,
+      "step": 2067
+    },
+    {
+      "epoch": 0.02068,
+      "grad_norm": 0.6723403011847539,
+      "learning_rate": 0.003,
+      "loss": 4.1662,
+      "step": 2068
+    },
+    {
+      "epoch": 0.02069,
+      "grad_norm": 0.6336511314262159,
+      "learning_rate": 0.003,
+      "loss": 4.1849,
+      "step": 2069
+    },
+    {
+      "epoch": 0.0207,
+      "grad_norm": 0.70657998006886,
+      "learning_rate": 0.003,
+      "loss": 4.1607,
+      "step": 2070
+    },
+    {
+      "epoch": 0.02071,
+      "grad_norm": 0.7487170820279913,
+      "learning_rate": 0.003,
+      "loss": 4.1912,
+      "step": 2071
+    },
+    {
+      "epoch": 0.02072,
+      "grad_norm": 0.7172563865038417,
+      "learning_rate": 0.003,
+      "loss": 4.1982,
+      "step": 2072
+    },
+    {
+      "epoch": 0.02073,
+      "grad_norm": 0.7815042091167561,
+      "learning_rate": 0.003,
+      "loss": 4.1796,
+      "step": 2073
+    },
+    {
+      "epoch": 0.02074,
+      "grad_norm": 0.7869404752550972,
+      "learning_rate": 0.003,
+      "loss": 4.1864,
+      "step": 2074
+    },
+    {
+      "epoch": 0.02075,
+      "grad_norm": 0.7115380813621578,
+      "learning_rate": 0.003,
+      "loss": 4.2091,
+      "step": 2075
+    },
+    {
+      "epoch": 0.02076,
+      "grad_norm": 0.6271616020332964,
+      "learning_rate": 0.003,
+      "loss": 4.199,
+      "step": 2076
+    },
+    {
+      "epoch": 0.02077,
+      "grad_norm": 0.5945516898335454,
+      "learning_rate": 0.003,
+      "loss": 4.1669,
+      "step": 2077
+    },
+    {
+      "epoch": 0.02078,
+      "grad_norm": 0.5807187113226505,
+      "learning_rate": 0.003,
+      "loss": 4.1615,
+      "step": 2078
+    },
+    {
+      "epoch": 0.02079,
+      "grad_norm": 0.4917157616397027,
+      "learning_rate": 0.003,
+      "loss": 4.1449,
+      "step": 2079
+    },
+    {
+      "epoch": 0.0208,
+      "grad_norm": 0.5388783899386764,
+      "learning_rate": 0.003,
+      "loss": 4.1563,
+      "step": 2080
+    },
+    {
+      "epoch": 0.02081,
+      "grad_norm": 0.6250429891561764,
+      "learning_rate": 0.003,
+      "loss": 4.1899,
+      "step": 2081
+    },
+    {
+      "epoch": 0.02082,
+      "grad_norm": 0.7539524872161132,
+      "learning_rate": 0.003,
+      "loss": 4.1879,
+      "step": 2082
+    },
+    {
+      "epoch": 0.02083,
+      "grad_norm": 0.9021795929932535,
+      "learning_rate": 0.003,
+      "loss": 4.189,
+      "step": 2083
+    },
+    {
+      "epoch": 0.02084,
+      "grad_norm": 0.9892236370325144,
+      "learning_rate": 0.003,
+      "loss": 4.1914,
+      "step": 2084
+    },
+    {
+      "epoch": 0.02085,
+      "grad_norm": 0.9279291435694554,
+      "learning_rate": 0.003,
+      "loss": 4.1653,
+      "step": 2085
+    },
+    {
+      "epoch": 0.02086,
+      "grad_norm": 0.7919688525399468,
+      "learning_rate": 0.003,
+      "loss": 4.1726,
+      "step": 2086
+    },
+    {
+      "epoch": 0.02087,
+      "grad_norm": 0.7965117787009012,
+      "learning_rate": 0.003,
+      "loss": 4.1719,
+      "step": 2087
+    },
+    {
+      "epoch": 0.02088,
+      "grad_norm": 0.882456029692519,
+      "learning_rate": 0.003,
+      "loss": 4.2097,
+      "step": 2088
+    },
+    {
+      "epoch": 0.02089,
+      "grad_norm": 0.8843676371103549,
+      "learning_rate": 0.003,
+      "loss": 4.1781,
+      "step": 2089
+    },
+    {
+      "epoch": 0.0209,
+      "grad_norm": 0.9095817089780983,
+      "learning_rate": 0.003,
+      "loss": 4.2024,
+      "step": 2090
+    },
+    {
+      "epoch": 0.02091,
+      "grad_norm": 0.8360707980502976,
+      "learning_rate": 0.003,
+      "loss": 4.1884,
+      "step": 2091
+    },
+    {
+      "epoch": 0.02092,
+      "grad_norm": 0.8954851906550106,
+      "learning_rate": 0.003,
+      "loss": 4.189,
+      "step": 2092
+    },
+    {
+      "epoch": 0.02093,
+      "grad_norm": 0.8146056324857043,
+      "learning_rate": 0.003,
+      "loss": 4.1544,
+      "step": 2093
+    },
+    {
+      "epoch": 0.02094,
+      "grad_norm": 0.7558490596404085,
+      "learning_rate": 0.003,
+      "loss": 4.1629,
+      "step": 2094
+    },
+    {
+      "epoch": 0.02095,
+      "grad_norm": 0.7384447141579936,
+      "learning_rate": 0.003,
+      "loss": 4.2013,
+      "step": 2095
+    },
+    {
+      "epoch": 0.02096,
+      "grad_norm": 0.7339651428587286,
+      "learning_rate": 0.003,
+      "loss": 4.1917,
+      "step": 2096
+    },
+    {
+      "epoch": 0.02097,
+      "grad_norm": 0.6671356121778469,
+      "learning_rate": 0.003,
+      "loss": 4.1698,
+      "step": 2097
+    },
+    {
+      "epoch": 0.02098,
+      "grad_norm": 0.653522008057512,
+      "learning_rate": 0.003,
+      "loss": 4.1857,
+      "step": 2098
+    },
+    {
+      "epoch": 0.02099,
+      "grad_norm": 0.6717296847888601,
+      "learning_rate": 0.003,
+      "loss": 4.1945,
+      "step": 2099
+    },
+    {
+      "epoch": 0.021,
+      "grad_norm": 0.6641683049026381,
+      "learning_rate": 0.003,
+      "loss": 4.206,
+      "step": 2100
+    },
+    {
+      "epoch": 0.02101,
+      "grad_norm": 0.8263568144502693,
+      "learning_rate": 0.003,
+      "loss": 4.2014,
+      "step": 2101
+    },
+    {
+      "epoch": 0.02102,
+      "grad_norm": 0.9182856357229072,
+      "learning_rate": 0.003,
+      "loss": 4.2116,
+      "step": 2102
+    },
+    {
+      "epoch": 0.02103,
+      "grad_norm": 0.8775651409452159,
+      "learning_rate": 0.003,
+      "loss": 4.1702,
+      "step": 2103
+    },
+    {
+      "epoch": 0.02104,
+      "grad_norm": 0.717248452996101,
+      "learning_rate": 0.003,
+      "loss": 4.1698,
+      "step": 2104
+    },
+    {
+      "epoch": 0.02105,
+      "grad_norm": 0.6147629172749789,
+      "learning_rate": 0.003,
+      "loss": 4.1709,
+      "step": 2105
+    },
+    {
+      "epoch": 0.02106,
+      "grad_norm": 0.5350819274969562,
+      "learning_rate": 0.003,
+      "loss": 4.1599,
+      "step": 2106
+    },
+    {
+      "epoch": 0.02107,
+      "grad_norm": 0.6022860581700974,
+      "learning_rate": 0.003,
+      "loss": 4.1732,
+      "step": 2107
+    },
+    {
+      "epoch": 0.02108,
+      "grad_norm": 0.6033513423948526,
+      "learning_rate": 0.003,
+      "loss": 4.1543,
+      "step": 2108
+    },
+    {
+      "epoch": 0.02109,
+      "grad_norm": 0.6042225440605065,
+      "learning_rate": 0.003,
+      "loss": 4.144,
+      "step": 2109
+    },
+    {
+      "epoch": 0.0211,
+      "grad_norm": 0.560183095906506,
+      "learning_rate": 0.003,
+      "loss": 4.1584,
+      "step": 2110
+    },
+    {
+      "epoch": 0.02111,
+      "grad_norm": 0.5249161566388668,
+      "learning_rate": 0.003,
+      "loss": 4.1663,
+      "step": 2111
+    },
+    {
+      "epoch": 0.02112,
+      "grad_norm": 0.5391312057228763,
+      "learning_rate": 0.003,
+      "loss": 4.1641,
+      "step": 2112
+    },
+    {
+      "epoch": 0.02113,
+      "grad_norm": 0.5260626981895078,
+      "learning_rate": 0.003,
+      "loss": 4.1597,
+      "step": 2113
+    },
+    {
+      "epoch": 0.02114,
+      "grad_norm": 0.5072590511188936,
+      "learning_rate": 0.003,
+      "loss": 4.1483,
+      "step": 2114
+    },
+    {
+      "epoch": 0.02115,
+      "grad_norm": 0.5460141991570172,
+      "learning_rate": 0.003,
+      "loss": 4.1417,
+      "step": 2115
+    },
+    {
+      "epoch": 0.02116,
+      "grad_norm": 0.5749891422597724,
+      "learning_rate": 0.003,
+      "loss": 4.1705,
+      "step": 2116
+    },
+    {
+      "epoch": 0.02117,
+      "grad_norm": 0.6781459947006083,
+      "learning_rate": 0.003,
+      "loss": 4.1668,
+      "step": 2117
+    },
+    {
+      "epoch": 0.02118,
+      "grad_norm": 0.797658364148174,
+      "learning_rate": 0.003,
+      "loss": 4.1692,
+      "step": 2118
+    },
+    {
+      "epoch": 0.02119,
+      "grad_norm": 0.8688435880924688,
+      "learning_rate": 0.003,
+      "loss": 4.1833,
+      "step": 2119
+    },
+    {
+      "epoch": 0.0212,
+      "grad_norm": 0.7833741674522176,
+      "learning_rate": 0.003,
+      "loss": 4.1545,
+      "step": 2120
+    },
+    {
+      "epoch": 0.02121,
+      "grad_norm": 0.791532521338522,
+      "learning_rate": 0.003,
+      "loss": 4.1508,
+      "step": 2121
+    },
+    {
+      "epoch": 0.02122,
+      "grad_norm": 0.7913660839492322,
+      "learning_rate": 0.003,
+      "loss": 4.1614,
+      "step": 2122
+    },
+    {
+      "epoch": 0.02123,
+      "grad_norm": 0.7506478823984966,
+      "learning_rate": 0.003,
+      "loss": 4.1402,
+      "step": 2123
+    },
+    {
+      "epoch": 0.02124,
+      "grad_norm": 0.7580020078198468,
+      "learning_rate": 0.003,
+      "loss": 4.1643,
+      "step": 2124
+    },
+    {
+      "epoch": 0.02125,
+      "grad_norm": 0.7006901754929752,
+      "learning_rate": 0.003,
+      "loss": 4.1792,
+      "step": 2125
+    },
+    {
+      "epoch": 0.02126,
+      "grad_norm": 0.7653616945644249,
+      "learning_rate": 0.003,
+      "loss": 4.1781,
+      "step": 2126
+    },
+    {
+      "epoch": 0.02127,
+      "grad_norm": 0.7492928706532664,
+      "learning_rate": 0.003,
+      "loss": 4.1859,
+      "step": 2127
+    },
+    {
+      "epoch": 0.02128,
+      "grad_norm": 0.7795854700268753,
+      "learning_rate": 0.003,
+      "loss": 4.1758,
+      "step": 2128
+    },
+    {
+      "epoch": 0.02129,
+      "grad_norm": 0.9805419044523191,
+      "learning_rate": 0.003,
+      "loss": 4.186,
+      "step": 2129
+    },
+    {
+      "epoch": 0.0213,
+      "grad_norm": 1.0343575534498912,
+      "learning_rate": 0.003,
+      "loss": 4.1641,
+      "step": 2130
+    },
+    {
+      "epoch": 0.02131,
+      "grad_norm": 0.9468793830055084,
+      "learning_rate": 0.003,
+      "loss": 4.1869,
+      "step": 2131
+    },
+    {
+      "epoch": 0.02132,
+      "grad_norm": 0.9021420581140779,
+      "learning_rate": 0.003,
+      "loss": 4.2032,
+      "step": 2132
+    },
+    {
+      "epoch": 0.02133,
+      "grad_norm": 0.8122842455378513,
+      "learning_rate": 0.003,
+      "loss": 4.1908,
+      "step": 2133
+    },
+    {
+      "epoch": 0.02134,
+      "grad_norm": 1.06138323809129,
+      "learning_rate": 0.003,
+      "loss": 4.2048,
+      "step": 2134
+    },
+    {
+      "epoch": 0.02135,
+      "grad_norm": 1.2270031550495697,
+      "learning_rate": 0.003,
+      "loss": 4.212,
+      "step": 2135
+    },
+    {
+      "epoch": 0.02136,
+      "grad_norm": 0.824636655199696,
+      "learning_rate": 0.003,
+      "loss": 4.1714,
+      "step": 2136
+    },
+    {
+      "epoch": 0.02137,
+      "grad_norm": 0.7977636532107003,
+      "learning_rate": 0.003,
+      "loss": 4.2074,
+      "step": 2137
+    },
+    {
+      "epoch": 0.02138,
+      "grad_norm": 0.8175212273276642,
+      "learning_rate": 0.003,
+      "loss": 4.2175,
+      "step": 2138
+    },
+    {
+      "epoch": 0.02139,
+      "grad_norm": 0.8041667304441814,
+      "learning_rate": 0.003,
+      "loss": 4.2081,
+      "step": 2139
+    },
+    {
+      "epoch": 0.0214,
+      "grad_norm": 0.9410719295719912,
+      "learning_rate": 0.003,
+      "loss": 4.2078,
+      "step": 2140
+    },
+    {
+      "epoch": 0.02141,
+      "grad_norm": 0.9128066534114356,
+      "learning_rate": 0.003,
+      "loss": 4.1891,
+      "step": 2141
+    },
+    {
+      "epoch": 0.02142,
+      "grad_norm": 0.7166876786866843,
+      "learning_rate": 0.003,
+      "loss": 4.1743,
+      "step": 2142
+    },
+    {
+      "epoch": 0.02143,
+      "grad_norm": 0.7553648404993994,
+      "learning_rate": 0.003,
+      "loss": 4.1879,
+      "step": 2143
+    },
+    {
+      "epoch": 0.02144,
+      "grad_norm": 0.9693569859499341,
+      "learning_rate": 0.003,
+      "loss": 4.1979,
+      "step": 2144
+    },
+    {
+      "epoch": 0.02145,
+      "grad_norm": 1.005405043081365,
+      "learning_rate": 0.003,
+      "loss": 4.1853,
+      "step": 2145
+    },
+    {
+      "epoch": 0.02146,
+      "grad_norm": 0.9174606002788802,
+      "learning_rate": 0.003,
+      "loss": 4.1967,
+      "step": 2146
+    },
+    {
+      "epoch": 0.02147,
+      "grad_norm": 0.7660679608210765,
+      "learning_rate": 0.003,
+      "loss": 4.2109,
+      "step": 2147
+    },
+    {
+      "epoch": 0.02148,
+      "grad_norm": 0.7067244008901828,
+      "learning_rate": 0.003,
+      "loss": 4.222,
+      "step": 2148
+    },
+    {
+      "epoch": 0.02149,
+      "grad_norm": 0.5597922211015285,
+      "learning_rate": 0.003,
+      "loss": 4.1982,
+      "step": 2149
+    },
+    {
+      "epoch": 0.0215,
+      "grad_norm": 0.5887191266363829,
+      "learning_rate": 0.003,
+      "loss": 4.1847,
+      "step": 2150
+    },
+    {
+      "epoch": 0.02151,
+      "grad_norm": 0.5623719175027401,
+      "learning_rate": 0.003,
+      "loss": 4.1889,
+      "step": 2151
+    },
+    {
+      "epoch": 0.02152,
+      "grad_norm": 0.5432643275971011,
+      "learning_rate": 0.003,
+      "loss": 4.1639,
+      "step": 2152
+    },
+    {
+      "epoch": 0.02153,
+      "grad_norm": 0.5447875969469528,
+      "learning_rate": 0.003,
+      "loss": 4.1717,
+      "step": 2153
+    },
+    {
+      "epoch": 0.02154,
+      "grad_norm": 0.64393428597859,
+      "learning_rate": 0.003,
+      "loss": 4.1635,
+      "step": 2154
+    },
+    {
+      "epoch": 0.02155,
+      "grad_norm": 0.836540801508149,
+      "learning_rate": 0.003,
+      "loss": 4.1972,
+      "step": 2155
+    },
+    {
+      "epoch": 0.02156,
+      "grad_norm": 0.9453076222627308,
+      "learning_rate": 0.003,
+      "loss": 4.1938,
+      "step": 2156
+    },
+    {
+      "epoch": 0.02157,
+      "grad_norm": 0.7417036890833684,
+      "learning_rate": 0.003,
+      "loss": 4.1515,
+      "step": 2157
+    },
+    {
+      "epoch": 0.02158,
+      "grad_norm": 0.5957713091881504,
+      "learning_rate": 0.003,
+      "loss": 4.1686,
+      "step": 2158
+    },
+    {
+      "epoch": 0.02159,
+      "grad_norm": 0.6018851123999939,
+      "learning_rate": 0.003,
+      "loss": 4.1807,
+      "step": 2159
+    },
+    {
+      "epoch": 0.0216,
+      "grad_norm": 0.6716376115123553,
+      "learning_rate": 0.003,
+      "loss": 4.1831,
+      "step": 2160
+    },
+    {
+      "epoch": 0.02161,
+      "grad_norm": 0.7819649346338547,
+      "learning_rate": 0.003,
+      "loss": 4.1755,
+      "step": 2161
+    },
+    {
+      "epoch": 0.02162,
+      "grad_norm": 0.8233828232110464,
+      "learning_rate": 0.003,
+      "loss": 4.1801,
+      "step": 2162
+    },
+    {
+      "epoch": 0.02163,
+      "grad_norm": 0.7824568090505709,
+      "learning_rate": 0.003,
+      "loss": 4.1982,
+      "step": 2163
+    },
+    {
+      "epoch": 0.02164,
+      "grad_norm": 0.8597315749146158,
+      "learning_rate": 0.003,
+      "loss": 4.1915,
+      "step": 2164
+    },
+    {
+      "epoch": 0.02165,
+      "grad_norm": 0.7979744235864884,
+      "learning_rate": 0.003,
+      "loss": 4.1562,
+      "step": 2165
+    },
+    {
+      "epoch": 0.02166,
+      "grad_norm": 0.7759936363363831,
+      "learning_rate": 0.003,
+      "loss": 4.1947,
+      "step": 2166
+    },
+    {
+      "epoch": 0.02167,
+      "grad_norm": 0.7097277369414112,
+      "learning_rate": 0.003,
+      "loss": 4.1708,
+      "step": 2167
+    },
+    {
+      "epoch": 0.02168,
+      "grad_norm": 0.6129275664005919,
+      "learning_rate": 0.003,
+      "loss": 4.1674,
+      "step": 2168
+    },
+    {
+      "epoch": 0.02169,
+      "grad_norm": 0.5617202579327312,
+      "learning_rate": 0.003,
+      "loss": 4.1636,
+      "step": 2169
+    },
+    {
+      "epoch": 0.0217,
+      "grad_norm": 0.4729981033232719,
+      "learning_rate": 0.003,
+      "loss": 4.1693,
+      "step": 2170
+    },
+    {
+      "epoch": 0.02171,
+      "grad_norm": 0.5127457916731437,
+      "learning_rate": 0.003,
+      "loss": 4.1819,
+      "step": 2171
+    },
+    {
+      "epoch": 0.02172,
+      "grad_norm": 0.6304303564787295,
+      "learning_rate": 0.003,
+      "loss": 4.1556,
+      "step": 2172
+    },
+    {
+      "epoch": 0.02173,
+      "grad_norm": 0.7235970129567234,
+      "learning_rate": 0.003,
+      "loss": 4.1636,
+      "step": 2173
+    },
+    {
+      "epoch": 0.02174,
+      "grad_norm": 0.9537416454234164,
+      "learning_rate": 0.003,
+      "loss": 4.1668,
+      "step": 2174
+    },
+    {
+      "epoch": 0.02175,
+      "grad_norm": 1.0772501316127283,
+      "learning_rate": 0.003,
+      "loss": 4.2082,
+      "step": 2175
+    },
+    {
+      "epoch": 0.02176,
+      "grad_norm": 0.9320344556958656,
+      "learning_rate": 0.003,
+      "loss": 4.1798,
+      "step": 2176
+    },
+    {
+      "epoch": 0.02177,
+      "grad_norm": 1.2306562612139564,
+      "learning_rate": 0.003,
+      "loss": 4.2022,
+      "step": 2177
+    },
+    {
+      "epoch": 0.02178,
+      "grad_norm": 0.9649462703992993,
+      "learning_rate": 0.003,
+      "loss": 4.1822,
+      "step": 2178
+    },
+    {
+      "epoch": 0.02179,
+      "grad_norm": 0.8364440339114567,
+      "learning_rate": 0.003,
+      "loss": 4.1982,
+      "step": 2179
+    },
+    {
+      "epoch": 0.0218,
+      "grad_norm": 0.8494241782855642,
+      "learning_rate": 0.003,
+      "loss": 4.2094,
+      "step": 2180
+    },
+    {
+      "epoch": 0.02181,
+      "grad_norm": 0.8028341199053393,
+      "learning_rate": 0.003,
+      "loss": 4.2103,
+      "step": 2181
+    },
+    {
+      "epoch": 0.02182,
+      "grad_norm": 0.8789575016731874,
+      "learning_rate": 0.003,
+      "loss": 4.1991,
+      "step": 2182
+    },
+    {
+      "epoch": 0.02183,
+      "grad_norm": 0.7339272838818488,
+      "learning_rate": 0.003,
+      "loss": 4.2018,
+      "step": 2183
+    },
+    {
+      "epoch": 0.02184,
+      "grad_norm": 0.6992146186501603,
+      "learning_rate": 0.003,
+      "loss": 4.1762,
+      "step": 2184
+    },
+    {
+      "epoch": 0.02185,
+      "grad_norm": 0.6859058633396239,
+      "learning_rate": 0.003,
+      "loss": 4.1699,
+      "step": 2185
+    },
+    {
+      "epoch": 0.02186,
+      "grad_norm": 0.671595445918346,
+      "learning_rate": 0.003,
+      "loss": 4.186,
+      "step": 2186
+    },
+    {
+      "epoch": 0.02187,
+      "grad_norm": 0.6535925166183008,
+      "learning_rate": 0.003,
+      "loss": 4.1812,
+      "step": 2187
+    },
+    {
+      "epoch": 0.02188,
+      "grad_norm": 0.6198661745230879,
+      "learning_rate": 0.003,
+      "loss": 4.1584,
+      "step": 2188
+    },
+    {
+      "epoch": 0.02189,
+      "grad_norm": 0.5776255514929863,
+      "learning_rate": 0.003,
+      "loss": 4.1869,
+      "step": 2189
+    },
+    {
+      "epoch": 0.0219,
+      "grad_norm": 0.7433015793364336,
+      "learning_rate": 0.003,
+      "loss": 4.1787,
+      "step": 2190
+    },
+    {
+      "epoch": 0.02191,
+      "grad_norm": 0.8514656528184604,
+      "learning_rate": 0.003,
+      "loss": 4.1806,
+      "step": 2191
+    },
+    {
+      "epoch": 0.02192,
+      "grad_norm": 0.9191452260055095,
+      "learning_rate": 0.003,
+      "loss": 4.2122,
+      "step": 2192
+    },
+    {
+      "epoch": 0.02193,
+      "grad_norm": 0.7405436618584794,
+      "learning_rate": 0.003,
+      "loss": 4.1726,
+      "step": 2193
+    },
+    {
+      "epoch": 0.02194,
+      "grad_norm": 0.7140573914846811,
+      "learning_rate": 0.003,
+      "loss": 4.2215,
+      "step": 2194
+    },
+    {
+      "epoch": 0.02195,
+      "grad_norm": 0.7595143608079454,
+      "learning_rate": 0.003,
+      "loss": 4.1859,
+      "step": 2195
+    },
+    {
+      "epoch": 0.02196,
+      "grad_norm": 0.7549194736868536,
+      "learning_rate": 0.003,
+      "loss": 4.1951,
+      "step": 2196
+    },
+    {
+      "epoch": 0.02197,
+      "grad_norm": 0.6975555957838411,
+      "learning_rate": 0.003,
+      "loss": 4.1831,
+      "step": 2197
+    },
+    {
+      "epoch": 0.02198,
+      "grad_norm": 0.7595090640426224,
+      "learning_rate": 0.003,
+      "loss": 4.1756,
+      "step": 2198
+    },
+    {
+      "epoch": 0.02199,
+      "grad_norm": 0.7642539057941399,
+      "learning_rate": 0.003,
+      "loss": 4.145,
+      "step": 2199
+    },
+    {
+      "epoch": 0.022,
+      "grad_norm": 0.7482736596797179,
+      "learning_rate": 0.003,
+      "loss": 4.1698,
+      "step": 2200
+    },
+    {
+      "epoch": 0.02201,
+      "grad_norm": 0.6545018698949331,
+      "learning_rate": 0.003,
+      "loss": 4.1824,
+      "step": 2201
+    },
+    {
+      "epoch": 0.02202,
+      "grad_norm": 0.6269777415335018,
+      "learning_rate": 0.003,
+      "loss": 4.1683,
+      "step": 2202
+    },
+    {
+      "epoch": 0.02203,
+      "grad_norm": 0.7202806939294419,
+      "learning_rate": 0.003,
+      "loss": 4.1965,
+      "step": 2203
+    },
+    {
+      "epoch": 0.02204,
+      "grad_norm": 0.7354393023905447,
+      "learning_rate": 0.003,
+      "loss": 4.1919,
+      "step": 2204
+    },
+    {
+      "epoch": 0.02205,
+      "grad_norm": 0.6658405305239573,
+      "learning_rate": 0.003,
+      "loss": 4.1888,
+      "step": 2205
+    },
+    {
+      "epoch": 0.02206,
+      "grad_norm": 0.5494433484311174,
+      "learning_rate": 0.003,
+      "loss": 4.1509,
+      "step": 2206
+    },
+    {
+      "epoch": 0.02207,
+      "grad_norm": 0.48191881483377974,
+      "learning_rate": 0.003,
+      "loss": 4.1617,
+      "step": 2207
+    },
+    {
+      "epoch": 0.02208,
+      "grad_norm": 0.5055031144775365,
+      "learning_rate": 0.003,
+      "loss": 4.1419,
+      "step": 2208
+    },
+    {
+      "epoch": 0.02209,
+      "grad_norm": 0.5328357421096614,
+      "learning_rate": 0.003,
+      "loss": 4.1587,
+      "step": 2209
+    },
+    {
+      "epoch": 0.0221,
+      "grad_norm": 0.580303567033314,
+      "learning_rate": 0.003,
+      "loss": 4.1426,
+      "step": 2210
+    },
+    {
+      "epoch": 0.02211,
+      "grad_norm": 0.627373370176424,
+      "learning_rate": 0.003,
+      "loss": 4.1541,
+      "step": 2211
+    },
+    {
+      "epoch": 0.02212,
+      "grad_norm": 0.8017723022310775,
+      "learning_rate": 0.003,
+      "loss": 4.1578,
+      "step": 2212
+    },
+    {
+      "epoch": 0.02213,
+      "grad_norm": 0.9755586674162295,
+      "learning_rate": 0.003,
+      "loss": 4.1762,
+      "step": 2213
+    },
+    {
+      "epoch": 0.02214,
+      "grad_norm": 0.8888070506530766,
+      "learning_rate": 0.003,
+      "loss": 4.1705,
+      "step": 2214
+    },
+    {
+      "epoch": 0.02215,
+      "grad_norm": 0.8341073466106758,
+      "learning_rate": 0.003,
+      "loss": 4.173,
+      "step": 2215
+    },
+    {
+      "epoch": 0.02216,
+      "grad_norm": 0.8721558185443227,
+      "learning_rate": 0.003,
+      "loss": 4.1552,
+      "step": 2216
+    },
+    {
+      "epoch": 0.02217,
+      "grad_norm": 0.9058617906836595,
+      "learning_rate": 0.003,
+      "loss": 4.1535,
+      "step": 2217
+    },
+    {
+      "epoch": 0.02218,
+      "grad_norm": 0.8239137730922835,
+      "learning_rate": 0.003,
+      "loss": 4.1865,
+      "step": 2218
+    },
+    {
+      "epoch": 0.02219,
+      "grad_norm": 0.7662188590247174,
+      "learning_rate": 0.003,
+      "loss": 4.1744,
+      "step": 2219
+    },
+    {
+      "epoch": 0.0222,
+      "grad_norm": 0.7276128620873287,
+      "learning_rate": 0.003,
+      "loss": 4.1721,
+      "step": 2220
+    },
+    {
+      "epoch": 0.02221,
+      "grad_norm": 0.8153838699769882,
+      "learning_rate": 0.003,
+      "loss": 4.1618,
+      "step": 2221
+    },
+    {
+      "epoch": 0.02222,
+      "grad_norm": 0.9239982401134911,
+      "learning_rate": 0.003,
+      "loss": 4.1621,
+      "step": 2222
+    },
+    {
+      "epoch": 0.02223,
+      "grad_norm": 0.9248509160902595,
+      "learning_rate": 0.003,
+      "loss": 4.1826,
+      "step": 2223
+    },
+    {
+      "epoch": 0.02224,
+      "grad_norm": 0.8778359853731296,
+      "learning_rate": 0.003,
+      "loss": 4.1633,
+      "step": 2224
+    },
+    {
+      "epoch": 0.02225,
+      "grad_norm": 1.1133790365277945,
+      "learning_rate": 0.003,
+      "loss": 4.1791,
+      "step": 2225
+    },
+    {
+      "epoch": 0.02226,
+      "grad_norm": 0.9357923626997176,
+      "learning_rate": 0.003,
+      "loss": 4.1895,
+      "step": 2226
+    },
+    {
+      "epoch": 0.02227,
+      "grad_norm": 0.8646361690518631,
+      "learning_rate": 0.003,
+      "loss": 4.183,
+      "step": 2227
+    },
+    {
+      "epoch": 0.02228,
+      "grad_norm": 0.938807129740724,
+      "learning_rate": 0.003,
+      "loss": 4.1718,
+      "step": 2228
+    },
+    {
+      "epoch": 0.02229,
+      "grad_norm": 0.8004751977002076,
+      "learning_rate": 0.003,
+      "loss": 4.149,
+      "step": 2229
+    },
+    {
+      "epoch": 0.0223,
+      "grad_norm": 0.7882101721343657,
+      "learning_rate": 0.003,
+      "loss": 4.1858,
+      "step": 2230
+    },
+    {
+      "epoch": 0.02231,
+      "grad_norm": 0.7274707205896823,
+      "learning_rate": 0.003,
+      "loss": 4.1896,
+      "step": 2231
+    },
+    {
+      "epoch": 0.02232,
+      "grad_norm": 0.8291412822325788,
+      "learning_rate": 0.003,
+      "loss": 4.2,
+      "step": 2232
+    },
+    {
+      "epoch": 0.02233,
+      "grad_norm": 0.8063025430085007,
+      "learning_rate": 0.003,
+      "loss": 4.166,
+      "step": 2233
+    },
+    {
+      "epoch": 0.02234,
+      "grad_norm": 0.6911537576664851,
+      "learning_rate": 0.003,
+      "loss": 4.1865,
+      "step": 2234
+    },
+    {
+      "epoch": 0.02235,
+      "grad_norm": 0.7448703016841659,
+      "learning_rate": 0.003,
+      "loss": 4.2113,
+      "step": 2235
+    },
+    {
+      "epoch": 0.02236,
+      "grad_norm": 0.8233234836151326,
+      "learning_rate": 0.003,
+      "loss": 4.1655,
+      "step": 2236
+    },
+    {
+      "epoch": 0.02237,
+      "grad_norm": 1.0093335974280213,
+      "learning_rate": 0.003,
+      "loss": 4.2052,
+      "step": 2237
+    },
+    {
+      "epoch": 0.02238,
+      "grad_norm": 1.1018651279861302,
+      "learning_rate": 0.003,
+      "loss": 4.196,
+      "step": 2238
+    },
+    {
+      "epoch": 0.02239,
+      "grad_norm": 0.7048103914820842,
+      "learning_rate": 0.003,
+      "loss": 4.1668,
+      "step": 2239
+    },
+    {
+      "epoch": 0.0224,
+      "grad_norm": 0.6705866907364436,
+      "learning_rate": 0.003,
+      "loss": 4.1477,
+      "step": 2240
+    },
+    {
+      "epoch": 0.02241,
+      "grad_norm": 0.768034661346772,
+      "learning_rate": 0.003,
+      "loss": 4.1807,
+      "step": 2241
+    },
+    {
+      "epoch": 0.02242,
+      "grad_norm": 0.7013024117659296,
+      "learning_rate": 0.003,
+      "loss": 4.1591,
+      "step": 2242
+    },
+    {
+      "epoch": 0.02243,
+      "grad_norm": 0.8407412837981002,
+      "learning_rate": 0.003,
+      "loss": 4.1937,
+      "step": 2243
+    },
+    {
+      "epoch": 0.02244,
+      "grad_norm": 0.8584973105281422,
+      "learning_rate": 0.003,
+      "loss": 4.1731,
+      "step": 2244
+    },
+    {
+      "epoch": 0.02245,
+      "grad_norm": 0.810849020894934,
+      "learning_rate": 0.003,
+      "loss": 4.1578,
+      "step": 2245
+    },
+    {
+      "epoch": 0.02246,
+      "grad_norm": 0.7441445696162579,
+      "learning_rate": 0.003,
+      "loss": 4.191,
+      "step": 2246
+    },
+    {
+      "epoch": 0.02247,
+      "grad_norm": 0.719895099061681,
+      "learning_rate": 0.003,
+      "loss": 4.162,
+      "step": 2247
+    },
+    {
+      "epoch": 0.02248,
+      "grad_norm": 0.6276605133960077,
+      "learning_rate": 0.003,
+      "loss": 4.1652,
+      "step": 2248
+    },
+    {
+      "epoch": 0.02249,
+      "grad_norm": 0.5820105403717251,
+      "learning_rate": 0.003,
+      "loss": 4.1561,
+      "step": 2249
+    },
+    {
+      "epoch": 0.0225,
+      "grad_norm": 0.494797181910464,
+      "learning_rate": 0.003,
+      "loss": 4.1646,
+      "step": 2250
+    },
+    {
+      "epoch": 0.02251,
+      "grad_norm": 0.45799261155941245,
+      "learning_rate": 0.003,
+      "loss": 4.1881,
+      "step": 2251
+    },
+    {
+      "epoch": 0.02252,
+      "grad_norm": 0.4312165334190654,
+      "learning_rate": 0.003,
+      "loss": 4.1745,
+      "step": 2252
+    },
+    {
+      "epoch": 0.02253,
+      "grad_norm": 0.4089956454491447,
+      "learning_rate": 0.003,
+      "loss": 4.1547,
+      "step": 2253
+    },
+    {
+      "epoch": 0.02254,
+      "grad_norm": 0.41711525641558334,
+      "learning_rate": 0.003,
+      "loss": 4.1591,
+      "step": 2254
+    },
+    {
+      "epoch": 0.02255,
+      "grad_norm": 0.39887312531471364,
+      "learning_rate": 0.003,
+      "loss": 4.1419,
+      "step": 2255
+    },
+    {
+      "epoch": 0.02256,
+      "grad_norm": 0.44922284802782086,
+      "learning_rate": 0.003,
+      "loss": 4.1321,
+      "step": 2256
+    },
+    {
+      "epoch": 0.02257,
+      "grad_norm": 0.43881289908027915,
+      "learning_rate": 0.003,
+      "loss": 4.1476,
+      "step": 2257
+    },
+    {
+      "epoch": 0.02258,
+      "grad_norm": 0.42711858762354266,
+      "learning_rate": 0.003,
+      "loss": 4.1569,
+      "step": 2258
+    },
+    {
+      "epoch": 0.02259,
+      "grad_norm": 0.4761674249696505,
+      "learning_rate": 0.003,
+      "loss": 4.1182,
+      "step": 2259
+    },
+    {
+      "epoch": 0.0226,
+      "grad_norm": 0.6172223177216527,
+      "learning_rate": 0.003,
+      "loss": 4.1745,
+      "step": 2260
+    },
+    {
+      "epoch": 0.02261,
+      "grad_norm": 0.9969811836278653,
+      "learning_rate": 0.003,
+      "loss": 4.1626,
+      "step": 2261
+    },
+    {
+      "epoch": 0.02262,
+      "grad_norm": 1.4207495716773118,
+      "learning_rate": 0.003,
+      "loss": 4.1955,
+      "step": 2262
+    },
+    {
+      "epoch": 0.02263,
+      "grad_norm": 0.5107196284950156,
+      "learning_rate": 0.003,
+      "loss": 4.1749,
+      "step": 2263
+    },
+    {
+      "epoch": 0.02264,
+      "grad_norm": 0.8124315452603879,
+      "learning_rate": 0.003,
+      "loss": 4.1794,
+      "step": 2264
+    },
+    {
+      "epoch": 0.02265,
+      "grad_norm": 0.7127120804095198,
+      "learning_rate": 0.003,
+      "loss": 4.1342,
+      "step": 2265
+    },
+    {
+      "epoch": 0.02266,
+      "grad_norm": 0.5120142831669139,
+      "learning_rate": 0.003,
+      "loss": 4.142,
+      "step": 2266
+    },
+    {
+      "epoch": 0.02267,
+      "grad_norm": 0.842715736214197,
+      "learning_rate": 0.003,
+      "loss": 4.1775,
+      "step": 2267
+    },
+    {
+      "epoch": 0.02268,
+      "grad_norm": 0.8370937405428989,
+      "learning_rate": 0.003,
+      "loss": 4.1593,
+      "step": 2268
+    },
+    {
+      "epoch": 0.02269,
+      "grad_norm": 0.7173617092756458,
+      "learning_rate": 0.003,
+      "loss": 4.174,
+      "step": 2269
+    },
+    {
+      "epoch": 0.0227,
+      "grad_norm": 0.8926880986379259,
+      "learning_rate": 0.003,
+      "loss": 4.1781,
+      "step": 2270
+    },
+    {
+      "epoch": 0.02271,
+      "grad_norm": 0.9394293424718215,
+      "learning_rate": 0.003,
+      "loss": 4.1873,
+      "step": 2271
+    },
+    {
+      "epoch": 0.02272,
+      "grad_norm": 0.9564097717489924,
+      "learning_rate": 0.003,
+      "loss": 4.1878,
+      "step": 2272
+    },
+    {
+      "epoch": 0.02273,
+      "grad_norm": 1.2112473190186792,
+      "learning_rate": 0.003,
+      "loss": 4.1998,
+      "step": 2273
+    },
+    {
+      "epoch": 0.02274,
+      "grad_norm": 0.9744415053378092,
+      "learning_rate": 0.003,
+      "loss": 4.1786,
+      "step": 2274
+    },
+    {
+      "epoch": 0.02275,
+      "grad_norm": 1.0078841940550498,
+      "learning_rate": 0.003,
+      "loss": 4.1796,
+      "step": 2275
+    },
+    {
+      "epoch": 0.02276,
+      "grad_norm": 0.8300538512961302,
+      "learning_rate": 0.003,
+      "loss": 4.1696,
+      "step": 2276
+    },
+    {
+      "epoch": 0.02277,
+      "grad_norm": 0.6643966250880426,
+      "learning_rate": 0.003,
+      "loss": 4.1719,
+      "step": 2277
+    },
+    {
+      "epoch": 0.02278,
+      "grad_norm": 0.6190627183036216,
+      "learning_rate": 0.003,
+      "loss": 4.1782,
+      "step": 2278
+    },
+    {
+      "epoch": 0.02279,
+      "grad_norm": 0.551131169249495,
+      "learning_rate": 0.003,
+      "loss": 4.1827,
+      "step": 2279
+    },
+    {
+      "epoch": 0.0228,
+      "grad_norm": 0.5572806608313633,
+      "learning_rate": 0.003,
+      "loss": 4.173,
+      "step": 2280
+    },
+    {
+      "epoch": 0.02281,
+      "grad_norm": 0.630640256246517,
+      "learning_rate": 0.003,
+      "loss": 4.1688,
+      "step": 2281
+    },
+    {
+      "epoch": 0.02282,
+      "grad_norm": 0.667161853354935,
+      "learning_rate": 0.003,
+      "loss": 4.181,
+      "step": 2282
+    },
+    {
+      "epoch": 0.02283,
+      "grad_norm": 0.7804313255347314,
+      "learning_rate": 0.003,
+      "loss": 4.1982,
+      "step": 2283
+    },
+    {
+      "epoch": 0.02284,
+      "grad_norm": 0.8174062173017085,
+      "learning_rate": 0.003,
+      "loss": 4.1998,
+      "step": 2284
+    },
+    {
+      "epoch": 0.02285,
+      "grad_norm": 0.8933970302021632,
+      "learning_rate": 0.003,
+      "loss": 4.147,
+      "step": 2285
+    },
+    {
+      "epoch": 0.02286,
+      "grad_norm": 0.8583328769415308,
+      "learning_rate": 0.003,
+      "loss": 4.2039,
+      "step": 2286
+    },
+    {
+      "epoch": 0.02287,
+      "grad_norm": 0.8734810759184407,
+      "learning_rate": 0.003,
+      "loss": 4.1855,
+      "step": 2287
+    },
+    {
+      "epoch": 0.02288,
+      "grad_norm": 0.8928624064077537,
+      "learning_rate": 0.003,
+      "loss": 4.2123,
+      "step": 2288
+    },
+    {
+      "epoch": 0.02289,
+      "grad_norm": 0.783879288737337,
+      "learning_rate": 0.003,
+      "loss": 4.2056,
+      "step": 2289
+    },
+    {
+      "epoch": 0.0229,
+      "grad_norm": 0.7701279568768081,
+      "learning_rate": 0.003,
+      "loss": 4.1809,
+      "step": 2290
+    },
+    {
+      "epoch": 0.02291,
+      "grad_norm": 0.7353904143222354,
+      "learning_rate": 0.003,
+      "loss": 4.1585,
+      "step": 2291
+    },
+    {
+      "epoch": 0.02292,
+      "grad_norm": 0.5673182764651727,
+      "learning_rate": 0.003,
+      "loss": 4.1467,
+      "step": 2292
+    },
+    {
+      "epoch": 0.02293,
+      "grad_norm": 0.6211445071442583,
+      "learning_rate": 0.003,
+      "loss": 4.1508,
+      "step": 2293
+    },
+    {
+      "epoch": 0.02294,
+      "grad_norm": 0.5897432424358218,
+      "learning_rate": 0.003,
+      "loss": 4.1497,
+      "step": 2294
+    },
+    {
+      "epoch": 0.02295,
+      "grad_norm": 0.6701151196414159,
+      "learning_rate": 0.003,
+      "loss": 4.147,
+      "step": 2295
+    },
+    {
+      "epoch": 0.02296,
+      "grad_norm": 1.0101216354387503,
+      "learning_rate": 0.003,
+      "loss": 4.1897,
+      "step": 2296
+    },
+    {
+      "epoch": 0.02297,
+      "grad_norm": 1.2406081319623274,
+      "learning_rate": 0.003,
+      "loss": 4.1822,
+      "step": 2297
+    },
+    {
+      "epoch": 0.02298,
+      "grad_norm": 0.5988120231143333,
+      "learning_rate": 0.003,
+      "loss": 4.1702,
+      "step": 2298
+    },
+    {
+      "epoch": 0.02299,
+      "grad_norm": 0.7795860072160454,
+      "learning_rate": 0.003,
+      "loss": 4.1804,
+      "step": 2299
+    },
+    {
+      "epoch": 0.023,
+      "grad_norm": 0.784959258488215,
+      "learning_rate": 0.003,
+      "loss": 4.1984,
+      "step": 2300
+    },
+    {
+      "epoch": 0.02301,
+      "grad_norm": 0.5879362613709453,
+      "learning_rate": 0.003,
+      "loss": 4.157,
+      "step": 2301
+    },
+    {
+      "epoch": 0.02302,
+      "grad_norm": 0.7106606864619079,
+      "learning_rate": 0.003,
+      "loss": 4.1649,
+      "step": 2302
+    },
+    {
+      "epoch": 0.02303,
+      "grad_norm": 0.7402995000062723,
+      "learning_rate": 0.003,
+      "loss": 4.1545,
+      "step": 2303
+    },
+    {
+      "epoch": 0.02304,
+      "grad_norm": 0.8291546359347778,
+      "learning_rate": 0.003,
+      "loss": 4.1898,
+      "step": 2304
+    },
+    {
+      "epoch": 0.02305,
+      "grad_norm": 0.753148855618092,
+      "learning_rate": 0.003,
+      "loss": 4.1615,
+      "step": 2305
+    },
+    {
+      "epoch": 0.02306,
+      "grad_norm": 0.8869246970763861,
+      "learning_rate": 0.003,
+      "loss": 4.1444,
+      "step": 2306
+    },
+    {
+      "epoch": 0.02307,
+      "grad_norm": 0.8925578818662347,
+      "learning_rate": 0.003,
+      "loss": 4.1758,
+      "step": 2307
+    },
+    {
+      "epoch": 0.02308,
+      "grad_norm": 0.9066759173397617,
+      "learning_rate": 0.003,
+      "loss": 4.1762,
+      "step": 2308
+    },
+    {
+      "epoch": 0.02309,
+      "grad_norm": 0.8201889707383639,
+      "learning_rate": 0.003,
+      "loss": 4.1952,
+      "step": 2309
+    },
+    {
+      "epoch": 0.0231,
+      "grad_norm": 0.7904644702923331,
+      "learning_rate": 0.003,
+      "loss": 4.1751,
+      "step": 2310
+    },
+    {
+      "epoch": 0.02311,
+      "grad_norm": 0.7060518287169474,
+      "learning_rate": 0.003,
+      "loss": 4.1776,
+      "step": 2311
+    },
+    {
+      "epoch": 0.02312,
+      "grad_norm": 0.7392874926033767,
+      "learning_rate": 0.003,
+      "loss": 4.1519,
+      "step": 2312
+    },
+    {
+      "epoch": 0.02313,
+      "grad_norm": 0.8538999440765905,
+      "learning_rate": 0.003,
+      "loss": 4.1529,
+      "step": 2313
+    },
+    {
+      "epoch": 0.02314,
+      "grad_norm": 0.9312222422709578,
+      "learning_rate": 0.003,
+      "loss": 4.1718,
+      "step": 2314
+    },
+    {
+      "epoch": 0.02315,
+      "grad_norm": 0.9687214234200097,
+      "learning_rate": 0.003,
+      "loss": 4.1771,
+      "step": 2315
+    },
+    {
+      "epoch": 0.02316,
+      "grad_norm": 0.891697328396861,
+      "learning_rate": 0.003,
+      "loss": 4.1909,
+      "step": 2316
+    },
+    {
+      "epoch": 0.02317,
+      "grad_norm": 0.897934373295407,
+      "learning_rate": 0.003,
+      "loss": 4.1753,
+      "step": 2317
+    },
+    {
+      "epoch": 0.02318,
+      "grad_norm": 0.8051331418091766,
+      "learning_rate": 0.003,
+      "loss": 4.1614,
+      "step": 2318
+    },
+    {
+      "epoch": 0.02319,
+      "grad_norm": 0.703806089932542,
+      "learning_rate": 0.003,
+      "loss": 4.1804,
+      "step": 2319
+    },
+    {
+      "epoch": 0.0232,
+      "grad_norm": 0.7388332818016227,
+      "learning_rate": 0.003,
+      "loss": 4.1898,
+      "step": 2320
+    },
+    {
+      "epoch": 0.02321,
+      "grad_norm": 0.7450344205679207,
+      "learning_rate": 0.003,
+      "loss": 4.1692,
+      "step": 2321
+    },
+    {
+      "epoch": 0.02322,
+      "grad_norm": 0.6660500555618327,
+      "learning_rate": 0.003,
+      "loss": 4.1746,
+      "step": 2322
+    },
+    {
+      "epoch": 0.02323,
+      "grad_norm": 0.6571661630697367,
+      "learning_rate": 0.003,
+      "loss": 4.1481,
+      "step": 2323
+    },
+    {
+      "epoch": 0.02324,
+      "grad_norm": 0.6161646162269099,
+      "learning_rate": 0.003,
+      "loss": 4.1462,
+      "step": 2324
+    },
+    {
+      "epoch": 0.02325,
+      "grad_norm": 0.5919512249258755,
+      "learning_rate": 0.003,
+      "loss": 4.1679,
+      "step": 2325
+    },
+    {
+      "epoch": 0.02326,
+      "grad_norm": 0.7224558633075708,
+      "learning_rate": 0.003,
+      "loss": 4.1831,
+      "step": 2326
+    },
+    {
+      "epoch": 0.02327,
+      "grad_norm": 0.8816178665411202,
+      "learning_rate": 0.003,
+      "loss": 4.1759,
+      "step": 2327
+    },
+    {
+      "epoch": 0.02328,
+      "grad_norm": 0.8699762426391761,
+      "learning_rate": 0.003,
+      "loss": 4.176,
+      "step": 2328
+    },
+    {
+      "epoch": 0.02329,
+      "grad_norm": 0.723960245924297,
+      "learning_rate": 0.003,
+      "loss": 4.1606,
+      "step": 2329
+    },
+    {
+      "epoch": 0.0233,
+      "grad_norm": 0.7628023637683098,
+      "learning_rate": 0.003,
+      "loss": 4.1563,
+      "step": 2330
+    },
+    {
+      "epoch": 0.02331,
+      "grad_norm": 0.7695226471244487,
+      "learning_rate": 0.003,
+      "loss": 4.1933,
+      "step": 2331
+    },
+    {
+      "epoch": 0.02332,
+      "grad_norm": 0.7712606828136085,
+      "learning_rate": 0.003,
+      "loss": 4.1567,
+      "step": 2332
+    },
+    {
+      "epoch": 0.02333,
+      "grad_norm": 0.7181228868994012,
+      "learning_rate": 0.003,
+      "loss": 4.1639,
+      "step": 2333
+    },
+    {
+      "epoch": 0.02334,
+      "grad_norm": 0.6385757918578437,
+      "learning_rate": 0.003,
+      "loss": 4.1304,
+      "step": 2334
+    },
+    {
+      "epoch": 0.02335,
+      "grad_norm": 0.5804913608133354,
+      "learning_rate": 0.003,
+      "loss": 4.1671,
+      "step": 2335
+    },
+    {
+      "epoch": 0.02336,
+      "grad_norm": 0.6406948035907543,
+      "learning_rate": 0.003,
+      "loss": 4.1601,
+      "step": 2336
+    },
+    {
+      "epoch": 0.02337,
+      "grad_norm": 0.7550291811921531,
+      "learning_rate": 0.003,
+      "loss": 4.1668,
+      "step": 2337
+    },
+    {
+      "epoch": 0.02338,
+      "grad_norm": 0.867264165349628,
+      "learning_rate": 0.003,
+      "loss": 4.1574,
+      "step": 2338
+    },
+    {
+      "epoch": 0.02339,
+      "grad_norm": 0.9370869698358649,
+      "learning_rate": 0.003,
+      "loss": 4.1596,
+      "step": 2339
+    },
+    {
+      "epoch": 0.0234,
+      "grad_norm": 1.04114552861465,
+      "learning_rate": 0.003,
+      "loss": 4.1762,
+      "step": 2340
+    },
+    {
+      "epoch": 0.02341,
+      "grad_norm": 0.8419443125713769,
+      "learning_rate": 0.003,
+      "loss": 4.1461,
+      "step": 2341
+    },
+    {
+      "epoch": 0.02342,
+      "grad_norm": 0.7516609856190137,
+      "learning_rate": 0.003,
+      "loss": 4.1608,
+      "step": 2342
+    },
+    {
+      "epoch": 0.02343,
+      "grad_norm": 0.7840218113502367,
+      "learning_rate": 0.003,
+      "loss": 4.1597,
+      "step": 2343
+    },
+    {
+      "epoch": 0.02344,
+      "grad_norm": 0.7106909207560747,
+      "learning_rate": 0.003,
+      "loss": 4.1579,
+      "step": 2344
+    },
+    {
+      "epoch": 0.02345,
+      "grad_norm": 0.695743479378358,
+      "learning_rate": 0.003,
+      "loss": 4.1505,
+      "step": 2345
+    },
+    {
+      "epoch": 0.02346,
+      "grad_norm": 0.7618556711999727,
+      "learning_rate": 0.003,
+      "loss": 4.1514,
+      "step": 2346
+    },
+    {
+      "epoch": 0.02347,
+      "grad_norm": 0.8494779389825688,
+      "learning_rate": 0.003,
+      "loss": 4.1923,
+      "step": 2347
+    },
+    {
+      "epoch": 0.02348,
+      "grad_norm": 0.8819304188600076,
+      "learning_rate": 0.003,
+      "loss": 4.1614,
+      "step": 2348
+    },
+    {
+      "epoch": 0.02349,
+      "grad_norm": 0.801874189029763,
+      "learning_rate": 0.003,
+      "loss": 4.1729,
+      "step": 2349
+    },
+    {
+      "epoch": 0.0235,
+      "grad_norm": 0.7502001194200574,
+      "learning_rate": 0.003,
+      "loss": 4.168,
+      "step": 2350
+    },
+    {
+      "epoch": 0.02351,
+      "grad_norm": 0.7987799293083814,
+      "learning_rate": 0.003,
+      "loss": 4.1781,
+      "step": 2351
+    },
+    {
+      "epoch": 0.02352,
+      "grad_norm": 0.9306705411998242,
+      "learning_rate": 0.003,
+      "loss": 4.142,
+      "step": 2352
+    },
+    {
+      "epoch": 0.02353,
+      "grad_norm": 0.8564501775467426,
+      "learning_rate": 0.003,
+      "loss": 4.1759,
+      "step": 2353
+    },
+    {
+      "epoch": 0.02354,
+      "grad_norm": 0.8345109850380096,
+      "learning_rate": 0.003,
+      "loss": 4.1813,
+      "step": 2354
+    },
+    {
+      "epoch": 0.02355,
+      "grad_norm": 0.7887172745864481,
+      "learning_rate": 0.003,
+      "loss": 4.1723,
+      "step": 2355
+    },
+    {
+      "epoch": 0.02356,
+      "grad_norm": 0.6582873563724112,
+      "learning_rate": 0.003,
+      "loss": 4.1515,
+      "step": 2356
+    },
+    {
+      "epoch": 0.02357,
+      "grad_norm": 0.6737457349000578,
+      "learning_rate": 0.003,
+      "loss": 4.1517,
+      "step": 2357
+    },
+    {
+      "epoch": 0.02358,
+      "grad_norm": 0.7705872627309706,
+      "learning_rate": 0.003,
+      "loss": 4.1742,
+      "step": 2358
+    },
+    {
+      "epoch": 0.02359,
+      "grad_norm": 0.8542556066081018,
+      "learning_rate": 0.003,
+      "loss": 4.1536,
+      "step": 2359
+    },
+    {
+      "epoch": 0.0236,
+      "grad_norm": 0.9842191585501165,
+      "learning_rate": 0.003,
+      "loss": 4.2032,
+      "step": 2360
+    },
+    {
+      "epoch": 0.02361,
+      "grad_norm": 0.9112511096880844,
+      "learning_rate": 0.003,
+      "loss": 4.1793,
+      "step": 2361
+    },
+    {
+      "epoch": 0.02362,
+      "grad_norm": 0.9278874197656558,
+      "learning_rate": 0.003,
+      "loss": 4.1712,
+      "step": 2362
+    },
+    {
+      "epoch": 0.02363,
+      "grad_norm": 0.7482616982968588,
+      "learning_rate": 0.003,
+      "loss": 4.1444,
+      "step": 2363
+    },
+    {
+      "epoch": 0.02364,
+      "grad_norm": 0.5981870727991454,
+      "learning_rate": 0.003,
+      "loss": 4.1597,
+      "step": 2364
+    },
+    {
+      "epoch": 0.02365,
+      "grad_norm": 0.6120042832874062,
+      "learning_rate": 0.003,
+      "loss": 4.1788,
+      "step": 2365
+    },
+    {
+      "epoch": 0.02366,
+      "grad_norm": 0.592965390221233,
+      "learning_rate": 0.003,
+      "loss": 4.1489,
+      "step": 2366
+    },
+    {
+      "epoch": 0.02367,
+      "grad_norm": 0.5967931209483315,
+      "learning_rate": 0.003,
+      "loss": 4.1579,
+      "step": 2367
+    },
+    {
+      "epoch": 0.02368,
+      "grad_norm": 0.6289890460021353,
+      "learning_rate": 0.003,
+      "loss": 4.1346,
+      "step": 2368
+    },
+    {
+      "epoch": 0.02369,
+      "grad_norm": 0.6580991252391354,
+      "learning_rate": 0.003,
+      "loss": 4.1441,
+      "step": 2369
+    },
+    {
+      "epoch": 0.0237,
+      "grad_norm": 0.6945790129958274,
+      "learning_rate": 0.003,
+      "loss": 4.1741,
+      "step": 2370
+    },
+    {
+      "epoch": 0.02371,
+      "grad_norm": 0.7107030962570803,
+      "learning_rate": 0.003,
+      "loss": 4.1569,
+      "step": 2371
+    },
+    {
+      "epoch": 0.02372,
+      "grad_norm": 0.8059296050081108,
+      "learning_rate": 0.003,
+      "loss": 4.149,
+      "step": 2372
+    },
+    {
+      "epoch": 0.02373,
+      "grad_norm": 0.921257026362051,
+      "learning_rate": 0.003,
+      "loss": 4.1414,
+      "step": 2373
+    },
+    {
+      "epoch": 0.02374,
+      "grad_norm": 0.955728704961619,
+      "learning_rate": 0.003,
+      "loss": 4.1544,
+      "step": 2374
+    },
+    {
+      "epoch": 0.02375,
+      "grad_norm": 0.7504286137015724,
+      "learning_rate": 0.003,
+      "loss": 4.15,
+      "step": 2375
+    },
+    {
+      "epoch": 0.02376,
+      "grad_norm": 0.7478781963883802,
+      "learning_rate": 0.003,
+      "loss": 4.1724,
+      "step": 2376
+    },
+    {
+      "epoch": 0.02377,
+      "grad_norm": 0.917428121872858,
+      "learning_rate": 0.003,
+      "loss": 4.1486,
+      "step": 2377
+    },
+    {
+      "epoch": 0.02378,
+      "grad_norm": 0.7822980127349399,
+      "learning_rate": 0.003,
+      "loss": 4.1546,
+      "step": 2378
+    },
+    {
+      "epoch": 0.02379,
+      "grad_norm": 0.5863463150077827,
+      "learning_rate": 0.003,
+      "loss": 4.136,
+      "step": 2379
+    },
+    {
+      "epoch": 0.0238,
+      "grad_norm": 0.5340016137538751,
+      "learning_rate": 0.003,
+      "loss": 4.109,
+      "step": 2380
+    },
+    {
+      "epoch": 0.02381,
+      "grad_norm": 0.6128857312553131,
+      "learning_rate": 0.003,
+      "loss": 4.1229,
+      "step": 2381
+    },
+    {
+      "epoch": 0.02382,
+      "grad_norm": 0.5658688824938058,
+      "learning_rate": 0.003,
+      "loss": 4.1779,
+      "step": 2382
+    },
+    {
+      "epoch": 0.02383,
+      "grad_norm": 0.5676379045904221,
+      "learning_rate": 0.003,
+      "loss": 4.1407,
+      "step": 2383
+    },
+    {
+      "epoch": 0.02384,
+      "grad_norm": 0.5238096521489247,
+      "learning_rate": 0.003,
+      "loss": 4.1258,
+      "step": 2384
+    },
+    {
+      "epoch": 0.02385,
+      "grad_norm": 0.5152751390365298,
+      "learning_rate": 0.003,
+      "loss": 4.1501,
+      "step": 2385
+    },
+    {
+      "epoch": 0.02386,
+      "grad_norm": 0.6091514932256747,
+      "learning_rate": 0.003,
+      "loss": 4.13,
+      "step": 2386
+    },
+    {
+      "epoch": 0.02387,
+      "grad_norm": 0.7793599922586826,
+      "learning_rate": 0.003,
+      "loss": 4.1323,
+      "step": 2387
+    },
+    {
+      "epoch": 0.02388,
+      "grad_norm": 0.9352617435026656,
+      "learning_rate": 0.003,
+      "loss": 4.1923,
+      "step": 2388
+    },
+    {
+      "epoch": 0.02389,
+      "grad_norm": 0.9609729763435508,
+      "learning_rate": 0.003,
+      "loss": 4.1586,
+      "step": 2389
+    },
+    {
+      "epoch": 0.0239,
+      "grad_norm": 0.8094018370976185,
+      "learning_rate": 0.003,
+      "loss": 4.1475,
+      "step": 2390
+    },
+    {
+      "epoch": 0.02391,
+      "grad_norm": 0.8506283143692317,
+      "learning_rate": 0.003,
+      "loss": 4.1734,
+      "step": 2391
+    },
+    {
+      "epoch": 0.02392,
+      "grad_norm": 0.9050332115497783,
+      "learning_rate": 0.003,
+      "loss": 4.1803,
+      "step": 2392
+    },
+    {
+      "epoch": 0.02393,
+      "grad_norm": 0.9634890728744777,
+      "learning_rate": 0.003,
+      "loss": 4.1413,
+      "step": 2393
+    },
+    {
+      "epoch": 0.02394,
+      "grad_norm": 0.8882373157340075,
+      "learning_rate": 0.003,
+      "loss": 4.1569,
+      "step": 2394
+    },
+    {
+      "epoch": 0.02395,
+      "grad_norm": 0.7565332153535935,
+      "learning_rate": 0.003,
+      "loss": 4.1675,
+      "step": 2395
+    },
+    {
+      "epoch": 0.02396,
+      "grad_norm": 0.8603458435447788,
+      "learning_rate": 0.003,
+      "loss": 4.1849,
+      "step": 2396
+    },
+    {
+      "epoch": 0.02397,
+      "grad_norm": 0.8261263361309694,
+      "learning_rate": 0.003,
+      "loss": 4.169,
+      "step": 2397
+    },
+    {
+      "epoch": 0.02398,
+      "grad_norm": 0.7794743587492478,
+      "learning_rate": 0.003,
+      "loss": 4.1744,
+      "step": 2398
+    },
+    {
+      "epoch": 0.02399,
+      "grad_norm": 0.8140982417839558,
+      "learning_rate": 0.003,
+      "loss": 4.1533,
+      "step": 2399
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 0.9091203275980858,
+      "learning_rate": 0.003,
+      "loss": 4.1719,
+      "step": 2400
+    },
+    {
+      "epoch": 0.02401,
+      "grad_norm": 0.9082248326953591,
+      "learning_rate": 0.003,
+      "loss": 4.1471,
+      "step": 2401
+    },
+    {
+      "epoch": 0.02402,
+      "grad_norm": 0.8996640369589112,
+      "learning_rate": 0.003,
+      "loss": 4.1798,
+      "step": 2402
+    },
+    {
+      "epoch": 0.02403,
+      "grad_norm": 0.9493494458526733,
+      "learning_rate": 0.003,
+      "loss": 4.1514,
+      "step": 2403
+    },
+    {
+      "epoch": 0.02404,
+      "grad_norm": 0.9827368493949332,
+      "learning_rate": 0.003,
+      "loss": 4.1965,
+      "step": 2404
+    },
+    {
+      "epoch": 0.02405,
+      "grad_norm": 0.9551905621174913,
+      "learning_rate": 0.003,
+      "loss": 4.2021,
+      "step": 2405
+    },
+    {
+      "epoch": 0.02406,
+      "grad_norm": 0.9364752365255031,
+      "learning_rate": 0.003,
+      "loss": 4.1854,
+      "step": 2406
+    },
+    {
+      "epoch": 0.02407,
+      "grad_norm": 0.9463612550147267,
+      "learning_rate": 0.003,
+      "loss": 4.1641,
+      "step": 2407
+    },
+    {
+      "epoch": 0.02408,
+      "grad_norm": 0.8719994888631714,
+      "learning_rate": 0.003,
+      "loss": 4.1732,
+      "step": 2408
+    },
+    {
+      "epoch": 0.02409,
+      "grad_norm": 0.864310092393717,
+      "learning_rate": 0.003,
+      "loss": 4.1757,
+      "step": 2409
+    },
+    {
+      "epoch": 0.0241,
+      "grad_norm": 0.8973500607718927,
+      "learning_rate": 0.003,
+      "loss": 4.1883,
+      "step": 2410
+    },
+    {
+      "epoch": 0.02411,
+      "grad_norm": 1.1693499034787715,
+      "learning_rate": 0.003,
+      "loss": 4.1693,
+      "step": 2411
+    },
+    {
+      "epoch": 0.02412,
+      "grad_norm": 0.8431604098452435,
+      "learning_rate": 0.003,
+      "loss": 4.164,
+      "step": 2412
+    },
+    {
+      "epoch": 0.02413,
+      "grad_norm": 0.7767055362680262,
+      "learning_rate": 0.003,
+      "loss": 4.1597,
+      "step": 2413
+    },
+    {
+      "epoch": 0.02414,
+      "grad_norm": 0.8330306566206577,
+      "learning_rate": 0.003,
+      "loss": 4.1744,
+      "step": 2414
+    },
+    {
+      "epoch": 0.02415,
+      "grad_norm": 0.8217708926095696,
+      "learning_rate": 0.003,
+      "loss": 4.1765,
+      "step": 2415
+    },
+    {
+      "epoch": 0.02416,
+      "grad_norm": 0.8597994699370953,
+      "learning_rate": 0.003,
+      "loss": 4.1589,
+      "step": 2416
+    },
+    {
+      "epoch": 0.02417,
+      "grad_norm": 0.8398506547738631,
+      "learning_rate": 0.003,
+      "loss": 4.1829,
+      "step": 2417
+    },
+    {
+      "epoch": 0.02418,
+      "grad_norm": 0.7704254084685477,
+      "learning_rate": 0.003,
+      "loss": 4.1714,
+      "step": 2418
+    },
+    {
+      "epoch": 0.02419,
+      "grad_norm": 0.752617109928811,
+      "learning_rate": 0.003,
+      "loss": 4.1691,
+      "step": 2419
+    },
+    {
+      "epoch": 0.0242,
+      "grad_norm": 0.681614467806083,
+      "learning_rate": 0.003,
+      "loss": 4.1781,
+      "step": 2420
+    },
+    {
+      "epoch": 0.02421,
+      "grad_norm": 0.57806703718961,
+      "learning_rate": 0.003,
+      "loss": 4.1374,
+      "step": 2421
+    },
+    {
+      "epoch": 0.02422,
+      "grad_norm": 0.5368829917602504,
+      "learning_rate": 0.003,
+      "loss": 4.1904,
+      "step": 2422
+    },
+    {
+      "epoch": 0.02423,
+      "grad_norm": 0.49926017966447833,
+      "learning_rate": 0.003,
+      "loss": 4.1767,
+      "step": 2423
+    },
+    {
+      "epoch": 0.02424,
+      "grad_norm": 0.44538511553678933,
+      "learning_rate": 0.003,
+      "loss": 4.1709,
+      "step": 2424
+    },
+    {
+      "epoch": 0.02425,
+      "grad_norm": 0.4455623217366007,
+      "learning_rate": 0.003,
+      "loss": 4.1546,
+      "step": 2425
+    },
+    {
+      "epoch": 0.02426,
+      "grad_norm": 0.45034256565793024,
+      "learning_rate": 0.003,
+      "loss": 4.1608,
+      "step": 2426
+    },
+    {
+      "epoch": 0.02427,
+      "grad_norm": 0.4277494354911392,
+      "learning_rate": 0.003,
+      "loss": 4.1831,
+      "step": 2427
+    },
+    {
+      "epoch": 0.02428,
+      "grad_norm": 0.49536930851083055,
+      "learning_rate": 0.003,
+      "loss": 4.1561,
+      "step": 2428
+    },
+    {
+      "epoch": 0.02429,
+      "grad_norm": 0.5349623201813287,
+      "learning_rate": 0.003,
+      "loss": 4.1417,
+      "step": 2429
+    },
+    {
+      "epoch": 0.0243,
+      "grad_norm": 0.6385700511917229,
+      "learning_rate": 0.003,
+      "loss": 4.1681,
+      "step": 2430
+    },
+    {
+      "epoch": 0.02431,
+      "grad_norm": 0.7962007873904552,
+      "learning_rate": 0.003,
+      "loss": 4.1506,
+      "step": 2431
+    },
+    {
+      "epoch": 0.02432,
+      "grad_norm": 0.7552982964141086,
+      "learning_rate": 0.003,
+      "loss": 4.1735,
+      "step": 2432
+    },
+    {
+      "epoch": 0.02433,
+      "grad_norm": 0.5576758209292167,
+      "learning_rate": 0.003,
+      "loss": 4.179,
+      "step": 2433
+    },
+    {
+      "epoch": 0.02434,
+      "grad_norm": 0.5731742455027719,
+      "learning_rate": 0.003,
+      "loss": 4.1457,
+      "step": 2434
+    },
+    {
+      "epoch": 0.02435,
+      "grad_norm": 0.6782357026569522,
+      "learning_rate": 0.003,
+      "loss": 4.1482,
+      "step": 2435
+    },
+    {
+      "epoch": 0.02436,
+      "grad_norm": 0.8071855900106416,
+      "learning_rate": 0.003,
+      "loss": 4.1587,
+      "step": 2436
+    },
+    {
+      "epoch": 0.02437,
+      "grad_norm": 0.9265059909484438,
+      "learning_rate": 0.003,
+      "loss": 4.1467,
+      "step": 2437
+    },
+    {
+      "epoch": 0.02438,
+      "grad_norm": 0.8793214152171194,
+      "learning_rate": 0.003,
+      "loss": 4.1644,
+      "step": 2438
+    },
+    {
+      "epoch": 0.02439,
+      "grad_norm": 0.8832749976596599,
+      "learning_rate": 0.003,
+      "loss": 4.1384,
+      "step": 2439
+    },
+    {
+      "epoch": 0.0244,
+      "grad_norm": 0.9746984844718865,
+      "learning_rate": 0.003,
+      "loss": 4.1934,
+      "step": 2440
+    },
+    {
+      "epoch": 0.02441,
+      "grad_norm": 1.0708150303834312,
+      "learning_rate": 0.003,
+      "loss": 4.1618,
+      "step": 2441
+    },
+    {
+      "epoch": 0.02442,
+      "grad_norm": 1.040411975678819,
+      "learning_rate": 0.003,
+      "loss": 4.2014,
+      "step": 2442
+    },
+    {
+      "epoch": 0.02443,
+      "grad_norm": 1.157601705274513,
+      "learning_rate": 0.003,
+      "loss": 4.1413,
+      "step": 2443
+    },
+    {
+      "epoch": 0.02444,
+      "grad_norm": 0.8998517717034221,
+      "learning_rate": 0.003,
+      "loss": 4.1633,
+      "step": 2444
+    },
+    {
+      "epoch": 0.02445,
+      "grad_norm": 0.7914422604882547,
+      "learning_rate": 0.003,
+      "loss": 4.1691,
+      "step": 2445
+    },
+    {
+      "epoch": 0.02446,
+      "grad_norm": 0.765570533672039,
+      "learning_rate": 0.003,
+      "loss": 4.168,
+      "step": 2446
+    },
+    {
+      "epoch": 0.02447,
+      "grad_norm": 0.7780408804262848,
+      "learning_rate": 0.003,
+      "loss": 4.1745,
+      "step": 2447
+    },
+    {
+      "epoch": 0.02448,
+      "grad_norm": 0.8355663969723807,
+      "learning_rate": 0.003,
+      "loss": 4.1748,
+      "step": 2448
+    },
+    {
+      "epoch": 0.02449,
+      "grad_norm": 0.8911004412381984,
+      "learning_rate": 0.003,
+      "loss": 4.173,
+      "step": 2449
+    },
+    {
+      "epoch": 0.0245,
+      "grad_norm": 0.8721571131136453,
+      "learning_rate": 0.003,
+      "loss": 4.1764,
+      "step": 2450
+    },
+    {
+      "epoch": 0.02451,
+      "grad_norm": 0.8424855685228627,
+      "learning_rate": 0.003,
+      "loss": 4.1549,
+      "step": 2451
+    },
+    {
+      "epoch": 0.02452,
+      "grad_norm": 0.71343486564028,
+      "learning_rate": 0.003,
+      "loss": 4.1589,
+      "step": 2452
+    },
+    {
+      "epoch": 0.02453,
+      "grad_norm": 0.830794139281049,
+      "learning_rate": 0.003,
+      "loss": 4.1386,
+      "step": 2453
+    },
+    {
+      "epoch": 0.02454,
+      "grad_norm": 0.8371964018463887,
+      "learning_rate": 0.003,
+      "loss": 4.1444,
+      "step": 2454
+    },
+    {
+      "epoch": 0.02455,
+      "grad_norm": 0.828173261775671,
+      "learning_rate": 0.003,
+      "loss": 4.1555,
+      "step": 2455
+    },
+    {
+      "epoch": 0.02456,
+      "grad_norm": 0.8111209228027585,
+      "learning_rate": 0.003,
+      "loss": 4.1205,
+      "step": 2456
+    },
+    {
+      "epoch": 0.02457,
+      "grad_norm": 0.7593406743031937,
+      "learning_rate": 0.003,
+      "loss": 4.175,
+      "step": 2457
+    },
+    {
+      "epoch": 0.02458,
+      "grad_norm": 0.6152494042836864,
+      "learning_rate": 0.003,
+      "loss": 4.1392,
+      "step": 2458
+    },
+    {
+      "epoch": 0.02459,
+      "grad_norm": 0.6726714704776762,
+      "learning_rate": 0.003,
+      "loss": 4.1704,
+      "step": 2459
+    },
+    {
+      "epoch": 0.0246,
+      "grad_norm": 0.6849708403280691,
+      "learning_rate": 0.003,
+      "loss": 4.1742,
+      "step": 2460
+    },
+    {
+      "epoch": 0.02461,
+      "grad_norm": 0.6926478544751591,
+      "learning_rate": 0.003,
+      "loss": 4.1489,
+      "step": 2461
+    },
+    {
+      "epoch": 0.02462,
+      "grad_norm": 0.6546290518877314,
+      "learning_rate": 0.003,
+      "loss": 4.1745,
+      "step": 2462
+    },
+    {
+      "epoch": 0.02463,
+      "grad_norm": 0.6160770029093054,
+      "learning_rate": 0.003,
+      "loss": 4.1691,
+      "step": 2463
+    },
+    {
+      "epoch": 0.02464,
+      "grad_norm": 0.5990408684539491,
+      "learning_rate": 0.003,
+      "loss": 4.1487,
+      "step": 2464
+    },
+    {
+      "epoch": 0.02465,
+      "grad_norm": 0.5034148559800348,
+      "learning_rate": 0.003,
+      "loss": 4.1619,
+      "step": 2465
+    },
+    {
+      "epoch": 0.02466,
+      "grad_norm": 0.4860347114545619,
+      "learning_rate": 0.003,
+      "loss": 4.1408,
+      "step": 2466
+    },
+    {
+      "epoch": 0.02467,
+      "grad_norm": 0.4781136004723512,
+      "learning_rate": 0.003,
+      "loss": 4.1183,
+      "step": 2467
+    },
+    {
+      "epoch": 0.02468,
+      "grad_norm": 0.4265902063645912,
+      "learning_rate": 0.003,
+      "loss": 4.1584,
+      "step": 2468
+    },
+    {
+      "epoch": 0.02469,
+      "grad_norm": 0.5508562887056804,
+      "learning_rate": 0.003,
+      "loss": 4.1409,
+      "step": 2469
+    },
+    {
+      "epoch": 0.0247,
+      "grad_norm": 0.8249536754461567,
+      "learning_rate": 0.003,
+      "loss": 4.1452,
+      "step": 2470
+    },
+    {
+      "epoch": 0.02471,
+      "grad_norm": 1.4359640347930054,
+      "learning_rate": 0.003,
+      "loss": 4.1561,
+      "step": 2471
+    },
+    {
+      "epoch": 0.02472,
+      "grad_norm": 0.6769644869683711,
+      "learning_rate": 0.003,
+      "loss": 4.1433,
+      "step": 2472
+    },
+    {
+      "epoch": 0.02473,
+      "grad_norm": 0.7658209871323785,
+      "learning_rate": 0.003,
+      "loss": 4.1289,
+      "step": 2473
+    },
+    {
+      "epoch": 0.02474,
+      "grad_norm": 0.9511612267134097,
+      "learning_rate": 0.003,
+      "loss": 4.1669,
+      "step": 2474
+    },
+    {
+      "epoch": 0.02475,
+      "grad_norm": 0.8892227903889872,
+      "learning_rate": 0.003,
+      "loss": 4.1391,
+      "step": 2475
+    },
+    {
+      "epoch": 0.02476,
+      "grad_norm": 0.9920143400330694,
+      "learning_rate": 0.003,
+      "loss": 4.152,
+      "step": 2476
+    },
+    {
+      "epoch": 0.02477,
+      "grad_norm": 0.9006606413119249,
+      "learning_rate": 0.003,
+      "loss": 4.1748,
+      "step": 2477
+    },
+    {
+      "epoch": 0.02478,
+      "grad_norm": 0.8088690703733978,
+      "learning_rate": 0.003,
+      "loss": 4.1471,
+      "step": 2478
+    },
+    {
+      "epoch": 0.02479,
+      "grad_norm": 0.8182856999501472,
+      "learning_rate": 0.003,
+      "loss": 4.1366,
+      "step": 2479
+    },
+    {
+      "epoch": 0.0248,
+      "grad_norm": 0.7355042387960451,
+      "learning_rate": 0.003,
+      "loss": 4.178,
+      "step": 2480
+    },
+    {
+      "epoch": 0.02481,
+      "grad_norm": 0.6962248057182712,
+      "learning_rate": 0.003,
+      "loss": 4.1578,
+      "step": 2481
+    },
+    {
+      "epoch": 0.02482,
+      "grad_norm": 0.6352228705173556,
+      "learning_rate": 0.003,
+      "loss": 4.1666,
+      "step": 2482
+    },
+    {
+      "epoch": 0.02483,
+      "grad_norm": 0.6906094612236757,
+      "learning_rate": 0.003,
+      "loss": 4.1183,
+      "step": 2483
+    },
+    {
+      "epoch": 0.02484,
+      "grad_norm": 0.66658401463766,
+      "learning_rate": 0.003,
+      "loss": 4.1474,
+      "step": 2484
+    },
+    {
+      "epoch": 0.02485,
+      "grad_norm": 0.6596750859147653,
+      "learning_rate": 0.003,
+      "loss": 4.1705,
+      "step": 2485
+    },
+    {
+      "epoch": 0.02486,
+      "grad_norm": 0.7066148867713012,
+      "learning_rate": 0.003,
+      "loss": 4.1929,
+      "step": 2486
+    },
+    {
+      "epoch": 0.02487,
+      "grad_norm": 0.754767630208959,
+      "learning_rate": 0.003,
+      "loss": 4.1498,
+      "step": 2487
+    },
+    {
+      "epoch": 0.02488,
+      "grad_norm": 0.7178651369815516,
+      "learning_rate": 0.003,
+      "loss": 4.1276,
+      "step": 2488
+    },
+    {
+      "epoch": 0.02489,
+      "grad_norm": 0.7388633013067267,
+      "learning_rate": 0.003,
+      "loss": 4.1159,
+      "step": 2489
+    },
+    {
+      "epoch": 0.0249,
+      "grad_norm": 0.7802405314573639,
+      "learning_rate": 0.003,
+      "loss": 4.1239,
+      "step": 2490
+    },
+    {
+      "epoch": 0.02491,
+      "grad_norm": 0.6722919878141813,
+      "learning_rate": 0.003,
+      "loss": 4.147,
+      "step": 2491
+    },
+    {
+      "epoch": 0.02492,
+      "grad_norm": 0.6231446299505073,
+      "learning_rate": 0.003,
+      "loss": 4.1546,
+      "step": 2492
+    },
+    {
+      "epoch": 0.02493,
+      "grad_norm": 0.6055678621905288,
+      "learning_rate": 0.003,
+      "loss": 4.1252,
+      "step": 2493
+    },
+    {
+      "epoch": 0.02494,
+      "grad_norm": 0.6270135330055409,
+      "learning_rate": 0.003,
+      "loss": 4.1374,
+      "step": 2494
+    },
+    {
+      "epoch": 0.02495,
+      "grad_norm": 0.6525517462302536,
+      "learning_rate": 0.003,
+      "loss": 4.1487,
+      "step": 2495
+    },
+    {
+      "epoch": 0.02496,
+      "grad_norm": 0.7329319033542416,
+      "learning_rate": 0.003,
+      "loss": 4.14,
+      "step": 2496
+    },
+    {
+      "epoch": 0.02497,
+      "grad_norm": 0.9619996867887689,
+      "learning_rate": 0.003,
+      "loss": 4.1515,
+      "step": 2497
+    },
+    {
+      "epoch": 0.02498,
+      "grad_norm": 1.141814403141371,
+      "learning_rate": 0.003,
+      "loss": 4.1545,
+      "step": 2498
+    },
+    {
+      "epoch": 0.02499,
+      "grad_norm": 0.9016540268347464,
+      "learning_rate": 0.003,
+      "loss": 4.1556,
+      "step": 2499
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 0.9065801171881926,
+      "learning_rate": 0.003,
+      "loss": 4.1446,
+      "step": 2500
+    },
+    {
+      "epoch": 0.02501,
+      "grad_norm": 0.8844338776736929,
+      "learning_rate": 0.003,
+      "loss": 4.1765,
+      "step": 2501
+    },
+    {
+      "epoch": 0.02502,
+      "grad_norm": 0.8046602370535848,
+      "learning_rate": 0.003,
+      "loss": 4.1579,
+      "step": 2502
+    },
+    {
+      "epoch": 0.02503,
+      "grad_norm": 0.7039592913476108,
+      "learning_rate": 0.003,
+      "loss": 4.1643,
+      "step": 2503
+    },
+    {
+      "epoch": 0.02504,
+      "grad_norm": 0.7419132370317111,
+      "learning_rate": 0.003,
+      "loss": 4.1429,
+      "step": 2504
+    },
+    {
+      "epoch": 0.02505,
+      "grad_norm": 0.7606498464541366,
+      "learning_rate": 0.003,
+      "loss": 4.1455,
+      "step": 2505
+    },
+    {
+      "epoch": 0.02506,
+      "grad_norm": 0.769560412914919,
+      "learning_rate": 0.003,
+      "loss": 4.1579,
+      "step": 2506
+    },
+    {
+      "epoch": 0.02507,
+      "grad_norm": 0.8361991263151639,
+      "learning_rate": 0.003,
+      "loss": 4.1464,
+      "step": 2507
+    },
+    {
+      "epoch": 0.02508,
+      "grad_norm": 1.0314436698214402,
+      "learning_rate": 0.003,
+      "loss": 4.1552,
+      "step": 2508
+    },
+    {
+      "epoch": 0.02509,
+      "grad_norm": 1.0808307469983176,
+      "learning_rate": 0.003,
+      "loss": 4.1638,
+      "step": 2509
+    },
+    {
+      "epoch": 0.0251,
+      "grad_norm": 0.9756326578065097,
+      "learning_rate": 0.003,
+      "loss": 4.1772,
+      "step": 2510
+    },
+    {
+      "epoch": 0.02511,
+      "grad_norm": 0.9782444756269447,
+      "learning_rate": 0.003,
+      "loss": 4.1853,
+      "step": 2511
+    },
+    {
+      "epoch": 0.02512,
+      "grad_norm": 0.9679666828238646,
+      "learning_rate": 0.003,
+      "loss": 4.1672,
+      "step": 2512
+    },
+    {
+      "epoch": 0.02513,
+      "grad_norm": 0.9141144321737233,
+      "learning_rate": 0.003,
+      "loss": 4.1647,
+      "step": 2513
+    },
+    {
+      "epoch": 0.02514,
+      "grad_norm": 0.870311596191751,
+      "learning_rate": 0.003,
+      "loss": 4.1521,
+      "step": 2514
+    },
+    {
+      "epoch": 0.02515,
+      "grad_norm": 0.718099567068549,
+      "learning_rate": 0.003,
+      "loss": 4.134,
+      "step": 2515
+    },
+    {
+      "epoch": 0.02516,
+      "grad_norm": 0.6485758792814954,
+      "learning_rate": 0.003,
+      "loss": 4.153,
+      "step": 2516
+    },
+    {
+      "epoch": 0.02517,
+      "grad_norm": 0.6631125888262853,
+      "learning_rate": 0.003,
+      "loss": 4.1528,
+      "step": 2517
+    },
+    {
+      "epoch": 0.02518,
+      "grad_norm": 0.62508603256109,
+      "learning_rate": 0.003,
+      "loss": 4.1629,
+      "step": 2518
+    },
+    {
+      "epoch": 0.02519,
+      "grad_norm": 0.6533648513332591,
+      "learning_rate": 0.003,
+      "loss": 4.1392,
+      "step": 2519
+    },
+    {
+      "epoch": 0.0252,
+      "grad_norm": 0.8730883164618287,
+      "learning_rate": 0.003,
+      "loss": 4.1603,
+      "step": 2520
+    },
+    {
+      "epoch": 0.02521,
+      "grad_norm": 1.1508715959612579,
+      "learning_rate": 0.003,
+      "loss": 4.1617,
+      "step": 2521
+    },
+    {
+      "epoch": 0.02522,
+      "grad_norm": 0.8851560060677419,
+      "learning_rate": 0.003,
+      "loss": 4.1564,
+      "step": 2522
+    },
+    {
+      "epoch": 0.02523,
+      "grad_norm": 0.7609503224943213,
+      "learning_rate": 0.003,
+      "loss": 4.1544,
+      "step": 2523
+    },
+    {
+      "epoch": 0.02524,
+      "grad_norm": 0.7874722612084045,
+      "learning_rate": 0.003,
+      "loss": 4.1632,
+      "step": 2524
+    },
+    {
+      "epoch": 0.02525,
+      "grad_norm": 0.9497267581795765,
+      "learning_rate": 0.003,
+      "loss": 4.1834,
+      "step": 2525
+    },
+    {
+      "epoch": 0.02526,
+      "grad_norm": 1.1487960136042998,
+      "learning_rate": 0.003,
+      "loss": 4.1381,
+      "step": 2526
+    },
+    {
+      "epoch": 0.02527,
+      "grad_norm": 0.749631180786954,
+      "learning_rate": 0.003,
+      "loss": 4.1602,
+      "step": 2527
+    },
+    {
+      "epoch": 0.02528,
+      "grad_norm": 0.6717432753530606,
+      "learning_rate": 0.003,
+      "loss": 4.1816,
+      "step": 2528
+    },
+    {
+      "epoch": 0.02529,
+      "grad_norm": 0.6421573788772453,
+      "learning_rate": 0.003,
+      "loss": 4.1481,
+      "step": 2529
+    },
+    {
+      "epoch": 0.0253,
+      "grad_norm": 0.7486612963376212,
+      "learning_rate": 0.003,
+      "loss": 4.1689,
+      "step": 2530
+    },
+    {
+      "epoch": 0.02531,
+      "grad_norm": 0.7949983435688287,
+      "learning_rate": 0.003,
+      "loss": 4.171,
+      "step": 2531
+    },
+    {
+      "epoch": 0.02532,
+      "grad_norm": 0.7618621411023159,
+      "learning_rate": 0.003,
+      "loss": 4.1364,
+      "step": 2532
+    },
+    {
+      "epoch": 0.02533,
+      "grad_norm": 0.8123389715874717,
+      "learning_rate": 0.003,
+      "loss": 4.1773,
+      "step": 2533
+    },
+    {
+      "epoch": 0.02534,
+      "grad_norm": 0.808321322590649,
+      "learning_rate": 0.003,
+      "loss": 4.1573,
+      "step": 2534
+    },
+    {
+      "epoch": 0.02535,
+      "grad_norm": 0.7051467031412186,
+      "learning_rate": 0.003,
+      "loss": 4.1477,
+      "step": 2535
+    },
+    {
+      "epoch": 0.02536,
+      "grad_norm": 0.6093069610094953,
+      "learning_rate": 0.003,
+      "loss": 4.142,
+      "step": 2536
+    },
+    {
+      "epoch": 0.02537,
+      "grad_norm": 0.5843395517845005,
+      "learning_rate": 0.003,
+      "loss": 4.1344,
+      "step": 2537
+    },
+    {
+      "epoch": 0.02538,
+      "grad_norm": 0.5674485238862751,
+      "learning_rate": 0.003,
+      "loss": 4.1098,
+      "step": 2538
+    },
+    {
+      "epoch": 0.02539,
+      "grad_norm": 0.5652766007864589,
+      "learning_rate": 0.003,
+      "loss": 4.1606,
+      "step": 2539
+    },
+    {
+      "epoch": 0.0254,
+      "grad_norm": 0.6570254628858063,
+      "learning_rate": 0.003,
+      "loss": 4.1514,
+      "step": 2540
+    },
+    {
+      "epoch": 0.02541,
+      "grad_norm": 0.8206242142708078,
+      "learning_rate": 0.003,
+      "loss": 4.1443,
+      "step": 2541
+    },
+    {
+      "epoch": 0.02542,
+      "grad_norm": 0.9677168967962686,
+      "learning_rate": 0.003,
+      "loss": 4.134,
+      "step": 2542
+    },
+    {
+      "epoch": 0.02543,
+      "grad_norm": 1.024805521305044,
+      "learning_rate": 0.003,
+      "loss": 4.1646,
+      "step": 2543
+    },
+    {
+      "epoch": 0.02544,
+      "grad_norm": 0.7591855178067561,
+      "learning_rate": 0.003,
+      "loss": 4.148,
+      "step": 2544
+    },
+    {
+      "epoch": 0.02545,
+      "grad_norm": 0.6449474624868338,
+      "learning_rate": 0.003,
+      "loss": 4.1255,
+      "step": 2545
+    },
+    {
+      "epoch": 0.02546,
+      "grad_norm": 0.6980586139707194,
+      "learning_rate": 0.003,
+      "loss": 4.136,
+      "step": 2546
+    },
+    {
+      "epoch": 0.02547,
+      "grad_norm": 0.7283106303300475,
+      "learning_rate": 0.003,
+      "loss": 4.1754,
+      "step": 2547
+    },
+    {
+      "epoch": 0.02548,
+      "grad_norm": 0.6850467395404413,
+      "learning_rate": 0.003,
+      "loss": 4.1691,
+      "step": 2548
+    },
+    {
+      "epoch": 0.02549,
+      "grad_norm": 0.7158743790784491,
+      "learning_rate": 0.003,
+      "loss": 4.1632,
+      "step": 2549
+    },
+    {
+      "epoch": 0.0255,
+      "grad_norm": 0.799200990441851,
+      "learning_rate": 0.003,
+      "loss": 4.1303,
+      "step": 2550
+    },
+    {
+      "epoch": 0.02551,
+      "grad_norm": 0.8852033835267075,
+      "learning_rate": 0.003,
+      "loss": 4.165,
+      "step": 2551
+    },
+    {
+      "epoch": 0.02552,
+      "grad_norm": 0.9381299460340671,
+      "learning_rate": 0.003,
+      "loss": 4.1394,
+      "step": 2552
+    },
+    {
+      "epoch": 0.02553,
+      "grad_norm": 0.8787078285398282,
+      "learning_rate": 0.003,
+      "loss": 4.2019,
+      "step": 2553
+    },
+    {
+      "epoch": 0.02554,
+      "grad_norm": 1.0064349811006572,
+      "learning_rate": 0.003,
+      "loss": 4.1601,
+      "step": 2554
+    },
+    {
+      "epoch": 0.02555,
+      "grad_norm": 0.8623285229523936,
+      "learning_rate": 0.003,
+      "loss": 4.1454,
+      "step": 2555
+    },
+    {
+      "epoch": 0.02556,
+      "grad_norm": 0.779791532210884,
+      "learning_rate": 0.003,
+      "loss": 4.154,
+      "step": 2556
+    },
+    {
+      "epoch": 0.02557,
+      "grad_norm": 0.8041071030856266,
+      "learning_rate": 0.003,
+      "loss": 4.1561,
+      "step": 2557
+    },
+    {
+      "epoch": 0.02558,
+      "grad_norm": 0.9046731898466632,
+      "learning_rate": 0.003,
+      "loss": 4.1797,
+      "step": 2558
+    },
+    {
+      "epoch": 0.02559,
+      "grad_norm": 0.9231060963177342,
+      "learning_rate": 0.003,
+      "loss": 4.1753,
+      "step": 2559
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 0.7942918059321874,
+      "learning_rate": 0.003,
+      "loss": 4.1509,
+      "step": 2560
+    },
+    {
+      "epoch": 0.02561,
+      "grad_norm": 0.7182994269745618,
+      "learning_rate": 0.003,
+      "loss": 4.1458,
+      "step": 2561
+    },
+    {
+      "epoch": 0.02562,
+      "grad_norm": 0.7377092426971169,
+      "learning_rate": 0.003,
+      "loss": 4.1635,
+      "step": 2562
+    },
+    {
+      "epoch": 0.02563,
+      "grad_norm": 0.8008807069379792,
+      "learning_rate": 0.003,
+      "loss": 4.1444,
+      "step": 2563
+    },
+    {
+      "epoch": 0.02564,
+      "grad_norm": 0.9654334894945076,
+      "learning_rate": 0.003,
+      "loss": 4.1721,
+      "step": 2564
+    },
+    {
+      "epoch": 0.02565,
+      "grad_norm": 1.1270177489997797,
+      "learning_rate": 0.003,
+      "loss": 4.1607,
+      "step": 2565
+    },
+    {
+      "epoch": 0.02566,
+      "grad_norm": 0.7790702826784099,
+      "learning_rate": 0.003,
+      "loss": 4.1704,
+      "step": 2566
+    },
+    {
+      "epoch": 0.02567,
+      "grad_norm": 0.7253981064969616,
+      "learning_rate": 0.003,
+      "loss": 4.1684,
+      "step": 2567
+    },
+    {
+      "epoch": 0.02568,
+      "grad_norm": 0.7766989633341348,
+      "learning_rate": 0.003,
+      "loss": 4.1634,
+      "step": 2568
+    },
+    {
+      "epoch": 0.02569,
+      "grad_norm": 0.6753943323614341,
+      "learning_rate": 0.003,
+      "loss": 4.1406,
+      "step": 2569
+    },
+    {
+      "epoch": 0.0257,
+      "grad_norm": 0.6053036326649146,
+      "learning_rate": 0.003,
+      "loss": 4.142,
+      "step": 2570
+    },
+    {
+      "epoch": 0.02571,
+      "grad_norm": 0.5412545714840278,
+      "learning_rate": 0.003,
+      "loss": 4.1611,
+      "step": 2571
+    },
+    {
+      "epoch": 0.02572,
+      "grad_norm": 0.618814983550634,
+      "learning_rate": 0.003,
+      "loss": 4.1399,
+      "step": 2572
+    },
+    {
+      "epoch": 0.02573,
+      "grad_norm": 0.6647990779688017,
+      "learning_rate": 0.003,
+      "loss": 4.1458,
+      "step": 2573
+    },
+    {
+      "epoch": 0.02574,
+      "grad_norm": 0.7052848934500006,
+      "learning_rate": 0.003,
+      "loss": 4.1737,
+      "step": 2574
+    },
+    {
+      "epoch": 0.02575,
+      "grad_norm": 0.7610884337795838,
+      "learning_rate": 0.003,
+      "loss": 4.1101,
+      "step": 2575
+    },
+    {
+      "epoch": 0.02576,
+      "grad_norm": 0.6692271813839591,
+      "learning_rate": 0.003,
+      "loss": 4.1159,
+      "step": 2576
+    },
+    {
+      "epoch": 0.02577,
+      "grad_norm": 0.5756800187751938,
+      "learning_rate": 0.003,
+      "loss": 4.1153,
+      "step": 2577
+    },
+    {
+      "epoch": 0.02578,
+      "grad_norm": 0.5445744087322336,
+      "learning_rate": 0.003,
+      "loss": 4.1401,
+      "step": 2578
+    },
+    {
+      "epoch": 0.02579,
+      "grad_norm": 0.5526066768211919,
+      "learning_rate": 0.003,
+      "loss": 4.0838,
+      "step": 2579
+    },
+    {
+      "epoch": 0.0258,
+      "grad_norm": 0.555186498872379,
+      "learning_rate": 0.003,
+      "loss": 4.1418,
+      "step": 2580
+    },
+    {
+      "epoch": 0.02581,
+      "grad_norm": 0.590974511132741,
+      "learning_rate": 0.003,
+      "loss": 4.1622,
+      "step": 2581
+    },
+    {
+      "epoch": 0.02582,
+      "grad_norm": 0.6134714053872987,
+      "learning_rate": 0.003,
+      "loss": 4.157,
+      "step": 2582
+    },
+    {
+      "epoch": 0.02583,
+      "grad_norm": 0.6785456864342533,
+      "learning_rate": 0.003,
+      "loss": 4.1327,
+      "step": 2583
+    },
+    {
+      "epoch": 0.02584,
+      "grad_norm": 0.7692608608699008,
+      "learning_rate": 0.003,
+      "loss": 4.1476,
+      "step": 2584
+    },
+    {
+      "epoch": 0.02585,
+      "grad_norm": 0.8096555829199314,
+      "learning_rate": 0.003,
+      "loss": 4.1338,
+      "step": 2585
+    },
+    {
+      "epoch": 0.02586,
+      "grad_norm": 0.8850658157806096,
+      "learning_rate": 0.003,
+      "loss": 4.1563,
+      "step": 2586
+    },
+    {
+      "epoch": 0.02587,
+      "grad_norm": 1.0102418517134517,
+      "learning_rate": 0.003,
+      "loss": 4.1487,
+      "step": 2587
+    },
+    {
+      "epoch": 0.02588,
+      "grad_norm": 0.9275963803575394,
+      "learning_rate": 0.003,
+      "loss": 4.1505,
+      "step": 2588
+    },
+    {
+      "epoch": 0.02589,
+      "grad_norm": 0.9403619185225363,
+      "learning_rate": 0.003,
+      "loss": 4.1571,
+      "step": 2589
+    },
+    {
+      "epoch": 0.0259,
+      "grad_norm": 1.0078495428012795,
+      "learning_rate": 0.003,
+      "loss": 4.1693,
+      "step": 2590
+    },
+    {
+      "epoch": 0.02591,
+      "grad_norm": 1.2743604633788357,
+      "learning_rate": 0.003,
+      "loss": 4.1191,
+      "step": 2591
+    },
+    {
+      "epoch": 0.02592,
+      "grad_norm": 0.8231355726319857,
+      "learning_rate": 0.003,
+      "loss": 4.155,
+      "step": 2592
+    },
+    {
+      "epoch": 0.02593,
+      "grad_norm": 0.6890041984971179,
+      "learning_rate": 0.003,
+      "loss": 4.1511,
+      "step": 2593
+    },
+    {
+      "epoch": 0.02594,
+      "grad_norm": 0.7454180549344623,
+      "learning_rate": 0.003,
+      "loss": 4.1728,
+      "step": 2594
+    },
+    {
+      "epoch": 0.02595,
+      "grad_norm": 0.6902489836421277,
+      "learning_rate": 0.003,
+      "loss": 4.1203,
+      "step": 2595
+    },
+    {
+      "epoch": 0.02596,
+      "grad_norm": 0.8183952841129141,
+      "learning_rate": 0.003,
+      "loss": 4.1633,
+      "step": 2596
+    },
+    {
+      "epoch": 0.02597,
+      "grad_norm": 0.9211443305771235,
+      "learning_rate": 0.003,
+      "loss": 4.1359,
+      "step": 2597
+    },
+    {
+      "epoch": 0.02598,
+      "grad_norm": 1.0292390798108857,
+      "learning_rate": 0.003,
+      "loss": 4.1827,
+      "step": 2598
+    },
+    {
+      "epoch": 0.02599,
+      "grad_norm": 0.9373288426158787,
+      "learning_rate": 0.003,
+      "loss": 4.1706,
+      "step": 2599
+    },
+    {
+      "epoch": 0.026,
+      "grad_norm": 0.920433494603301,
+      "learning_rate": 0.003,
+      "loss": 4.1747,
+      "step": 2600
+    },
+    {
+      "epoch": 0.02601,
+      "grad_norm": 0.8451864765965912,
+      "learning_rate": 0.003,
+      "loss": 4.1776,
+      "step": 2601
+    },
+    {
+      "epoch": 0.02602,
+      "grad_norm": 0.7170319795923626,
+      "learning_rate": 0.003,
+      "loss": 4.1455,
+      "step": 2602
+    },
+    {
+      "epoch": 0.02603,
+      "grad_norm": 0.6966464803396121,
+      "learning_rate": 0.003,
+      "loss": 4.1401,
+      "step": 2603
+    },
+    {
+      "epoch": 0.02604,
+      "grad_norm": 0.7612566166038212,
+      "learning_rate": 0.003,
+      "loss": 4.1308,
+      "step": 2604
+    },
+    {
+      "epoch": 0.02605,
+      "grad_norm": 0.6464701620756803,
+      "learning_rate": 0.003,
+      "loss": 4.1437,
+      "step": 2605
+    },
+    {
+      "epoch": 0.02606,
+      "grad_norm": 0.5737007841123123,
+      "learning_rate": 0.003,
+      "loss": 4.162,
+      "step": 2606
+    },
+    {
+      "epoch": 0.02607,
+      "grad_norm": 0.6131766565865462,
+      "learning_rate": 0.003,
+      "loss": 4.1509,
+      "step": 2607
+    },
+    {
+      "epoch": 0.02608,
+      "grad_norm": 0.8086035003985383,
+      "learning_rate": 0.003,
+      "loss": 4.136,
+      "step": 2608
+    },
+    {
+      "epoch": 0.02609,
+      "grad_norm": 0.9817909383598199,
+      "learning_rate": 0.003,
+      "loss": 4.1914,
+      "step": 2609
+    },
+    {
+      "epoch": 0.0261,
+      "grad_norm": 1.0720508985841146,
+      "learning_rate": 0.003,
+      "loss": 4.1421,
+      "step": 2610
+    },
+    {
+      "epoch": 0.02611,
+      "grad_norm": 0.8659326470213682,
+      "learning_rate": 0.003,
+      "loss": 4.1249,
+      "step": 2611
+    },
+    {
+      "epoch": 0.02612,
+      "grad_norm": 0.8732525902579722,
+      "learning_rate": 0.003,
+      "loss": 4.1648,
+      "step": 2612
+    },
+    {
+      "epoch": 0.02613,
+      "grad_norm": 0.9874637099516153,
+      "learning_rate": 0.003,
+      "loss": 4.1605,
+      "step": 2613
+    },
+    {
+      "epoch": 0.02614,
+      "grad_norm": 0.906401203843813,
+      "learning_rate": 0.003,
+      "loss": 4.1404,
+      "step": 2614
+    },
+    {
+      "epoch": 0.02615,
+      "grad_norm": 0.6462260762478055,
+      "learning_rate": 0.003,
+      "loss": 4.1238,
+      "step": 2615
+    },
+    {
+      "epoch": 0.02616,
+      "grad_norm": 0.6006231478500065,
+      "learning_rate": 0.003,
+      "loss": 4.1352,
+      "step": 2616
+    },
+    {
+      "epoch": 0.02617,
+      "grad_norm": 0.6102511179138815,
+      "learning_rate": 0.003,
+      "loss": 4.1214,
+      "step": 2617
+    },
+    {
+      "epoch": 0.02618,
+      "grad_norm": 0.6619768894826954,
+      "learning_rate": 0.003,
+      "loss": 4.15,
+      "step": 2618
+    },
+    {
+      "epoch": 0.02619,
+      "grad_norm": 0.7040985692908992,
+      "learning_rate": 0.003,
+      "loss": 4.1627,
+      "step": 2619
+    },
+    {
+      "epoch": 0.0262,
+      "grad_norm": 0.7390370888569476,
+      "learning_rate": 0.003,
+      "loss": 4.1255,
+      "step": 2620
+    },
+    {
+      "epoch": 0.02621,
+      "grad_norm": 0.8595720375733896,
+      "learning_rate": 0.003,
+      "loss": 4.1547,
+      "step": 2621
+    },
+    {
+      "epoch": 0.02622,
+      "grad_norm": 0.8747740097574618,
+      "learning_rate": 0.003,
+      "loss": 4.1571,
+      "step": 2622
+    },
+    {
+      "epoch": 0.02623,
+      "grad_norm": 0.7865985620123697,
+      "learning_rate": 0.003,
+      "loss": 4.1577,
+      "step": 2623
+    },
+    {
+      "epoch": 0.02624,
+      "grad_norm": 0.721745866206175,
+      "learning_rate": 0.003,
+      "loss": 4.1924,
+      "step": 2624
+    },
+    {
+      "epoch": 0.02625,
+      "grad_norm": 0.6369082515266815,
+      "learning_rate": 0.003,
+      "loss": 4.1844,
+      "step": 2625
+    },
+    {
+      "epoch": 0.02626,
+      "grad_norm": 0.6738431853903379,
+      "learning_rate": 0.003,
+      "loss": 4.1204,
+      "step": 2626
+    },
+    {
+      "epoch": 0.02627,
+      "grad_norm": 0.8070717121127621,
+      "learning_rate": 0.003,
+      "loss": 4.1402,
+      "step": 2627
+    },
+    {
+      "epoch": 0.02628,
+      "grad_norm": 0.9052615625467546,
+      "learning_rate": 0.003,
+      "loss": 4.1733,
+      "step": 2628
+    },
+    {
+      "epoch": 0.02629,
+      "grad_norm": 0.9064578842099213,
+      "learning_rate": 0.003,
+      "loss": 4.112,
+      "step": 2629
+    },
+    {
+      "epoch": 0.0263,
+      "grad_norm": 0.9572710825858507,
+      "learning_rate": 0.003,
+      "loss": 4.1442,
+      "step": 2630
+    },
+    {
+      "epoch": 0.02631,
+      "grad_norm": 0.9098415499268271,
+      "learning_rate": 0.003,
+      "loss": 4.1702,
+      "step": 2631
+    },
+    {
+      "epoch": 0.02632,
+      "grad_norm": 0.957961730332363,
+      "learning_rate": 0.003,
+      "loss": 4.1333,
+      "step": 2632
+    },
+    {
+      "epoch": 0.02633,
+      "grad_norm": 0.8910675740871853,
+      "learning_rate": 0.003,
+      "loss": 4.1452,
+      "step": 2633
+    },
+    {
+      "epoch": 0.02634,
+      "grad_norm": 0.8894027249202754,
+      "learning_rate": 0.003,
+      "loss": 4.1657,
+      "step": 2634
+    },
+    {
+      "epoch": 0.02635,
+      "grad_norm": 1.0919965764145694,
+      "learning_rate": 0.003,
+      "loss": 4.171,
+      "step": 2635
+    },
+    {
+      "epoch": 0.02636,
+      "grad_norm": 1.101775206670289,
+      "learning_rate": 0.003,
+      "loss": 4.1603,
+      "step": 2636
+    },
+    {
+      "epoch": 0.02637,
+      "grad_norm": 0.9585912224073222,
+      "learning_rate": 0.003,
+      "loss": 4.1888,
+      "step": 2637
+    },
+    {
+      "epoch": 0.02638,
+      "grad_norm": 0.7865200036359159,
+      "learning_rate": 0.003,
+      "loss": 4.1464,
+      "step": 2638
+    },
+    {
+      "epoch": 0.02639,
+      "grad_norm": 0.7526947533075404,
+      "learning_rate": 0.003,
+      "loss": 4.1366,
+      "step": 2639
+    },
+    {
+      "epoch": 0.0264,
+      "grad_norm": 0.8605449721173474,
+      "learning_rate": 0.003,
+      "loss": 4.134,
+      "step": 2640
+    },
+    {
+      "epoch": 0.02641,
+      "grad_norm": 0.9267463548326204,
+      "learning_rate": 0.003,
+      "loss": 4.1502,
+      "step": 2641
+    },
+    {
+      "epoch": 0.02642,
+      "grad_norm": 0.9486962646703484,
+      "learning_rate": 0.003,
+      "loss": 4.1611,
+      "step": 2642
+    },
+    {
+      "epoch": 0.02643,
+      "grad_norm": 0.8844618097245053,
+      "learning_rate": 0.003,
+      "loss": 4.15,
+      "step": 2643
+    },
+    {
+      "epoch": 0.02644,
+      "grad_norm": 0.7967251454981952,
+      "learning_rate": 0.003,
+      "loss": 4.1599,
+      "step": 2644
+    },
+    {
+      "epoch": 0.02645,
+      "grad_norm": 0.8225181053423539,
+      "learning_rate": 0.003,
+      "loss": 4.1718,
+      "step": 2645
+    },
+    {
+      "epoch": 0.02646,
+      "grad_norm": 0.7987479226871795,
+      "learning_rate": 0.003,
+      "loss": 4.1374,
+      "step": 2646
+    },
+    {
+      "epoch": 0.02647,
+      "grad_norm": 0.7821642664442445,
+      "learning_rate": 0.003,
+      "loss": 4.1612,
+      "step": 2647
+    },
+    {
+      "epoch": 0.02648,
+      "grad_norm": 0.9284024850106564,
+      "learning_rate": 0.003,
+      "loss": 4.1813,
+      "step": 2648
+    },
+    {
+      "epoch": 0.02649,
+      "grad_norm": 1.0907008227907322,
+      "learning_rate": 0.003,
+      "loss": 4.1777,
+      "step": 2649
+    },
+    {
+      "epoch": 0.0265,
+      "grad_norm": 1.125793480123405,
+      "learning_rate": 0.003,
+      "loss": 4.1386,
+      "step": 2650
+    },
+    {
+      "epoch": 0.02651,
+      "grad_norm": 0.7578496969583901,
+      "learning_rate": 0.003,
+      "loss": 4.1454,
+      "step": 2651
+    },
+    {
+      "epoch": 0.02652,
+      "grad_norm": 0.7052794942111215,
+      "learning_rate": 0.003,
+      "loss": 4.1738,
+      "step": 2652
+    },
+    {
+      "epoch": 0.02653,
+      "grad_norm": 0.7103196739746442,
+      "learning_rate": 0.003,
+      "loss": 4.1293,
+      "step": 2653
+    },
+    {
+      "epoch": 0.02654,
+      "grad_norm": 0.6925385503676881,
+      "learning_rate": 0.003,
+      "loss": 4.1383,
+      "step": 2654
+    },
+    {
+      "epoch": 0.02655,
+      "grad_norm": 0.781411208540797,
+      "learning_rate": 0.003,
+      "loss": 4.147,
+      "step": 2655
+    },
+    {
+      "epoch": 0.02656,
+      "grad_norm": 0.7579190054446218,
+      "learning_rate": 0.003,
+      "loss": 4.1226,
+      "step": 2656
+    },
+    {
+      "epoch": 0.02657,
+      "grad_norm": 0.7277248228326177,
+      "learning_rate": 0.003,
+      "loss": 4.1484,
+      "step": 2657
+    },
+    {
+      "epoch": 0.02658,
+      "grad_norm": 0.6753248968688912,
+      "learning_rate": 0.003,
+      "loss": 4.1533,
+      "step": 2658
+    },
+    {
+      "epoch": 0.02659,
+      "grad_norm": 0.6796448265031177,
+      "learning_rate": 0.003,
+      "loss": 4.1465,
+      "step": 2659
+    },
+    {
+      "epoch": 0.0266,
+      "grad_norm": 0.6424181529879253,
+      "learning_rate": 0.003,
+      "loss": 4.1271,
+      "step": 2660
+    },
+    {
+      "epoch": 0.02661,
+      "grad_norm": 0.7766396799485783,
+      "learning_rate": 0.003,
+      "loss": 4.1524,
+      "step": 2661
+    },
+    {
+      "epoch": 0.02662,
+      "grad_norm": 0.864583375609625,
+      "learning_rate": 0.003,
+      "loss": 4.1433,
+      "step": 2662
+    },
+    {
+      "epoch": 0.02663,
+      "grad_norm": 0.8284021457779097,
+      "learning_rate": 0.003,
+      "loss": 4.1285,
+      "step": 2663
+    },
+    {
+      "epoch": 0.02664,
+      "grad_norm": 0.7120779790048899,
+      "learning_rate": 0.003,
+      "loss": 4.1258,
+      "step": 2664
+    },
+    {
+      "epoch": 0.02665,
+      "grad_norm": 0.6851733303412342,
+      "learning_rate": 0.003,
+      "loss": 4.1719,
+      "step": 2665
+    },
+    {
+      "epoch": 0.02666,
+      "grad_norm": 0.6758809967977054,
+      "learning_rate": 0.003,
+      "loss": 4.1627,
+      "step": 2666
+    },
+    {
+      "epoch": 0.02667,
+      "grad_norm": 0.6361286175404186,
+      "learning_rate": 0.003,
+      "loss": 4.1421,
+      "step": 2667
+    },
+    {
+      "epoch": 0.02668,
+      "grad_norm": 0.5673144703469689,
+      "learning_rate": 0.003,
+      "loss": 4.1084,
+      "step": 2668
+    },
+    {
+      "epoch": 0.02669,
+      "grad_norm": 0.5094068244293505,
+      "learning_rate": 0.003,
+      "loss": 4.1289,
+      "step": 2669
+    },
+    {
+      "epoch": 0.0267,
+      "grad_norm": 0.5324820431990631,
+      "learning_rate": 0.003,
+      "loss": 4.1272,
+      "step": 2670
+    },
+    {
+      "epoch": 0.02671,
+      "grad_norm": 0.6250352979316671,
+      "learning_rate": 0.003,
+      "loss": 4.1447,
+      "step": 2671
+    },
+    {
+      "epoch": 0.02672,
+      "grad_norm": 0.6359363422291939,
+      "learning_rate": 0.003,
+      "loss": 4.1384,
+      "step": 2672
+    },
+    {
+      "epoch": 0.02673,
+      "grad_norm": 0.6102304515358753,
+      "learning_rate": 0.003,
+      "loss": 4.1336,
+      "step": 2673
+    },
+    {
+      "epoch": 0.02674,
+      "grad_norm": 0.7766569602853689,
+      "learning_rate": 0.003,
+      "loss": 4.1082,
+      "step": 2674
+    },
+    {
+      "epoch": 0.02675,
+      "grad_norm": 0.9731715722532488,
+      "learning_rate": 0.003,
+      "loss": 4.1138,
+      "step": 2675
+    },
+    {
+      "epoch": 0.02676,
+      "grad_norm": 1.0370179032200628,
+      "learning_rate": 0.003,
+      "loss": 4.1437,
+      "step": 2676
+    },
+    {
+      "epoch": 0.02677,
+      "grad_norm": 0.7623073096555095,
+      "learning_rate": 0.003,
+      "loss": 4.1614,
+      "step": 2677
+    },
+    {
+      "epoch": 0.02678,
+      "grad_norm": 0.7777692153557114,
+      "learning_rate": 0.003,
+      "loss": 4.1586,
+      "step": 2678
+    },
+    {
+      "epoch": 0.02679,
+      "grad_norm": 0.8733324106472917,
+      "learning_rate": 0.003,
+      "loss": 4.1359,
+      "step": 2679
+    },
+    {
+      "epoch": 0.0268,
+      "grad_norm": 0.8266517145073005,
+      "learning_rate": 0.003,
+      "loss": 4.1463,
+      "step": 2680
+    },
+    {
+      "epoch": 0.02681,
+      "grad_norm": 0.758773812684903,
+      "learning_rate": 0.003,
+      "loss": 4.1643,
+      "step": 2681
+    },
+    {
+      "epoch": 0.02682,
+      "grad_norm": 0.8283252875504241,
+      "learning_rate": 0.003,
+      "loss": 4.1476,
+      "step": 2682
+    },
+    {
+      "epoch": 0.02683,
+      "grad_norm": 0.8404974594062254,
+      "learning_rate": 0.003,
+      "loss": 4.1692,
+      "step": 2683
+    },
+    {
+      "epoch": 0.02684,
+      "grad_norm": 0.818507190423211,
+      "learning_rate": 0.003,
+      "loss": 4.1427,
+      "step": 2684
+    },
+    {
+      "epoch": 0.02685,
+      "grad_norm": 0.7649702763962748,
+      "learning_rate": 0.003,
+      "loss": 4.1484,
+      "step": 2685
+    },
+    {
+      "epoch": 0.02686,
+      "grad_norm": 0.7611427665837982,
+      "learning_rate": 0.003,
+      "loss": 4.1328,
+      "step": 2686
+    },
+    {
+      "epoch": 0.02687,
+      "grad_norm": 0.9628527455690736,
+      "learning_rate": 0.003,
+      "loss": 4.1512,
+      "step": 2687
+    },
+    {
+      "epoch": 0.02688,
+      "grad_norm": 0.8564202799506135,
+      "learning_rate": 0.003,
+      "loss": 4.1277,
+      "step": 2688
+    },
+    {
+      "epoch": 0.02689,
+      "grad_norm": 0.8557838178212854,
+      "learning_rate": 0.003,
+      "loss": 4.1562,
+      "step": 2689
+    },
+    {
+      "epoch": 0.0269,
+      "grad_norm": 0.9118623509339678,
+      "learning_rate": 0.003,
+      "loss": 4.164,
+      "step": 2690
+    },
+    {
+      "epoch": 0.02691,
+      "grad_norm": 0.8301540595445771,
+      "learning_rate": 0.003,
+      "loss": 4.1545,
+      "step": 2691
+    },
+    {
+      "epoch": 0.02692,
+      "grad_norm": 0.6228920440524642,
+      "learning_rate": 0.003,
+      "loss": 4.1516,
+      "step": 2692
+    },
+    {
+      "epoch": 0.02693,
+      "grad_norm": 0.6859665731725182,
+      "learning_rate": 0.003,
+      "loss": 4.1655,
+      "step": 2693
+    },
+    {
+      "epoch": 0.02694,
+      "grad_norm": 0.7138959075984226,
+      "learning_rate": 0.003,
+      "loss": 4.1259,
+      "step": 2694
+    },
+    {
+      "epoch": 0.02695,
+      "grad_norm": 0.7527603952329793,
+      "learning_rate": 0.003,
+      "loss": 4.1438,
+      "step": 2695
+    },
+    {
+      "epoch": 0.02696,
+      "grad_norm": 0.7052829196693109,
+      "learning_rate": 0.003,
+      "loss": 4.144,
+      "step": 2696
+    },
+    {
+      "epoch": 0.02697,
+      "grad_norm": 0.6398105779035383,
+      "learning_rate": 0.003,
+      "loss": 4.1511,
+      "step": 2697
+    },
+    {
+      "epoch": 0.02698,
+      "grad_norm": 0.7194666389048138,
+      "learning_rate": 0.003,
+      "loss": 4.1368,
+      "step": 2698
+    },
+    {
+      "epoch": 0.02699,
+      "grad_norm": 0.7132487376498455,
+      "learning_rate": 0.003,
+      "loss": 4.1522,
+      "step": 2699
+    },
+    {
+      "epoch": 0.027,
+      "grad_norm": 0.656063779780492,
+      "learning_rate": 0.003,
+      "loss": 4.124,
+      "step": 2700
+    },
+    {
+      "epoch": 0.02701,
+      "grad_norm": 0.629455348788436,
+      "learning_rate": 0.003,
+      "loss": 4.1115,
+      "step": 2701
+    },
+    {
+      "epoch": 0.02702,
+      "grad_norm": 0.7466120152268881,
+      "learning_rate": 0.003,
+      "loss": 4.1658,
+      "step": 2702
+    },
+    {
+      "epoch": 0.02703,
+      "grad_norm": 0.777360526823959,
+      "learning_rate": 0.003,
+      "loss": 4.1623,
+      "step": 2703
+    },
+    {
+      "epoch": 0.02704,
+      "grad_norm": 0.8782075840804512,
+      "learning_rate": 0.003,
+      "loss": 4.11,
+      "step": 2704
+    },
+    {
+      "epoch": 0.02705,
+      "grad_norm": 0.8925534153148256,
+      "learning_rate": 0.003,
+      "loss": 4.1656,
+      "step": 2705
+    },
+    {
+      "epoch": 0.02706,
+      "grad_norm": 0.9726075369304501,
+      "learning_rate": 0.003,
+      "loss": 4.1314,
+      "step": 2706
+    },
+    {
+      "epoch": 0.02707,
+      "grad_norm": 0.9439770827114989,
+      "learning_rate": 0.003,
+      "loss": 4.1304,
+      "step": 2707
+    },
+    {
+      "epoch": 0.02708,
+      "grad_norm": 0.8038542754161909,
+      "learning_rate": 0.003,
+      "loss": 4.1613,
+      "step": 2708
+    },
+    {
+      "epoch": 0.02709,
+      "grad_norm": 0.8203620131302694,
+      "learning_rate": 0.003,
+      "loss": 4.1282,
+      "step": 2709
+    },
+    {
+      "epoch": 0.0271,
+      "grad_norm": 0.8724888263473524,
+      "learning_rate": 0.003,
+      "loss": 4.1371,
+      "step": 2710
+    },
+    {
+      "epoch": 0.02711,
+      "grad_norm": 0.8653954860429884,
+      "learning_rate": 0.003,
+      "loss": 4.1256,
+      "step": 2711
+    },
+    {
+      "epoch": 0.02712,
+      "grad_norm": 0.8020656899710369,
+      "learning_rate": 0.003,
+      "loss": 4.1343,
+      "step": 2712
+    },
+    {
+      "epoch": 0.02713,
+      "grad_norm": 0.8083253547021919,
+      "learning_rate": 0.003,
+      "loss": 4.1308,
+      "step": 2713
+    },
+    {
+      "epoch": 0.02714,
+      "grad_norm": 0.8690387355838295,
+      "learning_rate": 0.003,
+      "loss": 4.1327,
+      "step": 2714
+    },
+    {
+      "epoch": 0.02715,
+      "grad_norm": 1.0593958700668973,
+      "learning_rate": 0.003,
+      "loss": 4.1347,
+      "step": 2715
+    },
+    {
+      "epoch": 0.02716,
+      "grad_norm": 1.0574588180443436,
+      "learning_rate": 0.003,
+      "loss": 4.1532,
+      "step": 2716
+    },
+    {
+      "epoch": 0.02717,
+      "grad_norm": 0.9164750014025292,
+      "learning_rate": 0.003,
+      "loss": 4.147,
+      "step": 2717
+    },
+    {
+      "epoch": 0.02718,
+      "grad_norm": 0.9137711318115498,
+      "learning_rate": 0.003,
+      "loss": 4.1696,
+      "step": 2718
+    },
+    {
+      "epoch": 0.02719,
+      "grad_norm": 0.7747921431969758,
+      "learning_rate": 0.003,
+      "loss": 4.1371,
+      "step": 2719
+    },
+    {
+      "epoch": 0.0272,
+      "grad_norm": 0.8252071711518221,
+      "learning_rate": 0.003,
+      "loss": 4.1305,
+      "step": 2720
+    },
+    {
+      "epoch": 0.02721,
+      "grad_norm": 0.9010417786683242,
+      "learning_rate": 0.003,
+      "loss": 4.1529,
+      "step": 2721
+    },
+    {
+      "epoch": 0.02722,
+      "grad_norm": 1.0430685055558186,
+      "learning_rate": 0.003,
+      "loss": 4.1578,
+      "step": 2722
+    },
+    {
+      "epoch": 0.02723,
+      "grad_norm": 0.8338446086253062,
+      "learning_rate": 0.003,
+      "loss": 4.1591,
+      "step": 2723
+    },
+    {
+      "epoch": 0.02724,
+      "grad_norm": 0.7926892406317515,
+      "learning_rate": 0.003,
+      "loss": 4.1604,
+      "step": 2724
+    },
+    {
+      "epoch": 0.02725,
+      "grad_norm": 0.9476051066490954,
+      "learning_rate": 0.003,
+      "loss": 4.1559,
+      "step": 2725
+    },
+    {
+      "epoch": 0.02726,
+      "grad_norm": 0.9718868135312356,
+      "learning_rate": 0.003,
+      "loss": 4.1298,
+      "step": 2726
+    },
+    {
+      "epoch": 0.02727,
+      "grad_norm": 0.8794969874731017,
+      "learning_rate": 0.003,
+      "loss": 4.1606,
+      "step": 2727
+    },
+    {
+      "epoch": 0.02728,
+      "grad_norm": 0.7500620026401651,
+      "learning_rate": 0.003,
+      "loss": 4.1435,
+      "step": 2728
+    },
+    {
+      "epoch": 0.02729,
+      "grad_norm": 0.682111991166883,
+      "learning_rate": 0.003,
+      "loss": 4.172,
+      "step": 2729
+    },
+    {
+      "epoch": 0.0273,
+      "grad_norm": 0.5938416056877477,
+      "learning_rate": 0.003,
+      "loss": 4.1397,
+      "step": 2730
+    },
+    {
+      "epoch": 0.02731,
+      "grad_norm": 0.5014324118447245,
+      "learning_rate": 0.003,
+      "loss": 4.1354,
+      "step": 2731
+    },
+    {
+      "epoch": 0.02732,
+      "grad_norm": 0.5566218730286413,
+      "learning_rate": 0.003,
+      "loss": 4.1285,
+      "step": 2732
+    },
+    {
+      "epoch": 0.02733,
+      "grad_norm": 0.656514130835437,
+      "learning_rate": 0.003,
+      "loss": 4.16,
+      "step": 2733
+    },
+    {
+      "epoch": 0.02734,
+      "grad_norm": 0.7092083749191737,
+      "learning_rate": 0.003,
+      "loss": 4.1115,
+      "step": 2734
+    },
+    {
+      "epoch": 0.02735,
+      "grad_norm": 0.7490351435778966,
+      "learning_rate": 0.003,
+      "loss": 4.1581,
+      "step": 2735
+    },
+    {
+      "epoch": 0.02736,
+      "grad_norm": 0.8057410040143573,
+      "learning_rate": 0.003,
+      "loss": 4.1338,
+      "step": 2736
+    },
+    {
+      "epoch": 0.02737,
+      "grad_norm": 0.7272389058341057,
+      "learning_rate": 0.003,
+      "loss": 4.1253,
+      "step": 2737
+    },
+    {
+      "epoch": 0.02738,
+      "grad_norm": 0.7181903456270696,
+      "learning_rate": 0.003,
+      "loss": 4.1447,
+      "step": 2738
+    },
+    {
+      "epoch": 0.02739,
+      "grad_norm": 0.8790015592136898,
+      "learning_rate": 0.003,
+      "loss": 4.1427,
+      "step": 2739
+    },
+    {
+      "epoch": 0.0274,
+      "grad_norm": 1.0911668104806873,
+      "learning_rate": 0.003,
+      "loss": 4.1395,
+      "step": 2740
+    },
+    {
+      "epoch": 0.02741,
+      "grad_norm": 1.0149366352031834,
+      "learning_rate": 0.003,
+      "loss": 4.1544,
+      "step": 2741
+    },
+    {
+      "epoch": 0.02742,
+      "grad_norm": 0.8648703789133807,
+      "learning_rate": 0.003,
+      "loss": 4.1381,
+      "step": 2742
+    },
+    {
+      "epoch": 0.02743,
+      "grad_norm": 0.8509616129669461,
+      "learning_rate": 0.003,
+      "loss": 4.1613,
+      "step": 2743
+    },
+    {
+      "epoch": 0.02744,
+      "grad_norm": 0.7514267415016833,
+      "learning_rate": 0.003,
+      "loss": 4.1399,
+      "step": 2744
+    },
+    {
+      "epoch": 0.02745,
+      "grad_norm": 0.666188856053761,
+      "learning_rate": 0.003,
+      "loss": 4.1482,
+      "step": 2745
+    },
+    {
+      "epoch": 0.02746,
+      "grad_norm": 0.6204707613465867,
+      "learning_rate": 0.003,
+      "loss": 4.1227,
+      "step": 2746
+    },
+    {
+      "epoch": 0.02747,
+      "grad_norm": 0.5745978206046833,
+      "learning_rate": 0.003,
+      "loss": 4.1542,
+      "step": 2747
+    },
+    {
+      "epoch": 0.02748,
+      "grad_norm": 0.5095037008788326,
+      "learning_rate": 0.003,
+      "loss": 4.106,
+      "step": 2748
+    },
+    {
+      "epoch": 0.02749,
+      "grad_norm": 0.4918061860336399,
+      "learning_rate": 0.003,
+      "loss": 4.1075,
+      "step": 2749
+    },
+    {
+      "epoch": 0.0275,
+      "grad_norm": 0.4689633122885494,
+      "learning_rate": 0.003,
+      "loss": 4.1398,
+      "step": 2750
+    },
+    {
+      "epoch": 0.02751,
+      "grad_norm": 0.5240834818389507,
+      "learning_rate": 0.003,
+      "loss": 4.1544,
+      "step": 2751
+    },
+    {
+      "epoch": 0.02752,
+      "grad_norm": 0.45737584283965704,
+      "learning_rate": 0.003,
+      "loss": 4.1373,
+      "step": 2752
+    },
+    {
+      "epoch": 0.02753,
+      "grad_norm": 0.4854554116683414,
+      "learning_rate": 0.003,
+      "loss": 4.1197,
+      "step": 2753
+    },
+    {
+      "epoch": 0.02754,
+      "grad_norm": 0.5451937888698382,
+      "learning_rate": 0.003,
+      "loss": 4.1039,
+      "step": 2754
+    },
+    {
+      "epoch": 0.02755,
+      "grad_norm": 0.6902713768999703,
+      "learning_rate": 0.003,
+      "loss": 4.1055,
+      "step": 2755
+    },
+    {
+      "epoch": 0.02756,
+      "grad_norm": 0.9172618359251131,
+      "learning_rate": 0.003,
+      "loss": 4.1411,
+      "step": 2756
+    },
+    {
+      "epoch": 0.02757,
+      "grad_norm": 1.264093714132492,
+      "learning_rate": 0.003,
+      "loss": 4.1713,
+      "step": 2757
+    },
+    {
+      "epoch": 0.02758,
+      "grad_norm": 0.6764873385501415,
+      "learning_rate": 0.003,
+      "loss": 4.1046,
+      "step": 2758
+    },
+    {
+      "epoch": 0.02759,
+      "grad_norm": 0.6735199753323651,
+      "learning_rate": 0.003,
+      "loss": 4.1384,
+      "step": 2759
+    },
+    {
+      "epoch": 0.0276,
+      "grad_norm": 0.7869359581922244,
+      "learning_rate": 0.003,
+      "loss": 4.143,
+      "step": 2760
+    },
+    {
+      "epoch": 0.02761,
+      "grad_norm": 0.8232479826802365,
+      "learning_rate": 0.003,
+      "loss": 4.1614,
+      "step": 2761
+    },
+    {
+      "epoch": 0.02762,
+      "grad_norm": 0.8094991367198255,
+      "learning_rate": 0.003,
+      "loss": 4.1272,
+      "step": 2762
+    },
+    {
+      "epoch": 0.02763,
+      "grad_norm": 0.808838083900847,
+      "learning_rate": 0.003,
+      "loss": 4.1393,
+      "step": 2763
+    },
+    {
+      "epoch": 0.02764,
+      "grad_norm": 0.9499840304387701,
+      "learning_rate": 0.003,
+      "loss": 4.1387,
+      "step": 2764
+    },
+    {
+      "epoch": 0.02765,
+      "grad_norm": 1.1661515206508581,
+      "learning_rate": 0.003,
+      "loss": 4.1535,
+      "step": 2765
+    },
+    {
+      "epoch": 0.02766,
+      "grad_norm": 0.8834816047449258,
+      "learning_rate": 0.003,
+      "loss": 4.1143,
+      "step": 2766
+    },
+    {
+      "epoch": 0.02767,
+      "grad_norm": 0.9679955840143722,
+      "learning_rate": 0.003,
+      "loss": 4.1737,
+      "step": 2767
+    },
+    {
+      "epoch": 0.02768,
+      "grad_norm": 1.0122550261242687,
+      "learning_rate": 0.003,
+      "loss": 4.1138,
+      "step": 2768
+    },
+    {
+      "epoch": 0.02769,
+      "grad_norm": 0.936589272381051,
+      "learning_rate": 0.003,
+      "loss": 4.1722,
+      "step": 2769
+    },
+    {
+      "epoch": 0.0277,
+      "grad_norm": 0.8812342595154814,
+      "learning_rate": 0.003,
+      "loss": 4.1383,
+      "step": 2770
+    },
+    {
+      "epoch": 0.02771,
+      "grad_norm": 0.8385099378903244,
+      "learning_rate": 0.003,
+      "loss": 4.1266,
+      "step": 2771
+    },
+    {
+      "epoch": 0.02772,
+      "grad_norm": 0.8615548972194946,
+      "learning_rate": 0.003,
+      "loss": 4.1184,
+      "step": 2772
+    },
+    {
+      "epoch": 0.02773,
+      "grad_norm": 0.7470925818947043,
+      "learning_rate": 0.003,
+      "loss": 4.1495,
+      "step": 2773
+    },
+    {
+      "epoch": 0.02774,
+      "grad_norm": 0.7432890220276933,
+      "learning_rate": 0.003,
+      "loss": 4.1713,
+      "step": 2774
+    },
+    {
+      "epoch": 0.02775,
+      "grad_norm": 0.714251751808306,
+      "learning_rate": 0.003,
+      "loss": 4.1377,
+      "step": 2775
+    },
+    {
+      "epoch": 0.02776,
+      "grad_norm": 0.6696685091947796,
+      "learning_rate": 0.003,
+      "loss": 4.1315,
+      "step": 2776
+    },
+    {
+      "epoch": 0.02777,
+      "grad_norm": 0.6601977907009066,
+      "learning_rate": 0.003,
+      "loss": 4.1109,
+      "step": 2777
+    },
+    {
+      "epoch": 0.02778,
+      "grad_norm": 0.6363902696248916,
+      "learning_rate": 0.003,
+      "loss": 4.1509,
+      "step": 2778
+    },
+    {
+      "epoch": 0.02779,
+      "grad_norm": 0.7347310050156719,
+      "learning_rate": 0.003,
+      "loss": 4.126,
+      "step": 2779
+    },
+    {
+      "epoch": 0.0278,
+      "grad_norm": 0.8193092944005641,
+      "learning_rate": 0.003,
+      "loss": 4.1233,
+      "step": 2780
+    },
+    {
+      "epoch": 0.02781,
+      "grad_norm": 0.9505772704924963,
+      "learning_rate": 0.003,
+      "loss": 4.113,
+      "step": 2781
+    },
+    {
+      "epoch": 0.02782,
+      "grad_norm": 1.09253758346844,
+      "learning_rate": 0.003,
+      "loss": 4.1421,
+      "step": 2782
+    },
+    {
+      "epoch": 0.02783,
+      "grad_norm": 0.7922623009085619,
+      "learning_rate": 0.003,
+      "loss": 4.1407,
+      "step": 2783
+    },
+    {
+      "epoch": 0.02784,
+      "grad_norm": 0.7730396671741616,
+      "learning_rate": 0.003,
+      "loss": 4.136,
+      "step": 2784
+    },
+    {
+      "epoch": 0.02785,
+      "grad_norm": 0.8111585358679776,
+      "learning_rate": 0.003,
+      "loss": 4.1529,
+      "step": 2785
+    },
+    {
+      "epoch": 0.02786,
+      "grad_norm": 0.7983751779273013,
+      "learning_rate": 0.003,
+      "loss": 4.1458,
+      "step": 2786
+    },
+    {
+      "epoch": 0.02787,
+      "grad_norm": 0.8522172170925804,
+      "learning_rate": 0.003,
+      "loss": 4.1468,
+      "step": 2787
+    },
+    {
+      "epoch": 0.02788,
+      "grad_norm": 0.8952452301797756,
+      "learning_rate": 0.003,
+      "loss": 4.1379,
+      "step": 2788
+    },
+    {
+      "epoch": 0.02789,
+      "grad_norm": 0.9315897488998124,
+      "learning_rate": 0.003,
+      "loss": 4.1439,
+      "step": 2789
+    },
+    {
+      "epoch": 0.0279,
+      "grad_norm": 0.7715461952206646,
+      "learning_rate": 0.003,
+      "loss": 4.1139,
+      "step": 2790
+    },
+    {
+      "epoch": 0.02791,
+      "grad_norm": 0.7419347495262759,
+      "learning_rate": 0.003,
+      "loss": 4.1266,
+      "step": 2791
+    },
+    {
+      "epoch": 0.02792,
+      "grad_norm": 0.804466637477434,
+      "learning_rate": 0.003,
+      "loss": 4.1446,
+      "step": 2792
+    },
+    {
+      "epoch": 0.02793,
+      "grad_norm": 0.9499312139027115,
+      "learning_rate": 0.003,
+      "loss": 4.1481,
+      "step": 2793
+    },
+    {
+      "epoch": 0.02794,
+      "grad_norm": 1.079651373139261,
+      "learning_rate": 0.003,
+      "loss": 4.1503,
+      "step": 2794
+    },
+    {
+      "epoch": 0.02795,
+      "grad_norm": 0.7784790735501392,
+      "learning_rate": 0.003,
+      "loss": 4.1562,
+      "step": 2795
+    },
+    {
+      "epoch": 0.02796,
+      "grad_norm": 0.7016883329409419,
+      "learning_rate": 0.003,
+      "loss": 4.1447,
+      "step": 2796
+    },
+    {
+      "epoch": 0.02797,
+      "grad_norm": 0.8527425055953903,
+      "learning_rate": 0.003,
+      "loss": 4.1615,
+      "step": 2797
+    },
+    {
+      "epoch": 0.02798,
+      "grad_norm": 0.8344419810312547,
+      "learning_rate": 0.003,
+      "loss": 4.1599,
+      "step": 2798
+    },
+    {
+      "epoch": 0.02799,
+      "grad_norm": 0.8118333253983723,
+      "learning_rate": 0.003,
+      "loss": 4.1456,
+      "step": 2799
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 0.8276964023163034,
+      "learning_rate": 0.003,
+      "loss": 4.154,
+      "step": 2800
+    },
+    {
+      "epoch": 0.02801,
+      "grad_norm": 0.8829595698308325,
+      "learning_rate": 0.003,
+      "loss": 4.1343,
+      "step": 2801
+    },
+    {
+      "epoch": 0.02802,
+      "grad_norm": 0.9178695537376129,
+      "learning_rate": 0.003,
+      "loss": 4.1357,
+      "step": 2802
+    },
+    {
+      "epoch": 0.02803,
+      "grad_norm": 0.773306657979122,
+      "learning_rate": 0.003,
+      "loss": 4.1566,
+      "step": 2803
+    },
+    {
+      "epoch": 0.02804,
+      "grad_norm": 0.6894758923555496,
+      "learning_rate": 0.003,
+      "loss": 4.1238,
+      "step": 2804
+    },
+    {
+      "epoch": 0.02805,
+      "grad_norm": 0.6917993567356406,
+      "learning_rate": 0.003,
+      "loss": 4.1566,
+      "step": 2805
+    },
+    {
+      "epoch": 0.02806,
+      "grad_norm": 0.6387085680780493,
+      "learning_rate": 0.003,
+      "loss": 4.1312,
+      "step": 2806
+    },
+    {
+      "epoch": 0.02807,
+      "grad_norm": 0.580517980379294,
+      "learning_rate": 0.003,
+      "loss": 4.1301,
+      "step": 2807
+    },
+    {
+      "epoch": 0.02808,
+      "grad_norm": 0.4980072648857084,
+      "learning_rate": 0.003,
+      "loss": 4.1101,
+      "step": 2808
+    },
+    {
+      "epoch": 0.02809,
+      "grad_norm": 0.5292621050422022,
+      "learning_rate": 0.003,
+      "loss": 4.1503,
+      "step": 2809
+    },
+    {
+      "epoch": 0.0281,
+      "grad_norm": 0.5662997833368786,
+      "learning_rate": 0.003,
+      "loss": 4.1045,
+      "step": 2810
+    },
+    {
+      "epoch": 0.02811,
+      "grad_norm": 0.6056304842841205,
+      "learning_rate": 0.003,
+      "loss": 4.1149,
+      "step": 2811
+    },
+    {
+      "epoch": 0.02812,
+      "grad_norm": 0.6510291438212791,
+      "learning_rate": 0.003,
+      "loss": 4.131,
+      "step": 2812
+    },
+    {
+      "epoch": 0.02813,
+      "grad_norm": 0.6856589618404986,
+      "learning_rate": 0.003,
+      "loss": 4.1212,
+      "step": 2813
+    },
+    {
+      "epoch": 0.02814,
+      "grad_norm": 0.7345790506941087,
+      "learning_rate": 0.003,
+      "loss": 4.1221,
+      "step": 2814
+    },
+    {
+      "epoch": 0.02815,
+      "grad_norm": 0.7816359381581716,
+      "learning_rate": 0.003,
+      "loss": 4.1614,
+      "step": 2815
+    },
+    {
+      "epoch": 0.02816,
+      "grad_norm": 0.8925647710841813,
+      "learning_rate": 0.003,
+      "loss": 4.1426,
+      "step": 2816
+    },
+    {
+      "epoch": 0.02817,
+      "grad_norm": 1.169228788902705,
+      "learning_rate": 0.003,
+      "loss": 4.1079,
+      "step": 2817
+    },
+    {
+      "epoch": 0.02818,
+      "grad_norm": 0.9618528564502903,
+      "learning_rate": 0.003,
+      "loss": 4.1194,
+      "step": 2818
+    },
+    {
+      "epoch": 0.02819,
+      "grad_norm": 1.0178604687945154,
+      "learning_rate": 0.003,
+      "loss": 4.15,
+      "step": 2819
+    },
+    {
+      "epoch": 0.0282,
+      "grad_norm": 0.8729007051855019,
+      "learning_rate": 0.003,
+      "loss": 4.1374,
+      "step": 2820
+    },
+    {
+      "epoch": 0.02821,
+      "grad_norm": 0.7594720998624757,
+      "learning_rate": 0.003,
+      "loss": 4.1557,
+      "step": 2821
+    },
+    {
+      "epoch": 0.02822,
+      "grad_norm": 0.7764641169305468,
+      "learning_rate": 0.003,
+      "loss": 4.1245,
+      "step": 2822
+    },
+    {
+      "epoch": 0.02823,
+      "grad_norm": 0.79863443820284,
+      "learning_rate": 0.003,
+      "loss": 4.154,
+      "step": 2823
+    },
+    {
+      "epoch": 0.02824,
+      "grad_norm": 0.7182103089348308,
+      "learning_rate": 0.003,
+      "loss": 4.1387,
+      "step": 2824
+    },
+    {
+      "epoch": 0.02825,
+      "grad_norm": 0.7529407236815215,
+      "learning_rate": 0.003,
+      "loss": 4.1425,
+      "step": 2825
+    },
+    {
+      "epoch": 0.02826,
+      "grad_norm": 0.8056220195019022,
+      "learning_rate": 0.003,
+      "loss": 4.1212,
+      "step": 2826
+    },
+    {
+      "epoch": 0.02827,
+      "grad_norm": 0.8893025905667793,
+      "learning_rate": 0.003,
+      "loss": 4.1113,
+      "step": 2827
+    },
+    {
+      "epoch": 0.02828,
+      "grad_norm": 0.9458155387571202,
+      "learning_rate": 0.003,
+      "loss": 4.1547,
+      "step": 2828
+    },
+    {
+      "epoch": 0.02829,
+      "grad_norm": 1.0816962232960836,
+      "learning_rate": 0.003,
+      "loss": 4.1852,
+      "step": 2829
+    },
+    {
+      "epoch": 0.0283,
+      "grad_norm": 0.85593992308399,
+      "learning_rate": 0.003,
+      "loss": 4.1432,
+      "step": 2830
+    },
+    {
+      "epoch": 0.02831,
+      "grad_norm": 0.8103547032302357,
+      "learning_rate": 0.003,
+      "loss": 4.1446,
+      "step": 2831
+    },
+    {
+      "epoch": 0.02832,
+      "grad_norm": 1.0004568820848851,
+      "learning_rate": 0.003,
+      "loss": 4.1611,
+      "step": 2832
+    },
+    {
+      "epoch": 0.02833,
+      "grad_norm": 1.2257623869393093,
+      "learning_rate": 0.003,
+      "loss": 4.1482,
+      "step": 2833
+    },
+    {
+      "epoch": 0.02834,
+      "grad_norm": 0.8959695287844174,
+      "learning_rate": 0.003,
+      "loss": 4.1264,
+      "step": 2834
+    },
+    {
+      "epoch": 0.02835,
+      "grad_norm": 0.9046726088007764,
+      "learning_rate": 0.003,
+      "loss": 4.1382,
+      "step": 2835
+    },
+    {
+      "epoch": 0.02836,
+      "grad_norm": 0.9388179555413667,
+      "learning_rate": 0.003,
+      "loss": 4.1634,
+      "step": 2836
+    },
+    {
+      "epoch": 0.02837,
+      "grad_norm": 0.8876864866399585,
+      "learning_rate": 0.003,
+      "loss": 4.1312,
+      "step": 2837
+    },
+    {
+      "epoch": 0.02838,
+      "grad_norm": 0.93231141838983,
+      "learning_rate": 0.003,
+      "loss": 4.1029,
+      "step": 2838
+    },
+    {
+      "epoch": 0.02839,
+      "grad_norm": 0.8358991135319658,
+      "learning_rate": 0.003,
+      "loss": 4.0981,
+      "step": 2839
+    },
+    {
+      "epoch": 0.0284,
+      "grad_norm": 0.8834561842613636,
+      "learning_rate": 0.003,
+      "loss": 4.1133,
+      "step": 2840
+    },
+    {
+      "epoch": 0.02841,
+      "grad_norm": 0.9006412651182929,
+      "learning_rate": 0.003,
+      "loss": 4.1522,
+      "step": 2841
+    },
+    {
+      "epoch": 0.02842,
+      "grad_norm": 0.9446862919101724,
+      "learning_rate": 0.003,
+      "loss": 4.1417,
+      "step": 2842
+    },
+    {
+      "epoch": 0.02843,
+      "grad_norm": 0.8098223520337203,
+      "learning_rate": 0.003,
+      "loss": 4.1599,
+      "step": 2843
+    },
+    {
+      "epoch": 0.02844,
+      "grad_norm": 0.7322010149090498,
+      "learning_rate": 0.003,
+      "loss": 4.1529,
+      "step": 2844
+    },
+    {
+      "epoch": 0.02845,
+      "grad_norm": 0.6471809615510085,
+      "learning_rate": 0.003,
+      "loss": 4.126,
+      "step": 2845
+    },
+    {
+      "epoch": 0.02846,
+      "grad_norm": 0.6878791837952257,
+      "learning_rate": 0.003,
+      "loss": 4.1218,
+      "step": 2846
+    },
+    {
+      "epoch": 0.02847,
+      "grad_norm": 0.7696793006326048,
+      "learning_rate": 0.003,
+      "loss": 4.1788,
+      "step": 2847
+    },
+    {
+      "epoch": 0.02848,
+      "grad_norm": 0.8913190904442756,
+      "learning_rate": 0.003,
+      "loss": 4.1422,
+      "step": 2848
+    },
+    {
+      "epoch": 0.02849,
+      "grad_norm": 1.018118471142102,
+      "learning_rate": 0.003,
+      "loss": 4.1327,
+      "step": 2849
+    },
+    {
+      "epoch": 0.0285,
+      "grad_norm": 0.9082110603942541,
+      "learning_rate": 0.003,
+      "loss": 4.1356,
+      "step": 2850
+    },
+    {
+      "epoch": 0.02851,
+      "grad_norm": 0.822250823175019,
+      "learning_rate": 0.003,
+      "loss": 4.1725,
+      "step": 2851
+    },
+    {
+      "epoch": 0.02852,
+      "grad_norm": 0.6052901464954267,
+      "learning_rate": 0.003,
+      "loss": 4.1424,
+      "step": 2852
+    },
+    {
+      "epoch": 0.02853,
+      "grad_norm": 0.5735422812437921,
+      "learning_rate": 0.003,
+      "loss": 4.1065,
+      "step": 2853
+    },
+    {
+      "epoch": 0.02854,
+      "grad_norm": 0.5429930679913405,
+      "learning_rate": 0.003,
+      "loss": 4.1448,
+      "step": 2854
+    },
+    {
+      "epoch": 0.02855,
+      "grad_norm": 0.5339097730360199,
+      "learning_rate": 0.003,
+      "loss": 4.1387,
+      "step": 2855
+    },
+    {
+      "epoch": 0.02856,
+      "grad_norm": 0.5860774205400787,
+      "learning_rate": 0.003,
+      "loss": 4.1428,
+      "step": 2856
+    },
+    {
+      "epoch": 0.02857,
+      "grad_norm": 0.6433890086945638,
+      "learning_rate": 0.003,
+      "loss": 4.1236,
+      "step": 2857
+    },
+    {
+      "epoch": 0.02858,
+      "grad_norm": 0.7879326106492056,
+      "learning_rate": 0.003,
+      "loss": 4.1162,
+      "step": 2858
+    },
+    {
+      "epoch": 0.02859,
+      "grad_norm": 0.8693860304588211,
+      "learning_rate": 0.003,
+      "loss": 4.1274,
+      "step": 2859
+    },
+    {
+      "epoch": 0.0286,
+      "grad_norm": 0.8079095294087945,
+      "learning_rate": 0.003,
+      "loss": 4.1417,
+      "step": 2860
+    },
+    {
+      "epoch": 0.02861,
+      "grad_norm": 0.6656764279644372,
+      "learning_rate": 0.003,
+      "loss": 4.1049,
+      "step": 2861
+    },
+    {
+      "epoch": 0.02862,
+      "grad_norm": 0.6696791424270172,
+      "learning_rate": 0.003,
+      "loss": 4.1232,
+      "step": 2862
+    },
+    {
+      "epoch": 0.02863,
+      "grad_norm": 0.7769070451321298,
+      "learning_rate": 0.003,
+      "loss": 4.1498,
+      "step": 2863
+    },
+    {
+      "epoch": 0.02864,
+      "grad_norm": 0.8224098496945286,
+      "learning_rate": 0.003,
+      "loss": 4.1241,
+      "step": 2864
+    },
+    {
+      "epoch": 0.02865,
+      "grad_norm": 0.8864904617013009,
+      "learning_rate": 0.003,
+      "loss": 4.1583,
+      "step": 2865
+    },
+    {
+      "epoch": 0.02866,
+      "grad_norm": 0.9583148110901643,
+      "learning_rate": 0.003,
+      "loss": 4.1437,
+      "step": 2866
+    },
+    {
+      "epoch": 0.02867,
+      "grad_norm": 0.8904807738138028,
+      "learning_rate": 0.003,
+      "loss": 4.1281,
+      "step": 2867
+    },
+    {
+      "epoch": 0.02868,
+      "grad_norm": 0.9293551806467387,
+      "learning_rate": 0.003,
+      "loss": 4.1391,
+      "step": 2868
+    },
+    {
+      "epoch": 0.02869,
+      "grad_norm": 1.0457657175233683,
+      "learning_rate": 0.003,
+      "loss": 4.1353,
+      "step": 2869
+    },
+    {
+      "epoch": 0.0287,
+      "grad_norm": 1.0953444171616145,
+      "learning_rate": 0.003,
+      "loss": 4.1636,
+      "step": 2870
+    },
+    {
+      "epoch": 0.02871,
+      "grad_norm": 0.7587885970460945,
+      "learning_rate": 0.003,
+      "loss": 4.1356,
+      "step": 2871
+    },
+    {
+      "epoch": 0.02872,
+      "grad_norm": 0.6777165872502906,
+      "learning_rate": 0.003,
+      "loss": 4.1681,
+      "step": 2872
+    },
+    {
+      "epoch": 0.02873,
+      "grad_norm": 0.8924775793037929,
+      "learning_rate": 0.003,
+      "loss": 4.1422,
+      "step": 2873
+    },
+    {
+      "epoch": 0.02874,
+      "grad_norm": 0.9191095226023763,
+      "learning_rate": 0.003,
+      "loss": 4.1633,
+      "step": 2874
+    },
+    {
+      "epoch": 0.02875,
+      "grad_norm": 1.0057201785085328,
+      "learning_rate": 0.003,
+      "loss": 4.1702,
+      "step": 2875
+    },
+    {
+      "epoch": 0.02876,
+      "grad_norm": 0.9084894421524936,
+      "learning_rate": 0.003,
+      "loss": 4.1504,
+      "step": 2876
+    },
+    {
+      "epoch": 0.02877,
+      "grad_norm": 0.9095891721863432,
+      "learning_rate": 0.003,
+      "loss": 4.1305,
+      "step": 2877
+    },
+    {
+      "epoch": 0.02878,
+      "grad_norm": 0.9172527890033496,
+      "learning_rate": 0.003,
+      "loss": 4.1286,
+      "step": 2878
+    },
+    {
+      "epoch": 0.02879,
+      "grad_norm": 0.8805567763002417,
+      "learning_rate": 0.003,
+      "loss": 4.1227,
+      "step": 2879
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 0.9996123727255034,
+      "learning_rate": 0.003,
+      "loss": 4.1332,
+      "step": 2880
+    },
+    {
+      "epoch": 0.02881,
+      "grad_norm": 1.046330847202451,
+      "learning_rate": 0.003,
+      "loss": 4.1514,
+      "step": 2881
+    },
+    {
+      "epoch": 0.02882,
+      "grad_norm": 0.772117255675699,
+      "learning_rate": 0.003,
+      "loss": 4.1606,
+      "step": 2882
+    },
+    {
+      "epoch": 0.02883,
+      "grad_norm": 0.7008407026763371,
+      "learning_rate": 0.003,
+      "loss": 4.1396,
+      "step": 2883
+    },
+    {
+      "epoch": 0.02884,
+      "grad_norm": 0.778782483892839,
+      "learning_rate": 0.003,
+      "loss": 4.177,
+      "step": 2884
+    },
+    {
+      "epoch": 0.02885,
+      "grad_norm": 0.7648524574296616,
+      "learning_rate": 0.003,
+      "loss": 4.1174,
+      "step": 2885
+    },
+    {
+      "epoch": 0.02886,
+      "grad_norm": 0.6944709668733141,
+      "learning_rate": 0.003,
+      "loss": 4.1469,
+      "step": 2886
+    },
+    {
+      "epoch": 0.02887,
+      "grad_norm": 0.6701833801380009,
+      "learning_rate": 0.003,
+      "loss": 4.1291,
+      "step": 2887
+    },
+    {
+      "epoch": 0.02888,
+      "grad_norm": 0.6522446873326756,
+      "learning_rate": 0.003,
+      "loss": 4.1256,
+      "step": 2888
+    },
+    {
+      "epoch": 0.02889,
+      "grad_norm": 0.5695684026681817,
+      "learning_rate": 0.003,
+      "loss": 4.1113,
+      "step": 2889
+    },
+    {
+      "epoch": 0.0289,
+      "grad_norm": 0.5387645552557849,
+      "learning_rate": 0.003,
+      "loss": 4.1434,
+      "step": 2890
+    },
+    {
+      "epoch": 0.02891,
+      "grad_norm": 0.6068073035716243,
+      "learning_rate": 0.003,
+      "loss": 4.1589,
+      "step": 2891
+    },
+    {
+      "epoch": 0.02892,
+      "grad_norm": 0.5642060774681064,
+      "learning_rate": 0.003,
+      "loss": 4.109,
+      "step": 2892
+    },
+    {
+      "epoch": 0.02893,
+      "grad_norm": 0.64295806093738,
+      "learning_rate": 0.003,
+      "loss": 4.1179,
+      "step": 2893
+    },
+    {
+      "epoch": 0.02894,
+      "grad_norm": 0.7755044700890192,
+      "learning_rate": 0.003,
+      "loss": 4.1282,
+      "step": 2894
+    },
+    {
+      "epoch": 0.02895,
+      "grad_norm": 0.7810569175434396,
+      "learning_rate": 0.003,
+      "loss": 4.1719,
+      "step": 2895
+    },
+    {
+      "epoch": 0.02896,
+      "grad_norm": 0.5425405989527838,
+      "learning_rate": 0.003,
+      "loss": 4.1197,
+      "step": 2896
+    },
+    {
+      "epoch": 0.02897,
+      "grad_norm": 0.469070597819537,
+      "learning_rate": 0.003,
+      "loss": 4.0995,
+      "step": 2897
+    },
+    {
+      "epoch": 0.02898,
+      "grad_norm": 0.5139238004799116,
+      "learning_rate": 0.003,
+      "loss": 4.1442,
+      "step": 2898
+    },
+    {
+      "epoch": 0.02899,
+      "grad_norm": 0.6544046694256337,
+      "learning_rate": 0.003,
+      "loss": 4.1296,
+      "step": 2899
+    },
+    {
+      "epoch": 0.029,
+      "grad_norm": 0.7180257646745092,
+      "learning_rate": 0.003,
+      "loss": 4.1219,
+      "step": 2900
+    },
+    {
+      "epoch": 0.02901,
+      "grad_norm": 0.6895214504299034,
+      "learning_rate": 0.003,
+      "loss": 4.1358,
+      "step": 2901
+    },
+    {
+      "epoch": 0.02902,
+      "grad_norm": 0.6936366493807695,
+      "learning_rate": 0.003,
+      "loss": 4.1524,
+      "step": 2902
+    },
+    {
+      "epoch": 0.02903,
+      "grad_norm": 0.7680000086672168,
+      "learning_rate": 0.003,
+      "loss": 4.1137,
+      "step": 2903
+    },
+    {
+      "epoch": 0.02904,
+      "grad_norm": 1.0523944817643214,
+      "learning_rate": 0.003,
+      "loss": 4.1402,
+      "step": 2904
+    },
+    {
+      "epoch": 0.02905,
+      "grad_norm": 1.051471635673892,
+      "learning_rate": 0.003,
+      "loss": 4.1582,
+      "step": 2905
+    },
+    {
+      "epoch": 0.02906,
+      "grad_norm": 0.8511171595587916,
+      "learning_rate": 0.003,
+      "loss": 4.1318,
+      "step": 2906
+    },
+    {
+      "epoch": 0.02907,
+      "grad_norm": 0.8224344496311664,
+      "learning_rate": 0.003,
+      "loss": 4.1093,
+      "step": 2907
+    },
+    {
+      "epoch": 0.02908,
+      "grad_norm": 0.6999459176877906,
+      "learning_rate": 0.003,
+      "loss": 4.1259,
+      "step": 2908
+    },
+    {
+      "epoch": 0.02909,
+      "grad_norm": 0.819486070194941,
+      "learning_rate": 0.003,
+      "loss": 4.1263,
+      "step": 2909
+    },
+    {
+      "epoch": 0.0291,
+      "grad_norm": 1.066905831023775,
+      "learning_rate": 0.003,
+      "loss": 4.1444,
+      "step": 2910
+    },
+    {
+      "epoch": 0.02911,
+      "grad_norm": 0.993942719913462,
+      "learning_rate": 0.003,
+      "loss": 4.1435,
+      "step": 2911
+    },
+    {
+      "epoch": 0.02912,
+      "grad_norm": 0.9395324885369105,
+      "learning_rate": 0.003,
+      "loss": 4.114,
+      "step": 2912
+    },
+    {
+      "epoch": 0.02913,
+      "grad_norm": 0.9563813373046008,
+      "learning_rate": 0.003,
+      "loss": 4.1513,
+      "step": 2913
+    },
+    {
+      "epoch": 0.02914,
+      "grad_norm": 1.204534731933291,
+      "learning_rate": 0.003,
+      "loss": 4.1464,
+      "step": 2914
+    },
+    {
+      "epoch": 0.02915,
+      "grad_norm": 0.9555386893347948,
+      "learning_rate": 0.003,
+      "loss": 4.1331,
+      "step": 2915
+    },
+    {
+      "epoch": 0.02916,
+      "grad_norm": 1.0824011254780639,
+      "learning_rate": 0.003,
+      "loss": 4.1482,
+      "step": 2916
+    },
+    {
+      "epoch": 0.02917,
+      "grad_norm": 1.0942512414631864,
+      "learning_rate": 0.003,
+      "loss": 4.1821,
+      "step": 2917
+    },
+    {
+      "epoch": 0.02918,
+      "grad_norm": 0.9071341774299415,
+      "learning_rate": 0.003,
+      "loss": 4.1353,
+      "step": 2918
+    },
+    {
+      "epoch": 0.02919,
+      "grad_norm": 0.9246562992704038,
+      "learning_rate": 0.003,
+      "loss": 4.1768,
+      "step": 2919
+    },
+    {
+      "epoch": 0.0292,
+      "grad_norm": 0.8971677038672106,
+      "learning_rate": 0.003,
+      "loss": 4.1129,
+      "step": 2920
+    },
+    {
+      "epoch": 0.02921,
+      "grad_norm": 0.926961421602634,
+      "learning_rate": 0.003,
+      "loss": 4.1553,
+      "step": 2921
+    },
+    {
+      "epoch": 0.02922,
+      "grad_norm": 1.0848027072453201,
+      "learning_rate": 0.003,
+      "loss": 4.161,
+      "step": 2922
+    },
+    {
+      "epoch": 0.02923,
+      "grad_norm": 0.912142779077174,
+      "learning_rate": 0.003,
+      "loss": 4.1303,
+      "step": 2923
+    },
+    {
+      "epoch": 0.02924,
+      "grad_norm": 0.9865161740552381,
+      "learning_rate": 0.003,
+      "loss": 4.1648,
+      "step": 2924
+    },
+    {
+      "epoch": 0.02925,
+      "grad_norm": 1.0463105360491507,
+      "learning_rate": 0.003,
+      "loss": 4.1748,
+      "step": 2925
+    },
+    {
+      "epoch": 0.02926,
+      "grad_norm": 0.9582122548230526,
+      "learning_rate": 0.003,
+      "loss": 4.1501,
+      "step": 2926
+    },
+    {
+      "epoch": 0.02927,
+      "grad_norm": 0.8403311811917105,
+      "learning_rate": 0.003,
+      "loss": 4.1524,
+      "step": 2927
+    },
+    {
+      "epoch": 0.02928,
+      "grad_norm": 0.7854536393801775,
+      "learning_rate": 0.003,
+      "loss": 4.1695,
+      "step": 2928
+    },
+    {
+      "epoch": 0.02929,
+      "grad_norm": 0.7778022366151639,
+      "learning_rate": 0.003,
+      "loss": 4.1542,
+      "step": 2929
+    },
+    {
+      "epoch": 0.0293,
+      "grad_norm": 0.7271690214257167,
+      "learning_rate": 0.003,
+      "loss": 4.1453,
+      "step": 2930
+    },
+    {
+      "epoch": 0.02931,
+      "grad_norm": 0.782379273286375,
+      "learning_rate": 0.003,
+      "loss": 4.1435,
+      "step": 2931
+    },
+    {
+      "epoch": 0.02932,
+      "grad_norm": 0.7974568484352181,
+      "learning_rate": 0.003,
+      "loss": 4.1304,
+      "step": 2932
+    },
+    {
+      "epoch": 0.02933,
+      "grad_norm": 0.7828005432560315,
+      "learning_rate": 0.003,
+      "loss": 4.1554,
+      "step": 2933
+    },
+    {
+      "epoch": 0.02934,
+      "grad_norm": 0.6614858211514127,
+      "learning_rate": 0.003,
+      "loss": 4.14,
+      "step": 2934
+    },
+    {
+      "epoch": 0.02935,
+      "grad_norm": 0.5873856004708764,
+      "learning_rate": 0.003,
+      "loss": 4.1239,
+      "step": 2935
+    },
+    {
+      "epoch": 0.02936,
+      "grad_norm": 0.578823462032298,
+      "learning_rate": 0.003,
+      "loss": 4.1409,
+      "step": 2936
+    },
+    {
+      "epoch": 0.02937,
+      "grad_norm": 0.6078930790219389,
+      "learning_rate": 0.003,
+      "loss": 4.1209,
+      "step": 2937
+    },
+    {
+      "epoch": 0.02938,
+      "grad_norm": 0.7146509387847667,
+      "learning_rate": 0.003,
+      "loss": 4.139,
+      "step": 2938
+    },
+    {
+      "epoch": 0.02939,
+      "grad_norm": 0.8250738871743942,
+      "learning_rate": 0.003,
+      "loss": 4.1295,
+      "step": 2939
+    },
+    {
+      "epoch": 0.0294,
+      "grad_norm": 0.8696648935923786,
+      "learning_rate": 0.003,
+      "loss": 4.1547,
+      "step": 2940
+    },
+    {
+      "epoch": 0.02941,
+      "grad_norm": 0.9090371360813282,
+      "learning_rate": 0.003,
+      "loss": 4.1568,
+      "step": 2941
+    },
+    {
+      "epoch": 0.02942,
+      "grad_norm": 0.8939350957264195,
+      "learning_rate": 0.003,
+      "loss": 4.1447,
+      "step": 2942
+    },
+    {
+      "epoch": 0.02943,
+      "grad_norm": 0.7023666348770743,
+      "learning_rate": 0.003,
+      "loss": 4.1276,
+      "step": 2943
+    },
+    {
+      "epoch": 0.02944,
+      "grad_norm": 0.7067049448574164,
+      "learning_rate": 0.003,
+      "loss": 4.1465,
+      "step": 2944
+    },
+    {
+      "epoch": 0.02945,
+      "grad_norm": 0.6812738232543556,
+      "learning_rate": 0.003,
+      "loss": 4.1255,
+      "step": 2945
+    },
+    {
+      "epoch": 0.02946,
+      "grad_norm": 0.6818360632430761,
+      "learning_rate": 0.003,
+      "loss": 4.1214,
+      "step": 2946
+    },
+    {
+      "epoch": 0.02947,
+      "grad_norm": 0.6754875767553201,
+      "learning_rate": 0.003,
+      "loss": 4.1523,
+      "step": 2947
+    },
+    {
+      "epoch": 0.02948,
+      "grad_norm": 0.6113651211997226,
+      "learning_rate": 0.003,
+      "loss": 4.1317,
+      "step": 2948
+    },
+    {
+      "epoch": 0.02949,
+      "grad_norm": 0.5812056588270224,
+      "learning_rate": 0.003,
+      "loss": 4.1034,
+      "step": 2949
+    },
+    {
+      "epoch": 0.0295,
+      "grad_norm": 0.6282215017847426,
+      "learning_rate": 0.003,
+      "loss": 4.1274,
+      "step": 2950
+    },
+    {
+      "epoch": 0.02951,
+      "grad_norm": 0.7817687295389203,
+      "learning_rate": 0.003,
+      "loss": 4.0846,
+      "step": 2951
+    },
+    {
+      "epoch": 0.02952,
+      "grad_norm": 1.0587479164041036,
+      "learning_rate": 0.003,
+      "loss": 4.1248,
+      "step": 2952
+    },
+    {
+      "epoch": 0.02953,
+      "grad_norm": 0.9887356887399129,
+      "learning_rate": 0.003,
+      "loss": 4.1373,
+      "step": 2953
+    },
+    {
+      "epoch": 0.02954,
+      "grad_norm": 0.7805100959975674,
+      "learning_rate": 0.003,
+      "loss": 4.1688,
+      "step": 2954
+    },
+    {
+      "epoch": 0.02955,
+      "grad_norm": 0.6651525603429118,
+      "learning_rate": 0.003,
+      "loss": 4.138,
+      "step": 2955
+    },
+    {
+      "epoch": 0.02956,
+      "grad_norm": 0.7699019886627679,
+      "learning_rate": 0.003,
+      "loss": 4.1312,
+      "step": 2956
+    },
+    {
+      "epoch": 0.02957,
+      "grad_norm": 0.816443587627682,
+      "learning_rate": 0.003,
+      "loss": 4.1404,
+      "step": 2957
+    },
+    {
+      "epoch": 0.02958,
+      "grad_norm": 0.7961877462112285,
+      "learning_rate": 0.003,
+      "loss": 4.1553,
+      "step": 2958
+    },
+    {
+      "epoch": 0.02959,
+      "grad_norm": 0.7550243768723249,
+      "learning_rate": 0.003,
+      "loss": 4.1431,
+      "step": 2959
+    },
+    {
+      "epoch": 0.0296,
+      "grad_norm": 0.7723125453491877,
+      "learning_rate": 0.003,
+      "loss": 4.1525,
+      "step": 2960
+    },
+    {
+      "epoch": 0.02961,
+      "grad_norm": 0.8583607117371619,
+      "learning_rate": 0.003,
+      "loss": 4.127,
+      "step": 2961
+    },
+    {
+      "epoch": 0.02962,
+      "grad_norm": 0.8691287903706768,
+      "learning_rate": 0.003,
+      "loss": 4.1333,
+      "step": 2962
+    },
+    {
+      "epoch": 0.02963,
+      "grad_norm": 0.8786112930172387,
+      "learning_rate": 0.003,
+      "loss": 4.1116,
+      "step": 2963
+    },
+    {
+      "epoch": 0.02964,
+      "grad_norm": 0.9167891542829973,
+      "learning_rate": 0.003,
+      "loss": 4.1027,
+      "step": 2964
+    },
+    {
+      "epoch": 0.02965,
+      "grad_norm": 0.9419888897278245,
+      "learning_rate": 0.003,
+      "loss": 4.1533,
+      "step": 2965
+    },
+    {
+      "epoch": 0.02966,
+      "grad_norm": 1.1891655776183911,
+      "learning_rate": 0.003,
+      "loss": 4.1783,
+      "step": 2966
+    },
+    {
+      "epoch": 0.02967,
+      "grad_norm": 1.0455487688038185,
+      "learning_rate": 0.003,
+      "loss": 4.1592,
+      "step": 2967
+    },
+    {
+      "epoch": 0.02968,
+      "grad_norm": 1.0164630073257412,
+      "learning_rate": 0.003,
+      "loss": 4.155,
+      "step": 2968
+    },
+    {
+      "epoch": 0.02969,
+      "grad_norm": 1.0185954831664046,
+      "learning_rate": 0.003,
+      "loss": 4.1548,
+      "step": 2969
+    },
+    {
+      "epoch": 0.0297,
+      "grad_norm": 1.043969964180144,
+      "learning_rate": 0.003,
+      "loss": 4.1421,
+      "step": 2970
+    },
+    {
+      "epoch": 0.02971,
+      "grad_norm": 0.8397020648022002,
+      "learning_rate": 0.003,
+      "loss": 4.1551,
+      "step": 2971
+    },
+    {
+      "epoch": 0.02972,
+      "grad_norm": 0.6978014599627598,
+      "learning_rate": 0.003,
+      "loss": 4.1188,
+      "step": 2972
+    },
+    {
+      "epoch": 0.02973,
+      "grad_norm": 0.7307935837042818,
+      "learning_rate": 0.003,
+      "loss": 4.1067,
+      "step": 2973
+    },
+    {
+      "epoch": 0.02974,
+      "grad_norm": 0.8618322772785041,
+      "learning_rate": 0.003,
+      "loss": 4.1462,
+      "step": 2974
+    },
+    {
+      "epoch": 0.02975,
+      "grad_norm": 0.9050029913314735,
+      "learning_rate": 0.003,
+      "loss": 4.1377,
+      "step": 2975
+    },
+    {
+      "epoch": 0.02976,
+      "grad_norm": 0.8860984382067472,
+      "learning_rate": 0.003,
+      "loss": 4.143,
+      "step": 2976
+    },
+    {
+      "epoch": 0.02977,
+      "grad_norm": 0.6971184630795387,
+      "learning_rate": 0.003,
+      "loss": 4.1269,
+      "step": 2977
+    },
+    {
+      "epoch": 0.02978,
+      "grad_norm": 0.7418889082376986,
+      "learning_rate": 0.003,
+      "loss": 4.1484,
+      "step": 2978
+    },
+    {
+      "epoch": 0.02979,
+      "grad_norm": 0.7888236204090991,
+      "learning_rate": 0.003,
+      "loss": 4.133,
+      "step": 2979
+    },
+    {
+      "epoch": 0.0298,
+      "grad_norm": 0.7867568676249742,
+      "learning_rate": 0.003,
+      "loss": 4.1354,
+      "step": 2980
+    },
+    {
+      "epoch": 0.02981,
+      "grad_norm": 0.7844697697074997,
+      "learning_rate": 0.003,
+      "loss": 4.1438,
+      "step": 2981
+    },
+    {
+      "epoch": 0.02982,
+      "grad_norm": 0.6873427232122642,
+      "learning_rate": 0.003,
+      "loss": 4.1364,
+      "step": 2982
+    },
+    {
+      "epoch": 0.02983,
+      "grad_norm": 0.7086227659108,
+      "learning_rate": 0.003,
+      "loss": 4.1508,
+      "step": 2983
+    },
+    {
+      "epoch": 0.02984,
+      "grad_norm": 0.7859850298348623,
+      "learning_rate": 0.003,
+      "loss": 4.1284,
+      "step": 2984
+    },
+    {
+      "epoch": 0.02985,
+      "grad_norm": 0.7729797989684547,
+      "learning_rate": 0.003,
+      "loss": 4.1572,
+      "step": 2985
+    },
+    {
+      "epoch": 0.02986,
+      "grad_norm": 0.6844032687054661,
+      "learning_rate": 0.003,
+      "loss": 4.1183,
+      "step": 2986
+    },
+    {
+      "epoch": 0.02987,
+      "grad_norm": 0.5762042827564682,
+      "learning_rate": 0.003,
+      "loss": 4.1051,
+      "step": 2987
+    },
+    {
+      "epoch": 0.02988,
+      "grad_norm": 0.5620617182133862,
+      "learning_rate": 0.003,
+      "loss": 4.1187,
+      "step": 2988
+    },
+    {
+      "epoch": 0.02989,
+      "grad_norm": 0.5894029162100797,
+      "learning_rate": 0.003,
+      "loss": 4.132,
+      "step": 2989
+    },
+    {
+      "epoch": 0.0299,
+      "grad_norm": 0.5837394519047691,
+      "learning_rate": 0.003,
+      "loss": 4.0884,
+      "step": 2990
+    },
+    {
+      "epoch": 0.02991,
+      "grad_norm": 0.5606386580984015,
+      "learning_rate": 0.003,
+      "loss": 4.1426,
+      "step": 2991
+    },
+    {
+      "epoch": 0.02992,
+      "grad_norm": 0.6147897819301783,
+      "learning_rate": 0.003,
+      "loss": 4.1358,
+      "step": 2992
+    },
+    {
+      "epoch": 0.02993,
+      "grad_norm": 0.6621349604754609,
+      "learning_rate": 0.003,
+      "loss": 4.1071,
+      "step": 2993
+    },
+    {
+      "epoch": 0.02994,
+      "grad_norm": 0.8082428932736804,
+      "learning_rate": 0.003,
+      "loss": 4.1211,
+      "step": 2994
+    },
+    {
+      "epoch": 0.02995,
+      "grad_norm": 0.9615508716688156,
+      "learning_rate": 0.003,
+      "loss": 4.1639,
+      "step": 2995
+    },
+    {
+      "epoch": 0.02996,
+      "grad_norm": 0.9867828831321791,
+      "learning_rate": 0.003,
+      "loss": 4.1577,
+      "step": 2996
+    },
+    {
+      "epoch": 0.02997,
+      "grad_norm": 0.6984846126670015,
+      "learning_rate": 0.003,
+      "loss": 4.112,
+      "step": 2997
+    },
+    {
+      "epoch": 0.02998,
+      "grad_norm": 0.7004936715134856,
+      "learning_rate": 0.003,
+      "loss": 4.1088,
+      "step": 2998
+    },
+    {
+      "epoch": 0.02999,
+      "grad_norm": 0.8087922821030329,
+      "learning_rate": 0.003,
+      "loss": 4.106,
+      "step": 2999
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8187211297372818,
+      "learning_rate": 0.003,
+      "loss": 4.1524,
+      "step": 3000
+    },
+    {
+      "epoch": 0.03001,
+      "grad_norm": 0.7682209061752201,
+      "learning_rate": 0.003,
+      "loss": 4.1192,
+      "step": 3001
+    },
+    {
+      "epoch": 0.03002,
+      "grad_norm": 0.7895156191994268,
+      "learning_rate": 0.003,
+      "loss": 4.1308,
+      "step": 3002
+    },
+    {
+      "epoch": 0.03003,
+      "grad_norm": 0.9473169165490495,
+      "learning_rate": 0.003,
+      "loss": 4.1162,
+      "step": 3003
+    },
+    {
+      "epoch": 0.03004,
+      "grad_norm": 0.9772459835659143,
+      "learning_rate": 0.003,
+      "loss": 4.0967,
+      "step": 3004
+    },
+    {
+      "epoch": 0.03005,
+      "grad_norm": 0.9271870841526649,
+      "learning_rate": 0.003,
+      "loss": 4.1515,
+      "step": 3005
+    },
+    {
+      "epoch": 0.03006,
+      "grad_norm": 0.81719364295006,
+      "learning_rate": 0.003,
+      "loss": 4.1395,
+      "step": 3006
+    },
+    {
+      "epoch": 0.03007,
+      "grad_norm": 0.8963225985425357,
+      "learning_rate": 0.003,
+      "loss": 4.1152,
+      "step": 3007
+    },
+    {
+      "epoch": 0.03008,
+      "grad_norm": 0.7818424581508167,
+      "learning_rate": 0.003,
+      "loss": 4.1409,
+      "step": 3008
+    },
+    {
+      "epoch": 0.03009,
+      "grad_norm": 0.8175800868725163,
+      "learning_rate": 0.003,
+      "loss": 4.1174,
+      "step": 3009
+    },
+    {
+      "epoch": 0.0301,
+      "grad_norm": 0.8915634122617913,
+      "learning_rate": 0.003,
+      "loss": 4.1397,
+      "step": 3010
+    },
+    {
+      "epoch": 0.03011,
+      "grad_norm": 0.9916010748888695,
+      "learning_rate": 0.003,
+      "loss": 4.1279,
+      "step": 3011
+    },
+    {
+      "epoch": 0.03012,
+      "grad_norm": 0.9009840506519443,
+      "learning_rate": 0.003,
+      "loss": 4.1303,
+      "step": 3012
+    },
+    {
+      "epoch": 0.03013,
+      "grad_norm": 0.8990400357883044,
+      "learning_rate": 0.003,
+      "loss": 4.0956,
+      "step": 3013
+    },
+    {
+      "epoch": 0.03014,
+      "grad_norm": 0.8701803132962008,
+      "learning_rate": 0.003,
+      "loss": 4.1241,
+      "step": 3014
+    },
+    {
+      "epoch": 0.03015,
+      "grad_norm": 0.9059467708321113,
+      "learning_rate": 0.003,
+      "loss": 4.1246,
+      "step": 3015
+    },
+    {
+      "epoch": 0.03016,
+      "grad_norm": 0.8139710481414894,
+      "learning_rate": 0.003,
+      "loss": 4.1273,
+      "step": 3016
+    },
+    {
+      "epoch": 0.03017,
+      "grad_norm": 0.9003169401719128,
+      "learning_rate": 0.003,
+      "loss": 4.1309,
+      "step": 3017
+    },
+    {
+      "epoch": 0.03018,
+      "grad_norm": 0.9148332662761703,
+      "learning_rate": 0.003,
+      "loss": 4.1282,
+      "step": 3018
+    },
+    {
+      "epoch": 0.03019,
+      "grad_norm": 0.8097272342713991,
+      "learning_rate": 0.003,
+      "loss": 4.1396,
+      "step": 3019
+    },
+    {
+      "epoch": 0.0302,
+      "grad_norm": 0.8482835840133595,
+      "learning_rate": 0.003,
+      "loss": 4.1312,
+      "step": 3020
+    },
+    {
+      "epoch": 0.03021,
+      "grad_norm": 0.7875221358593745,
+      "learning_rate": 0.003,
+      "loss": 4.1522,
+      "step": 3021
+    },
+    {
+      "epoch": 0.03022,
+      "grad_norm": 0.792143621479002,
+      "learning_rate": 0.003,
+      "loss": 4.1508,
+      "step": 3022
+    },
+    {
+      "epoch": 0.03023,
+      "grad_norm": 0.8897124061645492,
+      "learning_rate": 0.003,
+      "loss": 4.1675,
+      "step": 3023
+    },
+    {
+      "epoch": 0.03024,
+      "grad_norm": 1.016846130536088,
+      "learning_rate": 0.003,
+      "loss": 4.1643,
+      "step": 3024
+    },
+    {
+      "epoch": 0.03025,
+      "grad_norm": 0.9813514493806819,
+      "learning_rate": 0.003,
+      "loss": 4.1447,
+      "step": 3025
+    },
+    {
+      "epoch": 0.03026,
+      "grad_norm": 0.9856440178637326,
+      "learning_rate": 0.003,
+      "loss": 4.119,
+      "step": 3026
+    },
+    {
+      "epoch": 0.03027,
+      "grad_norm": 1.0170462610037305,
+      "learning_rate": 0.003,
+      "loss": 4.1467,
+      "step": 3027
+    },
+    {
+      "epoch": 0.03028,
+      "grad_norm": 0.9963929882829512,
+      "learning_rate": 0.003,
+      "loss": 4.1061,
+      "step": 3028
+    },
+    {
+      "epoch": 0.03029,
+      "grad_norm": 1.1230470374601593,
+      "learning_rate": 0.003,
+      "loss": 4.1539,
+      "step": 3029
+    },
+    {
+      "epoch": 0.0303,
+      "grad_norm": 0.9952469868078255,
+      "learning_rate": 0.003,
+      "loss": 4.1371,
+      "step": 3030
+    },
+    {
+      "epoch": 0.03031,
+      "grad_norm": 0.7974194139158707,
+      "learning_rate": 0.003,
+      "loss": 4.1117,
+      "step": 3031
+    },
+    {
+      "epoch": 0.03032,
+      "grad_norm": 0.813594268367442,
+      "learning_rate": 0.003,
+      "loss": 4.1409,
+      "step": 3032
+    },
+    {
+      "epoch": 0.03033,
+      "grad_norm": 0.8000080591024091,
+      "learning_rate": 0.003,
+      "loss": 4.1323,
+      "step": 3033
+    },
+    {
+      "epoch": 0.03034,
+      "grad_norm": 0.8702456358651927,
+      "learning_rate": 0.003,
+      "loss": 4.1388,
+      "step": 3034
+    },
+    {
+      "epoch": 0.03035,
+      "grad_norm": 0.972910117224573,
+      "learning_rate": 0.003,
+      "loss": 4.1701,
+      "step": 3035
+    },
+    {
+      "epoch": 0.03036,
+      "grad_norm": 0.946589194340261,
+      "learning_rate": 0.003,
+      "loss": 4.1449,
+      "step": 3036
+    },
+    {
+      "epoch": 0.03037,
+      "grad_norm": 1.0013129497511943,
+      "learning_rate": 0.003,
+      "loss": 4.13,
+      "step": 3037
+    },
+    {
+      "epoch": 0.03038,
+      "grad_norm": 1.0504414618218458,
+      "learning_rate": 0.003,
+      "loss": 4.1374,
+      "step": 3038
+    },
+    {
+      "epoch": 0.03039,
+      "grad_norm": 0.9552159328342077,
+      "learning_rate": 0.003,
+      "loss": 4.1371,
+      "step": 3039
+    },
+    {
+      "epoch": 0.0304,
+      "grad_norm": 0.8125546290590606,
+      "learning_rate": 0.003,
+      "loss": 4.1406,
+      "step": 3040
+    },
+    {
+      "epoch": 0.03041,
+      "grad_norm": 0.7829978292483236,
+      "learning_rate": 0.003,
+      "loss": 4.142,
+      "step": 3041
+    },
+    {
+      "epoch": 0.03042,
+      "grad_norm": 0.7611930258666472,
+      "learning_rate": 0.003,
+      "loss": 4.1671,
+      "step": 3042
+    },
+    {
+      "epoch": 0.03043,
+      "grad_norm": 0.7149717720566953,
+      "learning_rate": 0.003,
+      "loss": 4.118,
+      "step": 3043
+    },
+    {
+      "epoch": 0.03044,
+      "grad_norm": 0.7624230406515938,
+      "learning_rate": 0.003,
+      "loss": 4.12,
+      "step": 3044
+    },
+    {
+      "epoch": 0.03045,
+      "grad_norm": 0.8408403467918055,
+      "learning_rate": 0.003,
+      "loss": 4.1235,
+      "step": 3045
+    },
+    {
+      "epoch": 0.03046,
+      "grad_norm": 0.7851334779365589,
+      "learning_rate": 0.003,
+      "loss": 4.1328,
+      "step": 3046
+    },
+    {
+      "epoch": 0.03047,
+      "grad_norm": 0.7122809857507961,
+      "learning_rate": 0.003,
+      "loss": 4.1374,
+      "step": 3047
+    },
+    {
+      "epoch": 0.03048,
+      "grad_norm": 0.7964799855759337,
+      "learning_rate": 0.003,
+      "loss": 4.1208,
+      "step": 3048
+    },
+    {
+      "epoch": 0.03049,
+      "grad_norm": 0.887239863638981,
+      "learning_rate": 0.003,
+      "loss": 4.1105,
+      "step": 3049
+    },
+    {
+      "epoch": 0.0305,
+      "grad_norm": 0.9999215882595172,
+      "learning_rate": 0.003,
+      "loss": 4.1315,
+      "step": 3050
+    },
+    {
+      "epoch": 0.03051,
+      "grad_norm": 1.0055098406963119,
+      "learning_rate": 0.003,
+      "loss": 4.1455,
+      "step": 3051
+    },
+    {
+      "epoch": 0.03052,
+      "grad_norm": 0.9173091708203194,
+      "learning_rate": 0.003,
+      "loss": 4.146,
+      "step": 3052
+    },
+    {
+      "epoch": 0.03053,
+      "grad_norm": 0.8629407073466706,
+      "learning_rate": 0.003,
+      "loss": 4.1506,
+      "step": 3053
+    },
+    {
+      "epoch": 0.03054,
+      "grad_norm": 0.7744384974895352,
+      "learning_rate": 0.003,
+      "loss": 4.1267,
+      "step": 3054
+    },
+    {
+      "epoch": 0.03055,
+      "grad_norm": 0.6670797287560947,
+      "learning_rate": 0.003,
+      "loss": 4.1245,
+      "step": 3055
+    },
+    {
+      "epoch": 0.03056,
+      "grad_norm": 0.658467407296622,
+      "learning_rate": 0.003,
+      "loss": 4.134,
+      "step": 3056
+    },
+    {
+      "epoch": 0.03057,
+      "grad_norm": 0.6556069512553365,
+      "learning_rate": 0.003,
+      "loss": 4.1326,
+      "step": 3057
+    },
+    {
+      "epoch": 0.03058,
+      "grad_norm": 0.5717711332151091,
+      "learning_rate": 0.003,
+      "loss": 4.1029,
+      "step": 3058
+    },
+    {
+      "epoch": 0.03059,
+      "grad_norm": 0.493409526384602,
+      "learning_rate": 0.003,
+      "loss": 4.1103,
+      "step": 3059
+    },
+    {
+      "epoch": 0.0306,
+      "grad_norm": 0.5082424754021986,
+      "learning_rate": 0.003,
+      "loss": 4.0984,
+      "step": 3060
+    },
+    {
+      "epoch": 0.03061,
+      "grad_norm": 0.4898096783004193,
+      "learning_rate": 0.003,
+      "loss": 4.1212,
+      "step": 3061
+    },
+    {
+      "epoch": 0.03062,
+      "grad_norm": 0.5061719030039263,
+      "learning_rate": 0.003,
+      "loss": 4.144,
+      "step": 3062
+    },
+    {
+      "epoch": 0.03063,
+      "grad_norm": 0.6337804215976139,
+      "learning_rate": 0.003,
+      "loss": 4.0908,
+      "step": 3063
+    },
+    {
+      "epoch": 0.03064,
+      "grad_norm": 0.7363596095895648,
+      "learning_rate": 0.003,
+      "loss": 4.1252,
+      "step": 3064
+    },
+    {
+      "epoch": 0.03065,
+      "grad_norm": 0.7804431388454032,
+      "learning_rate": 0.003,
+      "loss": 4.0953,
+      "step": 3065
+    },
+    {
+      "epoch": 0.03066,
+      "grad_norm": 0.8048441855814717,
+      "learning_rate": 0.003,
+      "loss": 4.14,
+      "step": 3066
+    },
+    {
+      "epoch": 0.03067,
+      "grad_norm": 0.936660465754291,
+      "learning_rate": 0.003,
+      "loss": 4.1504,
+      "step": 3067
+    },
+    {
+      "epoch": 0.03068,
+      "grad_norm": 1.199111088765608,
+      "learning_rate": 0.003,
+      "loss": 4.1327,
+      "step": 3068
+    },
+    {
+      "epoch": 0.03069,
+      "grad_norm": 1.024080179977825,
+      "learning_rate": 0.003,
+      "loss": 4.1138,
+      "step": 3069
+    },
+    {
+      "epoch": 0.0307,
+      "grad_norm": 0.9087822982161698,
+      "learning_rate": 0.003,
+      "loss": 4.125,
+      "step": 3070
+    },
+    {
+      "epoch": 0.03071,
+      "grad_norm": 0.773169782429931,
+      "learning_rate": 0.003,
+      "loss": 4.121,
+      "step": 3071
+    },
+    {
+      "epoch": 0.03072,
+      "grad_norm": 0.7385627562002322,
+      "learning_rate": 0.003,
+      "loss": 4.1224,
+      "step": 3072
+    },
+    {
+      "epoch": 0.03073,
+      "grad_norm": 0.8590738611383513,
+      "learning_rate": 0.003,
+      "loss": 4.1272,
+      "step": 3073
+    },
+    {
+      "epoch": 0.03074,
+      "grad_norm": 0.7460812381201545,
+      "learning_rate": 0.003,
+      "loss": 4.1228,
+      "step": 3074
+    },
+    {
+      "epoch": 0.03075,
+      "grad_norm": 0.9060557817564706,
+      "learning_rate": 0.003,
+      "loss": 4.1505,
+      "step": 3075
+    },
+    {
+      "epoch": 0.03076,
+      "grad_norm": 0.9352871058483179,
+      "learning_rate": 0.003,
+      "loss": 4.1495,
+      "step": 3076
+    },
+    {
+      "epoch": 0.03077,
+      "grad_norm": 0.8648002758853153,
+      "learning_rate": 0.003,
+      "loss": 4.1446,
+      "step": 3077
+    },
+    {
+      "epoch": 0.03078,
+      "grad_norm": 0.7596804078061197,
+      "learning_rate": 0.003,
+      "loss": 4.1244,
+      "step": 3078
+    },
+    {
+      "epoch": 0.03079,
+      "grad_norm": 0.7757175523238533,
+      "learning_rate": 0.003,
+      "loss": 4.1133,
+      "step": 3079
+    },
+    {
+      "epoch": 0.0308,
+      "grad_norm": 0.8209364936522558,
+      "learning_rate": 0.003,
+      "loss": 4.1425,
+      "step": 3080
+    },
+    {
+      "epoch": 0.03081,
+      "grad_norm": 0.8477812326348159,
+      "learning_rate": 0.003,
+      "loss": 4.1152,
+      "step": 3081
+    },
+    {
+      "epoch": 0.03082,
+      "grad_norm": 0.8013577673715985,
+      "learning_rate": 0.003,
+      "loss": 4.1191,
+      "step": 3082
+    },
+    {
+      "epoch": 0.03083,
+      "grad_norm": 0.7134216891930999,
+      "learning_rate": 0.003,
+      "loss": 4.1183,
+      "step": 3083
+    },
+    {
+      "epoch": 0.03084,
+      "grad_norm": 0.7257190995117091,
+      "learning_rate": 0.003,
+      "loss": 4.1299,
+      "step": 3084
+    },
+    {
+      "epoch": 0.03085,
+      "grad_norm": 0.8088783726503086,
+      "learning_rate": 0.003,
+      "loss": 4.1259,
+      "step": 3085
+    },
+    {
+      "epoch": 0.03086,
+      "grad_norm": 0.8932889669749948,
+      "learning_rate": 0.003,
+      "loss": 4.1502,
+      "step": 3086
+    },
+    {
+      "epoch": 0.03087,
+      "grad_norm": 0.9271283171010463,
+      "learning_rate": 0.003,
+      "loss": 4.1256,
+      "step": 3087
+    },
+    {
+      "epoch": 0.03088,
+      "grad_norm": 1.0610338130344832,
+      "learning_rate": 0.003,
+      "loss": 4.1534,
+      "step": 3088
+    },
+    {
+      "epoch": 0.03089,
+      "grad_norm": 0.9082960513513103,
+      "learning_rate": 0.003,
+      "loss": 4.1137,
+      "step": 3089
+    },
+    {
+      "epoch": 0.0309,
+      "grad_norm": 0.8382419044332593,
+      "learning_rate": 0.003,
+      "loss": 4.1152,
+      "step": 3090
+    },
+    {
+      "epoch": 0.03091,
+      "grad_norm": 0.7721347908154246,
+      "learning_rate": 0.003,
+      "loss": 4.0786,
+      "step": 3091
+    },
+    {
+      "epoch": 0.03092,
+      "grad_norm": 0.5770289909808661,
+      "learning_rate": 0.003,
+      "loss": 4.0988,
+      "step": 3092
+    },
+    {
+      "epoch": 0.03093,
+      "grad_norm": 0.6048583748826258,
+      "learning_rate": 0.003,
+      "loss": 4.1335,
+      "step": 3093
+    },
+    {
+      "epoch": 0.03094,
+      "grad_norm": 0.5039228690347607,
+      "learning_rate": 0.003,
+      "loss": 4.1124,
+      "step": 3094
+    },
+    {
+      "epoch": 0.03095,
+      "grad_norm": 0.4932619648402882,
+      "learning_rate": 0.003,
+      "loss": 4.1254,
+      "step": 3095
+    },
+    {
+      "epoch": 0.03096,
+      "grad_norm": 0.44699630710548827,
+      "learning_rate": 0.003,
+      "loss": 4.1011,
+      "step": 3096
+    },
+    {
+      "epoch": 0.03097,
+      "grad_norm": 0.42800774052172613,
+      "learning_rate": 0.003,
+      "loss": 4.1233,
+      "step": 3097
+    },
+    {
+      "epoch": 0.03098,
+      "grad_norm": 0.5150974229730493,
+      "learning_rate": 0.003,
+      "loss": 4.1199,
+      "step": 3098
+    },
+    {
+      "epoch": 0.03099,
+      "grad_norm": 0.6808104155412404,
+      "learning_rate": 0.003,
+      "loss": 4.1152,
+      "step": 3099
+    },
+    {
+      "epoch": 0.031,
+      "grad_norm": 1.1124941338393186,
+      "learning_rate": 0.003,
+      "loss": 4.1238,
+      "step": 3100
+    },
+    {
+      "epoch": 0.03101,
+      "grad_norm": 1.0220222334230853,
+      "learning_rate": 0.003,
+      "loss": 4.1232,
+      "step": 3101
+    },
+    {
+      "epoch": 0.03102,
+      "grad_norm": 0.6635321325690858,
+      "learning_rate": 0.003,
+      "loss": 4.1121,
+      "step": 3102
+    },
+    {
+      "epoch": 0.03103,
+      "grad_norm": 0.7879413374703326,
+      "learning_rate": 0.003,
+      "loss": 4.1018,
+      "step": 3103
+    },
+    {
+      "epoch": 0.03104,
+      "grad_norm": 1.0336622472450878,
+      "learning_rate": 0.003,
+      "loss": 4.1177,
+      "step": 3104
+    },
+    {
+      "epoch": 0.03105,
+      "grad_norm": 0.8225297439648239,
+      "learning_rate": 0.003,
+      "loss": 4.0889,
+      "step": 3105
+    },
+    {
+      "epoch": 0.03106,
+      "grad_norm": 0.9200113560338348,
+      "learning_rate": 0.003,
+      "loss": 4.1106,
+      "step": 3106
+    },
+    {
+      "epoch": 0.03107,
+      "grad_norm": 0.8693837130689374,
+      "learning_rate": 0.003,
+      "loss": 4.0948,
+      "step": 3107
+    },
+    {
+      "epoch": 0.03108,
+      "grad_norm": 0.7481880861084276,
+      "learning_rate": 0.003,
+      "loss": 4.1016,
+      "step": 3108
+    },
+    {
+      "epoch": 0.03109,
+      "grad_norm": 0.7161901524718123,
+      "learning_rate": 0.003,
+      "loss": 4.1063,
+      "step": 3109
+    },
+    {
+      "epoch": 0.0311,
+      "grad_norm": 0.7717771970106354,
+      "learning_rate": 0.003,
+      "loss": 4.1301,
+      "step": 3110
+    },
+    {
+      "epoch": 0.03111,
+      "grad_norm": 0.7918589418817704,
+      "learning_rate": 0.003,
+      "loss": 4.1105,
+      "step": 3111
+    },
+    {
+      "epoch": 0.03112,
+      "grad_norm": 0.8112295219654945,
+      "learning_rate": 0.003,
+      "loss": 4.1307,
+      "step": 3112
+    },
+    {
+      "epoch": 0.03113,
+      "grad_norm": 0.8724658411048615,
+      "learning_rate": 0.003,
+      "loss": 4.1204,
+      "step": 3113
+    },
+    {
+      "epoch": 0.03114,
+      "grad_norm": 1.0040711142290948,
+      "learning_rate": 0.003,
+      "loss": 4.1346,
+      "step": 3114
+    },
+    {
+      "epoch": 0.03115,
+      "grad_norm": 1.2108105696861353,
+      "learning_rate": 0.003,
+      "loss": 4.1509,
+      "step": 3115
+    },
+    {
+      "epoch": 0.03116,
+      "grad_norm": 0.9386810822379548,
+      "learning_rate": 0.003,
+      "loss": 4.1225,
+      "step": 3116
+    },
+    {
+      "epoch": 0.03117,
+      "grad_norm": 0.7805193907050799,
+      "learning_rate": 0.003,
+      "loss": 4.1247,
+      "step": 3117
+    },
+    {
+      "epoch": 0.03118,
+      "grad_norm": 0.9100639956456111,
+      "learning_rate": 0.003,
+      "loss": 4.1304,
+      "step": 3118
+    },
+    {
+      "epoch": 0.03119,
+      "grad_norm": 0.9031494351027202,
+      "learning_rate": 0.003,
+      "loss": 4.1434,
+      "step": 3119
+    },
+    {
+      "epoch": 0.0312,
+      "grad_norm": 0.9920122086984918,
+      "learning_rate": 0.003,
+      "loss": 4.154,
+      "step": 3120
+    },
+    {
+      "epoch": 0.03121,
+      "grad_norm": 1.24679349546566,
+      "learning_rate": 0.003,
+      "loss": 4.1376,
+      "step": 3121
+    },
+    {
+      "epoch": 0.03122,
+      "grad_norm": 0.8914719736220875,
+      "learning_rate": 0.003,
+      "loss": 4.1494,
+      "step": 3122
+    },
+    {
+      "epoch": 0.03123,
+      "grad_norm": 0.7579991423334604,
+      "learning_rate": 0.003,
+      "loss": 4.1403,
+      "step": 3123
+    },
+    {
+      "epoch": 0.03124,
+      "grad_norm": 0.7621871567104618,
+      "learning_rate": 0.003,
+      "loss": 4.1182,
+      "step": 3124
+    },
+    {
+      "epoch": 0.03125,
+      "grad_norm": 0.8062917259017726,
+      "learning_rate": 0.003,
+      "loss": 4.1181,
+      "step": 3125
+    },
+    {
+      "epoch": 0.03126,
+      "grad_norm": 0.833712338155859,
+      "learning_rate": 0.003,
+      "loss": 4.1244,
+      "step": 3126
+    },
+    {
+      "epoch": 0.03127,
+      "grad_norm": 0.8998971933566047,
+      "learning_rate": 0.003,
+      "loss": 4.1339,
+      "step": 3127
+    },
+    {
+      "epoch": 0.03128,
+      "grad_norm": 0.9380372829323406,
+      "learning_rate": 0.003,
+      "loss": 4.126,
+      "step": 3128
+    },
+    {
+      "epoch": 0.03129,
+      "grad_norm": 0.9521516498255859,
+      "learning_rate": 0.003,
+      "loss": 4.1452,
+      "step": 3129
+    },
+    {
+      "epoch": 0.0313,
+      "grad_norm": 0.8951772330842159,
+      "learning_rate": 0.003,
+      "loss": 4.1496,
+      "step": 3130
+    },
+    {
+      "epoch": 0.03131,
+      "grad_norm": 0.7755621391919874,
+      "learning_rate": 0.003,
+      "loss": 4.1279,
+      "step": 3131
+    },
+    {
+      "epoch": 0.03132,
+      "grad_norm": 0.8129807814076891,
+      "learning_rate": 0.003,
+      "loss": 4.1478,
+      "step": 3132
+    },
+    {
+      "epoch": 0.03133,
+      "grad_norm": 0.9303049881214197,
+      "learning_rate": 0.003,
+      "loss": 4.1402,
+      "step": 3133
+    },
+    {
+      "epoch": 0.03134,
+      "grad_norm": 0.9415721268309895,
+      "learning_rate": 0.003,
+      "loss": 4.1412,
+      "step": 3134
+    },
+    {
+      "epoch": 0.03135,
+      "grad_norm": 0.9104183542988432,
+      "learning_rate": 0.003,
+      "loss": 4.1062,
+      "step": 3135
+    },
+    {
+      "epoch": 0.03136,
+      "grad_norm": 0.9100236222536682,
+      "learning_rate": 0.003,
+      "loss": 4.1242,
+      "step": 3136
+    },
+    {
+      "epoch": 0.03137,
+      "grad_norm": 0.8798690651639479,
+      "learning_rate": 0.003,
+      "loss": 4.1248,
+      "step": 3137
+    },
+    {
+      "epoch": 0.03138,
+      "grad_norm": 0.7903392795080382,
+      "learning_rate": 0.003,
+      "loss": 4.1249,
+      "step": 3138
+    },
+    {
+      "epoch": 0.03139,
+      "grad_norm": 0.7667246215003086,
+      "learning_rate": 0.003,
+      "loss": 4.1279,
+      "step": 3139
+    },
+    {
+      "epoch": 0.0314,
+      "grad_norm": 0.7544018295001089,
+      "learning_rate": 0.003,
+      "loss": 4.1255,
+      "step": 3140
+    },
+    {
+      "epoch": 0.03141,
+      "grad_norm": 0.6716524130054017,
+      "learning_rate": 0.003,
+      "loss": 4.154,
+      "step": 3141
+    },
+    {
+      "epoch": 0.03142,
+      "grad_norm": 0.6382378457514613,
+      "learning_rate": 0.003,
+      "loss": 4.1317,
+      "step": 3142
+    },
+    {
+      "epoch": 0.03143,
+      "grad_norm": 0.7526757290066967,
+      "learning_rate": 0.003,
+      "loss": 4.1206,
+      "step": 3143
+    },
+    {
+      "epoch": 0.03144,
+      "grad_norm": 0.938489146233584,
+      "learning_rate": 0.003,
+      "loss": 4.0931,
+      "step": 3144
+    },
+    {
+      "epoch": 0.03145,
+      "grad_norm": 1.0493538668141305,
+      "learning_rate": 0.003,
+      "loss": 4.1262,
+      "step": 3145
+    },
+    {
+      "epoch": 0.03146,
+      "grad_norm": 0.9048504750556206,
+      "learning_rate": 0.003,
+      "loss": 4.1308,
+      "step": 3146
+    },
+    {
+      "epoch": 0.03147,
+      "grad_norm": 0.9152634263998236,
+      "learning_rate": 0.003,
+      "loss": 4.1297,
+      "step": 3147
+    },
+    {
+      "epoch": 0.03148,
+      "grad_norm": 0.878607908387141,
+      "learning_rate": 0.003,
+      "loss": 4.1341,
+      "step": 3148
+    },
+    {
+      "epoch": 0.03149,
+      "grad_norm": 0.6944713264838394,
+      "learning_rate": 0.003,
+      "loss": 4.1322,
+      "step": 3149
+    },
+    {
+      "epoch": 0.0315,
+      "grad_norm": 0.6811030653109197,
+      "learning_rate": 0.003,
+      "loss": 4.1195,
+      "step": 3150
+    },
+    {
+      "epoch": 0.03151,
+      "grad_norm": 0.7166494028947639,
+      "learning_rate": 0.003,
+      "loss": 4.1292,
+      "step": 3151
+    },
+    {
+      "epoch": 0.03152,
+      "grad_norm": 0.7180332750616866,
+      "learning_rate": 0.003,
+      "loss": 4.1294,
+      "step": 3152
+    },
+    {
+      "epoch": 0.03153,
+      "grad_norm": 0.8734661321822886,
+      "learning_rate": 0.003,
+      "loss": 4.1407,
+      "step": 3153
+    },
+    {
+      "epoch": 0.03154,
+      "grad_norm": 0.9738879775883073,
+      "learning_rate": 0.003,
+      "loss": 4.1387,
+      "step": 3154
+    },
+    {
+      "epoch": 0.03155,
+      "grad_norm": 0.8618576527781467,
+      "learning_rate": 0.003,
+      "loss": 4.127,
+      "step": 3155
+    },
+    {
+      "epoch": 0.03156,
+      "grad_norm": 0.7300214083601039,
+      "learning_rate": 0.003,
+      "loss": 4.1063,
+      "step": 3156
+    },
+    {
+      "epoch": 0.03157,
+      "grad_norm": 0.6260105425384149,
+      "learning_rate": 0.003,
+      "loss": 4.1296,
+      "step": 3157
+    },
+    {
+      "epoch": 0.03158,
+      "grad_norm": 0.682384994546209,
+      "learning_rate": 0.003,
+      "loss": 4.1132,
+      "step": 3158
+    },
+    {
+      "epoch": 0.03159,
+      "grad_norm": 0.7718937672456789,
+      "learning_rate": 0.003,
+      "loss": 4.1181,
+      "step": 3159
+    },
+    {
+      "epoch": 0.0316,
+      "grad_norm": 0.6586793613223971,
+      "learning_rate": 0.003,
+      "loss": 4.1277,
+      "step": 3160
+    },
+    {
+      "epoch": 0.03161,
+      "grad_norm": 0.6187219141678427,
+      "learning_rate": 0.003,
+      "loss": 4.133,
+      "step": 3161
+    },
+    {
+      "epoch": 0.03162,
+      "grad_norm": 0.7732103677060469,
+      "learning_rate": 0.003,
+      "loss": 4.1389,
+      "step": 3162
+    },
+    {
+      "epoch": 0.03163,
+      "grad_norm": 0.9681852590593839,
+      "learning_rate": 0.003,
+      "loss": 4.129,
+      "step": 3163
+    },
+    {
+      "epoch": 0.03164,
+      "grad_norm": 0.983612124215567,
+      "learning_rate": 0.003,
+      "loss": 4.1291,
+      "step": 3164
+    },
+    {
+      "epoch": 0.03165,
+      "grad_norm": 0.8828238997897765,
+      "learning_rate": 0.003,
+      "loss": 4.0995,
+      "step": 3165
+    },
+    {
+      "epoch": 0.03166,
+      "grad_norm": 0.9082253980909,
+      "learning_rate": 0.003,
+      "loss": 4.1181,
+      "step": 3166
+    },
+    {
+      "epoch": 0.03167,
+      "grad_norm": 0.8883995374606223,
+      "learning_rate": 0.003,
+      "loss": 4.11,
+      "step": 3167
+    },
+    {
+      "epoch": 0.03168,
+      "grad_norm": 0.8814873185877611,
+      "learning_rate": 0.003,
+      "loss": 4.1367,
+      "step": 3168
+    },
+    {
+      "epoch": 0.03169,
+      "grad_norm": 0.8358140037799937,
+      "learning_rate": 0.003,
+      "loss": 4.1428,
+      "step": 3169
+    },
+    {
+      "epoch": 0.0317,
+      "grad_norm": 0.9091839519331014,
+      "learning_rate": 0.003,
+      "loss": 4.1008,
+      "step": 3170
+    },
+    {
+      "epoch": 0.03171,
+      "grad_norm": 0.7528676805145117,
+      "learning_rate": 0.003,
+      "loss": 4.1362,
+      "step": 3171
+    },
+    {
+      "epoch": 0.03172,
+      "grad_norm": 0.6184044517416146,
+      "learning_rate": 0.003,
+      "loss": 4.1235,
+      "step": 3172
+    },
+    {
+      "epoch": 0.03173,
+      "grad_norm": 0.7308731124601682,
+      "learning_rate": 0.003,
+      "loss": 4.0974,
+      "step": 3173
+    },
+    {
+      "epoch": 0.03174,
+      "grad_norm": 0.8588565114301099,
+      "learning_rate": 0.003,
+      "loss": 4.1133,
+      "step": 3174
+    },
+    {
+      "epoch": 0.03175,
+      "grad_norm": 1.2162741354980293,
+      "learning_rate": 0.003,
+      "loss": 4.1527,
+      "step": 3175
+    },
+    {
+      "epoch": 0.03176,
+      "grad_norm": 0.9397594133644137,
+      "learning_rate": 0.003,
+      "loss": 4.1148,
+      "step": 3176
+    },
+    {
+      "epoch": 0.03177,
+      "grad_norm": 0.9189809109610272,
+      "learning_rate": 0.003,
+      "loss": 4.1275,
+      "step": 3177
+    },
+    {
+      "epoch": 0.03178,
+      "grad_norm": 0.8047272875332512,
+      "learning_rate": 0.003,
+      "loss": 4.1657,
+      "step": 3178
+    },
+    {
+      "epoch": 0.03179,
+      "grad_norm": 0.932150949680668,
+      "learning_rate": 0.003,
+      "loss": 4.1268,
+      "step": 3179
+    },
+    {
+      "epoch": 0.0318,
+      "grad_norm": 0.9592877689147434,
+      "learning_rate": 0.003,
+      "loss": 4.1293,
+      "step": 3180
+    },
+    {
+      "epoch": 0.03181,
+      "grad_norm": 0.9595136242555029,
+      "learning_rate": 0.003,
+      "loss": 4.1419,
+      "step": 3181
+    },
+    {
+      "epoch": 0.03182,
+      "grad_norm": 0.9565837686401868,
+      "learning_rate": 0.003,
+      "loss": 4.142,
+      "step": 3182
+    },
+    {
+      "epoch": 0.03183,
+      "grad_norm": 0.9741938675445059,
+      "learning_rate": 0.003,
+      "loss": 4.1478,
+      "step": 3183
+    },
+    {
+      "epoch": 0.03184,
+      "grad_norm": 0.9857805137742832,
+      "learning_rate": 0.003,
+      "loss": 4.1174,
+      "step": 3184
+    },
+    {
+      "epoch": 0.03185,
+      "grad_norm": 0.9176062885733086,
+      "learning_rate": 0.003,
+      "loss": 4.1379,
+      "step": 3185
+    },
+    {
+      "epoch": 0.03186,
+      "grad_norm": 0.8884334076590751,
+      "learning_rate": 0.003,
+      "loss": 4.1367,
+      "step": 3186
+    },
+    {
+      "epoch": 0.03187,
+      "grad_norm": 0.8658125563673901,
+      "learning_rate": 0.003,
+      "loss": 4.0961,
+      "step": 3187
+    },
+    {
+      "epoch": 0.03188,
+      "grad_norm": 0.8999051178552305,
+      "learning_rate": 0.003,
+      "loss": 4.1334,
+      "step": 3188
+    },
+    {
+      "epoch": 0.03189,
+      "grad_norm": 0.7859126574348619,
+      "learning_rate": 0.003,
+      "loss": 4.1212,
+      "step": 3189
+    },
+    {
+      "epoch": 0.0319,
+      "grad_norm": 0.8606531029730619,
+      "learning_rate": 0.003,
+      "loss": 4.1147,
+      "step": 3190
+    },
+    {
+      "epoch": 0.03191,
+      "grad_norm": 0.9918147629707751,
+      "learning_rate": 0.003,
+      "loss": 4.149,
+      "step": 3191
+    },
+    {
+      "epoch": 0.03192,
+      "grad_norm": 1.0056390969187974,
+      "learning_rate": 0.003,
+      "loss": 4.132,
+      "step": 3192
+    },
+    {
+      "epoch": 0.03193,
+      "grad_norm": 0.8860275310714892,
+      "learning_rate": 0.003,
+      "loss": 4.1258,
+      "step": 3193
+    },
+    {
+      "epoch": 0.03194,
+      "grad_norm": 0.9173243755728898,
+      "learning_rate": 0.003,
+      "loss": 4.1534,
+      "step": 3194
+    },
+    {
+      "epoch": 0.03195,
+      "grad_norm": 0.8392077257777418,
+      "learning_rate": 0.003,
+      "loss": 4.1203,
+      "step": 3195
+    },
+    {
+      "epoch": 0.03196,
+      "grad_norm": 0.7352847590623333,
+      "learning_rate": 0.003,
+      "loss": 4.1274,
+      "step": 3196
+    },
+    {
+      "epoch": 0.03197,
+      "grad_norm": 0.715497912918627,
+      "learning_rate": 0.003,
+      "loss": 4.0961,
+      "step": 3197
+    },
+    {
+      "epoch": 0.03198,
+      "grad_norm": 0.6376443032868672,
+      "learning_rate": 0.003,
+      "loss": 4.1222,
+      "step": 3198
+    },
+    {
+      "epoch": 0.03199,
+      "grad_norm": 0.5811704510301426,
+      "learning_rate": 0.003,
+      "loss": 4.0959,
+      "step": 3199
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 0.5381272704276707,
+      "learning_rate": 0.003,
+      "loss": 4.116,
+      "step": 3200
+    },
+    {
+      "epoch": 0.03201,
+      "grad_norm": 0.5120575482476429,
+      "learning_rate": 0.003,
+      "loss": 4.0758,
+      "step": 3201
+    },
+    {
+      "epoch": 0.03202,
+      "grad_norm": 0.5958579649120576,
+      "learning_rate": 0.003,
+      "loss": 4.1164,
+      "step": 3202
+    },
+    {
+      "epoch": 0.03203,
+      "grad_norm": 0.7890366994930119,
+      "learning_rate": 0.003,
+      "loss": 4.1041,
+      "step": 3203
+    },
+    {
+      "epoch": 0.03204,
+      "grad_norm": 0.9077676416657651,
+      "learning_rate": 0.003,
+      "loss": 4.1246,
+      "step": 3204
+    },
+    {
+      "epoch": 0.03205,
+      "grad_norm": 0.754000654803963,
+      "learning_rate": 0.003,
+      "loss": 4.1157,
+      "step": 3205
+    },
+    {
+      "epoch": 0.03206,
+      "grad_norm": 0.600502725716816,
+      "learning_rate": 0.003,
+      "loss": 4.1078,
+      "step": 3206
+    },
+    {
+      "epoch": 0.03207,
+      "grad_norm": 0.6048683504208053,
+      "learning_rate": 0.003,
+      "loss": 4.0876,
+      "step": 3207
+    },
+    {
+      "epoch": 0.03208,
+      "grad_norm": 0.7129517255787963,
+      "learning_rate": 0.003,
+      "loss": 4.1136,
+      "step": 3208
+    },
+    {
+      "epoch": 0.03209,
+      "grad_norm": 0.8158420961472652,
+      "learning_rate": 0.003,
+      "loss": 4.1516,
+      "step": 3209
+    },
+    {
+      "epoch": 0.0321,
+      "grad_norm": 0.7005791064338225,
+      "learning_rate": 0.003,
+      "loss": 4.1273,
+      "step": 3210
+    },
+    {
+      "epoch": 0.03211,
+      "grad_norm": 0.6098399298288999,
+      "learning_rate": 0.003,
+      "loss": 4.0984,
+      "step": 3211
+    },
+    {
+      "epoch": 0.03212,
+      "grad_norm": 0.6538561751290226,
+      "learning_rate": 0.003,
+      "loss": 4.1089,
+      "step": 3212
+    },
+    {
+      "epoch": 0.03213,
+      "grad_norm": 0.5790618052777453,
+      "learning_rate": 0.003,
+      "loss": 4.1087,
+      "step": 3213
+    },
+    {
+      "epoch": 0.03214,
+      "grad_norm": 0.6875003076630353,
+      "learning_rate": 0.003,
+      "loss": 4.0786,
+      "step": 3214
+    },
+    {
+      "epoch": 0.03215,
+      "grad_norm": 0.799314724709517,
+      "learning_rate": 0.003,
+      "loss": 4.0949,
+      "step": 3215
+    },
+    {
+      "epoch": 0.03216,
+      "grad_norm": 0.8890107745271224,
+      "learning_rate": 0.003,
+      "loss": 4.1075,
+      "step": 3216
+    },
+    {
+      "epoch": 0.03217,
+      "grad_norm": 0.7725727639400009,
+      "learning_rate": 0.003,
+      "loss": 4.1044,
+      "step": 3217
+    },
+    {
+      "epoch": 0.03218,
+      "grad_norm": 0.7089931384963014,
+      "learning_rate": 0.003,
+      "loss": 4.096,
+      "step": 3218
+    },
+    {
+      "epoch": 0.03219,
+      "grad_norm": 0.755051345678096,
+      "learning_rate": 0.003,
+      "loss": 4.1025,
+      "step": 3219
+    },
+    {
+      "epoch": 0.0322,
+      "grad_norm": 0.9521310432872768,
+      "learning_rate": 0.003,
+      "loss": 4.1255,
+      "step": 3220
+    },
+    {
+      "epoch": 0.03221,
+      "grad_norm": 0.8616808278278768,
+      "learning_rate": 0.003,
+      "loss": 4.1285,
+      "step": 3221
+    },
+    {
+      "epoch": 0.03222,
+      "grad_norm": 0.873993324764959,
+      "learning_rate": 0.003,
+      "loss": 4.1188,
+      "step": 3222
+    },
+    {
+      "epoch": 0.03223,
+      "grad_norm": 0.8264507288043698,
+      "learning_rate": 0.003,
+      "loss": 4.1232,
+      "step": 3223
+    },
+    {
+      "epoch": 0.03224,
+      "grad_norm": 0.8289275480200801,
+      "learning_rate": 0.003,
+      "loss": 4.0847,
+      "step": 3224
+    },
+    {
+      "epoch": 0.03225,
+      "grad_norm": 0.7514528620660997,
+      "learning_rate": 0.003,
+      "loss": 4.1167,
+      "step": 3225
+    },
+    {
+      "epoch": 0.03226,
+      "grad_norm": 0.7197038673889892,
+      "learning_rate": 0.003,
+      "loss": 4.0974,
+      "step": 3226
+    },
+    {
+      "epoch": 0.03227,
+      "grad_norm": 0.8892719165128083,
+      "learning_rate": 0.003,
+      "loss": 4.1198,
+      "step": 3227
+    },
+    {
+      "epoch": 0.03228,
+      "grad_norm": 0.9570294472379692,
+      "learning_rate": 0.003,
+      "loss": 4.1042,
+      "step": 3228
+    },
+    {
+      "epoch": 0.03229,
+      "grad_norm": 0.9375751530879919,
+      "learning_rate": 0.003,
+      "loss": 4.12,
+      "step": 3229
+    },
+    {
+      "epoch": 0.0323,
+      "grad_norm": 0.9755322805010285,
+      "learning_rate": 0.003,
+      "loss": 4.1181,
+      "step": 3230
+    },
+    {
+      "epoch": 0.03231,
+      "grad_norm": 1.2740697032530537,
+      "learning_rate": 0.003,
+      "loss": 4.1456,
+      "step": 3231
+    },
+    {
+      "epoch": 0.03232,
+      "grad_norm": 0.8205568087154713,
+      "learning_rate": 0.003,
+      "loss": 4.1234,
+      "step": 3232
+    },
+    {
+      "epoch": 0.03233,
+      "grad_norm": 0.7719838274157066,
+      "learning_rate": 0.003,
+      "loss": 4.1376,
+      "step": 3233
+    },
+    {
+      "epoch": 0.03234,
+      "grad_norm": 0.7880067935874491,
+      "learning_rate": 0.003,
+      "loss": 4.1178,
+      "step": 3234
+    },
+    {
+      "epoch": 0.03235,
+      "grad_norm": 0.8025776008313478,
+      "learning_rate": 0.003,
+      "loss": 4.1267,
+      "step": 3235
+    },
+    {
+      "epoch": 0.03236,
+      "grad_norm": 1.0153689767602323,
+      "learning_rate": 0.003,
+      "loss": 4.124,
+      "step": 3236
+    },
+    {
+      "epoch": 0.03237,
+      "grad_norm": 1.0960896554896313,
+      "learning_rate": 0.003,
+      "loss": 4.1491,
+      "step": 3237
+    },
+    {
+      "epoch": 0.03238,
+      "grad_norm": 0.9653438379897337,
+      "learning_rate": 0.003,
+      "loss": 4.1538,
+      "step": 3238
+    },
+    {
+      "epoch": 0.03239,
+      "grad_norm": 0.9477216778661465,
+      "learning_rate": 0.003,
+      "loss": 4.1309,
+      "step": 3239
+    },
+    {
+      "epoch": 0.0324,
+      "grad_norm": 1.0000644526629416,
+      "learning_rate": 0.003,
+      "loss": 4.1522,
+      "step": 3240
+    },
+    {
+      "epoch": 0.03241,
+      "grad_norm": 0.9623250673117251,
+      "learning_rate": 0.003,
+      "loss": 4.1424,
+      "step": 3241
+    },
+    {
+      "epoch": 0.03242,
+      "grad_norm": 1.00080805991717,
+      "learning_rate": 0.003,
+      "loss": 4.1214,
+      "step": 3242
+    },
+    {
+      "epoch": 0.03243,
+      "grad_norm": 0.864922721509246,
+      "learning_rate": 0.003,
+      "loss": 4.1035,
+      "step": 3243
+    },
+    {
+      "epoch": 0.03244,
+      "grad_norm": 0.7896866879394736,
+      "learning_rate": 0.003,
+      "loss": 4.13,
+      "step": 3244
+    },
+    {
+      "epoch": 0.03245,
+      "grad_norm": 0.723475186710102,
+      "learning_rate": 0.003,
+      "loss": 4.1345,
+      "step": 3245
+    },
+    {
+      "epoch": 0.03246,
+      "grad_norm": 0.6786037554737262,
+      "learning_rate": 0.003,
+      "loss": 4.1199,
+      "step": 3246
+    },
+    {
+      "epoch": 0.03247,
+      "grad_norm": 0.6637920051672874,
+      "learning_rate": 0.003,
+      "loss": 4.1275,
+      "step": 3247
+    },
+    {
+      "epoch": 0.03248,
+      "grad_norm": 0.7161721376620257,
+      "learning_rate": 0.003,
+      "loss": 4.1439,
+      "step": 3248
+    },
+    {
+      "epoch": 0.03249,
+      "grad_norm": 0.7558986889375228,
+      "learning_rate": 0.003,
+      "loss": 4.1108,
+      "step": 3249
+    },
+    {
+      "epoch": 0.0325,
+      "grad_norm": 0.7122471010381649,
+      "learning_rate": 0.003,
+      "loss": 4.1412,
+      "step": 3250
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 100000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 25,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.28841839280128e+17,
+  "train_batch_size": 256,
+  "trial_name": null,
+  "trial_params": null
+}