diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,20860 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2974,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.145509046049739,
+      "learning_rate": 2.2222222222222224e-07,
+      "loss": 1.711,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.689988589471009,
+      "learning_rate": 4.444444444444445e-07,
+      "loss": 1.4669,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.780364756585921,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 1.6715,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.048605937437529,
+      "learning_rate": 8.88888888888889e-07,
+      "loss": 1.4211,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.16240700862966,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 1.5633,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.693656283744568,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 1.591,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.3397309091797345,
+      "learning_rate": 1.5555555555555558e-06,
+      "loss": 1.466,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.7069966875335805,
+      "learning_rate": 1.777777777777778e-06,
+      "loss": 1.4321,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.733533899795211,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.4301,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.30186223465512,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 1.5644,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.87079668439928,
+      "learning_rate": 2.4444444444444447e-06,
+      "loss": 1.1974,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.947066344397533,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.2685,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6074783313318366,
+      "learning_rate": 2.888888888888889e-06,
+      "loss": 1.4666,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0068550118525916,
+      "learning_rate": 3.1111111111111116e-06,
+      "loss": 1.2364,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.098100275582049,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.363,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.677203397909001,
+      "learning_rate": 3.555555555555556e-06,
+      "loss": 1.3066,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8199466923873038,
+      "learning_rate": 3.777777777777778e-06,
+      "loss": 1.2832,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.720002239921566,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.3814,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.781764070037142,
+      "learning_rate": 4.222222222222223e-06,
+      "loss": 1.1279,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1153153133901594,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 1.2338,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1174611850277554,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 1.1997,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.128355570743166,
+      "learning_rate": 4.888888888888889e-06,
+      "loss": 1.0005,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.109359427001423,
+      "learning_rate": 5.1111111111111115e-06,
+      "loss": 1.2416,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.670604968615722,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.1705,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4938343375472387,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 1.3773,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7328180037138103,
+      "learning_rate": 5.777777777777778e-06,
+      "loss": 1.1309,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6105361193449466,
+      "learning_rate": 6e-06,
+      "loss": 1.2843,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.219908527469652,
+      "learning_rate": 6.222222222222223e-06,
+      "loss": 1.1766,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.752011203914987,
+      "learning_rate": 6.444444444444445e-06,
+      "loss": 1.189,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.910491967197568,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.1235,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4774967222492834,
+      "learning_rate": 6.88888888888889e-06,
+      "loss": 1.0784,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8241113956357555,
+      "learning_rate": 7.111111111111112e-06,
+      "loss": 1.2289,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6752937519404,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 1.0088,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5606750065147825,
+      "learning_rate": 7.555555555555556e-06,
+      "loss": 1.3267,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1761953917152856,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 0.9905,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5608183166677263,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.0039,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7096513185013038,
+      "learning_rate": 8.222222222222222e-06,
+      "loss": 1.2876,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.381369408549807,
+      "learning_rate": 8.444444444444446e-06,
+      "loss": 0.9766,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.554403101991628,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 1.266,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7507001512887992,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 1.1901,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1071645545346411,
+      "learning_rate": 9.111111111111112e-06,
+      "loss": 1.1278,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.274826558543676,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 1.1951,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0484429249918603,
+      "learning_rate": 9.555555555555556e-06,
+      "loss": 1.0017,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2029512872260628,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 1.3877,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6669283438305957,
+      "learning_rate": 1e-05,
+      "loss": 1.1071,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1185827448769006,
+      "learning_rate": 1.0222222222222223e-05,
+      "loss": 1.1338,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5480507726410204,
+      "learning_rate": 1.0444444444444445e-05,
+      "loss": 1.0086,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.226752132086547,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 1.1507,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9973959645497247,
+      "learning_rate": 1.088888888888889e-05,
+      "loss": 1.3136,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3472944599433774,
+      "learning_rate": 1.1111111111111113e-05,
+      "loss": 0.9927,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1922239126921468,
+      "learning_rate": 1.1333333333333334e-05,
+      "loss": 1.2235,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.510333951696827,
+      "learning_rate": 1.1555555555555556e-05,
+      "loss": 1.1597,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8934009954827016,
+      "learning_rate": 1.177777777777778e-05,
+      "loss": 1.1583,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0354787221956951,
+      "learning_rate": 1.2e-05,
+      "loss": 1.0694,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9383151717164119,
+      "learning_rate": 1.2222222222222224e-05,
+      "loss": 1.0453,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.166212974761446,
+      "learning_rate": 1.2444444444444446e-05,
+      "loss": 1.1256,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.383558951762265,
+      "learning_rate": 1.2666666666666667e-05,
+      "loss": 1.4557,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8799718513766942,
+      "learning_rate": 1.288888888888889e-05,
+      "loss": 1.2921,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9443728136297348,
+      "learning_rate": 1.3111111111111113e-05,
+      "loss": 0.9132,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.132606824935557,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.2668,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.892355796013468,
+      "learning_rate": 1.3555555555555557e-05,
+      "loss": 0.9113,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.823815938370713,
+      "learning_rate": 1.377777777777778e-05,
+      "loss": 0.9135,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9522975357249323,
+      "learning_rate": 1.4e-05,
+      "loss": 1.2335,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0282129648179574,
+      "learning_rate": 1.4222222222222224e-05,
+      "loss": 1.1095,
+      "step": 64
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.276456052661155,
+      "learning_rate": 1.4444444444444446e-05,
+      "loss": 1.1339,
+      "step": 65
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9321663219408367,
+      "learning_rate": 1.4666666666666666e-05,
+      "loss": 1.0119,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9710810912034431,
+      "learning_rate": 1.488888888888889e-05,
+      "loss": 0.9126,
+      "step": 67
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9254847351230349,
+      "learning_rate": 1.5111111111111112e-05,
+      "loss": 1.1444,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0565393389838067,
+      "learning_rate": 1.5333333333333334e-05,
+      "loss": 1.082,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0128153522000864,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 1.1593,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1263836302868957,
+      "learning_rate": 1.577777777777778e-05,
+      "loss": 1.0389,
+      "step": 71
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0332549591012339,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.0365,
+      "step": 72
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1184436916754281,
+      "learning_rate": 1.6222222222222223e-05,
+      "loss": 1.0116,
+      "step": 73
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.990063730156102,
+      "learning_rate": 1.6444444444444444e-05,
+      "loss": 1.0325,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8645684240872512,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.0486,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9780485991917567,
+      "learning_rate": 1.688888888888889e-05,
+      "loss": 1.019,
+      "step": 76
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0712577324920656,
+      "learning_rate": 1.7111111111111112e-05,
+      "loss": 1.0647,
+      "step": 77
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0696953714327513,
+      "learning_rate": 1.7333333333333336e-05,
+      "loss": 0.8645,
+      "step": 78
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1131535335147456,
+      "learning_rate": 1.7555555555555556e-05,
+      "loss": 1.0354,
+      "step": 79
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0371604587346046,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 1.1331,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8342733483892693,
+      "learning_rate": 1.8e-05,
+      "loss": 1.1346,
+      "step": 81
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9681899565433489,
+      "learning_rate": 1.8222222222222224e-05,
+      "loss": 1.06,
+      "step": 82
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.854677393423223,
+      "learning_rate": 1.8444444444444448e-05,
+      "loss": 1.049,
+      "step": 83
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3419293207378742,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 0.9382,
+      "step": 84
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0126347655378625,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 1.111,
+      "step": 85
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1741588785383885,
+      "learning_rate": 1.9111111111111113e-05,
+      "loss": 1.0366,
+      "step": 86
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9461290749963134,
+      "learning_rate": 1.9333333333333333e-05,
+      "loss": 1.1457,
+      "step": 87
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2312063210933593,
+      "learning_rate": 1.9555555555555557e-05,
+      "loss": 0.8975,
+      "step": 88
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9976444606067569,
+      "learning_rate": 1.977777777777778e-05,
+      "loss": 1.1427,
+      "step": 89
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8818802891884793,
+      "learning_rate": 2e-05,
+      "loss": 0.9452,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 28.07310862926811,
+      "learning_rate": 1.9999994066933796e-05,
+      "loss": 1.3344,
+      "step": 91
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.939174652850251,
+      "learning_rate": 1.999997626774221e-05,
+      "loss": 1.2119,
+      "step": 92
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.959822413849329,
+      "learning_rate": 1.9999946602446377e-05,
+      "loss": 0.9324,
+      "step": 93
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0746493714558114,
+      "learning_rate": 1.9999905071081485e-05,
+      "loss": 1.0919,
+      "step": 94
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9056540699256841,
+      "learning_rate": 1.9999851673696827e-05,
+      "loss": 1.0376,
+      "step": 95
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9504495677266392,
+      "learning_rate": 1.999978641035576e-05,
+      "loss": 1.1545,
+      "step": 96
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0740299603209316,
+      "learning_rate": 1.9999709281135723e-05,
+      "loss": 1.1174,
+      "step": 97
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9788856780821036,
+      "learning_rate": 1.9999620286128243e-05,
+      "loss": 0.9033,
+      "step": 98
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5712339273878735,
+      "learning_rate": 1.9999519425438922e-05,
+      "loss": 1.2003,
+      "step": 99
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.517478775904822,
+      "learning_rate": 1.9999406699187443e-05,
+      "loss": 0.9131,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.039319380461107,
+      "learning_rate": 1.9999282107507567e-05,
+      "loss": 1.4205,
+      "step": 101
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0634370884859843,
+      "learning_rate": 1.9999145650547137e-05,
+      "loss": 1.0192,
+      "step": 102
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3045433361021674,
+      "learning_rate": 1.9998997328468075e-05,
+      "loss": 1.1243,
+      "step": 103
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0621539679699104,
+      "learning_rate": 1.9998837141446378e-05,
+      "loss": 1.1697,
+      "step": 104
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7989150094411379,
+      "learning_rate": 1.9998665089672132e-05,
+      "loss": 0.9719,
+      "step": 105
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0467529493566348,
+      "learning_rate": 1.9998481173349497e-05,
+      "loss": 1.1218,
+      "step": 106
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1558607512119348,
+      "learning_rate": 1.9998285392696702e-05,
+      "loss": 0.856,
+      "step": 107
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9394648481056181,
+      "learning_rate": 1.999807774794607e-05,
+      "loss": 1.1423,
+      "step": 108
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9827663783604025,
+      "learning_rate": 1.999785823934399e-05,
+      "loss": 1.0188,
+      "step": 109
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.029727828548156,
+      "learning_rate": 1.999762686715094e-05,
+      "loss": 1.2326,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9496869401101034,
+      "learning_rate": 1.9997383631641464e-05,
+      "loss": 1.1756,
+      "step": 111
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2087187358127875,
+      "learning_rate": 1.9997128533104192e-05,
+      "loss": 0.9469,
+      "step": 112
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9890842630953784,
+      "learning_rate": 1.9996861571841824e-05,
+      "loss": 1.1544,
+      "step": 113
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0371487141137545,
+      "learning_rate": 1.999658274817114e-05,
+      "loss": 0.9698,
+      "step": 114
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.160300182258626,
+      "learning_rate": 1.9996292062423e-05,
+      "loss": 1.059,
+      "step": 115
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0422134035850719,
+      "learning_rate": 1.999598951494233e-05,
+      "loss": 0.8791,
+      "step": 116
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9560080103157398,
+      "learning_rate": 1.9995675106088143e-05,
+      "loss": 0.8508,
+      "step": 117
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.765020296603597,
+      "learning_rate": 1.9995348836233517e-05,
+      "loss": 1.6636,
+      "step": 118
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2617236248853447,
+      "learning_rate": 1.9995010705765607e-05,
+      "loss": 1.0192,
+      "step": 119
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9393966137706408,
+      "learning_rate": 1.9994660715085645e-05,
+      "loss": 1.0995,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.025904059501948,
+      "learning_rate": 1.9994298864608936e-05,
+      "loss": 0.9876,
+      "step": 121
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.249882787474103,
+      "learning_rate": 1.9993925154764856e-05,
+      "loss": 1.0285,
+      "step": 122
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0365764632823515,
+      "learning_rate": 1.999353958599685e-05,
+      "loss": 1.0074,
+      "step": 123
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9478582145125564,
+      "learning_rate": 1.9993142158762443e-05,
+      "loss": 0.991,
+      "step": 124
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8866803192034362,
+      "learning_rate": 1.9992732873533223e-05,
+      "loss": 1.2045,
+      "step": 125
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9535580837376233,
+      "learning_rate": 1.999231173079486e-05,
+      "loss": 0.9507,
+      "step": 126
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.465863915214437,
+      "learning_rate": 1.9991878731047083e-05,
+      "loss": 0.9225,
+      "step": 127
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.016798204265439,
+      "learning_rate": 1.99914338748037e-05,
+      "loss": 0.9703,
+      "step": 128
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8911325155490301,
+      "learning_rate": 1.9990977162592575e-05,
+      "loss": 0.9629,
+      "step": 129
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0236158605809704,
+      "learning_rate": 1.9990508594955654e-05,
+      "loss": 1.0533,
+      "step": 130
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9661678372714788,
+      "learning_rate": 1.9990028172448945e-05,
+      "loss": 1.0432,
+      "step": 131
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.027634530516263,
+      "learning_rate": 1.9989535895642523e-05,
+      "loss": 1.1097,
+      "step": 132
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.955520702669643,
+      "learning_rate": 1.9989031765120534e-05,
+      "loss": 1.0219,
+      "step": 133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8118577539358406,
+      "learning_rate": 1.998851578148118e-05,
+      "loss": 1.0469,
+      "step": 134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8717765502605094,
+      "learning_rate": 1.998798794533674e-05,
+      "loss": 0.9876,
+      "step": 135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.17960281004929,
+      "learning_rate": 1.9987448257313545e-05,
+      "loss": 1.1367,
+      "step": 136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8416563893120645,
+      "learning_rate": 1.9986896718052e-05,
+      "loss": 1.1154,
+      "step": 137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.858690892725861,
+      "learning_rate": 1.9986333328206567e-05,
+      "loss": 0.9012,
+      "step": 138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9332102449560893,
+      "learning_rate": 1.9985758088445774e-05,
+      "loss": 1.0137,
+      "step": 139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2347571221209575,
+      "learning_rate": 1.9985170999452207e-05,
+      "loss": 1.0305,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.340758429287578,
+      "learning_rate": 1.9984572061922515e-05,
+      "loss": 0.8832,
+      "step": 141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9407134647691217,
+      "learning_rate": 1.99839612765674e-05,
+      "loss": 1.1197,
+      "step": 142
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1920521846803322,
+      "learning_rate": 1.9983338644111635e-05,
+      "loss": 1.042,
+      "step": 143
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.001834471097025,
+      "learning_rate": 1.998270416529404e-05,
+      "loss": 1.1142,
+      "step": 144
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.73409478965739,
+      "learning_rate": 1.9982057840867495e-05,
+      "loss": 0.9457,
+      "step": 145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8249642425651692,
+      "learning_rate": 1.998139967159894e-05,
+      "loss": 0.9988,
+      "step": 146
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2566797079917007,
+      "learning_rate": 1.9980729658269364e-05,
+      "loss": 1.0385,
+      "step": 147
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9609974920840131,
+      "learning_rate": 1.9980047801673818e-05,
+      "loss": 0.9153,
+      "step": 148
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7668631759787643,
+      "learning_rate": 1.99793541026214e-05,
+      "loss": 0.9022,
+      "step": 149
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8076438260099845,
+      "learning_rate": 1.997864856193526e-05,
+      "loss": 0.9951,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8534821358649156,
+      "learning_rate": 1.9977931180452606e-05,
+      "loss": 1.0849,
+      "step": 151
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8815388383929056,
+      "learning_rate": 1.997720195902469e-05,
+      "loss": 0.9589,
+      "step": 152
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8596179465433073,
+      "learning_rate": 1.9976460898516817e-05,
+      "loss": 1.059,
+      "step": 153
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3785460844786657,
+      "learning_rate": 1.9975707999808338e-05,
+      "loss": 0.7652,
+      "step": 154
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9994156405822782,
+      "learning_rate": 1.997494326379265e-05,
+      "loss": 1.025,
+      "step": 155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9134317702035045,
+      "learning_rate": 1.997416669137721e-05,
+      "loss": 1.1192,
+      "step": 156
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.768350727163631,
+      "learning_rate": 1.9973378283483497e-05,
+      "loss": 0.9397,
+      "step": 157
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8345967460733459,
+      "learning_rate": 1.997257804104705e-05,
+      "loss": 1.0718,
+      "step": 158
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.982582694315392,
+      "learning_rate": 1.9971765965017446e-05,
+      "loss": 0.9381,
+      "step": 159
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8719372000872521,
+      "learning_rate": 1.997094205635831e-05,
+      "loss": 0.9063,
+      "step": 160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7928597450415266,
+      "learning_rate": 1.9970106316047295e-05,
+      "loss": 1.072,
+      "step": 161
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7937821573988728,
+      "learning_rate": 1.996925874507611e-05,
+      "loss": 0.8513,
+      "step": 162
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2300780949685524,
+      "learning_rate": 1.9968399344450492e-05,
+      "loss": 0.9218,
+      "step": 163
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.908693322180365,
+      "learning_rate": 1.996752811519021e-05,
+      "loss": 0.8755,
+      "step": 164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8140528332020808,
+      "learning_rate": 1.9966645058329087e-05,
+      "loss": 0.9224,
+      "step": 165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7818763734282062,
+      "learning_rate": 1.996575017491496e-05,
+      "loss": 1.0764,
+      "step": 166
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9319644992271967,
+      "learning_rate": 1.9964843466009714e-05,
+      "loss": 1.004,
+      "step": 167
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7999957287230275,
+      "learning_rate": 1.9963924932689265e-05,
+      "loss": 1.2018,
+      "step": 168
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8204209219616087,
+      "learning_rate": 1.9962994576043555e-05,
+      "loss": 1.0093,
+      "step": 169
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8721952396533811,
+      "learning_rate": 1.9962052397176553e-05,
+      "loss": 1.1051,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7860413415691008,
+      "learning_rate": 1.9961098397206264e-05,
+      "loss": 0.7301,
+      "step": 171
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8823148215195609,
+      "learning_rate": 1.996013257726472e-05,
+      "loss": 0.9603,
+      "step": 172
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6932235885319473,
+      "learning_rate": 1.9959154938497968e-05,
+      "loss": 0.8932,
+      "step": 173
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.788176905022184,
+      "learning_rate": 1.9958165482066094e-05,
+      "loss": 0.895,
+      "step": 174
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9093841147394239,
+      "learning_rate": 1.99571642091432e-05,
+      "loss": 1.1527,
+      "step": 175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8271354697572706,
+      "learning_rate": 1.9956151120917405e-05,
+      "loss": 0.8722,
+      "step": 176
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8073734199165604,
+      "learning_rate": 1.9955126218590858e-05,
+      "loss": 1.0911,
+      "step": 177
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8421611992846783,
+      "learning_rate": 1.9954089503379716e-05,
+      "loss": 0.8116,
+      "step": 178
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1630900568755946,
+      "learning_rate": 1.9953040976514164e-05,
+      "loss": 1.1317,
+      "step": 179
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8666849048396474,
+      "learning_rate": 1.9951980639238398e-05,
+      "loss": 0.8519,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8111653141684296,
+      "learning_rate": 1.9950908492810624e-05,
+      "loss": 0.8803,
+      "step": 181
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8916190836513918,
+      "learning_rate": 1.994982453850307e-05,
+      "loss": 1.1622,
+      "step": 182
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9163981406037838,
+      "learning_rate": 1.9948728777601966e-05,
+      "loss": 1.0444,
+      "step": 183
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8347434874048546,
+      "learning_rate": 1.994762121140756e-05,
+      "loss": 0.9855,
+      "step": 184
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8187470252833041,
+      "learning_rate": 1.99465018412341e-05,
+      "loss": 0.8782,
+      "step": 185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3573581082901731,
+      "learning_rate": 1.994537066840985e-05,
+      "loss": 0.8165,
+      "step": 186
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8092197273373762,
+      "learning_rate": 1.9944227694277076e-05,
+      "loss": 1.0054,
+      "step": 187
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.898125390537374,
+      "learning_rate": 1.9943072920192042e-05,
+      "loss": 0.9791,
+      "step": 188
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8001053121928392,
+      "learning_rate": 1.994190634752502e-05,
+      "loss": 1.1427,
+      "step": 189
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7244139887619134,
+      "learning_rate": 1.9940727977660275e-05,
+      "loss": 0.7282,
+      "step": 190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8260279197783507,
+      "learning_rate": 1.993953781199609e-05,
+      "loss": 1.0319,
+      "step": 191
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8531307174691874,
+      "learning_rate": 1.993833585194472e-05,
+      "loss": 1.1516,
+      "step": 192
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8322244620944192,
+      "learning_rate": 1.9937122098932428e-05,
+      "loss": 0.9177,
+      "step": 193
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4558810448015924,
+      "learning_rate": 1.9935896554399472e-05,
+      "loss": 1.0415,
+      "step": 194
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8820778194852221,
+      "learning_rate": 1.99346592198001e-05,
+      "loss": 0.9487,
+      "step": 195
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8825507297224408,
+      "learning_rate": 1.9933410096602544e-05,
+      "loss": 1.0812,
+      "step": 196
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.864019065789547,
+      "learning_rate": 1.993214918628904e-05,
+      "loss": 0.8798,
+      "step": 197
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7470402950013341,
+      "learning_rate": 1.993087649035579e-05,
+      "loss": 0.7039,
+      "step": 198
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0654505851435063,
+      "learning_rate": 1.9929592010312997e-05,
+      "loss": 1.1054,
+      "step": 199
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8809331559897549,
+      "learning_rate": 1.9928295747684846e-05,
+      "loss": 0.9162,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8654552244827939,
+      "learning_rate": 1.992698770400949e-05,
+      "loss": 1.2065,
+      "step": 201
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9077140269564298,
+      "learning_rate": 1.992566788083908e-05,
+      "loss": 0.9333,
+      "step": 202
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2957232170976485,
+      "learning_rate": 1.9924336279739732e-05,
+      "loss": 1.0172,
+      "step": 203
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7991376030279981,
+      "learning_rate": 1.992299290229154e-05,
+      "loss": 1.0731,
+      "step": 204
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4680976813532676,
+      "learning_rate": 1.9921637750088577e-05,
+      "loss": 1.0206,
+      "step": 205
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7013837457658997,
+      "learning_rate": 1.9920270824738882e-05,
+      "loss": 0.7849,
+      "step": 206
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8699165923787442,
+      "learning_rate": 1.9918892127864463e-05,
+      "loss": 0.9154,
+      "step": 207
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.916550033458954,
+      "learning_rate": 1.9917501661101307e-05,
+      "loss": 1.0381,
+      "step": 208
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7724702010056272,
+      "learning_rate": 1.991609942609936e-05,
+      "loss": 0.8242,
+      "step": 209
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9950145348363237,
+      "learning_rate": 1.9914685424522525e-05,
+      "loss": 0.9021,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8561436534277924,
+      "learning_rate": 1.9913259658048684e-05,
+      "loss": 1.0552,
+      "step": 211
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8755524226206489,
+      "learning_rate": 1.9911822128369668e-05,
+      "loss": 1.0664,
+      "step": 212
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7981014609334808,
+      "learning_rate": 1.991037283719126e-05,
+      "loss": 0.9429,
+      "step": 213
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0662306788203761,
+      "learning_rate": 1.990891178623322e-05,
+      "loss": 0.9915,
+      "step": 214
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2528502541708226,
+      "learning_rate": 1.990743897722925e-05,
+      "loss": 1.086,
+      "step": 215
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7761532741303336,
+      "learning_rate": 1.9905954411926992e-05,
+      "loss": 0.7258,
+      "step": 216
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8822814724011605,
+      "learning_rate": 1.990445809208806e-05,
+      "loss": 0.8426,
+      "step": 217
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.221973086228616,
+      "learning_rate": 1.990295001948801e-05,
+      "loss": 1.0827,
+      "step": 218
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6489958624789225,
+      "learning_rate": 1.9901430195916336e-05,
+      "loss": 0.8003,
+      "step": 219
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1010789359188058,
+      "learning_rate": 1.989989862317648e-05,
+      "loss": 1.1315,
+      "step": 220
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9421037718680494,
+      "learning_rate": 1.9898355303085833e-05,
+      "loss": 0.8515,
+      "step": 221
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8520131678355319,
+      "learning_rate": 1.989680023747571e-05,
+      "loss": 0.8449,
+      "step": 222
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8235167261620184,
+      "learning_rate": 1.989523342819138e-05,
+      "loss": 1.0571,
+      "step": 223
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7740191433040625,
+      "learning_rate": 1.9893654877092032e-05,
+      "loss": 0.6982,
+      "step": 224
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0640993145594082,
+      "learning_rate": 1.9892064586050804e-05,
+      "loss": 1.0489,
+      "step": 225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9119434159980002,
+      "learning_rate": 1.989046255695475e-05,
+      "loss": 1.0142,
+      "step": 226
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8337901133120824,
+      "learning_rate": 1.9888848791704863e-05,
+      "loss": 1.0728,
+      "step": 227
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8210292555380148,
+      "learning_rate": 1.9887223292216058e-05,
+      "loss": 0.8802,
+      "step": 228
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8525973966223237,
+      "learning_rate": 1.988558606041717e-05,
+      "loss": 0.9086,
+      "step": 229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8906629139683332,
+      "learning_rate": 1.9883937098250962e-05,
+      "loss": 1.1219,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3130849476436859,
+      "learning_rate": 1.988227640767412e-05,
+      "loss": 0.9624,
+      "step": 231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8157736700302571,
+      "learning_rate": 1.988060399065723e-05,
+      "loss": 1.0526,
+      "step": 232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8506343438966375,
+      "learning_rate": 1.9878919849184813e-05,
+      "loss": 0.8392,
+      "step": 233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1068055695946513,
+      "learning_rate": 1.9877223985255293e-05,
+      "loss": 0.7682,
+      "step": 234
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0247938061138755,
+      "learning_rate": 1.9875516400881002e-05,
+      "loss": 1.0343,
+      "step": 235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7960429659812753,
+      "learning_rate": 1.9873797098088185e-05,
+      "loss": 0.8162,
+      "step": 236
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8121638640653978,
+      "learning_rate": 1.987206607891699e-05,
+      "loss": 1.0591,
+      "step": 237
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9850099724701252,
+      "learning_rate": 1.9870323345421457e-05,
+      "loss": 0.9953,
+      "step": 238
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8839214718269696,
+      "learning_rate": 1.986856889966955e-05,
+      "loss": 1.0472,
+      "step": 239
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.909285977108231,
+      "learning_rate": 1.986680274374311e-05,
+      "loss": 0.9201,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.944508194200787,
+      "learning_rate": 1.9865024879737885e-05,
+      "loss": 0.9749,
+      "step": 241
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8585690920357815,
+      "learning_rate": 1.9863235309763505e-05,
+      "loss": 0.9847,
+      "step": 242
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.890577440367318,
+      "learning_rate": 1.9861434035943507e-05,
+      "loss": 0.8751,
+      "step": 243
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8789811133418142,
+      "learning_rate": 1.9859621060415302e-05,
+      "loss": 1.1749,
+      "step": 244
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0155817782439478,
+      "learning_rate": 1.985779638533019e-05,
+      "loss": 0.8749,
+      "step": 245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8645408950378959,
+      "learning_rate": 1.9855960012853354e-05,
+      "loss": 1.0873,
+      "step": 246
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0672227510572743,
+      "learning_rate": 1.985411194516386e-05,
+      "loss": 0.8203,
+      "step": 247
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9928509075839661,
+      "learning_rate": 1.9852252184454646e-05,
+      "loss": 0.8518,
+      "step": 248
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0316778453821742,
+      "learning_rate": 1.9850380732932538e-05,
+      "loss": 1.1259,
+      "step": 249
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9129669757243072,
+      "learning_rate": 1.9848497592818214e-05,
+      "loss": 1.0012,
+      "step": 250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8125186441529946,
+      "learning_rate": 1.9846602766346238e-05,
+      "loss": 0.791,
+      "step": 251
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0693554534167662,
+      "learning_rate": 1.9844696255765036e-05,
+      "loss": 0.9263,
+      "step": 252
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9157037078612111,
+      "learning_rate": 1.9842778063336898e-05,
+      "loss": 0.8893,
+      "step": 253
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8067316276149176,
+      "learning_rate": 1.9840848191337976e-05,
+      "loss": 1.0045,
+      "step": 254
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.956961614610815,
+      "learning_rate": 1.9838906642058283e-05,
+      "loss": 1.0261,
+      "step": 255
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6869647859497687,
+      "learning_rate": 1.9836953417801686e-05,
+      "loss": 1.0184,
+      "step": 256
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0325185421584318,
+      "learning_rate": 1.9834988520885912e-05,
+      "loss": 0.9315,
+      "step": 257
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7122509230220453,
+      "learning_rate": 1.9833011953642525e-05,
+      "loss": 0.9195,
+      "step": 258
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8474919989225517,
+      "learning_rate": 1.983102371841695e-05,
+      "loss": 0.9226,
+      "step": 259
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9853376880762219,
+      "learning_rate": 1.9829023817568455e-05,
+      "loss": 0.8125,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9486031874364855,
+      "learning_rate": 1.982701225347015e-05,
+      "loss": 0.9933,
+      "step": 261
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7538525510033952,
+      "learning_rate": 1.9824989028508975e-05,
+      "loss": 0.9058,
+      "step": 262
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8348405956438738,
+      "learning_rate": 1.9822954145085723e-05,
+      "loss": 1.0449,
+      "step": 263
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7933560320802892,
+      "learning_rate": 1.9820907605615013e-05,
+      "loss": 0.818,
+      "step": 264
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8228340120983578,
+      "learning_rate": 1.9818849412525294e-05,
+      "loss": 0.9363,
+      "step": 265
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8506044224291006,
+      "learning_rate": 1.981677956825885e-05,
+      "loss": 1.0201,
+      "step": 266
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.029496849315608,
+      "learning_rate": 1.9814698075271776e-05,
+      "loss": 1.0858,
+      "step": 267
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8215549968005228,
+      "learning_rate": 1.9812604936034004e-05,
+      "loss": 0.9446,
+      "step": 268
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7756394069275098,
+      "learning_rate": 1.9810500153029285e-05,
+      "loss": 0.6926,
+      "step": 269
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.003213754153978,
+      "learning_rate": 1.9808383728755176e-05,
+      "loss": 1.1303,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8300225576207885,
+      "learning_rate": 1.980625566572306e-05,
+      "loss": 0.838,
+      "step": 271
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.913511357077783,
+      "learning_rate": 1.9804115966458118e-05,
+      "loss": 0.9633,
+      "step": 272
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.708960544387244,
+      "learning_rate": 1.9801964633499352e-05,
+      "loss": 1.0457,
+      "step": 273
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8612709984776217,
+      "learning_rate": 1.9799801669399556e-05,
+      "loss": 0.8057,
+      "step": 274
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.245851108422822,
+      "learning_rate": 1.9797627076725338e-05,
+      "loss": 0.9489,
+      "step": 275
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7804341778759951,
+      "learning_rate": 1.9795440858057094e-05,
+      "loss": 0.9727,
+      "step": 276
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.763175849140008,
+      "learning_rate": 1.979324301598902e-05,
+      "loss": 0.8061,
+      "step": 277
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7613582177979776,
+      "learning_rate": 1.979103355312911e-05,
+      "loss": 1.0272,
+      "step": 278
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8936166637216687,
+      "learning_rate": 1.9788812472099138e-05,
+      "loss": 0.8395,
+      "step": 279
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8470466692587256,
+      "learning_rate": 1.9786579775534663e-05,
+      "loss": 1.0632,
+      "step": 280
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9954925920221578,
+      "learning_rate": 1.978433546608504e-05,
+      "loss": 0.9843,
+      "step": 281
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.829258532800488,
+      "learning_rate": 1.9782079546413396e-05,
+      "loss": 1.044,
+      "step": 282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9490085281537983,
+      "learning_rate": 1.9779812019196632e-05,
+      "loss": 1.072,
+      "step": 283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7679789482502705,
+      "learning_rate": 1.9777532887125425e-05,
+      "loss": 0.8417,
+      "step": 284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9376699552547922,
+      "learning_rate": 1.9775242152904228e-05,
+      "loss": 1.1225,
+      "step": 285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3126084569214906,
+      "learning_rate": 1.9772939819251247e-05,
+      "loss": 0.8214,
+      "step": 286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.060174190073838,
+      "learning_rate": 1.9770625888898477e-05,
+      "loss": 0.6928,
+      "step": 287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8718995052378542,
+      "learning_rate": 1.9768300364591643e-05,
+      "loss": 1.0416,
+      "step": 288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8689088456145415,
+      "learning_rate": 1.9765963249090254e-05,
+      "loss": 1.0778,
+      "step": 289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9898070519633684,
+      "learning_rate": 1.9763614545167552e-05,
+      "loss": 0.925,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9493097193079656,
+      "learning_rate": 1.9761254255610553e-05,
+      "loss": 0.9008,
+      "step": 291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8087048560789514,
+      "learning_rate": 1.9758882383219995e-05,
+      "loss": 0.9977,
+      "step": 292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9544054598758029,
+      "learning_rate": 1.975649893081038e-05,
+      "loss": 0.8586,
+      "step": 293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8909722244814637,
+      "learning_rate": 1.9754103901209943e-05,
+      "loss": 1.0622,
+      "step": 294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.102247689920816,
+      "learning_rate": 1.975169729726066e-05,
+      "loss": 0.9679,
+      "step": 295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4517908393008148,
+      "learning_rate": 1.9749279121818235e-05,
+      "loss": 0.7955,
+      "step": 296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7793660285185525,
+      "learning_rate": 1.9746849377752113e-05,
+      "loss": 0.897,
+      "step": 297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9350917653593643,
+      "learning_rate": 1.9744408067945455e-05,
+      "loss": 0.8636,
+      "step": 298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7692891672181558,
+      "learning_rate": 1.974195519529515e-05,
+      "loss": 1.1026,
+      "step": 299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.040165574019115,
+      "learning_rate": 1.9739490762711812e-05,
+      "loss": 0.8506,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7874647187250486,
+      "learning_rate": 1.9737014773119774e-05,
+      "loss": 0.9727,
+      "step": 301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.021397964664765,
+      "learning_rate": 1.973452722945707e-05,
+      "loss": 0.8288,
+      "step": 302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9701193689212246,
+      "learning_rate": 1.9732028134675457e-05,
+      "loss": 0.9873,
+      "step": 303
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2257655330480595,
+      "learning_rate": 1.972951749174039e-05,
+      "loss": 0.7961,
+      "step": 304
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8188623565512945,
+      "learning_rate": 1.9726995303631038e-05,
+      "loss": 0.9722,
+      "step": 305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8873237788327742,
+      "learning_rate": 1.9724461573340258e-05,
+      "loss": 1.0153,
+      "step": 306
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9514822311880522,
+      "learning_rate": 1.9721916303874607e-05,
+      "loss": 0.9885,
+      "step": 307
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9402587113472795,
+      "learning_rate": 1.971935949825434e-05,
+      "loss": 1.0062,
+      "step": 308
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.264625109132174,
+      "learning_rate": 1.971679115951339e-05,
+      "loss": 0.8365,
+      "step": 309
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8182699448541717,
+      "learning_rate": 1.971421129069939e-05,
+      "loss": 0.969,
+      "step": 310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.810054243806555,
+      "learning_rate": 1.9711619894873637e-05,
+      "loss": 1.2192,
+      "step": 311
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8253441165769226,
+      "learning_rate": 1.9709016975111123e-05,
+      "loss": 0.7947,
+      "step": 312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.021959747179416,
+      "learning_rate": 1.9706402534500504e-05,
+      "loss": 0.7134,
+      "step": 313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.202809059576842,
+      "learning_rate": 1.9703776576144106e-05,
+      "loss": 0.954,
+      "step": 314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8219706935941108,
+      "learning_rate": 1.9701139103157937e-05,
+      "loss": 1.0629,
+      "step": 315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9994475151476777,
+      "learning_rate": 1.9698490118671647e-05,
+      "loss": 1.0246,
+      "step": 316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0534433990993741,
+      "learning_rate": 1.9695829625828563e-05,
+      "loss": 0.8811,
+      "step": 317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9142063592896705,
+      "learning_rate": 1.9693157627785654e-05,
+      "loss": 0.9574,
+      "step": 318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0310472492067906,
+      "learning_rate": 1.9690474127713554e-05,
+      "loss": 0.9441,
+      "step": 319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.813227638404784,
+      "learning_rate": 1.9687779128796537e-05,
+      "loss": 0.9759,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7601429215779439,
+      "learning_rate": 1.9685072634232524e-05,
+      "loss": 0.9447,
+      "step": 321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.440774905457145,
+      "learning_rate": 1.9682354647233078e-05,
+      "loss": 0.7429,
+      "step": 322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8697328605865349,
+      "learning_rate": 1.96796251710234e-05,
+      "loss": 1.0626,
+      "step": 323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9959478346389822,
+      "learning_rate": 1.9676884208842323e-05,
+      "loss": 0.8934,
+      "step": 324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8718999626944196,
+      "learning_rate": 1.9674131763942305e-05,
+      "loss": 0.9757,
+      "step": 325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1299377797979526,
+      "learning_rate": 1.9671367839589438e-05,
+      "loss": 0.8568,
+      "step": 326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8901234970529214,
+      "learning_rate": 1.9668592439063426e-05,
+      "loss": 0.9395,
+      "step": 327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7724400466321772,
+      "learning_rate": 1.9665805565657602e-05,
+      "loss": 1.0139,
+      "step": 328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.005643061448652,
+      "learning_rate": 1.9663007222678904e-05,
+      "loss": 1.0145,
+      "step": 329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8329474543961511,
+      "learning_rate": 1.966019741344788e-05,
+      "loss": 1.1101,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7196949875513099,
+      "learning_rate": 1.9657376141298692e-05,
+      "loss": 0.6523,
+      "step": 331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7826727211315921,
+      "learning_rate": 1.96545434095791e-05,
+      "loss": 0.9805,
+      "step": 332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9623788066702109,
+      "learning_rate": 1.9651699221650454e-05,
+      "loss": 1.0347,
+      "step": 333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0264367247822503,
+      "learning_rate": 1.964884358088771e-05,
+      "loss": 1.0774,
+      "step": 334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6533916516386217,
+      "learning_rate": 1.9645976490679402e-05,
+      "loss": 0.7609,
+      "step": 335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7445641202607042,
+      "learning_rate": 1.964309795442767e-05,
+      "loss": 0.9396,
+      "step": 336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.779350542948613,
+      "learning_rate": 1.9640207975548216e-05,
+      "loss": 0.9764,
+      "step": 337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9760463573766373,
+      "learning_rate": 1.9637306557470325e-05,
+      "loss": 0.8578,
+      "step": 338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7566577994250876,
+      "learning_rate": 1.963439370363686e-05,
+      "loss": 0.7715,
+      "step": 339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7571208108274197,
+      "learning_rate": 1.9631469417504253e-05,
+      "loss": 0.8684,
+      "step": 340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9747404483038127,
+      "learning_rate": 1.96285337025425e-05,
+      "loss": 1.0465,
+      "step": 341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8839794708603397,
+      "learning_rate": 1.962558656223516e-05,
+      "loss": 1.1049,
+      "step": 342
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7992367772760274,
+      "learning_rate": 1.9622628000079347e-05,
+      "loss": 0.924,
+      "step": 343
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9470595808220512,
+      "learning_rate": 1.9619658019585736e-05,
+      "loss": 1.0498,
+      "step": 344
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.776659772656201,
+      "learning_rate": 1.9616676624278535e-05,
+      "loss": 0.9,
+      "step": 345
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9546744734912924,
+      "learning_rate": 1.9613683817695516e-05,
+      "loss": 1.0773,
+      "step": 346
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9800564643153523,
+      "learning_rate": 1.9610679603387977e-05,
+      "loss": 0.9666,
+      "step": 347
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8393893094123767,
+      "learning_rate": 1.960766398492076e-05,
+      "loss": 0.9028,
+      "step": 348
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6729378438770978,
+      "learning_rate": 1.9604636965872243e-05,
+      "loss": 0.7619,
+      "step": 349
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.954095597483001,
+      "learning_rate": 1.9601598549834318e-05,
+      "loss": 0.8636,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8866908655702838,
+      "learning_rate": 1.9598548740412414e-05,
+      "loss": 1.0199,
+      "step": 351
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8631039745772586,
+      "learning_rate": 1.959548754122548e-05,
+      "loss": 0.8632,
+      "step": 352
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.073891009080767,
+      "learning_rate": 1.959241495590597e-05,
+      "loss": 0.9709,
+      "step": 353
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7329531107566277,
+      "learning_rate": 1.958933098809985e-05,
+      "loss": 1.1408,
+      "step": 354
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0035744724919882,
+      "learning_rate": 1.9586235641466607e-05,
+      "loss": 0.9866,
+      "step": 355
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9510324838472417,
+      "learning_rate": 1.9583128919679218e-05,
+      "loss": 1.0279,
+      "step": 356
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9511000508938103,
+      "learning_rate": 1.958001082642415e-05,
+      "loss": 0.7051,
+      "step": 357
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1260201556750307,
+      "learning_rate": 1.957688136540139e-05,
+      "loss": 1.0213,
+      "step": 358
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7636637833209765,
+      "learning_rate": 1.9573740540324385e-05,
+      "loss": 0.7138,
+      "step": 359
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7907778315038314,
+      "learning_rate": 1.9570588354920085e-05,
+      "loss": 0.9242,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8941558354146251,
+      "learning_rate": 1.9567424812928918e-05,
+      "loss": 1.0871,
+      "step": 361
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9786121023248314,
+      "learning_rate": 1.9564249918104777e-05,
+      "loss": 0.8639,
+      "step": 362
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8557430292372328,
+      "learning_rate": 1.9561063674215038e-05,
+      "loss": 1.0127,
+      "step": 363
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7455354143454493,
+      "learning_rate": 1.955786608504054e-05,
+      "loss": 1.0134,
+      "step": 364
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.027601559515422,
+      "learning_rate": 1.955465715437559e-05,
+      "loss": 0.9384,
+      "step": 365
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1406141875757172,
+      "learning_rate": 1.9551436886027938e-05,
+      "loss": 0.7997,
+      "step": 366
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7852406663914134,
+      "learning_rate": 1.9548205283818803e-05,
+      "loss": 0.9374,
+      "step": 367
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.80913184392625,
+      "learning_rate": 1.9544962351582844e-05,
+      "loss": 1.0634,
+      "step": 368
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6820221784584387,
+      "learning_rate": 1.954170809316817e-05,
+      "loss": 0.9866,
+      "step": 369
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8990661003319323,
+      "learning_rate": 1.953844251243633e-05,
+      "loss": 0.9586,
+      "step": 370
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7974356822240232,
+      "learning_rate": 1.9535165613262296e-05,
+      "loss": 0.7558,
+      "step": 371
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9156743942246397,
+      "learning_rate": 1.953187739953449e-05,
+      "loss": 0.7877,
+      "step": 372
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7806162487447357,
+      "learning_rate": 1.952857787515474e-05,
+      "loss": 0.9774,
+      "step": 373
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.890887179251295,
+      "learning_rate": 1.9525267044038318e-05,
+      "loss": 1.0052,
+      "step": 374
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8299728032205362,
+      "learning_rate": 1.9521944910113887e-05,
+      "loss": 0.9872,
+      "step": 375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7949860781305633,
+      "learning_rate": 1.951861147732354e-05,
+      "loss": 0.6586,
+      "step": 376
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8739204174477981,
+      "learning_rate": 1.951526674962278e-05,
+      "loss": 1.0139,
+      "step": 377
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0197485641787933,
+      "learning_rate": 1.951191073098049e-05,
+      "loss": 1.0806,
+      "step": 378
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.862558451869881,
+      "learning_rate": 1.9508543425378983e-05,
+      "loss": 0.8719,
+      "step": 379
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8380290849441115,
+      "learning_rate": 1.9505164836813938e-05,
+      "loss": 1.0484,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7646094551007471,
+      "learning_rate": 1.950177496929443e-05,
+      "loss": 0.8421,
+      "step": 381
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8123862223181021,
+      "learning_rate": 1.949837382684293e-05,
+      "loss": 1.0146,
+      "step": 382
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4192392096752982,
+      "learning_rate": 1.9494961413495273e-05,
+      "loss": 0.6722,
+      "step": 383
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8297271680960732,
+      "learning_rate": 1.9491537733300677e-05,
+      "loss": 0.9411,
+      "step": 384
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8187540675313177,
+      "learning_rate": 1.9488102790321726e-05,
+      "loss": 0.9907,
+      "step": 385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9060779879804844,
+      "learning_rate": 1.9484656588634362e-05,
+      "loss": 0.8756,
+      "step": 386
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9125530758206074,
+      "learning_rate": 1.94811991323279e-05,
+      "loss": 1.1912,
+      "step": 387
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9677616954035728,
+      "learning_rate": 1.947773042550501e-05,
+      "loss": 0.9993,
+      "step": 388
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7590912094643,
+      "learning_rate": 1.947425047228169e-05,
+      "loss": 0.8917,
+      "step": 389
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0271895301930498,
+      "learning_rate": 1.947075927678731e-05,
+      "loss": 1.0288,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5706227474488768,
+      "learning_rate": 1.946725684316456e-05,
+      "loss": 0.9839,
+      "step": 391
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7617868805116956,
+      "learning_rate": 1.946374317556948e-05,
+      "loss": 0.6964,
+      "step": 392
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9018215643832843,
+      "learning_rate": 1.946021827817144e-05,
+      "loss": 0.8789,
+      "step": 393
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.890662919291753,
+      "learning_rate": 1.945668215515312e-05,
+      "loss": 1.028,
+      "step": 394
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9602579241465282,
+      "learning_rate": 1.9453134810710536e-05,
+      "loss": 0.7556,
+      "step": 395
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9072519231280196,
+      "learning_rate": 1.9449576249053008e-05,
+      "loss": 1.0539,
+      "step": 396
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9941647214232195,
+      "learning_rate": 1.9446006474403177e-05,
+      "loss": 0.8858,
+      "step": 397
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9717266112844459,
+      "learning_rate": 1.9442425490996987e-05,
+      "loss": 1.0057,
+      "step": 398
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7884381929537104,
+      "learning_rate": 1.9438833303083677e-05,
+      "loss": 0.9338,
+      "step": 399
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8235567536666165,
+      "learning_rate": 1.9435229914925785e-05,
+      "loss": 0.972,
+      "step": 400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8374948337925906,
+      "learning_rate": 1.943161533079914e-05,
+      "loss": 1.0043,
+      "step": 401
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7870714312145136,
+      "learning_rate": 1.9427989554992854e-05,
+      "loss": 0.6401,
+      "step": 402
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4104335401284773,
+      "learning_rate": 1.9424352591809322e-05,
+      "loss": 0.9807,
+      "step": 403
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.825797402086701,
+      "learning_rate": 1.942070444556421e-05,
+      "loss": 0.9433,
+      "step": 404
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.027167442370408,
+      "learning_rate": 1.941704512058646e-05,
+      "loss": 0.8146,
+      "step": 405
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9631464384810691,
+      "learning_rate": 1.9413374621218275e-05,
+      "loss": 1.0656,
+      "step": 406
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.933614821672341,
+      "learning_rate": 1.9409692951815117e-05,
+      "loss": 0.8076,
+      "step": 407
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8556188135400654,
+      "learning_rate": 1.94060001167457e-05,
+      "loss": 1.0653,
+      "step": 408
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9997443393230303,
+      "learning_rate": 1.9402296120392003e-05,
+      "loss": 0.9524,
+      "step": 409
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7492733104612114,
+      "learning_rate": 1.9398580967149222e-05,
+      "loss": 0.6375,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9688229906877118,
+      "learning_rate": 1.9394854661425816e-05,
+      "loss": 1.1128,
+      "step": 411
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9010827956339198,
+      "learning_rate": 1.939111720764347e-05,
+      "loss": 0.9208,
+      "step": 412
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8390016918850447,
+      "learning_rate": 1.938736861023709e-05,
+      "loss": 1.0563,
+      "step": 413
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9759495682609128,
+      "learning_rate": 1.9383608873654817e-05,
+      "loss": 0.8522,
+      "step": 414
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9337861743785696,
+      "learning_rate": 1.9379838002357995e-05,
+      "loss": 0.8894,
+      "step": 415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8787320505806754,
+      "learning_rate": 1.93760560008212e-05,
+      "loss": 1.0724,
+      "step": 416
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7882793295569603,
+      "learning_rate": 1.9372262873532207e-05,
+      "loss": 0.9523,
+      "step": 417
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8905840384652451,
+      "learning_rate": 1.936845862499198e-05,
+      "loss": 0.9807,
+      "step": 418
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7970399066097561,
+      "learning_rate": 1.9364643259714696e-05,
+      "loss": 0.8139,
+      "step": 419
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.671952613858416,
+      "learning_rate": 1.936081678222772e-05,
+      "loss": 0.8395,
+      "step": 420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8210540643689855,
+      "learning_rate": 1.9356979197071597e-05,
+      "loss": 0.9134,
+      "step": 421
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7778960575240391,
+      "learning_rate": 1.935313050880006e-05,
+      "loss": 0.767,
+      "step": 422
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9076788187274797,
+      "learning_rate": 1.934927072198001e-05,
+      "loss": 1.0964,
+      "step": 423
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9820158200019157,
+      "learning_rate": 1.9345399841191527e-05,
+      "loss": 0.8864,
+      "step": 424
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8305705514329524,
+      "learning_rate": 1.9341517871027843e-05,
+      "loss": 0.9678,
+      "step": 425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8442256453997006,
+      "learning_rate": 1.933762481609536e-05,
+      "loss": 1.0022,
+      "step": 426
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.807093202541644,
+      "learning_rate": 1.9333720681013618e-05,
+      "loss": 1.0125,
+      "step": 427
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9109606068318379,
+      "learning_rate": 1.932980547041533e-05,
+      "loss": 0.8184,
+      "step": 428
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.234295394802666,
+      "learning_rate": 1.9325879188946328e-05,
+      "loss": 0.8536,
+      "step": 429
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8089608175530729,
+      "learning_rate": 1.9321941841265596e-05,
+      "loss": 1.1113,
+      "step": 430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7964030962002292,
+      "learning_rate": 1.9317993432045236e-05,
+      "loss": 0.9221,
+      "step": 431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7249858840203752,
+      "learning_rate": 1.9314033965970483e-05,
+      "loss": 0.9795,
+      "step": 432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9490825624241788,
+      "learning_rate": 1.9310063447739697e-05,
+      "loss": 0.8361,
+      "step": 433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0649774808494774,
+      "learning_rate": 1.9306081882064346e-05,
+      "loss": 0.8959,
+      "step": 434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8438280469481267,
+      "learning_rate": 1.9302089273669002e-05,
+      "loss": 1.0903,
+      "step": 435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6282692734015429,
+      "learning_rate": 1.929808562729136e-05,
+      "loss": 0.7903,
+      "step": 436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.798739081607902,
+      "learning_rate": 1.9294070947682186e-05,
+      "loss": 0.9781,
+      "step": 437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9471680839977574,
+      "learning_rate": 1.929004523960536e-05,
+      "loss": 0.9871,
+      "step": 438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8745550925705619,
+      "learning_rate": 1.9286008507837838e-05,
+      "loss": 1.0259,
+      "step": 439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7530510493414099,
+      "learning_rate": 1.928196075716966e-05,
+      "loss": 0.8382,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.035908864024238,
+      "learning_rate": 1.9277901992403943e-05,
+      "loss": 0.9818,
+      "step": 441
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7375816923739792,
+      "learning_rate": 1.9273832218356867e-05,
+      "loss": 1.107,
+      "step": 442
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8308296202100398,
+      "learning_rate": 1.926975143985768e-05,
+      "loss": 0.7771,
+      "step": 443
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6604903113542485,
+      "learning_rate": 1.926565966174869e-05,
+      "loss": 0.8915,
+      "step": 444
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8386818963424485,
+      "learning_rate": 1.9261556888885257e-05,
+      "loss": 0.8821,
+      "step": 445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6132662469666368,
+      "learning_rate": 1.9257443126135776e-05,
+      "loss": 0.834,
+      "step": 446
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7654183245584124,
+      "learning_rate": 1.9253318378381703e-05,
+      "loss": 0.8499,
+      "step": 447
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7352116226185325,
+      "learning_rate": 1.924918265051752e-05,
+      "loss": 0.9164,
+      "step": 448
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7419584929709256,
+      "learning_rate": 1.924503594745072e-05,
+      "loss": 0.9538,
+      "step": 449
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9437480643054456,
+      "learning_rate": 1.9240878274101854e-05,
+      "loss": 0.8695,
+      "step": 450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7636808642010086,
+      "learning_rate": 1.9236709635404462e-05,
+      "loss": 0.9317,
+      "step": 451
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7242004321317458,
+      "learning_rate": 1.923253003630511e-05,
+      "loss": 0.9601,
+      "step": 452
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8542338843290717,
+      "learning_rate": 1.9228339481763365e-05,
+      "loss": 1.0535,
+      "step": 453
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6305953521996935,
+      "learning_rate": 1.9224137976751797e-05,
+      "loss": 0.6755,
+      "step": 454
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.941999235564769,
+      "learning_rate": 1.9219925526255956e-05,
+      "loss": 0.9317,
+      "step": 455
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.738306625962872,
+      "learning_rate": 1.921570213527441e-05,
+      "loss": 0.9518,
+      "step": 456
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9681171410531718,
+      "learning_rate": 1.921146780881867e-05,
+      "loss": 0.8339,
+      "step": 457
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9044825105066895,
+      "learning_rate": 1.9207222551913263e-05,
+      "loss": 0.9709,
+      "step": 458
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8010859845445788,
+      "learning_rate": 1.9202966369595653e-05,
+      "loss": 1.0361,
+      "step": 459
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1015390771739801,
+      "learning_rate": 1.919869926691629e-05,
+      "loss": 0.8121,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8049787448411682,
+      "learning_rate": 1.9194421248938576e-05,
+      "loss": 0.9457,
+      "step": 461
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7003891801682062,
+      "learning_rate": 1.9190132320738856e-05,
+      "loss": 0.9779,
+      "step": 462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8424269049213262,
+      "learning_rate": 1.9185832487406433e-05,
+      "loss": 0.8141,
+      "step": 463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8657291308083285,
+      "learning_rate": 1.918152175404355e-05,
+      "loss": 0.8777,
+      "step": 464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9519798244048794,
+      "learning_rate": 1.917720012576537e-05,
+      "loss": 0.8586,
+      "step": 465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7597675670106401,
+      "learning_rate": 1.9172867607700004e-05,
+      "loss": 0.9548,
+      "step": 466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1283659859719528,
+      "learning_rate": 1.9168524204988475e-05,
+      "loss": 0.9571,
+      "step": 467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.794787598906403,
+      "learning_rate": 1.9164169922784717e-05,
+      "loss": 1.0521,
+      "step": 468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8667914664446931,
+      "learning_rate": 1.9159804766255583e-05,
+      "loss": 0.9477,
+      "step": 469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7509505505393578,
+      "learning_rate": 1.915542874058082e-05,
+      "loss": 1.0496,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.764807049641001,
+      "learning_rate": 1.9151041850953086e-05,
+      "loss": 0.8207,
+      "step": 471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8370190641710725,
+      "learning_rate": 1.9146644102577917e-05,
+      "loss": 0.8024,
+      "step": 472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9475629156173441,
+      "learning_rate": 1.914223550067374e-05,
+      "loss": 0.6694,
+      "step": 473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7781504778556082,
+      "learning_rate": 1.913781605047186e-05,
+      "loss": 1.0145,
+      "step": 474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7642388759007182,
+      "learning_rate": 1.9133385757216458e-05,
+      "loss": 1.0272,
+      "step": 475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7713740221315241,
+      "learning_rate": 1.9128944626164576e-05,
+      "loss": 0.9521,
+      "step": 476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7658433524863524,
+      "learning_rate": 1.912449266258612e-05,
+      "loss": 0.7938,
+      "step": 477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8998411651037341,
+      "learning_rate": 1.912002987176385e-05,
+      "loss": 1.0284,
+      "step": 478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.79527207105774,
+      "learning_rate": 1.9115556258993368e-05,
+      "loss": 0.8447,
+      "step": 479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7778678481498615,
+      "learning_rate": 1.911107182958313e-05,
+      "loss": 1.0378,
+      "step": 480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.828240463203576,
+      "learning_rate": 1.910657658885441e-05,
+      "loss": 0.9775,
+      "step": 481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0176789748448587,
+      "learning_rate": 1.910207054214133e-05,
+      "loss": 0.7574,
+      "step": 482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8286908864689141,
+      "learning_rate": 1.9097553694790816e-05,
+      "loss": 0.9022,
+      "step": 483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8022240898962768,
+      "learning_rate": 1.9093026052162623e-05,
+      "loss": 0.8228,
+      "step": 484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8358371072702176,
+      "learning_rate": 1.9088487619629314e-05,
+      "loss": 1.1086,
+      "step": 485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.795995260853623,
+      "learning_rate": 1.9083938402576246e-05,
+      "loss": 0.8352,
+      "step": 486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7684997311876186,
+      "learning_rate": 1.907937840640159e-05,
+      "loss": 0.9844,
+      "step": 487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8673021612065877,
+      "learning_rate": 1.9074807636516286e-05,
+      "loss": 0.822,
+      "step": 488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8535398638046439,
+      "learning_rate": 1.907022609834408e-05,
+      "loss": 1.1414,
+      "step": 489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7681238508920412,
+      "learning_rate": 1.9065633797321484e-05,
+      "loss": 0.6777,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7878043040791883,
+      "learning_rate": 1.9061030738897778e-05,
+      "loss": 0.9152,
+      "step": 491
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7963561292949844,
+      "learning_rate": 1.9056416928535016e-05,
+      "loss": 1.0017,
+      "step": 492
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9181925487850104,
+      "learning_rate": 1.905179237170801e-05,
+      "loss": 1.0826,
+      "step": 493
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7064843938712639,
+      "learning_rate": 1.904715707390431e-05,
+      "loss": 0.8866,
+      "step": 494
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7631317626204064,
+      "learning_rate": 1.9042511040624235e-05,
+      "loss": 0.8885,
+      "step": 495
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7648798917709139,
+      "learning_rate": 1.903785427738082e-05,
+      "loss": 0.8475,
+      "step": 496
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8269401266098649,
+      "learning_rate": 1.9033186789699846e-05,
+      "loss": 1.169,
+      "step": 497
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8170063147790075,
+      "learning_rate": 1.902850858311981e-05,
+      "loss": 0.8569,
+      "step": 498
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6305855328493665,
+      "learning_rate": 1.902381966319194e-05,
+      "loss": 0.6749,
+      "step": 499
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8349195411996646,
+      "learning_rate": 1.9019120035480174e-05,
+      "loss": 0.8478,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8036086015824385,
+      "learning_rate": 1.9014409705561142e-05,
+      "loss": 1.0752,
+      "step": 501
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8785779218991436,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 0.9954,
+      "step": 502
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8244989719540887,
+      "learning_rate": 1.9004956961471355e-05,
+      "loss": 0.8919,
+      "step": 503
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7557091448614597,
+      "learning_rate": 1.9000214558517343e-05,
+      "loss": 0.9142,
+      "step": 504
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9568414373194462,
+      "learning_rate": 1.899546147578956e-05,
+      "loss": 0.9919,
+      "step": 505
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7507648941962622,
+      "learning_rate": 1.8990697718928084e-05,
+      "loss": 0.9887,
+      "step": 506
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5830076238003876,
+      "learning_rate": 1.8985923293585638e-05,
+      "loss": 0.7451,
+      "step": 507
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.049418049866393,
+      "learning_rate": 1.898113820542763e-05,
+      "loss": 1.0719,
+      "step": 508
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8649589628106087,
+      "learning_rate": 1.8976342460132097e-05,
+      "loss": 0.8444,
+      "step": 509
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7743475213194182,
+      "learning_rate": 1.8971536063389745e-05,
+      "loss": 0.8715,
+      "step": 510
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9044061930073355,
+      "learning_rate": 1.89667190209039e-05,
+      "loss": 0.935,
+      "step": 511
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9406369207081097,
+      "learning_rate": 1.8961891338390534e-05,
+      "loss": 0.9099,
+      "step": 512
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0480954203860386,
+      "learning_rate": 1.895705302157824e-05,
+      "loss": 1.0464,
+      "step": 513
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8834985144283968,
+      "learning_rate": 1.8952204076208224e-05,
+      "loss": 0.8579,
+      "step": 514
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8317052053996427,
+      "learning_rate": 1.894734450803431e-05,
+      "loss": 1.0077,
+      "step": 515
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9682623973694197,
+      "learning_rate": 1.8942474322822924e-05,
+      "loss": 1.0652,
+      "step": 516
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7984938240131658,
+      "learning_rate": 1.8937593526353098e-05,
+      "loss": 0.6089,
+      "step": 517
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0686079063080491,
+      "learning_rate": 1.8932702124416446e-05,
+      "loss": 1.0045,
+      "step": 518
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.829307892221941,
+      "learning_rate": 1.892780012281717e-05,
+      "loss": 0.963,
+      "step": 519
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9172582322979269,
+      "learning_rate": 1.8922887527372053e-05,
+      "loss": 1.0235,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.817360292952608,
+      "learning_rate": 1.891796434391044e-05,
+      "loss": 0.759,
+      "step": 521
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0827702441416898,
+      "learning_rate": 1.8913030578274254e-05,
+      "loss": 0.9377,
+      "step": 522
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9602680047399479,
+      "learning_rate": 1.8908086236317958e-05,
+      "loss": 1.0066,
+      "step": 523
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7290989773729804,
+      "learning_rate": 1.890313132390858e-05,
+      "loss": 0.8591,
+      "step": 524
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7769144914751788,
+      "learning_rate": 1.889816584692568e-05,
+      "loss": 0.7926,
+      "step": 525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9806351846990633,
+      "learning_rate": 1.8893189811261364e-05,
+      "loss": 0.7754,
+      "step": 526
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1585076359620554,
+      "learning_rate": 1.8888203222820255e-05,
+      "loss": 1.0508,
+      "step": 527
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8504968509068221,
+      "learning_rate": 1.8883206087519512e-05,
+      "loss": 0.9001,
+      "step": 528
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8054861141226577,
+      "learning_rate": 1.88781984112888e-05,
+      "loss": 0.959,
+      "step": 529
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9617419067776709,
+      "learning_rate": 1.887318020007029e-05,
+      "loss": 1.006,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4526232721250611,
+      "learning_rate": 1.886815145981866e-05,
+      "loss": 0.8172,
+      "step": 531
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8549272979513417,
+      "learning_rate": 1.886311219650108e-05,
+      "loss": 1.026,
+      "step": 532
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.81239615099935,
+      "learning_rate": 1.8858062416097208e-05,
+      "loss": 0.8624,
+      "step": 533
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8867014761694884,
+      "learning_rate": 1.885300212459918e-05,
+      "loss": 0.9517,
+      "step": 534
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8689410413183054,
+      "learning_rate": 1.8847931328011602e-05,
+      "loss": 0.721,
+      "step": 535
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9897530426445178,
+      "learning_rate": 1.884285003235155e-05,
+      "loss": 0.8151,
+      "step": 536
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9243105320690729,
+      "learning_rate": 1.883775824364856e-05,
+      "loss": 0.9864,
+      "step": 537
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7476739289674776,
+      "learning_rate": 1.8832655967944607e-05,
+      "loss": 0.857,
+      "step": 538
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.079271851876113,
+      "learning_rate": 1.8827543211294134e-05,
+      "loss": 0.9562,
+      "step": 539
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1339796387452818,
+      "learning_rate": 1.882241997976399e-05,
+      "loss": 1.056,
+      "step": 540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7889398156216972,
+      "learning_rate": 1.8817286279433474e-05,
+      "loss": 0.8347,
+      "step": 541
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9623015500001476,
+      "learning_rate": 1.8812142116394308e-05,
+      "loss": 0.7613,
+      "step": 542
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1169992505028887,
+      "learning_rate": 1.8806987496750623e-05,
+      "loss": 0.8529,
+      "step": 543
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.071589419834099,
+      "learning_rate": 1.8801822426618956e-05,
+      "loss": 1.0211,
+      "step": 544
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7342734777674703,
+      "learning_rate": 1.8796646912128247e-05,
+      "loss": 0.6838,
+      "step": 545
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7717331188975163,
+      "learning_rate": 1.8791460959419832e-05,
+      "loss": 0.8859,
+      "step": 546
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.914054126717051,
+      "learning_rate": 1.8786264574647428e-05,
+      "loss": 1.1067,
+      "step": 547
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0053771220067544,
+      "learning_rate": 1.878105776397714e-05,
+      "loss": 0.829,
+      "step": 548
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7697891440809641,
+      "learning_rate": 1.8775840533587436e-05,
+      "loss": 0.9684,
+      "step": 549
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9029868575643235,
+      "learning_rate": 1.8770612889669146e-05,
+      "loss": 1.0556,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0649493405808745,
+      "learning_rate": 1.8765374838425467e-05,
+      "loss": 0.8485,
+      "step": 551
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8166718676337571,
+      "learning_rate": 1.8760126386071937e-05,
+      "loss": 0.7715,
+      "step": 552
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2874792314405863,
+      "learning_rate": 1.8754867538836437e-05,
+      "loss": 0.9504,
+      "step": 553
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7226548672030846,
+      "learning_rate": 1.8749598302959195e-05,
+      "loss": 1.0514,
+      "step": 554
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8001802854579623,
+      "learning_rate": 1.8744318684692742e-05,
+      "loss": 0.7869,
+      "step": 555
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8656985519405953,
+      "learning_rate": 1.8739028690301953e-05,
+      "loss": 1.0064,
+      "step": 556
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7969547126727694,
+      "learning_rate": 1.8733728326064e-05,
+      "loss": 0.6896,
+      "step": 557
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.778729266941782,
+      "learning_rate": 1.8728417598268375e-05,
+      "loss": 0.713,
+      "step": 558
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8006093821977203,
+      "learning_rate": 1.8723096513216842e-05,
+      "loss": 0.9659,
+      "step": 559
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7760964991529317,
+      "learning_rate": 1.8717765077223485e-05,
+      "loss": 0.8831,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8297961954362056,
+      "learning_rate": 1.871242329661465e-05,
+      "loss": 0.7653,
+      "step": 561
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8591220776192616,
+      "learning_rate": 1.8707071177728962e-05,
+      "loss": 0.8329,
+      "step": 562
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8483048510638795,
+      "learning_rate": 1.8701708726917325e-05,
+      "loss": 1.0763,
+      "step": 563
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9370785533258552,
+      "learning_rate": 1.8696335950542886e-05,
+      "loss": 0.9484,
+      "step": 564
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8275147490281236,
+      "learning_rate": 1.869095285498106e-05,
+      "loss": 0.9822,
+      "step": 565
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8282584888220358,
+      "learning_rate": 1.868555944661949e-05,
+      "loss": 0.9898,
+      "step": 566
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8107688565315581,
+      "learning_rate": 1.868015573185807e-05,
+      "loss": 0.8488,
+      "step": 567
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8200646705975873,
+      "learning_rate": 1.8674741717108923e-05,
+      "loss": 0.9716,
+      "step": 568
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6997293956092064,
+      "learning_rate": 1.8669317408796387e-05,
+      "loss": 0.6137,
+      "step": 569
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8426986160088773,
+      "learning_rate": 1.866388281335702e-05,
+      "loss": 0.8667,
+      "step": 570
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8280394987112549,
+      "learning_rate": 1.8658437937239586e-05,
+      "loss": 0.9846,
+      "step": 571
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8231793803084683,
+      "learning_rate": 1.8652982786905043e-05,
+      "loss": 0.8993,
+      "step": 572
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7816371569113563,
+      "learning_rate": 1.8647517368826545e-05,
+      "loss": 1.2043,
+      "step": 573
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9515978356716752,
+      "learning_rate": 1.8642041689489435e-05,
+      "loss": 0.9756,
+      "step": 574
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8285215090181695,
+      "learning_rate": 1.863655575539122e-05,
+      "loss": 1.0181,
+      "step": 575
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7586463998534619,
+      "learning_rate": 1.8631059573041583e-05,
+      "loss": 0.8272,
+      "step": 576
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8390430070542286,
+      "learning_rate": 1.8625553148962373e-05,
+      "loss": 0.9556,
+      "step": 577
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8069085940343184,
+      "learning_rate": 1.8620036489687578e-05,
+      "loss": 0.6809,
+      "step": 578
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8130237799562006,
+      "learning_rate": 1.8614509601763343e-05,
+      "loss": 0.9045,
+      "step": 579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7970207712384925,
+      "learning_rate": 1.8608972491747946e-05,
+      "loss": 0.8542,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7755284125904147,
+      "learning_rate": 1.8603425166211795e-05,
+      "loss": 0.8532,
+      "step": 581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8842937498553658,
+      "learning_rate": 1.859786763173742e-05,
+      "loss": 1.0869,
+      "step": 582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.72086153400602,
+      "learning_rate": 1.8592299894919462e-05,
+      "loss": 0.8969,
+      "step": 583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8000537109151721,
+      "learning_rate": 1.8586721962364676e-05,
+      "loss": 0.9522,
+      "step": 584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7955760733523876,
+      "learning_rate": 1.8581133840691907e-05,
+      "loss": 0.9301,
+      "step": 585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7234885576556337,
+      "learning_rate": 1.8575535536532093e-05,
+      "loss": 0.9348,
+      "step": 586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8318653462072466,
+      "learning_rate": 1.8569927056528265e-05,
+      "loss": 0.9933,
+      "step": 587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0348590572473806,
+      "learning_rate": 1.8564308407335508e-05,
+      "loss": 0.6092,
+      "step": 588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.061044392515714,
+      "learning_rate": 1.8558679595620993e-05,
+      "loss": 0.9892,
+      "step": 589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.709774407417619,
+      "learning_rate": 1.855304062806394e-05,
+      "loss": 0.8922,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9825189412387342,
+      "learning_rate": 1.854739151135562e-05,
+      "loss": 0.865,
+      "step": 591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8019531673594383,
+      "learning_rate": 1.8541732252199357e-05,
+      "loss": 1.0017,
+      "step": 592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8596994428540136,
+      "learning_rate": 1.8536062857310496e-05,
+      "loss": 0.8179,
+      "step": 593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7865630594067082,
+      "learning_rate": 1.853038333341642e-05,
+      "loss": 0.9197,
+      "step": 594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9352582794048222,
+      "learning_rate": 1.852469368725652e-05,
+      "loss": 1.0421,
+      "step": 595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1297483564333295,
+      "learning_rate": 1.8518993925582217e-05,
+      "loss": 0.6431,
+      "step": 596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6887416929239897,
+      "learning_rate": 1.8513284055156916e-05,
+      "loss": 1.0139,
+      "step": 597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9141491642246028,
+      "learning_rate": 1.8507564082756024e-05,
+      "loss": 0.9273,
+      "step": 598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8432297090753629,
+      "learning_rate": 1.850183401516694e-05,
+      "loss": 1.1031,
+      "step": 599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8196266178305026,
+      "learning_rate": 1.8496093859189037e-05,
+      "loss": 0.8379,
+      "step": 600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9848650261339109,
+      "learning_rate": 1.849034362163366e-05,
+      "loss": 0.8317,
+      "step": 601
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8173771111864856,
+      "learning_rate": 1.848458330932412e-05,
+      "loss": 1.0146,
+      "step": 602
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.058789897159261,
+      "learning_rate": 1.847881292909567e-05,
+      "loss": 0.9167,
+      "step": 603
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6335796790649482,
+      "learning_rate": 1.847303248779553e-05,
+      "loss": 0.7959,
+      "step": 604
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2840462032038535,
+      "learning_rate": 1.8467241992282842e-05,
+      "loss": 0.9203,
+      "step": 605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9142719013480949,
+      "learning_rate": 1.846144144942869e-05,
+      "loss": 0.8276,
+      "step": 606
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9478579077044378,
+      "learning_rate": 1.8455630866116073e-05,
+      "loss": 0.8418,
+      "step": 607
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0255943997529289,
+      "learning_rate": 1.84498102492399e-05,
+      "loss": 0.7894,
+      "step": 608
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9127809312861257,
+      "learning_rate": 1.8443979605707e-05,
+      "loss": 1.094,
+      "step": 609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9507096397085456,
+      "learning_rate": 1.843813894243609e-05,
+      "loss": 0.8269,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0072805174904753,
+      "learning_rate": 1.8432288266357778e-05,
+      "loss": 0.9924,
+      "step": 611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0559742455667238,
+      "learning_rate": 1.842642758441455e-05,
+      "loss": 0.9611,
+      "step": 612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8116646102988814,
+      "learning_rate": 1.8420556903560773e-05,
+      "loss": 0.994,
+      "step": 613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6613928443180463,
+      "learning_rate": 1.8414676230762674e-05,
+      "loss": 0.6114,
+      "step": 614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.603433456463681,
+      "learning_rate": 1.8408785572998335e-05,
+      "loss": 0.9457,
+      "step": 615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9884737323513365,
+      "learning_rate": 1.8402884937257693e-05,
+      "loss": 1.114,
+      "step": 616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8004548150713351,
+      "learning_rate": 1.8396974330542518e-05,
+      "loss": 0.7637,
+      "step": 617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8693699933189984,
+      "learning_rate": 1.8391053759866412e-05,
+      "loss": 1.0448,
+      "step": 618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9094139320030791,
+      "learning_rate": 1.8385123232254805e-05,
+      "loss": 0.8214,
+      "step": 619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8787632708153642,
+      "learning_rate": 1.8379182754744936e-05,
+      "loss": 0.874,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8321253685682574,
+      "learning_rate": 1.837323233438586e-05,
+      "loss": 0.9718,
+      "step": 621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7319371215128208,
+      "learning_rate": 1.8367271978238422e-05,
+      "loss": 0.765,
+      "step": 622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7741926890868392,
+      "learning_rate": 1.836130169337526e-05,
+      "loss": 0.9853,
+      "step": 623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9683686061213077,
+      "learning_rate": 1.835532148688079e-05,
+      "loss": 0.9079,
+      "step": 624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7354801708414044,
+      "learning_rate": 1.834933136585121e-05,
+      "loss": 0.9913,
+      "step": 625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7603706563724042,
+      "learning_rate": 1.8343331337394472e-05,
+      "loss": 0.8346,
+      "step": 626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3380648007826994,
+      "learning_rate": 1.8337321408630292e-05,
+      "loss": 1.0194,
+      "step": 627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7880676534298809,
+      "learning_rate": 1.833130158669013e-05,
+      "loss": 0.9932,
+      "step": 628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8259240657552208,
+      "learning_rate": 1.8325271878717187e-05,
+      "loss": 0.7296,
+      "step": 629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7508252501018285,
+      "learning_rate": 1.8319232291866388e-05,
+      "loss": 0.8917,
+      "step": 630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.906230488289512,
+      "learning_rate": 1.8313182833304397e-05,
+      "loss": 0.9179,
+      "step": 631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.609296387121373,
+      "learning_rate": 1.830712351020958e-05,
+      "loss": 0.8377,
+      "step": 632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7522743599116025,
+      "learning_rate": 1.8301054329772e-05,
+      "loss": 0.7958,
+      "step": 633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8792575702824518,
+      "learning_rate": 1.829497529919344e-05,
+      "loss": 0.8756,
+      "step": 634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6949971866425552,
+      "learning_rate": 1.828888642568735e-05,
+      "loss": 0.9221,
+      "step": 635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8875879078481215,
+      "learning_rate": 1.828278771647887e-05,
+      "loss": 0.8289,
+      "step": 636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8436699545224912,
+      "learning_rate": 1.827667917880481e-05,
+      "loss": 1.0129,
+      "step": 637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7385927072781265,
+      "learning_rate": 1.8270560819913637e-05,
+      "loss": 0.8418,
+      "step": 638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1352517639472552,
+      "learning_rate": 1.8264432647065485e-05,
+      "loss": 1.0138,
+      "step": 639
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5316686628632747,
+      "learning_rate": 1.825829466753212e-05,
+      "loss": 0.7054,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8511204679869904,
+      "learning_rate": 1.8252146888596946e-05,
+      "loss": 0.8547,
+      "step": 641
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7936783340253832,
+      "learning_rate": 1.8245989317555008e-05,
+      "loss": 0.9197,
+      "step": 642
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7678900687896442,
+      "learning_rate": 1.8239821961712953e-05,
+      "loss": 0.8849,
+      "step": 643
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7950978244140323,
+      "learning_rate": 1.8233644828389056e-05,
+      "loss": 0.9101,
+      "step": 644
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8385473342315476,
+      "learning_rate": 1.8227457924913175e-05,
+      "loss": 0.9862,
+      "step": 645
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7328440102614739,
+      "learning_rate": 1.8221261258626777e-05,
+      "loss": 0.7499,
+      "step": 646
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8089043769067935,
+      "learning_rate": 1.821505483688291e-05,
+      "loss": 0.9265,
+      "step": 647
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7978437858779638,
+      "learning_rate": 1.8208838667046193e-05,
+      "loss": 1.0026,
+      "step": 648
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7650170872089084,
+      "learning_rate": 1.8202612756492817e-05,
+      "loss": 0.7804,
+      "step": 649
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8118572109691062,
+      "learning_rate": 1.8196377112610524e-05,
+      "loss": 0.9211,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0496645065100711,
+      "learning_rate": 1.8190131742798623e-05,
+      "loss": 0.8757,
+      "step": 651
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.873906670577279,
+      "learning_rate": 1.8183876654467947e-05,
+      "loss": 1.0033,
+      "step": 652
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7848595743896681,
+      "learning_rate": 1.8177611855040865e-05,
+      "loss": 0.7562,
+      "step": 653
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8455998322466037,
+      "learning_rate": 1.8171337351951268e-05,
+      "loss": 1.0077,
+      "step": 654
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9344114194205986,
+      "learning_rate": 1.816505315264457e-05,
+      "loss": 0.9495,
+      "step": 655
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.744454467920045,
+      "learning_rate": 1.8158759264577684e-05,
+      "loss": 1.031,
+      "step": 656
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8870669525426801,
+      "learning_rate": 1.8152455695219025e-05,
+      "loss": 0.8169,
+      "step": 657
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8703443929939522,
+      "learning_rate": 1.8146142452048484e-05,
+      "loss": 0.8062,
+      "step": 658
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5829694910631769,
+      "learning_rate": 1.813981954255744e-05,
+      "loss": 0.6544,
+      "step": 659
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9275913643997038,
+      "learning_rate": 1.8133486974248748e-05,
+      "loss": 1.0811,
+      "step": 660
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8436205774449693,
+      "learning_rate": 1.812714475463671e-05,
+      "loss": 0.9144,
+      "step": 661
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8075107490773799,
+      "learning_rate": 1.812079289124709e-05,
+      "loss": 0.9156,
+      "step": 662
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8751342342646846,
+      "learning_rate": 1.8114431391617096e-05,
+      "loss": 0.8813,
+      "step": 663
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9377939422716164,
+      "learning_rate": 1.8108060263295364e-05,
+      "loss": 1.0093,
+      "step": 664
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.758959553012563,
+      "learning_rate": 1.810167951384196e-05,
+      "loss": 0.8589,
+      "step": 665
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9227711750671249,
+      "learning_rate": 1.8095289150828368e-05,
+      "loss": 0.9374,
+      "step": 666
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7191830565713468,
+      "learning_rate": 1.8088889181837474e-05,
+      "loss": 0.7451,
+      "step": 667
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8821820912757701,
+      "learning_rate": 1.808247961446357e-05,
+      "loss": 1.0023,
+      "step": 668
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9219067683543885,
+      "learning_rate": 1.807606045631233e-05,
+      "loss": 0.9388,
+      "step": 669
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8405379665072232,
+      "learning_rate": 1.8069631715000815e-05,
+      "loss": 0.8149,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6871607728295709,
+      "learning_rate": 1.8063193398157453e-05,
+      "loss": 1.1074,
+      "step": 671
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7907408297321608,
+      "learning_rate": 1.8056745513422034e-05,
+      "loss": 0.7938,
+      "step": 672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8457308555782176,
+      "learning_rate": 1.8050288068445706e-05,
+      "loss": 0.9719,
+      "step": 673
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8402425090762534,
+      "learning_rate": 1.8043821070890958e-05,
+      "loss": 0.7909,
+      "step": 674
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8197923212782555,
+      "learning_rate": 1.8037344528431616e-05,
+      "loss": 1.0647,
+      "step": 675
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6035802782791444,
+      "learning_rate": 1.803085844875283e-05,
+      "loss": 0.6309,
+      "step": 676
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7854399891547433,
+      "learning_rate": 1.802436283955107e-05,
+      "loss": 0.9265,
+      "step": 677
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.87160136272218,
+      "learning_rate": 1.8017857708534107e-05,
+      "loss": 1.0332,
+      "step": 678
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9153375791152376,
+      "learning_rate": 1.8011343063421022e-05,
+      "loss": 0.9817,
+      "step": 679
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7932573603368148,
+      "learning_rate": 1.8004818911942173e-05,
+      "loss": 0.9363,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1516998266083844,
+      "learning_rate": 1.7998285261839208e-05,
+      "loss": 0.9354,
+      "step": 681
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9010303349366862,
+      "learning_rate": 1.799174212086504e-05,
+      "loss": 0.7207,
+      "step": 682
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7683500663777423,
+      "learning_rate": 1.7985189496783853e-05,
+      "loss": 1.1771,
+      "step": 683
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.936603964309903,
+      "learning_rate": 1.7978627397371072e-05,
+      "loss": 0.776,
+      "step": 684
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7067439626411914,
+      "learning_rate": 1.7972055830413372e-05,
+      "loss": 0.6924,
+      "step": 685
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1285802429762517,
+      "learning_rate": 1.7965474803708657e-05,
+      "loss": 0.8253,
+      "step": 686
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7399187193515173,
+      "learning_rate": 1.7958884325066073e-05,
+      "loss": 1.0326,
+      "step": 687
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1172737898599372,
+      "learning_rate": 1.7952284402305956e-05,
+      "loss": 0.9901,
+      "step": 688
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7291615375813696,
+      "learning_rate": 1.7945675043259868e-05,
+      "loss": 0.7909,
+      "step": 689
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7564880556978405,
+      "learning_rate": 1.7939056255770562e-05,
+      "loss": 0.9401,
+      "step": 690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9332865526750138,
+      "learning_rate": 1.7932428047691978e-05,
+      "loss": 0.9645,
+      "step": 691
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8644732393049999,
+      "learning_rate": 1.7925790426889237e-05,
+      "loss": 1.0079,
+      "step": 692
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6946084803061833,
+      "learning_rate": 1.7919143401238623e-05,
+      "loss": 0.6892,
+      "step": 693
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8751422602313949,
+      "learning_rate": 1.791248697862759e-05,
+      "loss": 1.0255,
+      "step": 694
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7336586373624162,
+      "learning_rate": 1.7905821166954732e-05,
+      "loss": 0.7598,
+      "step": 695
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7789507247018542,
+      "learning_rate": 1.7899145974129793e-05,
+      "loss": 0.8298,
+      "step": 696
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7345992979167735,
+      "learning_rate": 1.789246140807365e-05,
+      "loss": 0.9623,
+      "step": 697
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9491307781097194,
+      "learning_rate": 1.7885767476718285e-05,
+      "loss": 0.8757,
+      "step": 698
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8759090700756539,
+      "learning_rate": 1.7879064188006818e-05,
+      "loss": 1.0338,
+      "step": 699
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8004617528557186,
+      "learning_rate": 1.7872351549893452e-05,
+      "loss": 0.8361,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.989320268883639,
+      "learning_rate": 1.7865629570343496e-05,
+      "loss": 1.0638,
+      "step": 701
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7108461404049543,
+      "learning_rate": 1.785889825733334e-05,
+      "loss": 0.7513,
+      "step": 702
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9835919811234671,
+      "learning_rate": 1.7852157618850447e-05,
+      "loss": 0.7768,
+      "step": 703
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8881871267633705,
+      "learning_rate": 1.784540766289335e-05,
+      "loss": 0.9968,
+      "step": 704
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8158337158445435,
+      "learning_rate": 1.7838648397471634e-05,
+      "loss": 0.92,
+      "step": 705
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8074289424363337,
+      "learning_rate": 1.783187983060594e-05,
+      "loss": 0.9611,
+      "step": 706
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8065125783104768,
+      "learning_rate": 1.7825101970327926e-05,
+      "loss": 0.7727,
+      "step": 707
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9207715886854075,
+      "learning_rate": 1.78183148246803e-05,
+      "loss": 0.896,
+      "step": 708
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8415956009968806,
+      "learning_rate": 1.7811518401716776e-05,
+      "loss": 0.9507,
+      "step": 709
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7363872904009363,
+      "learning_rate": 1.7804712709502083e-05,
+      "loss": 0.8665,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6329621386563533,
+      "learning_rate": 1.7797897756111942e-05,
+      "loss": 0.744,
+      "step": 711
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1125887283998672,
+      "learning_rate": 1.7791073549633065e-05,
+      "loss": 0.8568,
+      "step": 712
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.878085736174585,
+      "learning_rate": 1.7784240098163154e-05,
+      "loss": 1.0596,
+      "step": 713
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7735732649202102,
+      "learning_rate": 1.7777397409810863e-05,
+      "loss": 0.8957,
+      "step": 714
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7735721697225134,
+      "learning_rate": 1.7770545492695825e-05,
+      "loss": 0.8821,
+      "step": 715
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9382829844794944,
+      "learning_rate": 1.7763684354948608e-05,
+      "loss": 0.9473,
+      "step": 716
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8501401238854098,
+      "learning_rate": 1.7756814004710734e-05,
+      "loss": 0.8789,
+      "step": 717
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8260951351982395,
+      "learning_rate": 1.774993445013465e-05,
+      "loss": 1.0165,
+      "step": 718
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8223018880589215,
+      "learning_rate": 1.7743045699383726e-05,
+      "loss": 0.836,
+      "step": 719
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9337748119848038,
+      "learning_rate": 1.773614776063225e-05,
+      "loss": 0.911,
+      "step": 720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6613459802364837,
+      "learning_rate": 1.7729240642065396e-05,
+      "loss": 0.674,
+      "step": 721
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.790250697037028,
+      "learning_rate": 1.7722324351879252e-05,
+      "loss": 0.7504,
+      "step": 722
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7711215326248821,
+      "learning_rate": 1.771539889828078e-05,
+      "loss": 0.9836,
+      "step": 723
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7420988723888557,
+      "learning_rate": 1.770846428948781e-05,
+      "loss": 0.8854,
+      "step": 724
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0688762426867753,
+      "learning_rate": 1.770152053372904e-05,
+      "loss": 0.9763,
+      "step": 725
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7045471982485659,
+      "learning_rate": 1.769456763924403e-05,
+      "loss": 1.0655,
+      "step": 726
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6966364799483906,
+      "learning_rate": 1.7687605614283167e-05,
+      "loss": 0.7992,
+      "step": 727
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8225255896783794,
+      "learning_rate": 1.768063446710769e-05,
+      "loss": 0.7435,
+      "step": 728
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0263888182535368,
+      "learning_rate": 1.767365420598965e-05,
+      "loss": 0.8926,
+      "step": 729
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9407462166797562,
+      "learning_rate": 1.766666483921192e-05,
+      "loss": 0.9085,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6780083350149689,
+      "learning_rate": 1.765966637506817e-05,
+      "loss": 0.7233,
+      "step": 731
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7779319569372881,
+      "learning_rate": 1.765265882186288e-05,
+      "loss": 0.8846,
+      "step": 732
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8496093240453721,
+      "learning_rate": 1.7645642187911292e-05,
+      "loss": 1.0591,
+      "step": 733
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7992398249228102,
+      "learning_rate": 1.763861648153945e-05,
+      "loss": 0.8792,
+      "step": 734
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7444282665758954,
+      "learning_rate": 1.763158171108414e-05,
+      "loss": 0.9545,
+      "step": 735
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1090717096644573,
+      "learning_rate": 1.7624537884892923e-05,
+      "loss": 1.0919,
+      "step": 736
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8141339720277483,
+      "learning_rate": 1.761748501132409e-05,
+      "loss": 1.0354,
+      "step": 737
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9838085185271663,
+      "learning_rate": 1.761042309874667e-05,
+      "loss": 0.6666,
+      "step": 738
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8046022505838206,
+      "learning_rate": 1.7603352155540433e-05,
+      "loss": 0.8578,
+      "step": 739
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7779370663948793,
+      "learning_rate": 1.759627219009585e-05,
+      "loss": 1.0355,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7123623181573308,
+      "learning_rate": 1.7589183210814094e-05,
+      "loss": 0.7949,
+      "step": 741
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8016732076355079,
+      "learning_rate": 1.7582085226107054e-05,
+      "loss": 0.9879,
+      "step": 742
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.00147107464386,
+      "learning_rate": 1.7574978244397286e-05,
+      "loss": 0.72,
+      "step": 743
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8352198692421553,
+      "learning_rate": 1.7567862274118027e-05,
+      "loss": 0.7976,
+      "step": 744
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7536438454909684,
+      "learning_rate": 1.7560737323713184e-05,
+      "loss": 0.9288,
+      "step": 745
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7814631020793061,
+      "learning_rate": 1.7553603401637317e-05,
+      "loss": 0.9098,
+      "step": 746
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8936694868080809,
+      "learning_rate": 1.754646051635563e-05,
+      "loss": 0.7521,
+      "step": 747
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.795825185620166,
+      "learning_rate": 1.7539308676343972e-05,
+      "loss": 0.8364,
+      "step": 748
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.726424233341848,
+      "learning_rate": 1.7532147890088807e-05,
+      "loss": 1.0881,
+      "step": 749
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5152815379842386,
+      "learning_rate": 1.752497816608722e-05,
+      "loss": 0.9634,
+      "step": 750
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9563327832023625,
+      "learning_rate": 1.7517799512846897e-05,
+      "loss": 0.9666,
+      "step": 751
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7898598496189634,
+      "learning_rate": 1.7510611938886123e-05,
+      "loss": 0.9319,
+      "step": 752
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7519620654828607,
+      "learning_rate": 1.7503415452733774e-05,
+      "loss": 0.8332,
+      "step": 753
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8463024604757508,
+      "learning_rate": 1.7496210062929287e-05,
+      "loss": 0.9867,
+      "step": 754
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7084810372074458,
+      "learning_rate": 1.7488995778022687e-05,
+      "loss": 0.6206,
+      "step": 755
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9058373986088494,
+      "learning_rate": 1.7481772606574526e-05,
+      "loss": 0.9818,
+      "step": 756
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9026994424796992,
+      "learning_rate": 1.7474540557155917e-05,
+      "loss": 0.8292,
+      "step": 757
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7201441808479074,
+      "learning_rate": 1.7467299638348514e-05,
+      "loss": 0.8234,
+      "step": 758
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9157353929868001,
+      "learning_rate": 1.7460049858744482e-05,
+      "loss": 1.1801,
+      "step": 759
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.436932094872689,
+      "learning_rate": 1.7452791226946502e-05,
+      "loss": 0.9624,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7396457457492303,
+      "learning_rate": 1.744552375156777e-05,
+      "loss": 1.0051,
+      "step": 761
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8417560567505239,
+      "learning_rate": 1.743824744123196e-05,
+      "loss": 0.7646,
+      "step": 762
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1701082801016527,
+      "learning_rate": 1.743096230457325e-05,
+      "loss": 0.9458,
+      "step": 763
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6911344516913996,
+      "learning_rate": 1.7423668350236265e-05,
+      "loss": 0.7034,
+      "step": 764
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7545181016399742,
+      "learning_rate": 1.7416365586876118e-05,
+      "loss": 0.8406,
+      "step": 765
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7004605587170897,
+      "learning_rate": 1.7409054023158362e-05,
+      "loss": 0.8668,
+      "step": 766
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0842017715858998,
+      "learning_rate": 1.7401733667758993e-05,
+      "loss": 0.8463,
+      "step": 767
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7570185832253038,
+      "learning_rate": 1.7394404529364447e-05,
+      "loss": 0.9552,
+      "step": 768
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9124479403734435,
+      "learning_rate": 1.7387066616671572e-05,
+      "loss": 0.8466,
+      "step": 769
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8462269001721561,
+      "learning_rate": 1.7379719938387634e-05,
+      "loss": 0.8926,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8655711452174912,
+      "learning_rate": 1.73723645032303e-05,
+      "loss": 0.9967,
+      "step": 771
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8052992936115821,
+      "learning_rate": 1.7365000319927624e-05,
+      "loss": 0.9965,
+      "step": 772
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7482308025931095,
+      "learning_rate": 1.735762739721804e-05,
+      "loss": 0.7047,
+      "step": 773
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.80781288165983,
+      "learning_rate": 1.7350245743850367e-05,
+      "loss": 0.8481,
+      "step": 774
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9718824474847136,
+      "learning_rate": 1.7342855368583764e-05,
+      "loss": 0.9794,
+      "step": 775
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8597837971092859,
+      "learning_rate": 1.7335456280187752e-05,
+      "loss": 0.9071,
+      "step": 776
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5535418764970395,
+      "learning_rate": 1.7328048487442184e-05,
+      "loss": 0.8923,
+      "step": 777
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8890317225493944,
+      "learning_rate": 1.7320631999137245e-05,
+      "loss": 0.9949,
+      "step": 778
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8168121533350582,
+      "learning_rate": 1.7313206824073444e-05,
+      "loss": 0.7658,
+      "step": 779
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0414455712162651,
+      "learning_rate": 1.7305772971061583e-05,
+      "loss": 1.0603,
+      "step": 780
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.837990714122133,
+      "learning_rate": 1.729833044892278e-05,
+      "loss": 0.612,
+      "step": 781
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8952976541350629,
+      "learning_rate": 1.729087926648842e-05,
+      "loss": 0.832,
+      "step": 782
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8220527631767338,
+      "learning_rate": 1.7283419432600184e-05,
+      "loss": 1.0586,
+      "step": 783
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0915411720322319,
+      "learning_rate": 1.7275950956110003e-05,
+      "loss": 0.9458,
+      "step": 784
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8810281846212868,
+      "learning_rate": 1.7268473845880076e-05,
+      "loss": 0.9866,
+      "step": 785
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7270233845899853,
+      "learning_rate": 1.7260988110782836e-05,
+      "loss": 0.8091,
+      "step": 786
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8675359375696707,
+      "learning_rate": 1.7253493759700958e-05,
+      "loss": 0.9881,
+      "step": 787
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0995854580476723,
+      "learning_rate": 1.724599080152734e-05,
+      "loss": 0.8347,
+      "step": 788
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8465868229792299,
+      "learning_rate": 1.7238479245165084e-05,
+      "loss": 0.9242,
+      "step": 789
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7221605797839197,
+      "learning_rate": 1.7230959099527512e-05,
+      "loss": 0.8427,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8302820819749986,
+      "learning_rate": 1.7223430373538125e-05,
+      "loss": 0.9232,
+      "step": 791
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6932430287691834,
+      "learning_rate": 1.7215893076130603e-05,
+      "loss": 0.8041,
+      "step": 792
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8640081709165142,
+      "learning_rate": 1.720834721624881e-05,
+      "loss": 0.8391,
+      "step": 793
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7336591441625315,
+      "learning_rate": 1.7200792802846763e-05,
+      "loss": 0.8018,
+      "step": 794
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.917266954919667,
+      "learning_rate": 1.7193229844888625e-05,
+      "loss": 1.0678,
+      "step": 795
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8250664156645172,
+      "learning_rate": 1.7185658351348703e-05,
+      "loss": 0.7889,
+      "step": 796
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7142560139210092,
+      "learning_rate": 1.717807833121143e-05,
+      "loss": 1.0226,
+      "step": 797
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8091575789994654,
+      "learning_rate": 1.7170489793471365e-05,
+      "loss": 0.9077,
+      "step": 798
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8127247294464867,
+      "learning_rate": 1.7162892747133163e-05,
+      "loss": 0.9228,
+      "step": 799
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6380555276239949,
+      "learning_rate": 1.7155287201211578e-05,
+      "loss": 0.5964,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.861728436606956,
+      "learning_rate": 1.7147673164731453e-05,
+      "loss": 0.9472,
+      "step": 801
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8216982679857924,
+      "learning_rate": 1.7140050646727706e-05,
+      "loss": 1.0563,
+      "step": 802
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.162533340357871,
+      "learning_rate": 1.7132419656245316e-05,
+      "loss": 0.9678,
+      "step": 803
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9123759506552687,
+      "learning_rate": 1.712478020233932e-05,
+      "loss": 0.9863,
+      "step": 804
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.805729301638304,
+      "learning_rate": 1.7117132294074793e-05,
+      "loss": 0.7882,
+      "step": 805
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.045566170126961,
+      "learning_rate": 1.710947594052684e-05,
+      "loss": 0.9831,
+      "step": 806
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1109012501567257,
+      "learning_rate": 1.7101811150780602e-05,
+      "loss": 0.9383,
+      "step": 807
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6516468257056686,
+      "learning_rate": 1.709413793393121e-05,
+      "loss": 0.6283,
+      "step": 808
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0653741270461896,
+      "learning_rate": 1.708645629908381e-05,
+      "loss": 1.004,
+      "step": 809
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.122342603511113,
+      "learning_rate": 1.707876625535353e-05,
+      "loss": 0.9084,
+      "step": 810
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8922456895910541,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 1.059,
+      "step": 811
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8814995299804417,
+      "learning_rate": 1.7063360977754726e-05,
+      "loss": 0.8194,
+      "step": 812
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1755414161347353,
+      "learning_rate": 1.7055645762166314e-05,
+      "loss": 1.0223,
+      "step": 813
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1542662857832553,
+      "learning_rate": 1.704792217425521e-05,
+      "loss": 0.8808,
+      "step": 814
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8782605499124044,
+      "learning_rate": 1.7040190223186327e-05,
+      "loss": 0.8062,
+      "step": 815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9325163944414536,
+      "learning_rate": 1.7032449918134506e-05,
+      "loss": 0.952,
+      "step": 816
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3063717931607546,
+      "learning_rate": 1.7024701268284492e-05,
+      "loss": 0.7965,
+      "step": 817
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.922100714971386,
+      "learning_rate": 1.7016944282830935e-05,
+      "loss": 0.8238,
+      "step": 818
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7555996799463895,
+      "learning_rate": 1.7009178970978373e-05,
+      "loss": 0.8023,
+      "step": 819
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2203089879322946,
+      "learning_rate": 1.700140534194124e-05,
+      "loss": 0.8492,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0160473426277221,
+      "learning_rate": 1.699362340494381e-05,
+      "loss": 0.9305,
+      "step": 821
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9614660592388611,
+      "learning_rate": 1.698583316922025e-05,
+      "loss": 0.8092,
+      "step": 822
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7676697916687426,
+      "learning_rate": 1.6978034644014544e-05,
+      "loss": 1.0268,
+      "step": 823
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.067037827759109,
+      "learning_rate": 1.697022783858053e-05,
+      "loss": 0.969,
+      "step": 824
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7493538359342498,
+      "learning_rate": 1.6962412762181867e-05,
+      "loss": 0.9576,
+      "step": 825
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6076587188183725,
+      "learning_rate": 1.695458942409203e-05,
+      "loss": 0.6692,
+      "step": 826
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3167286864935341,
+      "learning_rate": 1.6946757833594293e-05,
+      "loss": 0.8795,
+      "step": 827
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9133470174645718,
+      "learning_rate": 1.693891799998172e-05,
+      "loss": 0.8979,
+      "step": 828
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9474550683196258,
+      "learning_rate": 1.6931069932557173e-05,
+      "loss": 0.8453,
+      "step": 829
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.677166211061645,
+      "learning_rate": 1.692321364063326e-05,
+      "loss": 1.0496,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1205265283947512,
+      "learning_rate": 1.691534913353237e-05,
+      "loss": 0.8891,
+      "step": 831
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8331011857720567,
+      "learning_rate": 1.6907476420586632e-05,
+      "loss": 0.7113,
+      "step": 832
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7204343779663089,
+      "learning_rate": 1.6899595511137905e-05,
+      "loss": 0.8159,
+      "step": 833
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6426171518420113,
+      "learning_rate": 1.689170641453778e-05,
+      "loss": 0.7224,
+      "step": 834
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7838557482206585,
+      "learning_rate": 1.6883809140147575e-05,
+      "loss": 1.0519,
+      "step": 835
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7466545469169208,
+      "learning_rate": 1.687590369733829e-05,
+      "loss": 0.8521,
+      "step": 836
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9044991422874638,
+      "learning_rate": 1.6867990095490626e-05,
+      "loss": 0.8613,
+      "step": 837
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0528002288672123,
+      "learning_rate": 1.6860068343994975e-05,
+      "loss": 1.0193,
+      "step": 838
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7310690610983246,
+      "learning_rate": 1.685213845225139e-05,
+      "loss": 0.7134,
+      "step": 839
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8812373559253033,
+      "learning_rate": 1.6844200429669582e-05,
+      "loss": 1.0184,
+      "step": 840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1413620670136257,
+      "learning_rate": 1.683625428566892e-05,
+      "loss": 0.9303,
+      "step": 841
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7768448622424368,
+      "learning_rate": 1.68283000296784e-05,
+      "loss": 1.0536,
+      "step": 842
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7936953107315399,
+      "learning_rate": 1.682033767113665e-05,
+      "loss": 0.7943,
+      "step": 843
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7834490514409316,
+      "learning_rate": 1.6812367219491907e-05,
+      "loss": 0.7799,
+      "step": 844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.693084754393923,
+      "learning_rate": 1.6804388684202012e-05,
+      "loss": 0.6909,
+      "step": 845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8175667740441278,
+      "learning_rate": 1.6796402074734404e-05,
+      "loss": 1.0087,
+      "step": 846
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8741959629952243,
+      "learning_rate": 1.67884074005661e-05,
+      "loss": 1.09,
+      "step": 847
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7731054816125622,
+      "learning_rate": 1.678040467118369e-05,
+      "loss": 0.8876,
+      "step": 848
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7846142919708455,
+      "learning_rate": 1.6772393896083314e-05,
+      "loss": 0.808,
+      "step": 849
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6357542951578113,
+      "learning_rate": 1.6764375084770657e-05,
+      "loss": 1.0003,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8010871453503373,
+      "learning_rate": 1.675634824676096e-05,
+      "loss": 0.7869,
+      "step": 851
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7734763101721908,
+      "learning_rate": 1.6748313391578973e-05,
+      "loss": 0.9774,
+      "step": 852
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6920966192266278,
+      "learning_rate": 1.674027052875895e-05,
+      "loss": 0.7266,
+      "step": 853
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.777358280193258,
+      "learning_rate": 1.673221966784467e-05,
+      "loss": 0.9722,
+      "step": 854
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7521109546294623,
+      "learning_rate": 1.672416081838939e-05,
+      "loss": 0.9267,
+      "step": 855
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.709228868650214,
+      "learning_rate": 1.671609398995584e-05,
+      "loss": 0.9102,
+      "step": 856
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7929921708645303,
+      "learning_rate": 1.6708019192116233e-05,
+      "loss": 1.0254,
+      "step": 857
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7078115557994459,
+      "learning_rate": 1.669993643445223e-05,
+      "loss": 0.8044,
+      "step": 858
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9232559989785296,
+      "learning_rate": 1.6691845726554933e-05,
+      "loss": 1.0358,
+      "step": 859
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7788564121203436,
+      "learning_rate": 1.6683747078024887e-05,
+      "loss": 0.7622,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7374839138107102,
+      "learning_rate": 1.6675640498472055e-05,
+      "loss": 1.0363,
+      "step": 861
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5154096312238047,
+      "learning_rate": 1.6667525997515815e-05,
+      "loss": 0.6764,
+      "step": 862
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6785720233014382,
+      "learning_rate": 1.6659403584784937e-05,
+      "loss": 0.9379,
+      "step": 863
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8040175702825757,
+      "learning_rate": 1.6651273269917586e-05,
+      "loss": 0.9724,
+      "step": 864
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7015520232647929,
+      "learning_rate": 1.66431350625613e-05,
+      "loss": 0.9954,
+      "step": 865
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.761671878399352,
+      "learning_rate": 1.663498897237298e-05,
+      "loss": 0.8782,
+      "step": 866
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8492361558455176,
+      "learning_rate": 1.6626835009018892e-05,
+      "loss": 0.9278,
+      "step": 867
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7871648557176816,
+      "learning_rate": 1.6618673182174635e-05,
+      "loss": 0.7268,
+      "step": 868
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7796809277979989,
+      "learning_rate": 1.6610503501525136e-05,
+      "loss": 1.1369,
+      "step": 869
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6390949436245977,
+      "learning_rate": 1.660232597676465e-05,
+      "loss": 0.6427,
+      "step": 870
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5454810353585549,
+      "learning_rate": 1.6594140617596733e-05,
+      "loss": 0.7315,
+      "step": 871
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8765887696138821,
+      "learning_rate": 1.6585947433734247e-05,
+      "loss": 0.834,
+      "step": 872
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7502376076568613,
+      "learning_rate": 1.6577746434899324e-05,
+      "loss": 1.0404,
+      "step": 873
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6079654185812325,
+      "learning_rate": 1.6569537630823385e-05,
+      "loss": 0.9825,
+      "step": 874
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6747111851834838,
+      "learning_rate": 1.65613210312471e-05,
+      "loss": 0.7437,
+      "step": 875
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9266831165182088,
+      "learning_rate": 1.65530966459204e-05,
+      "loss": 1.0767,
+      "step": 876
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8075077300119082,
+      "learning_rate": 1.6544864484602446e-05,
+      "loss": 0.8547,
+      "step": 877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6839602746456831,
+      "learning_rate": 1.6536624557061632e-05,
+      "loss": 0.9611,
+      "step": 878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5834071226967482,
+      "learning_rate": 1.652837687307556e-05,
+      "loss": 0.694,
+      "step": 879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7086235085741724,
+      "learning_rate": 1.6520121442431053e-05,
+      "loss": 1.0323,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6733966885112628,
+      "learning_rate": 1.6511858274924098e-05,
+      "loss": 0.7784,
+      "step": 881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7232168200085507,
+      "learning_rate": 1.6503587380359894e-05,
+      "loss": 0.862,
+      "step": 882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8484812504611454,
+      "learning_rate": 1.6495308768552783e-05,
+      "loss": 0.9153,
+      "step": 883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7042421033837928,
+      "learning_rate": 1.6487022449326282e-05,
+      "loss": 0.866,
+      "step": 884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7361571758777307,
+      "learning_rate": 1.6478728432513043e-05,
+      "loss": 1.011,
+      "step": 885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6058481155323943,
+      "learning_rate": 1.647042672795486e-05,
+      "loss": 0.9536,
+      "step": 886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7890296263548905,
+      "learning_rate": 1.6462117345502638e-05,
+      "loss": 1.0438,
+      "step": 887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5618604018475466,
+      "learning_rate": 1.645380029501641e-05,
+      "loss": 0.7491,
+      "step": 888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.788329388141288,
+      "learning_rate": 1.6445475586365295e-05,
+      "loss": 0.7301,
+      "step": 889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7557617225261046,
+      "learning_rate": 1.64371432294275e-05,
+      "loss": 0.9634,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8204713499036317,
+      "learning_rate": 1.642880323409031e-05,
+      "loss": 0.9849,
+      "step": 891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7566007170197683,
+      "learning_rate": 1.6420455610250077e-05,
+      "loss": 0.9405,
+      "step": 892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6847653020310005,
+      "learning_rate": 1.6412100367812204e-05,
+      "loss": 0.766,
+      "step": 893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6961550088752534,
+      "learning_rate": 1.6403737516691122e-05,
+      "loss": 0.845,
+      "step": 894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.715812371730346,
+      "learning_rate": 1.6395367066810312e-05,
+      "loss": 1.0064,
+      "step": 895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7811163321587572,
+      "learning_rate": 1.6386989028102257e-05,
+      "loss": 0.8286,
+      "step": 896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.633762160878455,
+      "learning_rate": 1.6378603410508444e-05,
+      "loss": 0.7736,
+      "step": 897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8230265046151389,
+      "learning_rate": 1.6370210223979366e-05,
+      "loss": 0.8217,
+      "step": 898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6930263559307966,
+      "learning_rate": 1.6361809478474483e-05,
+      "loss": 1.0334,
+      "step": 899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7710003609684137,
+      "learning_rate": 1.6353401183962235e-05,
+      "loss": 0.8907,
+      "step": 900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6725998736708801,
+      "learning_rate": 1.634498535042001e-05,
+      "loss": 0.8361,
+      "step": 901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8071273188862474,
+      "learning_rate": 1.6336561987834155e-05,
+      "loss": 0.9733,
+      "step": 902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8438061385566864,
+      "learning_rate": 1.6328131106199936e-05,
+      "loss": 0.8782,
+      "step": 903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7830798237488775,
+      "learning_rate": 1.6319692715521552e-05,
+      "loss": 0.9795,
+      "step": 904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6988975984477284,
+      "learning_rate": 1.6311246825812115e-05,
+      "loss": 0.8616,
+      "step": 905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7918122122943023,
+      "learning_rate": 1.630279344709362e-05,
+      "loss": 0.8484,
+      "step": 906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7242768284833926,
+      "learning_rate": 1.6294332589396965e-05,
+      "loss": 0.6663,
+      "step": 907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7037181822771138,
+      "learning_rate": 1.628586426276191e-05,
+      "loss": 0.7391,
+      "step": 908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8315617714043437,
+      "learning_rate": 1.6277388477237087e-05,
+      "loss": 1.0743,
+      "step": 909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7467230426366882,
+      "learning_rate": 1.6268905242879975e-05,
+      "loss": 0.8356,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7046327052863598,
+      "learning_rate": 1.6260414569756894e-05,
+      "loss": 1.0592,
+      "step": 911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6099212264408972,
+      "learning_rate": 1.625191646794299e-05,
+      "loss": 1.0055,
+      "step": 912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8270192254659644,
+      "learning_rate": 1.6243410947522214e-05,
+      "loss": 0.7486,
+      "step": 913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7005001879998711,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.6865,
+      "step": 914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9392896753926776,
+      "learning_rate": 1.6226377691239913e-05,
+      "loss": 0.8898,
+      "step": 915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6603371567564035,
+      "learning_rate": 1.6217849975590275e-05,
+      "loss": 0.9288,
+      "step": 916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6473465284617047,
+      "learning_rate": 1.620931488175752e-05,
+      "loss": 0.727,
+      "step": 917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7940530657074946,
+      "learning_rate": 1.6200772419869508e-05,
+      "loss": 1.0197,
+      "step": 918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8115523419277061,
+      "learning_rate": 1.6192222600062833e-05,
+      "loss": 0.8911,
+      "step": 919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7392602690466112,
+      "learning_rate": 1.618366543248283e-05,
+      "loss": 0.8353,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.692573299234057,
+      "learning_rate": 1.617510092728354e-05,
+      "loss": 0.8867,
+      "step": 921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9536051322294732,
+      "learning_rate": 1.6166529094627727e-05,
+      "loss": 1.0762,
+      "step": 922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.781141858268823,
+      "learning_rate": 1.6157949944686827e-05,
+      "loss": 0.9931,
+      "step": 923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7988155611599281,
+      "learning_rate": 1.614936348764099e-05,
+      "loss": 0.6543,
+      "step": 924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8555340114322555,
+      "learning_rate": 1.614076973367901e-05,
+      "loss": 0.8893,
+      "step": 925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7999875843792692,
+      "learning_rate": 1.6132168692998353e-05,
+      "loss": 1.0086,
+      "step": 926
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8323008095581319,
+      "learning_rate": 1.6123560375805123e-05,
+      "loss": 0.7504,
+      "step": 927
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8538282438693405,
+      "learning_rate": 1.6114944792314067e-05,
+      "loss": 0.9845,
+      "step": 928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6910997043702324,
+      "learning_rate": 1.610632195274855e-05,
+      "loss": 0.7141,
+      "step": 929
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7203753824199999,
+      "learning_rate": 1.6097691867340547e-05,
+      "loss": 0.6741,
+      "step": 930
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7845788645460976,
+      "learning_rate": 1.608905454633063e-05,
+      "loss": 0.9909,
+      "step": 931
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.82826595691395,
+      "learning_rate": 1.6080409999967962e-05,
+      "loss": 0.9101,
+      "step": 932
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8085901702030814,
+      "learning_rate": 1.6071758238510267e-05,
+      "loss": 0.7434,
+      "step": 933
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7114530913674436,
+      "learning_rate": 1.6063099272223856e-05,
+      "loss": 0.7963,
+      "step": 934
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7578048513623175,
+      "learning_rate": 1.6054433111383562e-05,
+      "loss": 1.1454,
+      "step": 935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7866286848579276,
+      "learning_rate": 1.604575976627277e-05,
+      "loss": 0.7805,
+      "step": 936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.748147971522313,
+      "learning_rate": 1.603707924718338e-05,
+      "loss": 0.9894,
+      "step": 937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6894713296031452,
+      "learning_rate": 1.602839156441582e-05,
+      "loss": 0.8428,
+      "step": 938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8667254177794428,
+      "learning_rate": 1.6019696728279e-05,
+      "loss": 0.8243,
+      "step": 939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7239192530226252,
+      "learning_rate": 1.6010994749090338e-05,
+      "loss": 0.8948,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7270221018642831,
+      "learning_rate": 1.6002285637175706e-05,
+      "loss": 0.5745,
+      "step": 941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7823441922291158,
+      "learning_rate": 1.599356940286946e-05,
+      "loss": 0.9705,
+      "step": 942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7814222773622909,
+      "learning_rate": 1.59848460565144e-05,
+      "loss": 0.7869,
+      "step": 943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7968723540157404,
+      "learning_rate": 1.5976115608461756e-05,
+      "loss": 0.9228,
+      "step": 944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8154038621787826,
+      "learning_rate": 1.59673780690712e-05,
+      "loss": 1.2026,
+      "step": 945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8491475317701689,
+      "learning_rate": 1.595863344871081e-05,
+      "loss": 0.9248,
+      "step": 946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8284219773919698,
+      "learning_rate": 1.5949881757757073e-05,
+      "loss": 0.973,
+      "step": 947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8209428367645729,
+      "learning_rate": 1.594112300659485e-05,
+      "loss": 0.7891,
+      "step": 948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7334706137336199,
+      "learning_rate": 1.5932357205617406e-05,
+      "loss": 0.966,
+      "step": 949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6441645490197562,
+      "learning_rate": 1.592358436522634e-05,
+      "loss": 0.6879,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6910026919817118,
+      "learning_rate": 1.5914804495831633e-05,
+      "loss": 0.834,
+      "step": 951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7218255739906858,
+      "learning_rate": 1.5906017607851595e-05,
+      "loss": 0.8859,
+      "step": 952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6901209849483055,
+      "learning_rate": 1.589722371171285e-05,
+      "loss": 0.7872,
+      "step": 953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8101564967102394,
+      "learning_rate": 1.5888422817850367e-05,
+      "loss": 0.9656,
+      "step": 954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7845334934328133,
+      "learning_rate": 1.5879614936707395e-05,
+      "loss": 0.9489,
+      "step": 955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7846744847190611,
+      "learning_rate": 1.5870800078735484e-05,
+      "loss": 0.8007,
+      "step": 956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7419985463922827,
+      "learning_rate": 1.5861978254394465e-05,
+      "loss": 0.9965,
+      "step": 957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8854786330902192,
+      "learning_rate": 1.585314947415242e-05,
+      "loss": 0.9663,
+      "step": 958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6194473460781366,
+      "learning_rate": 1.5844313748485714e-05,
+      "loss": 0.6811,
+      "step": 959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7481928557691698,
+      "learning_rate": 1.583547108787892e-05,
+      "loss": 0.7631,
+      "step": 960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8195157489028261,
+      "learning_rate": 1.5826621502824862e-05,
+      "loss": 1.0356,
+      "step": 961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7344036874993363,
+      "learning_rate": 1.5817765003824578e-05,
+      "loss": 0.7608,
+      "step": 962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8262150524166191,
+      "learning_rate": 1.5808901601387304e-05,
+      "loss": 0.8687,
+      "step": 963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7622888277578065,
+      "learning_rate": 1.5800031306030476e-05,
+      "loss": 1.031,
+      "step": 964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7372368085374054,
+      "learning_rate": 1.5791154128279694e-05,
+      "loss": 0.753,
+      "step": 965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6838635548418247,
+      "learning_rate": 1.578227007866874e-05,
+      "loss": 0.9769,
+      "step": 966
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7345272787314996,
+      "learning_rate": 1.5773379167739547e-05,
+      "loss": 0.6098,
+      "step": 967
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7650900001167809,
+      "learning_rate": 1.5764481406042183e-05,
+      "loss": 0.8064,
+      "step": 968
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6743351247964544,
+      "learning_rate": 1.5755576804134855e-05,
+      "loss": 1.0518,
+      "step": 969
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7953891226378719,
+      "learning_rate": 1.5746665372583878e-05,
+      "loss": 0.9282,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8036596745202361,
+      "learning_rate": 1.5737747121963673e-05,
+      "loss": 1.0129,
+      "step": 971
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8387872814999655,
+      "learning_rate": 1.5728822062856758e-05,
+      "loss": 0.8099,
+      "step": 972
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7968636188700214,
+      "learning_rate": 1.5719890205853724e-05,
+      "loss": 1.0212,
+      "step": 973
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7855537399589521,
+      "learning_rate": 1.5710951561553237e-05,
+      "loss": 0.8018,
+      "step": 974
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6202779707400545,
+      "learning_rate": 1.5702006140562e-05,
+      "loss": 0.9538,
+      "step": 975
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6919743164732264,
+      "learning_rate": 1.5693053953494776e-05,
+      "loss": 0.7079,
+      "step": 976
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7081221390384608,
+      "learning_rate": 1.5684095010974346e-05,
+      "loss": 0.6662,
+      "step": 977
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.733802574874526,
+      "learning_rate": 1.567512932363151e-05,
+      "loss": 1.0957,
+      "step": 978
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6599406719409543,
+      "learning_rate": 1.5666156902105073e-05,
+      "loss": 0.8228,
+      "step": 979
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7717553323077113,
+      "learning_rate": 1.5657177757041823e-05,
+      "loss": 0.9386,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7670068644588712,
+      "learning_rate": 1.5648191899096538e-05,
+      "loss": 0.9613,
+      "step": 981
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7957898556185731,
+      "learning_rate": 1.5639199338931958e-05,
+      "loss": 0.8779,
+      "step": 982
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7459377231572725,
+      "learning_rate": 1.563020008721877e-05,
+      "loss": 1.041,
+      "step": 983
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7039082688345556,
+      "learning_rate": 1.5621194154635604e-05,
+      "loss": 0.9044,
+      "step": 984
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.742344579966378,
+      "learning_rate": 1.5612181551869027e-05,
+      "loss": 0.924,
+      "step": 985
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6024372169441903,
+      "learning_rate": 1.5603162289613503e-05,
+      "loss": 0.825,
+      "step": 986
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.737263958051855,
+      "learning_rate": 1.559413637857141e-05,
+      "loss": 0.8641,
+      "step": 987
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7826235654917905,
+      "learning_rate": 1.5585103829453018e-05,
+      "loss": 0.9496,
+      "step": 988
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7638898909110585,
+      "learning_rate": 1.5576064652976465e-05,
+      "loss": 0.9035,
+      "step": 989
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7145409188434416,
+      "learning_rate": 1.556701885986776e-05,
+      "loss": 0.9516,
+      "step": 990
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7256112640585685,
+      "learning_rate": 1.5557966460860763e-05,
+      "loss": 0.8362,
+      "step": 991
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7255874506973344,
+      "learning_rate": 1.554890746669716e-05,
+      "loss": 1.0159,
+      "step": 992
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9183052127224383,
+      "learning_rate": 1.553984188812649e-05,
+      "loss": 0.9674,
+      "step": 993
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5530416002352779,
+      "learning_rate": 1.5530769735906076e-05,
+      "loss": 0.6135,
+      "step": 994
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7574031702524219,
+      "learning_rate": 1.552169102080106e-05,
+      "loss": 0.9188,
+      "step": 995
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8475856873324751,
+      "learning_rate": 1.551260575358436e-05,
+      "loss": 0.9213,
+      "step": 996
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7986347458100779,
+      "learning_rate": 1.5503513945036682e-05,
+      "loss": 1.0056,
+      "step": 997
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8244590904982243,
+      "learning_rate": 1.549441560594648e-05,
+      "loss": 0.7996,
+      "step": 998
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6362334538213847,
+      "learning_rate": 1.5485310747109964e-05,
+      "loss": 0.9953,
+      "step": 999
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7499826167456226,
+      "learning_rate": 1.547619937933108e-05,
+      "loss": 0.9225,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7515529369779602,
+      "learning_rate": 1.5467081513421504e-05,
+      "loss": 0.8067,
+      "step": 1001
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7052450043807378,
+      "learning_rate": 1.5457957160200608e-05,
+      "loss": 0.9224,
+      "step": 1002
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6961599076825891,
+      "learning_rate": 1.544882633049548e-05,
+      "loss": 0.5898,
+      "step": 1003
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7852110146860738,
+      "learning_rate": 1.5439689035140873e-05,
+      "loss": 1.0005,
+      "step": 1004
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7237991392703117,
+      "learning_rate": 1.5430545284979225e-05,
+      "loss": 0.8284,
+      "step": 1005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8185248423762147,
+      "learning_rate": 1.542139509086064e-05,
+      "loss": 0.9245,
+      "step": 1006
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6443900019016031,
+      "learning_rate": 1.5412238463642845e-05,
+      "loss": 0.8363,
+      "step": 1007
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7573966370362879,
+      "learning_rate": 1.540307541419123e-05,
+      "loss": 0.8061,
+      "step": 1008
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.669704022634805,
+      "learning_rate": 1.539390595337878e-05,
+      "loss": 1.0209,
+      "step": 1009
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6950785416299625,
+      "learning_rate": 1.5384730092086106e-05,
+      "loss": 1.0058,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6993080736123284,
+      "learning_rate": 1.5375547841201397e-05,
+      "loss": 0.9771,
+      "step": 1011
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7393208779179037,
+      "learning_rate": 1.5366359211620442e-05,
+      "loss": 0.5074,
+      "step": 1012
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8331846445052292,
+      "learning_rate": 1.535716421424659e-05,
+      "loss": 0.8916,
+      "step": 1013
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6876833672942964,
+      "learning_rate": 1.5347962859990744e-05,
+      "loss": 0.9164,
+      "step": 1014
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7155177269538452,
+      "learning_rate": 1.533875515977135e-05,
+      "loss": 0.8252,
+      "step": 1015
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7262462381043328,
+      "learning_rate": 1.5329541124514394e-05,
+      "loss": 1.0375,
+      "step": 1016
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7749969918247179,
+      "learning_rate": 1.5320320765153367e-05,
+      "loss": 0.876,
+      "step": 1017
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7574956498554806,
+      "learning_rate": 1.531109409262927e-05,
+      "loss": 0.7098,
+      "step": 1018
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6370469341815794,
+      "learning_rate": 1.5301861117890598e-05,
+      "loss": 0.8086,
+      "step": 1019
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8113931295330084,
+      "learning_rate": 1.5292621851893318e-05,
+      "loss": 0.7468,
+      "step": 1020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7310726801536405,
+      "learning_rate": 1.5283376305600866e-05,
+      "loss": 1.0767,
+      "step": 1021
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.69759840796116,
+      "learning_rate": 1.527412448998413e-05,
+      "loss": 0.8617,
+      "step": 1022
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7741834794463082,
+      "learning_rate": 1.526486641602144e-05,
+      "loss": 0.9354,
+      "step": 1023
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9284169619852171,
+      "learning_rate": 1.525560209469854e-05,
+      "loss": 0.9398,
+      "step": 1024
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6862536126862043,
+      "learning_rate": 1.5246331537008607e-05,
+      "loss": 0.6747,
+      "step": 1025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7818398509063521,
+      "learning_rate": 1.5237054753952203e-05,
+      "loss": 1.0382,
+      "step": 1026
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8389526673583257,
+      "learning_rate": 1.522777175653728e-05,
+      "loss": 0.8384,
+      "step": 1027
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7770221380950589,
+      "learning_rate": 1.5218482555779167e-05,
+      "loss": 1.0275,
+      "step": 1028
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7712455448079656,
+      "learning_rate": 1.5209187162700551e-05,
+      "loss": 0.8311,
+      "step": 1029
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7555867584885227,
+      "learning_rate": 1.5199885588331476e-05,
+      "loss": 0.9537,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6649922536577427,
+      "learning_rate": 1.5190577843709302e-05,
+      "loss": 0.5565,
+      "step": 1031
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7017918535105656,
+      "learning_rate": 1.5181263939878728e-05,
+      "loss": 0.9226,
+      "step": 1032
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.90993428165647,
+      "learning_rate": 1.517194388789176e-05,
+      "loss": 1.0486,
+      "step": 1033
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7077849974919667,
+      "learning_rate": 1.5162617698807686e-05,
+      "loss": 0.8076,
+      "step": 1034
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7175245901167262,
+      "learning_rate": 1.5153285383693091e-05,
+      "loss": 0.9587,
+      "step": 1035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.00892123489755,
+      "learning_rate": 1.5143946953621825e-05,
+      "loss": 0.9419,
+      "step": 1036
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6738780631722774,
+      "learning_rate": 1.5134602419674986e-05,
+      "loss": 0.8443,
+      "step": 1037
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7758906971909708,
+      "learning_rate": 1.5125251792940927e-05,
+      "loss": 0.9593,
+      "step": 1038
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5991926215956831,
+      "learning_rate": 1.5115895084515226e-05,
+      "loss": 0.7164,
+      "step": 1039
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8342345894478881,
+      "learning_rate": 1.5106532305500677e-05,
+      "loss": 0.9418,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8269873284111129,
+      "learning_rate": 1.5097163467007272e-05,
+      "loss": 0.9565,
+      "step": 1041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.691329913447092,
+      "learning_rate": 1.5087788580152207e-05,
+      "loss": 1.0611,
+      "step": 1042
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8059828082272708,
+      "learning_rate": 1.5078407656059839e-05,
+      "loss": 0.8851,
+      "step": 1043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7218259468748448,
+      "learning_rate": 1.5069020705861705e-05,
+      "loss": 0.7877,
+      "step": 1044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7670236222389332,
+      "learning_rate": 1.5059627740696476e-05,
+      "loss": 1.0065,
+      "step": 1045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6938246394284109,
+      "learning_rate": 1.5050228771709975e-05,
+      "loss": 0.7229,
+      "step": 1046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7274716793914339,
+      "learning_rate": 1.5040823810055143e-05,
+      "loss": 1.0815,
+      "step": 1047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.705518446670655,
+      "learning_rate": 1.5031412866892025e-05,
+      "loss": 0.8256,
+      "step": 1048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7130042658089357,
+      "learning_rate": 1.502199595338778e-05,
+      "loss": 0.9151,
+      "step": 1049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8010087708958203,
+      "learning_rate": 1.5012573080716636e-05,
+      "loss": 0.7931,
+      "step": 1050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7833448632298003,
+      "learning_rate": 1.50031442600599e-05,
+      "loss": 0.9561,
+      "step": 1051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7303642149935582,
+      "learning_rate": 1.4993709502605937e-05,
+      "loss": 0.8519,
+      "step": 1052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.613865211856125,
+      "learning_rate": 1.4984268819550154e-05,
+      "loss": 0.8723,
+      "step": 1053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.761893658292256,
+      "learning_rate": 1.4974822222094991e-05,
+      "loss": 0.9456,
+      "step": 1054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7092264387152523,
+      "learning_rate": 1.4965369721449902e-05,
+      "loss": 0.9833,
+      "step": 1055
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4807966238060191,
+      "learning_rate": 1.4955911328831357e-05,
+      "loss": 0.5369,
+      "step": 1056
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7266167591447288,
+      "learning_rate": 1.4946447055462803e-05,
+      "loss": 0.9366,
+      "step": 1057
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8485010005232518,
+      "learning_rate": 1.4936976912574675e-05,
+      "loss": 0.8447,
+      "step": 1058
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6753145837736161,
+      "learning_rate": 1.492750091140437e-05,
+      "loss": 0.9554,
+      "step": 1059
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5964237019932189,
+      "learning_rate": 1.4918019063196236e-05,
+      "loss": 0.9386,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7099124306116348,
+      "learning_rate": 1.4908531379201561e-05,
+      "loss": 0.7532,
+      "step": 1061
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8076555333794316,
+      "learning_rate": 1.4899037870678552e-05,
+      "loss": 1.069,
+      "step": 1062
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8046830321638706,
+      "learning_rate": 1.4889538548892336e-05,
+      "loss": 0.7969,
+      "step": 1063
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6502994234343192,
+      "learning_rate": 1.4880033425114937e-05,
+      "loss": 0.9136,
+      "step": 1064
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6014325100278229,
+      "learning_rate": 1.4870522510625252e-05,
+      "loss": 0.6562,
+      "step": 1065
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8517150254038642,
+      "learning_rate": 1.4861005816709064e-05,
+      "loss": 0.9957,
+      "step": 1066
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7382912449486626,
+      "learning_rate": 1.4851483354659009e-05,
+      "loss": 0.7526,
+      "step": 1067
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7172353438696498,
+      "learning_rate": 1.4841955135774562e-05,
+      "loss": 0.8117,
+      "step": 1068
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7753780175659826,
+      "learning_rate": 1.4832421171362037e-05,
+      "loss": 0.9983,
+      "step": 1069
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8451928330487922,
+      "learning_rate": 1.4822881472734563e-05,
+      "loss": 0.8421,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.728208093184136,
+      "learning_rate": 1.481333605121207e-05,
+      "loss": 0.9504,
+      "step": 1071
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7987579717907854,
+      "learning_rate": 1.4803784918121282e-05,
+      "loss": 0.9017,
+      "step": 1072
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7234732380342227,
+      "learning_rate": 1.4794228084795704e-05,
+      "loss": 1.0107,
+      "step": 1073
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6890850095390785,
+      "learning_rate": 1.4784665562575596e-05,
+      "loss": 0.7929,
+      "step": 1074
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7486911687431321,
+      "learning_rate": 1.4775097362807973e-05,
+      "loss": 0.6791,
+      "step": 1075
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8321396327195957,
+      "learning_rate": 1.4765523496846592e-05,
+      "loss": 0.9853,
+      "step": 1076
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7924527703516977,
+      "learning_rate": 1.4755943976051928e-05,
+      "loss": 1.0448,
+      "step": 1077
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6781816000663969,
+      "learning_rate": 1.4746358811791164e-05,
+      "loss": 0.8331,
+      "step": 1078
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.724920009409107,
+      "learning_rate": 1.4736768015438185e-05,
+      "loss": 0.8235,
+      "step": 1079
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8850310320689315,
+      "learning_rate": 1.472717159837356e-05,
+      "loss": 0.8825,
+      "step": 1080
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7243389210912141,
+      "learning_rate": 1.471756957198452e-05,
+      "loss": 0.9505,
+      "step": 1081
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8220931186402518,
+      "learning_rate": 1.4707961947664958e-05,
+      "loss": 0.8019,
+      "step": 1082
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6021070031475277,
+      "learning_rate": 1.4698348736815407e-05,
+      "loss": 0.758,
+      "step": 1083
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.802422829288466,
+      "learning_rate": 1.4688729950843035e-05,
+      "loss": 0.8279,
+      "step": 1084
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6933241074366427,
+      "learning_rate": 1.4679105601161614e-05,
+      "loss": 1.0316,
+      "step": 1085
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2113550559719253,
+      "learning_rate": 1.4669475699191531e-05,
+      "loss": 0.8126,
+      "step": 1086
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6762678589309667,
+      "learning_rate": 1.465984025635975e-05,
+      "loss": 0.8723,
+      "step": 1087
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7998514245837691,
+      "learning_rate": 1.4650199284099817e-05,
+      "loss": 1.0006,
+      "step": 1088
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8113278432232134,
+      "learning_rate": 1.4640552793851842e-05,
+      "loss": 0.8679,
+      "step": 1089
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7642118543115265,
+      "learning_rate": 1.4630900797062473e-05,
+      "loss": 0.9507,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7519874091138785,
+      "learning_rate": 1.4621243305184898e-05,
+      "loss": 0.8758,
+      "step": 1091
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6999615137834931,
+      "learning_rate": 1.4611580329678822e-05,
+      "loss": 0.9784,
+      "step": 1092
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7460968145401363,
+      "learning_rate": 1.4601911882010465e-05,
+      "loss": 0.5269,
+      "step": 1093
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8455030117642025,
+      "learning_rate": 1.4592237973652532e-05,
+      "loss": 0.8086,
+      "step": 1094
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7389891970613135,
+      "learning_rate": 1.458255861608421e-05,
+      "loss": 0.9661,
+      "step": 1095
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7093502613464686,
+      "learning_rate": 1.4572873820791159e-05,
+      "loss": 0.8589,
+      "step": 1096
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8210800621789142,
+      "learning_rate": 1.4563183599265476e-05,
+      "loss": 1.0655,
+      "step": 1097
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6982333675817944,
+      "learning_rate": 1.4553487963005712e-05,
+      "loss": 0.9979,
+      "step": 1098
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6889054484153586,
+      "learning_rate": 1.4543786923516835e-05,
+      "loss": 0.8898,
+      "step": 1099
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8954096466237704,
+      "learning_rate": 1.453408049231023e-05,
+      "loss": 0.5867,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9641512831779538,
+      "learning_rate": 1.4524368680903676e-05,
+      "loss": 0.8842,
+      "step": 1101
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.66350831659718,
+      "learning_rate": 1.4514651500821333e-05,
+      "loss": 0.8518,
+      "step": 1102
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7511635538821679,
+      "learning_rate": 1.4504928963593738e-05,
+      "loss": 0.7863,
+      "step": 1103
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7744526944166783,
+      "learning_rate": 1.4495201080757786e-05,
+      "loss": 1.0565,
+      "step": 1104
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9138153337065161,
+      "learning_rate": 1.4485467863856704e-05,
+      "loss": 0.8234,
+      "step": 1105
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7492078947638284,
+      "learning_rate": 1.4475729324440063e-05,
+      "loss": 0.8216,
+      "step": 1106
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6853487863587396,
+      "learning_rate": 1.4465985474063741e-05,
+      "loss": 0.8999,
+      "step": 1107
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9654563916947938,
+      "learning_rate": 1.4456236324289919e-05,
+      "loss": 1.0705,
+      "step": 1108
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7857167610233978,
+      "learning_rate": 1.4446481886687067e-05,
+      "loss": 0.7087,
+      "step": 1109
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8500586058072618,
+      "learning_rate": 1.4436722172829931e-05,
+      "loss": 0.8108,
+      "step": 1110
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7553763529073516,
+      "learning_rate": 1.4426957194299516e-05,
+      "loss": 1.0292,
+      "step": 1111
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6339049392072063,
+      "learning_rate": 1.4417186962683072e-05,
+      "loss": 0.9225,
+      "step": 1112
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6616083433856973,
+      "learning_rate": 1.4407411489574088e-05,
+      "loss": 0.7512,
+      "step": 1113
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.859620054660873,
+      "learning_rate": 1.4397630786572272e-05,
+      "loss": 1.0532,
+      "step": 1114
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6620611297983571,
+      "learning_rate": 1.4387844865283534e-05,
+      "loss": 0.661,
+      "step": 1115
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7551771532376446,
+      "learning_rate": 1.4378053737319975e-05,
+      "loss": 0.8533,
+      "step": 1116
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7117607818329134,
+      "learning_rate": 1.436825741429988e-05,
+      "loss": 0.8195,
+      "step": 1117
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8888079385700399,
+      "learning_rate": 1.4358455907847695e-05,
+      "loss": 0.9258,
+      "step": 1118
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6785001811365153,
+      "learning_rate": 1.4348649229594019e-05,
+      "loss": 0.6705,
+      "step": 1119
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7216973871380711,
+      "learning_rate": 1.4338837391175582e-05,
+      "loss": 0.9187,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0000458854228198,
+      "learning_rate": 1.4329020404235245e-05,
+      "loss": 1.0878,
+      "step": 1121
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8154377609496678,
+      "learning_rate": 1.4319198280421973e-05,
+      "loss": 0.798,
+      "step": 1122
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5466408292008748,
+      "learning_rate": 1.430937103139083e-05,
+      "loss": 0.9662,
+      "step": 1123
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7105232753685735,
+      "learning_rate": 1.429953866880296e-05,
+      "loss": 0.8759,
+      "step": 1124
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.013400552722546,
+      "learning_rate": 1.428970120432557e-05,
+      "loss": 0.7506,
+      "step": 1125
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6914220420835987,
+      "learning_rate": 1.427985864963193e-05,
+      "loss": 0.907,
+      "step": 1126
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.549411682756462,
+      "learning_rate": 1.4270011016401345e-05,
+      "loss": 0.579,
+      "step": 1127
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7787137463055716,
+      "learning_rate": 1.4260158316319146e-05,
+      "loss": 0.9728,
+      "step": 1128
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8350286510655481,
+      "learning_rate": 1.425030056107668e-05,
+      "loss": 0.7976,
+      "step": 1129
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.76525603891466,
+      "learning_rate": 1.4240437762371287e-05,
+      "loss": 1.0294,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8973233521224101,
+      "learning_rate": 1.4230569931906295e-05,
+      "loss": 1.0086,
+      "step": 1131
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8982562759580802,
+      "learning_rate": 1.4220697081391002e-05,
+      "loss": 0.894,
+      "step": 1132
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6185527880064313,
+      "learning_rate": 1.4210819222540665e-05,
+      "loss": 0.973,
+      "step": 1133
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8095433243558583,
+      "learning_rate": 1.4200936367076481e-05,
+      "loss": 0.7619,
+      "step": 1134
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8401991821740782,
+      "learning_rate": 1.4191048526725575e-05,
+      "loss": 0.9877,
+      "step": 1135
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6332624732914807,
+      "learning_rate": 1.4181155713220994e-05,
+      "loss": 0.5114,
+      "step": 1136
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6995291907056385,
+      "learning_rate": 1.4171257938301679e-05,
+      "loss": 0.856,
+      "step": 1137
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6422281325482256,
+      "learning_rate": 1.416135521371246e-05,
+      "loss": 0.8563,
+      "step": 1138
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7548528312412408,
+      "learning_rate": 1.415144755120404e-05,
+      "loss": 0.7394,
+      "step": 1139
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8606131194037356,
+      "learning_rate": 1.4141534962532986e-05,
+      "loss": 1.044,
+      "step": 1140
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7063757862957443,
+      "learning_rate": 1.4131617459461706e-05,
+      "loss": 0.9277,
+      "step": 1141
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7114497001182538,
+      "learning_rate": 1.4121695053758437e-05,
+      "loss": 0.8946,
+      "step": 1142
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6855662010094433,
+      "learning_rate": 1.4111767757197243e-05,
+      "loss": 0.8458,
+      "step": 1143
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6308383235777157,
+      "learning_rate": 1.4101835581557983e-05,
+      "loss": 0.954,
+      "step": 1144
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6808661820552,
+      "learning_rate": 1.4091898538626306e-05,
+      "loss": 0.6218,
+      "step": 1145
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7803496673271412,
+      "learning_rate": 1.4081956640193639e-05,
+      "loss": 0.8159,
+      "step": 1146
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6668545893029373,
+      "learning_rate": 1.4072009898057173e-05,
+      "loss": 0.9926,
+      "step": 1147
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7445303003133904,
+      "learning_rate": 1.4062058324019842e-05,
+      "loss": 0.7684,
+      "step": 1148
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7211521084225674,
+      "learning_rate": 1.4052101929890317e-05,
+      "loss": 0.9149,
+      "step": 1149
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7729164540795318,
+      "learning_rate": 1.4042140727482985e-05,
+      "loss": 0.9484,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7866120289325282,
+      "learning_rate": 1.4032174728617943e-05,
+      "loss": 0.763,
+      "step": 1151
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6109328522829994,
+      "learning_rate": 1.4022203945120971e-05,
+      "loss": 0.982,
+      "step": 1152
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7596747281088297,
+      "learning_rate": 1.4012228388823543e-05,
+      "loss": 0.6196,
+      "step": 1153
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7789233722401135,
+      "learning_rate": 1.400224807156278e-05,
+      "loss": 0.9948,
+      "step": 1154
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6496781071214587,
+      "learning_rate": 1.3992263005181458e-05,
+      "loss": 0.9549,
+      "step": 1155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6839044360848114,
+      "learning_rate": 1.3982273201527994e-05,
+      "loss": 0.9153,
+      "step": 1156
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7586607346922905,
+      "learning_rate": 1.3972278672456416e-05,
+      "loss": 0.9901,
+      "step": 1157
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7345845540777466,
+      "learning_rate": 1.3962279429826367e-05,
+      "loss": 0.8488,
+      "step": 1158
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7722326914242615,
+      "learning_rate": 1.3952275485503077e-05,
+      "loss": 0.9682,
+      "step": 1159
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8342267512413897,
+      "learning_rate": 1.3942266851357365e-05,
+      "loss": 0.9529,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7138863372119542,
+      "learning_rate": 1.3932253539265603e-05,
+      "loss": 0.9213,
+      "step": 1161
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6259182735631825,
+      "learning_rate": 1.3922235561109725e-05,
+      "loss": 0.7357,
+      "step": 1162
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7162032200216009,
+      "learning_rate": 1.3912212928777192e-05,
+      "loss": 0.5948,
+      "step": 1163
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0302287626061204,
+      "learning_rate": 1.3902185654160995e-05,
+      "loss": 1.0414,
+      "step": 1164
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6747023525923188,
+      "learning_rate": 1.3892153749159629e-05,
+      "loss": 0.8207,
+      "step": 1165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7401830194976567,
+      "learning_rate": 1.3882117225677087e-05,
+      "loss": 0.959,
+      "step": 1166
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8507977645754743,
+      "learning_rate": 1.3872076095622842e-05,
+      "loss": 0.9433,
+      "step": 1167
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8354294859624324,
+      "learning_rate": 1.386203037091183e-05,
+      "loss": 0.8859,
+      "step": 1168
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6805907386654584,
+      "learning_rate": 1.3851980063464436e-05,
+      "loss": 0.9916,
+      "step": 1169
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8066292802604099,
+      "learning_rate": 1.3841925185206497e-05,
+      "loss": 0.9022,
+      "step": 1170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6509489040303578,
+      "learning_rate": 1.383186574806926e-05,
+      "loss": 0.9473,
+      "step": 1171
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9750911584684134,
+      "learning_rate": 1.3821801763989385e-05,
+      "loss": 0.8066,
+      "step": 1172
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7471742747142969,
+      "learning_rate": 1.3811733244908932e-05,
+      "loss": 1.0432,
+      "step": 1173
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8067780458775917,
+      "learning_rate": 1.3801660202775335e-05,
+      "loss": 0.7267,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7560064387363743,
+      "learning_rate": 1.3791582649541404e-05,
+      "loss": 0.8953,
+      "step": 1175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7083772890693635,
+      "learning_rate": 1.3781500597165294e-05,
+      "loss": 0.8979,
+      "step": 1176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8222250719820199,
+      "learning_rate": 1.3771414057610505e-05,
+      "loss": 0.8909,
+      "step": 1177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6647309515698221,
+      "learning_rate": 1.3761323042845855e-05,
+      "loss": 0.9431,
+      "step": 1178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7042478090370445,
+      "learning_rate": 1.3751227564845475e-05,
+      "loss": 0.9338,
+      "step": 1179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.562765862033568,
+      "learning_rate": 1.3741127635588795e-05,
+      "loss": 0.6793,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7835373770998655,
+      "learning_rate": 1.3731023267060526e-05,
+      "loss": 0.8666,
+      "step": 1181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7009491013245143,
+      "learning_rate": 1.3720914471250644e-05,
+      "loss": 0.8476,
+      "step": 1182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7461269455756311,
+      "learning_rate": 1.3710801260154378e-05,
+      "loss": 0.9512,
+      "step": 1183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6996301152849353,
+      "learning_rate": 1.37006836457722e-05,
+      "loss": 0.7488,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7286886687163431,
+      "learning_rate": 1.3690561640109807e-05,
+      "loss": 1.0065,
+      "step": 1185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7662334862242761,
+      "learning_rate": 1.3680435255178104e-05,
+      "loss": 0.8318,
+      "step": 1186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6257497507333832,
+      "learning_rate": 1.3670304502993192e-05,
+      "loss": 0.6905,
+      "step": 1187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7156059779496787,
+      "learning_rate": 1.3660169395576357e-05,
+      "loss": 1.0135,
+      "step": 1188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5692026172191501,
+      "learning_rate": 1.3650029944954048e-05,
+      "loss": 0.5167,
+      "step": 1189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.769506804220159,
+      "learning_rate": 1.3639886163157878e-05,
+      "loss": 0.9827,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7542838173134176,
+      "learning_rate": 1.3629738062224589e-05,
+      "loss": 0.828,
+      "step": 1191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7769667263109054,
+      "learning_rate": 1.361958565419605e-05,
+      "loss": 0.8185,
+      "step": 1192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6641199227366559,
+      "learning_rate": 1.3609428951119248e-05,
+      "loss": 0.8552,
+      "step": 1193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7960795384698918,
+      "learning_rate": 1.3599267965046257e-05,
+      "loss": 0.9232,
+      "step": 1194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7227085157884554,
+      "learning_rate": 1.3589102708034236e-05,
+      "loss": 0.978,
+      "step": 1195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6280672941963308,
+      "learning_rate": 1.3578933192145421e-05,
+      "loss": 0.979,
+      "step": 1196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7218187302393909,
+      "learning_rate": 1.3568759429447087e-05,
+      "loss": 0.9554,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6162141472733976,
+      "learning_rate": 1.3558581432011559e-05,
+      "loss": 0.5578,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7954362857790177,
+      "learning_rate": 1.3548399211916183e-05,
+      "loss": 0.8538,
+      "step": 1199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6375226601967073,
+      "learning_rate": 1.3538212781243317e-05,
+      "loss": 0.8307,
+      "step": 1200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9094032030402622,
+      "learning_rate": 1.3528022152080314e-05,
+      "loss": 0.9966,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7138205194661355,
+      "learning_rate": 1.3517827336519505e-05,
+      "loss": 0.9625,
+      "step": 1202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8186002154599256,
+      "learning_rate": 1.35076283466582e-05,
+      "loss": 0.884,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7081858434471008,
+      "learning_rate": 1.3497425194598654e-05,
+      "loss": 0.826,
+      "step": 1204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.694274255197322,
+      "learning_rate": 1.3487217892448059e-05,
+      "loss": 0.6835,
+      "step": 1205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6476950515047137,
+      "learning_rate": 1.3477006452318538e-05,
+      "loss": 0.6813,
+      "step": 1206
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8139119360098364,
+      "learning_rate": 1.346679088632712e-05,
+      "loss": 0.9988,
+      "step": 1207
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6482993486093173,
+      "learning_rate": 1.345657120659573e-05,
+      "loss": 0.9088,
+      "step": 1208
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7842476413229573,
+      "learning_rate": 1.3446347425251176e-05,
+      "loss": 0.9541,
+      "step": 1209
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7177115476957531,
+      "learning_rate": 1.3436119554425133e-05,
+      "loss": 0.9045,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7088183368960808,
+      "learning_rate": 1.3425887606254126e-05,
+      "loss": 0.6432,
+      "step": 1211
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.20974562921467,
+      "learning_rate": 1.3415651592879522e-05,
+      "loss": 1.0642,
+      "step": 1212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7791616470263661,
+      "learning_rate": 1.3405411526447511e-05,
+      "loss": 0.8272,
+      "step": 1213
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7598298565406084,
+      "learning_rate": 1.339516741910909e-05,
+      "loss": 1.0615,
+      "step": 1214
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8175061788081083,
+      "learning_rate": 1.338491928302005e-05,
+      "loss": 0.7969,
+      "step": 1215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6569626087225867,
+      "learning_rate": 1.3374667130340968e-05,
+      "loss": 0.672,
+      "step": 1216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8178620701885475,
+      "learning_rate": 1.3364410973237185e-05,
+      "loss": 0.6793,
+      "step": 1217
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7221553717885957,
+      "learning_rate": 1.3354150823878792e-05,
+      "loss": 0.9533,
+      "step": 1218
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7338334020147146,
+      "learning_rate": 1.3343886694440615e-05,
+      "loss": 0.9657,
+      "step": 1219
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7716685194794929,
+      "learning_rate": 1.333361859710221e-05,
+      "loss": 0.8748,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7885030246429685,
+      "learning_rate": 1.3323346544047835e-05,
+      "loss": 0.9437,
+      "step": 1221
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7711199476095912,
+      "learning_rate": 1.3313070547466445e-05,
+      "loss": 0.9153,
+      "step": 1222
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.745593101305048,
+      "learning_rate": 1.3302790619551673e-05,
+      "loss": 0.9575,
+      "step": 1223
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.706311357586275,
+      "learning_rate": 1.3292506772501819e-05,
+      "loss": 0.7987,
+      "step": 1224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7877923300348607,
+      "learning_rate": 1.3282219018519832e-05,
+      "loss": 0.7027,
+      "step": 1225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7371638283509776,
+      "learning_rate": 1.3271927369813292e-05,
+      "loss": 0.9898,
+      "step": 1226
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7999528996202798,
+      "learning_rate": 1.326163183859441e-05,
+      "loss": 0.8338,
+      "step": 1227
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6622327879404688,
+      "learning_rate": 1.3251332437080003e-05,
+      "loss": 1.0712,
+      "step": 1228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7280299695611689,
+      "learning_rate": 1.3241029177491472e-05,
+      "loss": 0.8589,
+      "step": 1229
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7585534341666522,
+      "learning_rate": 1.3230722072054803e-05,
+      "loss": 0.8109,
+      "step": 1230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7744402514010759,
+      "learning_rate": 1.3220411133000544e-05,
+      "loss": 0.9238,
+      "step": 1231
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.697355358565349,
+      "learning_rate": 1.321009637256379e-05,
+      "loss": 0.781,
+      "step": 1232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5805147661026007,
+      "learning_rate": 1.3199777802984177e-05,
+      "loss": 0.7633,
+      "step": 1233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8854018552920748,
+      "learning_rate": 1.318945543650585e-05,
+      "loss": 0.9158,
+      "step": 1234
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7391556202744867,
+      "learning_rate": 1.3179129285377472e-05,
+      "loss": 0.9896,
+      "step": 1235
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8226685151616795,
+      "learning_rate": 1.3168799361852184e-05,
+      "loss": 0.8098,
+      "step": 1236
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8284000742092257,
+      "learning_rate": 1.3158465678187616e-05,
+      "loss": 0.9385,
+      "step": 1237
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7129324588628093,
+      "learning_rate": 1.314812824664585e-05,
+      "loss": 0.8621,
+      "step": 1238
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6687380084991045,
+      "learning_rate": 1.313778707949342e-05,
+      "loss": 0.8781,
+      "step": 1239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7401787780503182,
+      "learning_rate": 1.3127442189001296e-05,
+      "loss": 0.942,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8318116824479628,
+      "learning_rate": 1.3117093587444858e-05,
+      "loss": 1.0046,
+      "step": 1241
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5000426918304754,
+      "learning_rate": 1.3106741287103891e-05,
+      "loss": 0.618,
+      "step": 1242
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7118863444055111,
+      "learning_rate": 1.3096385300262578e-05,
+      "loss": 0.7721,
+      "step": 1243
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8234526683355539,
+      "learning_rate": 1.3086025639209465e-05,
+      "loss": 0.844,
+      "step": 1244
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7506496486181399,
+      "learning_rate": 1.3075662316237466e-05,
+      "loss": 0.9742,
+      "step": 1245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7858885598094232,
+      "learning_rate": 1.3065295343643833e-05,
+      "loss": 0.9527,
+      "step": 1246
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6701075306262199,
+      "learning_rate": 1.3054924733730161e-05,
+      "loss": 0.8594,
+      "step": 1247
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7387127878422741,
+      "learning_rate": 1.3044550498802346e-05,
+      "loss": 0.9084,
+      "step": 1248
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7228412886809369,
+      "learning_rate": 1.3034172651170591e-05,
+      "loss": 0.7659,
+      "step": 1249
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6716158099551267,
+      "learning_rate": 1.3023791203149396e-05,
+      "loss": 0.9538,
+      "step": 1250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5603722903522841,
+      "learning_rate": 1.3013406167057521e-05,
+      "loss": 0.6399,
+      "step": 1251
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7838873704707887,
+      "learning_rate": 1.300301755521798e-05,
+      "loss": 1.0077,
+      "step": 1252
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7332802560233432,
+      "learning_rate": 1.2992625379958048e-05,
+      "loss": 0.7756,
+      "step": 1253
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6772730650470918,
+      "learning_rate": 1.2982229653609215e-05,
+      "loss": 0.9068,
+      "step": 1254
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7601271954647116,
+      "learning_rate": 1.2971830388507183e-05,
+      "loss": 0.8317,
+      "step": 1255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8145601137112645,
+      "learning_rate": 1.296142759699186e-05,
+      "loss": 0.7801,
+      "step": 1256
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7825705365831194,
+      "learning_rate": 1.2951021291407338e-05,
+      "loss": 0.9879,
+      "step": 1257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6075657405043862,
+      "learning_rate": 1.2940611484101875e-05,
+      "loss": 0.9249,
+      "step": 1258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7733912068628034,
+      "learning_rate": 1.2930198187427888e-05,
+      "loss": 1.1108,
+      "step": 1259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6975883907709338,
+      "learning_rate": 1.291978141374193e-05,
+      "loss": 0.7458,
+      "step": 1260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.737255907927765,
+      "learning_rate": 1.2909361175404684e-05,
+      "loss": 0.8599,
+      "step": 1261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6884686457565306,
+      "learning_rate": 1.2898937484780944e-05,
+      "loss": 0.9298,
+      "step": 1262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8308508871264243,
+      "learning_rate": 1.2888510354239598e-05,
+      "loss": 0.775,
+      "step": 1263
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6449617851114762,
+      "learning_rate": 1.287807979615362e-05,
+      "loss": 0.7783,
+      "step": 1264
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9870969861161099,
+      "learning_rate": 1.2867645822900043e-05,
+      "loss": 0.8658,
+      "step": 1265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7683965467784423,
+      "learning_rate": 1.285720844685996e-05,
+      "loss": 0.9699,
+      "step": 1266
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.718583746309167,
+      "learning_rate": 1.2846767680418502e-05,
+      "loss": 0.803,
+      "step": 1267
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8274677832477049,
+      "learning_rate": 1.283632353596482e-05,
+      "loss": 0.858,
+      "step": 1268
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.67358224473455,
+      "learning_rate": 1.282587602589207e-05,
+      "loss": 0.6892,
+      "step": 1269
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8815086256650224,
+      "learning_rate": 1.281542516259741e-05,
+      "loss": 0.7996,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7994657512315484,
+      "learning_rate": 1.2804970958481973e-05,
+      "loss": 1.0349,
+      "step": 1271
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7271287753017769,
+      "learning_rate": 1.2794513425950851e-05,
+      "loss": 0.7212,
+      "step": 1272
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7396697282998779,
+      "learning_rate": 1.2784052577413097e-05,
+      "loss": 1.0279,
+      "step": 1273
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9681945480165919,
+      "learning_rate": 1.2773588425281692e-05,
+      "loss": 0.7926,
+      "step": 1274
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0008773341585384,
+      "learning_rate": 1.2763120981973532e-05,
+      "loss": 0.8474,
+      "step": 1275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.770844115809097,
+      "learning_rate": 1.2752650259909425e-05,
+      "loss": 0.9537,
+      "step": 1276
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7943339393823831,
+      "learning_rate": 1.2742176271514076e-05,
+      "loss": 0.714,
+      "step": 1277
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6273948184957681,
+      "learning_rate": 1.273169902921605e-05,
+      "loss": 0.7143,
+      "step": 1278
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8102344795614105,
+      "learning_rate": 1.272121854544778e-05,
+      "loss": 0.7267,
+      "step": 1279
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0025238880072758,
+      "learning_rate": 1.2710734832645557e-05,
+      "loss": 0.7634,
+      "step": 1280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7400630980735202,
+      "learning_rate": 1.2700247903249485e-05,
+      "loss": 0.9661,
+      "step": 1281
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8279521766878795,
+      "learning_rate": 1.2689757769703495e-05,
+      "loss": 0.7931,
+      "step": 1282
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.077796561193155,
+      "learning_rate": 1.2679264444455322e-05,
+      "loss": 1.0634,
+      "step": 1283
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9164944023513445,
+      "learning_rate": 1.266876793995648e-05,
+      "loss": 0.9944,
+      "step": 1284
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7310445982982227,
+      "learning_rate": 1.2658268268662259e-05,
+      "loss": 0.9066,
+      "step": 1285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.661986766528897,
+      "learning_rate": 1.2647765443031715e-05,
+      "loss": 0.5913,
+      "step": 1286
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0186281603693028,
+      "learning_rate": 1.2637259475527635e-05,
+      "loss": 0.8525,
+      "step": 1287
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8211116409997725,
+      "learning_rate": 1.2626750378616539e-05,
+      "loss": 0.7983,
+      "step": 1288
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7879246966190182,
+      "learning_rate": 1.2616238164768659e-05,
+      "loss": 0.8286,
+      "step": 1289
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6926242246139822,
+      "learning_rate": 1.2605722846457934e-05,
+      "loss": 1.0196,
+      "step": 1290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8256931259384713,
+      "learning_rate": 1.2595204436161976e-05,
+      "loss": 0.7806,
+      "step": 1291
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8610196888314808,
+      "learning_rate": 1.2584682946362067e-05,
+      "loss": 0.8132,
+      "step": 1292
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.804965480705006,
+      "learning_rate": 1.2574158389543152e-05,
+      "loss": 0.8761,
+      "step": 1293
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7391154303962527,
+      "learning_rate": 1.2563630778193805e-05,
+      "loss": 1.0428,
+      "step": 1294
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5537098312050998,
+      "learning_rate": 1.255310012480623e-05,
+      "loss": 0.6768,
+      "step": 1295
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7975139577091548,
+      "learning_rate": 1.2542566441876237e-05,
+      "loss": 0.8022,
+      "step": 1296
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7850766676601497,
+      "learning_rate": 1.253202974190324e-05,
+      "loss": 0.9922,
+      "step": 1297
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.868541998841001,
+      "learning_rate": 1.2521490037390224e-05,
+      "loss": 0.9546,
+      "step": 1298
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6719762629506857,
+      "learning_rate": 1.251094734084374e-05,
+      "loss": 0.7424,
+      "step": 1299
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8102881910035892,
+      "learning_rate": 1.2500401664773894e-05,
+      "loss": 1.0057,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7579672603079809,
+      "learning_rate": 1.2489853021694323e-05,
+      "loss": 0.6609,
+      "step": 1301
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8229500546366435,
+      "learning_rate": 1.2479301424122186e-05,
+      "loss": 0.9033,
+      "step": 1302
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8271613570275586,
+      "learning_rate": 1.2468746884578149e-05,
+      "loss": 0.8897,
+      "step": 1303
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9872469593322497,
+      "learning_rate": 1.245818941558637e-05,
+      "loss": 0.851,
+      "step": 1304
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5634642252709411,
+      "learning_rate": 1.2447629029674482e-05,
+      "loss": 0.6798,
+      "step": 1305
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7632817926302065,
+      "learning_rate": 1.2437065739373573e-05,
+      "loss": 0.918,
+      "step": 1306
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8894130824026574,
+      "learning_rate": 1.2426499557218191e-05,
+      "loss": 1.0601,
+      "step": 1307
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8136872872883449,
+      "learning_rate": 1.2415930495746301e-05,
+      "loss": 0.8148,
+      "step": 1308
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6811051726120316,
+      "learning_rate": 1.2405358567499295e-05,
+      "loss": 0.9802,
+      "step": 1309
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.797451844617788,
+      "learning_rate": 1.2394783785021964e-05,
+      "loss": 0.8955,
+      "step": 1310
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.15297420883346,
+      "learning_rate": 1.2384206160862484e-05,
+      "loss": 0.7398,
+      "step": 1311
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.700444190914931,
+      "learning_rate": 1.2373625707572401e-05,
+      "loss": 0.9008,
+      "step": 1312
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.654730669536977,
+      "learning_rate": 1.2363042437706623e-05,
+      "loss": 0.576,
+      "step": 1313
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7762200310735561,
+      "learning_rate": 1.2352456363823397e-05,
+      "loss": 0.9933,
+      "step": 1314
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8003945196278123,
+      "learning_rate": 1.2341867498484303e-05,
+      "loss": 0.8012,
+      "step": 1315
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7389033896113799,
+      "learning_rate": 1.2331275854254222e-05,
+      "loss": 1.0536,
+      "step": 1316
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7816355437837621,
+      "learning_rate": 1.2320681443701345e-05,
+      "loss": 0.9651,
+      "step": 1317
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8953280750458812,
+      "learning_rate": 1.2310084279397137e-05,
+      "loss": 0.9695,
+      "step": 1318
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6979866501666386,
+      "learning_rate": 1.2299484373916333e-05,
+      "loss": 0.9569,
+      "step": 1319
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9292695973291779,
+      "learning_rate": 1.2288881739836925e-05,
+      "loss": 0.7536,
+      "step": 1320
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7174555810791096,
+      "learning_rate": 1.2278276389740134e-05,
+      "loss": 0.9654,
+      "step": 1321
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8141272941659201,
+      "learning_rate": 1.2267668336210411e-05,
+      "loss": 0.6129,
+      "step": 1322
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7444590749996792,
+      "learning_rate": 1.2257057591835416e-05,
+      "loss": 0.7752,
+      "step": 1323
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6459218272730434,
+      "learning_rate": 1.2246444169205994e-05,
+      "loss": 0.8227,
+      "step": 1324
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8313696265166877,
+      "learning_rate": 1.2235828080916174e-05,
+      "loss": 0.8228,
+      "step": 1325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7114133446441637,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.9477,
+      "step": 1326
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7646993862280282,
+      "learning_rate": 1.2214587957747253e-05,
+      "loss": 0.9545,
+      "step": 1327
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7676074950512275,
+      "learning_rate": 1.2203963948071964e-05,
+      "loss": 0.8766,
+      "step": 1328
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7341906249767792,
+      "learning_rate": 1.2193337323143867e-05,
+      "loss": 0.8646,
+      "step": 1329
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8114274260643364,
+      "learning_rate": 1.2182708095572667e-05,
+      "loss": 0.9182,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8607466202760129,
+      "learning_rate": 1.2172076277971132e-05,
+      "loss": 0.6583,
+      "step": 1331
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7811665342031313,
+      "learning_rate": 1.2161441882955127e-05,
+      "loss": 0.8153,
+      "step": 1332
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7377753768461437,
+      "learning_rate": 1.2150804923143562e-05,
+      "loss": 0.9801,
+      "step": 1333
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.768533230605561,
+      "learning_rate": 1.2140165411158395e-05,
+      "loss": 0.7054,
+      "step": 1334
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7793438282036099,
+      "learning_rate": 1.2129523359624615e-05,
+      "loss": 1.041,
+      "step": 1335
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7204227839589812,
+      "learning_rate": 1.2118878781170213e-05,
+      "loss": 0.7754,
+      "step": 1336
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7802851314269371,
+      "learning_rate": 1.2108231688426197e-05,
+      "loss": 0.7633,
+      "step": 1337
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.646461783599246,
+      "learning_rate": 1.2097582094026544e-05,
+      "loss": 0.9797,
+      "step": 1338
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6390213657577729,
+      "learning_rate": 1.2086930010608198e-05,
+      "loss": 0.6172,
+      "step": 1339
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8421554123583941,
+      "learning_rate": 1.207627545081107e-05,
+      "loss": 1.0056,
+      "step": 1340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7540701983344708,
+      "learning_rate": 1.2065618427278e-05,
+      "loss": 0.9341,
+      "step": 1341
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6917839606875729,
+      "learning_rate": 1.2054958952654751e-05,
+      "loss": 0.9083,
+      "step": 1342
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7853747841596916,
+      "learning_rate": 1.2044297039589999e-05,
+      "loss": 0.9442,
+      "step": 1343
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7291121727808876,
+      "learning_rate": 1.203363270073531e-05,
+      "loss": 0.7374,
+      "step": 1344
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8076232052910601,
+      "learning_rate": 1.2022965948745129e-05,
+      "loss": 0.9803,
+      "step": 1345
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6137444824606897,
+      "learning_rate": 1.2012296796276765e-05,
+      "loss": 0.9142,
+      "step": 1346
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7296442036214944,
+      "learning_rate": 1.200162525599038e-05,
+      "loss": 0.9474,
+      "step": 1347
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.830825833019242,
+      "learning_rate": 1.1990951340548961e-05,
+      "loss": 0.7494,
+      "step": 1348
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5697907137143726,
+      "learning_rate": 1.1980275062618316e-05,
+      "loss": 0.5783,
+      "step": 1349
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6788801092447339,
+      "learning_rate": 1.1969596434867063e-05,
+      "loss": 0.9912,
+      "step": 1350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7400127607700471,
+      "learning_rate": 1.19589154699666e-05,
+      "loss": 0.8185,
+      "step": 1351
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7960092832604626,
+      "learning_rate": 1.1948232180591097e-05,
+      "loss": 1.0045,
+      "step": 1352
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6354945035533864,
+      "learning_rate": 1.1937546579417492e-05,
+      "loss": 0.9266,
+      "step": 1353
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7774411229080882,
+      "learning_rate": 1.1926858679125461e-05,
+      "loss": 0.8166,
+      "step": 1354
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7338962541273654,
+      "learning_rate": 1.1916168492397408e-05,
+      "loss": 0.9953,
+      "step": 1355
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8562291175480741,
+      "learning_rate": 1.190547603191845e-05,
+      "loss": 0.8526,
+      "step": 1356
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6729573565383916,
+      "learning_rate": 1.1894781310376397e-05,
+      "loss": 0.8785,
+      "step": 1357
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5602936113054424,
+      "learning_rate": 1.1884084340461757e-05,
+      "loss": 0.7993,
+      "step": 1358
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6579774690474998,
+      "learning_rate": 1.1873385134867687e-05,
+      "loss": 0.9959,
+      "step": 1359
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8948410265683073,
+      "learning_rate": 1.186268370629001e-05,
+      "loss": 0.8234,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8095601497032602,
+      "learning_rate": 1.1851980067427182e-05,
+      "loss": 0.9614,
+      "step": 1361
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9048144326319856,
+      "learning_rate": 1.1841274230980284e-05,
+      "loss": 0.8584,
+      "step": 1362
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7612682321391326,
+      "learning_rate": 1.1830566209653001e-05,
+      "loss": 0.8543,
+      "step": 1363
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6820874090154995,
+      "learning_rate": 1.1819856016151617e-05,
+      "loss": 0.9546,
+      "step": 1364
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6915085238045725,
+      "learning_rate": 1.1809143663184983e-05,
+      "loss": 0.8978,
+      "step": 1365
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.49575020862808555,
+      "learning_rate": 1.1798429163464524e-05,
+      "loss": 0.6814,
+      "step": 1366
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7072088035895112,
+      "learning_rate": 1.1787712529704205e-05,
+      "loss": 0.8767,
+      "step": 1367
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8129629813974341,
+      "learning_rate": 1.177699377462053e-05,
+      "loss": 0.7811,
+      "step": 1368
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.658712756864093,
+      "learning_rate": 1.1766272910932508e-05,
+      "loss": 0.9116,
+      "step": 1369
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7214677656512932,
+      "learning_rate": 1.1755549951361662e-05,
+      "loss": 0.7663,
+      "step": 1370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6281165707243047,
+      "learning_rate": 1.1744824908631997e-05,
+      "loss": 0.9904,
+      "step": 1371
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8193671506058152,
+      "learning_rate": 1.1734097795469995e-05,
+      "loss": 0.8049,
+      "step": 1372
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6027316694371374,
+      "learning_rate": 1.1723368624604585e-05,
+      "loss": 0.8207,
+      "step": 1373
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.727501035839299,
+      "learning_rate": 1.1712637408767145e-05,
+      "loss": 0.8981,
+      "step": 1374
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5058153151636343,
+      "learning_rate": 1.1701904160691477e-05,
+      "loss": 0.5143,
+      "step": 1375
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7032552723954945,
+      "learning_rate": 1.1691168893113798e-05,
+      "loss": 1.0291,
+      "step": 1376
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8932656231665945,
+      "learning_rate": 1.1680431618772714e-05,
+      "loss": 0.7724,
+      "step": 1377
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0218514218490864,
+      "learning_rate": 1.1669692350409223e-05,
+      "loss": 0.8317,
+      "step": 1378
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6934129147378003,
+      "learning_rate": 1.1658951100766677e-05,
+      "loss": 0.8538,
+      "step": 1379
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7832269376829784,
+      "learning_rate": 1.1648207882590788e-05,
+      "loss": 0.9008,
+      "step": 1380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8182681169534534,
+      "learning_rate": 1.1637462708629603e-05,
+      "loss": 0.9884,
+      "step": 1381
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.640191594971987,
+      "learning_rate": 1.1626715591633484e-05,
+      "loss": 1.0,
+      "step": 1382
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7199274571751357,
+      "learning_rate": 1.1615966544355106e-05,
+      "loss": 0.9238,
+      "step": 1383
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.656544857421239,
+      "learning_rate": 1.1605215579549426e-05,
+      "loss": 0.5645,
+      "step": 1384
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1670877025163326,
+      "learning_rate": 1.1594462709973684e-05,
+      "loss": 0.8617,
+      "step": 1385
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7411394031671346,
+      "learning_rate": 1.1583707948387376e-05,
+      "loss": 0.8808,
+      "step": 1386
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8736710796131284,
+      "learning_rate": 1.1572951307552249e-05,
+      "loss": 0.9246,
+      "step": 1387
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7267391842026819,
+      "learning_rate": 1.156219280023227e-05,
+      "loss": 1.0159,
+      "step": 1388
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8244490877720176,
+      "learning_rate": 1.155143243919363e-05,
+      "loss": 0.7613,
+      "step": 1389
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7439871651563,
+      "learning_rate": 1.1540670237204715e-05,
+      "loss": 0.8996,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7044380007327473,
+      "learning_rate": 1.1529906207036095e-05,
+      "loss": 0.6979,
+      "step": 1391
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7035370794274857,
+      "learning_rate": 1.151914036146051e-05,
+      "loss": 0.6737,
+      "step": 1392
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9862009774930434,
+      "learning_rate": 1.1508372713252858e-05,
+      "loss": 1.0307,
+      "step": 1393
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7138575267950122,
+      "learning_rate": 1.149760327519017e-05,
+      "loss": 0.8122,
+      "step": 1394
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8350927402787927,
+      "learning_rate": 1.1486832060051608e-05,
+      "loss": 0.9417,
+      "step": 1395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6473042696077634,
+      "learning_rate": 1.147605908061843e-05,
+      "loss": 0.9219,
+      "step": 1396
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7242163822989973,
+      "learning_rate": 1.1465284349674006e-05,
+      "loss": 0.8326,
+      "step": 1397
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7604773115071917,
+      "learning_rate": 1.1454507880003768e-05,
+      "loss": 1.027,
+      "step": 1398
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7292363833969846,
+      "learning_rate": 1.1443729684395225e-05,
+      "loss": 0.8082,
+      "step": 1399
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8138999711444417,
+      "learning_rate": 1.1432949775637913e-05,
+      "loss": 1.0862,
+      "step": 1400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7952416014939017,
+      "learning_rate": 1.1422168166523431e-05,
+      "loss": 0.7591,
+      "step": 1401
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6053074106628248,
+      "learning_rate": 1.1411384869845367e-05,
+      "loss": 0.614,
+      "step": 1402
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7654299309332647,
+      "learning_rate": 1.1400599898399329e-05,
+      "loss": 0.8257,
+      "step": 1403
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6727814184227147,
+      "learning_rate": 1.1389813264982903e-05,
+      "loss": 0.9086,
+      "step": 1404
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8126985290828174,
+      "learning_rate": 1.1379024982395654e-05,
+      "loss": 1.0262,
+      "step": 1405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7397304859881121,
+      "learning_rate": 1.1368235063439103e-05,
+      "loss": 0.8359,
+      "step": 1406
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8143804564859841,
+      "learning_rate": 1.1357443520916704e-05,
+      "loss": 0.9921,
+      "step": 1407
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7190668161916098,
+      "learning_rate": 1.1346650367633854e-05,
+      "loss": 0.8987,
+      "step": 1408
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6674985092239988,
+      "learning_rate": 1.1335855616397844e-05,
+      "loss": 0.9546,
+      "step": 1409
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7778298708319246,
+      "learning_rate": 1.1325059280017866e-05,
+      "loss": 0.8016,
+      "step": 1410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5735947674998021,
+      "learning_rate": 1.1314261371305006e-05,
+      "loss": 0.7094,
+      "step": 1411
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7941066723172566,
+      "learning_rate": 1.1303461903072201e-05,
+      "loss": 0.9873,
+      "step": 1412
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8663219599691195,
+      "learning_rate": 1.1292660888134242e-05,
+      "loss": 0.8407,
+      "step": 1413
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7030920540501334,
+      "learning_rate": 1.1281858339307753e-05,
+      "loss": 1.0279,
+      "step": 1414
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8193587699568144,
+      "learning_rate": 1.1271054269411186e-05,
+      "loss": 0.8398,
+      "step": 1415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6826643868508016,
+      "learning_rate": 1.1260248691264792e-05,
+      "loss": 0.7035,
+      "step": 1416
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8310627323357849,
+      "learning_rate": 1.1249441617690613e-05,
+      "loss": 0.9551,
+      "step": 1417
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7467432774772956,
+      "learning_rate": 1.1238633061512467e-05,
+      "loss": 0.7834,
+      "step": 1418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5362841741924542,
+      "learning_rate": 1.122782303555593e-05,
+      "loss": 0.7285,
+      "step": 1419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0103374974690789,
+      "learning_rate": 1.1217011552648316e-05,
+      "loss": 0.9186,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7667351257194586,
+      "learning_rate": 1.1206198625618683e-05,
+      "loss": 0.9738,
+      "step": 1421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.649564488427806,
+      "learning_rate": 1.1195384267297788e-05,
+      "loss": 0.777,
+      "step": 1422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8057292378010696,
+      "learning_rate": 1.1184568490518095e-05,
+      "loss": 0.9311,
+      "step": 1423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7362456545445054,
+      "learning_rate": 1.1173751308113741e-05,
+      "loss": 0.8867,
+      "step": 1424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7523511367458553,
+      "learning_rate": 1.1162932732920547e-05,
+      "loss": 0.8371,
+      "step": 1425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7342498788605681,
+      "learning_rate": 1.115211277777597e-05,
+      "loss": 1.0644,
+      "step": 1426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8738579981506678,
+      "learning_rate": 1.1141291455519116e-05,
+      "loss": 0.8306,
+      "step": 1427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.47844571849655376,
+      "learning_rate": 1.1130468778990708e-05,
+      "loss": 0.5784,
+      "step": 1428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7813091708144514,
+      "learning_rate": 1.1119644761033079e-05,
+      "loss": 0.8261,
+      "step": 1429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.800408115559751,
+      "learning_rate": 1.1108819414490151e-05,
+      "loss": 0.8331,
+      "step": 1430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6482817315989068,
+      "learning_rate": 1.1097992752207423e-05,
+      "loss": 0.9428,
+      "step": 1431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9632155421120483,
+      "learning_rate": 1.1087164787031957e-05,
+      "loss": 0.9576,
+      "step": 1432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7426370614843596,
+      "learning_rate": 1.1076335531812357e-05,
+      "loss": 0.8494,
+      "step": 1433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7790973231796181,
+      "learning_rate": 1.1065504999398762e-05,
+      "loss": 0.9211,
+      "step": 1434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7334129437823987,
+      "learning_rate": 1.1054673202642828e-05,
+      "loss": 0.9198,
+      "step": 1435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7165421855580107,
+      "learning_rate": 1.1043840154397706e-05,
+      "loss": 0.9418,
+      "step": 1436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5496695027814007,
+      "learning_rate": 1.1033005867518034e-05,
+      "loss": 0.528,
+      "step": 1437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7007631325077428,
+      "learning_rate": 1.1022170354859922e-05,
+      "loss": 1.0095,
+      "step": 1438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8096912253277259,
+      "learning_rate": 1.1011333629280932e-05,
+      "loss": 0.8087,
+      "step": 1439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6738302092777735,
+      "learning_rate": 1.1000495703640063e-05,
+      "loss": 0.8865,
+      "step": 1440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6998891156968557,
+      "learning_rate": 1.0989656590797748e-05,
+      "loss": 0.8295,
+      "step": 1441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7756468432480083,
+      "learning_rate": 1.0978816303615815e-05,
+      "loss": 0.7342,
+      "step": 1442
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.802872412334687,
+      "learning_rate": 1.0967974854957497e-05,
+      "loss": 0.9601,
+      "step": 1443
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8731630863189139,
+      "learning_rate": 1.0957132257687397e-05,
+      "loss": 0.9605,
+      "step": 1444
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6888149652471947,
+      "learning_rate": 1.0946288524671489e-05,
+      "loss": 0.9276,
+      "step": 1445
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5936381282421894,
+      "learning_rate": 1.0935443668777086e-05,
+      "loss": 0.705,
+      "step": 1446
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7210845567474229,
+      "learning_rate": 1.0924597702872836e-05,
+      "loss": 0.8881,
+      "step": 1447
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.771233119077489,
+      "learning_rate": 1.091375063982871e-05,
+      "loss": 0.8954,
+      "step": 1448
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7672316723317599,
+      "learning_rate": 1.0902902492515978e-05,
+      "loss": 0.7292,
+      "step": 1449
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6371382061150561,
+      "learning_rate": 1.0892053273807193e-05,
+      "loss": 0.8322,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7047627738723445,
+      "learning_rate": 1.0881202996576176e-05,
+      "loss": 0.8902,
+      "step": 1451
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.690089540405703,
+      "learning_rate": 1.087035167369802e-05,
+      "loss": 0.9724,
+      "step": 1452
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6914693823437812,
+      "learning_rate": 1.0859499318049043e-05,
+      "loss": 0.8155,
+      "step": 1453
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7620412823915277,
+      "learning_rate": 1.084864594250679e-05,
+      "loss": 0.8099,
+      "step": 1454
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5331299099400754,
+      "learning_rate": 1.0837791559950029e-05,
+      "loss": 0.6838,
+      "step": 1455
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7757075585655383,
+      "learning_rate": 1.0826936183258705e-05,
+      "loss": 0.7687,
+      "step": 1456
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6644633398963062,
+      "learning_rate": 1.081607982531396e-05,
+      "loss": 1.0365,
+      "step": 1457
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6541801755521327,
+      "learning_rate": 1.0805222498998085e-05,
+      "loss": 0.6801,
+      "step": 1458
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7142929064700833,
+      "learning_rate": 1.0794364217194533e-05,
+      "loss": 1.0567,
+      "step": 1459
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8146345937619995,
+      "learning_rate": 1.0783504992787882e-05,
+      "loss": 0.8192,
+      "step": 1460
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7534100799949432,
+      "learning_rate": 1.0772644838663828e-05,
+      "loss": 0.8293,
+      "step": 1461
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6654279194402299,
+      "learning_rate": 1.0761783767709182e-05,
+      "loss": 0.9656,
+      "step": 1462
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8244050291337461,
+      "learning_rate": 1.0750921792811829e-05,
+      "loss": 0.7379,
+      "step": 1463
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5617436164952762,
+      "learning_rate": 1.074005892686073e-05,
+      "loss": 0.744,
+      "step": 1464
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7407037161599137,
+      "learning_rate": 1.0729195182745914e-05,
+      "loss": 0.7095,
+      "step": 1465
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7261603437434939,
+      "learning_rate": 1.0718330573358436e-05,
+      "loss": 0.7494,
+      "step": 1466
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7072590894304079,
+      "learning_rate": 1.0707465111590385e-05,
+      "loss": 0.9472,
+      "step": 1467
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8688342473026618,
+      "learning_rate": 1.0696598810334868e-05,
+      "loss": 0.9516,
+      "step": 1468
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8500505977849132,
+      "learning_rate": 1.068573168248598e-05,
+      "loss": 1.0929,
+      "step": 1469
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5648700358997044,
+      "learning_rate": 1.0674863740938797e-05,
+      "loss": 0.941,
+      "step": 1470
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7276729213768136,
+      "learning_rate": 1.066399499858936e-05,
+      "loss": 0.931,
+      "step": 1471
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5621995757284318,
+      "learning_rate": 1.0653125468334667e-05,
+      "loss": 0.5528,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7463810987205793,
+      "learning_rate": 1.0642255163072646e-05,
+      "loss": 0.8332,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6652906652352826,
+      "learning_rate": 1.063138409570214e-05,
+      "loss": 0.8199,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7454172169518675,
+      "learning_rate": 1.062051227912291e-05,
+      "loss": 0.7551,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7154629083144596,
+      "learning_rate": 1.0609639726235592e-05,
+      "loss": 1.0458,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7167918806983917,
+      "learning_rate": 1.05987664499417e-05,
+      "loss": 0.7985,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7507996406612581,
+      "learning_rate": 1.0587892463143611e-05,
+      "loss": 0.8444,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.739053947481401,
+      "learning_rate": 1.0577017778744545e-05,
+      "loss": 0.8729,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7425237063390285,
+      "learning_rate": 1.0566142409648539e-05,
+      "loss": 1.0516,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6763657331697125,
+      "learning_rate": 1.0555266368760451e-05,
+      "loss": 0.7169,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6626402432222167,
+      "learning_rate": 1.054438966898594e-05,
+      "loss": 0.7374,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8367409875880206,
+      "learning_rate": 1.0533512323231438e-05,
+      "loss": 1.0866,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6167553115360243,
+      "learning_rate": 1.0522634344404148e-05,
+      "loss": 0.94,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6857755797830202,
+      "learning_rate": 1.0511755745412026e-05,
+      "loss": 0.7227,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7569823601202339,
+      "learning_rate": 1.050087653916376e-05,
+      "loss": 0.9539,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.631331045748264,
+      "learning_rate": 1.0489996738568757e-05,
+      "loss": 0.5474,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7654067261928684,
+      "learning_rate": 1.047911635653714e-05,
+      "loss": 0.9646,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8035551031038103,
+      "learning_rate": 1.046823540597971e-05,
+      "loss": 0.8499,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.846113449900642,
+      "learning_rate": 1.0457353899807947e-05,
+      "loss": 0.9611,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7323122931368441,
+      "learning_rate": 1.0446471850933988e-05,
+      "loss": 0.5922,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.785700361108295,
+      "learning_rate": 1.0435589272270622e-05,
+      "loss": 0.9664,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7309013441127912,
+      "learning_rate": 1.0424706176731254e-05,
+      "loss": 1.0528,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.823096896204593,
+      "learning_rate": 1.0413822577229913e-05,
+      "loss": 0.7803,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5777822519113438,
+      "learning_rate": 1.0402938486681224e-05,
+      "loss": 0.946,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7009713569692715,
+      "learning_rate": 1.0392053918000393e-05,
+      "loss": 0.8458,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8393830771296926,
+      "learning_rate": 1.0381168884103187e-05,
+      "loss": 0.9254,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6652453368800452,
+      "learning_rate": 1.0370283397905941e-05,
+      "loss": 0.7946,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7157119872551941,
+      "learning_rate": 1.035939747232551e-05,
+      "loss": 0.541,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7176409680349055,
+      "learning_rate": 1.034851112027928e-05,
+      "loss": 0.9641,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8062776667448205,
+      "learning_rate": 1.0337624354685136e-05,
+      "loss": 0.8226,
+      "step": 1501
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.745037803374207,
+      "learning_rate": 1.0326737188461465e-05,
+      "loss": 0.9713,
+      "step": 1502
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.704155934620944,
+      "learning_rate": 1.031584963452712e-05,
+      "loss": 0.9184,
+      "step": 1503
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7172459707310884,
+      "learning_rate": 1.0304961705801415e-05,
+      "loss": 0.9289,
+      "step": 1504
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6271707253930564,
+      "learning_rate": 1.0294073415204114e-05,
+      "loss": 0.9038,
+      "step": 1505
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8083170318258965,
+      "learning_rate": 1.0283184775655401e-05,
+      "loss": 0.7392,
+      "step": 1506
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5714503173223557,
+      "learning_rate": 1.0272295800075885e-05,
+      "loss": 0.8266,
+      "step": 1507
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7987984009081422,
+      "learning_rate": 1.0261406501386568e-05,
+      "loss": 0.9223,
+      "step": 1508
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7900377199672082,
+      "learning_rate": 1.0250516892508835e-05,
+      "loss": 0.726,
+      "step": 1509
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6119573773422947,
+      "learning_rate": 1.0239626986364441e-05,
+      "loss": 0.8339,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7001284756064661,
+      "learning_rate": 1.0228736795875489e-05,
+      "loss": 0.8128,
+      "step": 1511
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7366163154954504,
+      "learning_rate": 1.021784633396443e-05,
+      "loss": 0.9766,
+      "step": 1512
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7503662759805028,
+      "learning_rate": 1.0206955613554024e-05,
+      "loss": 0.9059,
+      "step": 1513
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7097701432332376,
+      "learning_rate": 1.0196064647567346e-05,
+      "loss": 0.8586,
+      "step": 1514
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7232127917804818,
+      "learning_rate": 1.0185173448927762e-05,
+      "loss": 0.8915,
+      "step": 1515
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6733806736925138,
+      "learning_rate": 1.017428203055891e-05,
+      "loss": 0.8881,
+      "step": 1516
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6065171599371298,
+      "learning_rate": 1.0163390405384694e-05,
+      "loss": 0.6495,
+      "step": 1517
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7786775135453665,
+      "learning_rate": 1.015249858632926e-05,
+      "loss": 0.798,
+      "step": 1518
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.558002897260543,
+      "learning_rate": 1.0141606586316985e-05,
+      "loss": 0.9942,
+      "step": 1519
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6609950130688277,
+      "learning_rate": 1.013071441827246e-05,
+      "loss": 0.7016,
+      "step": 1520
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7417583017942826,
+      "learning_rate": 1.0119822095120471e-05,
+      "loss": 0.9828,
+      "step": 1521
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7208828430507275,
+      "learning_rate": 1.0108929629786001e-05,
+      "loss": 0.7782,
+      "step": 1522
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7402750998669178,
+      "learning_rate": 1.009803703519419e-05,
+      "loss": 0.7346,
+      "step": 1523
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6200206453981619,
+      "learning_rate": 1.0087144324270332e-05,
+      "loss": 0.9412,
+      "step": 1524
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.580641631801443,
+      "learning_rate": 1.0076251509939867e-05,
+      "loss": 0.594,
+      "step": 1525
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7130057145525144,
+      "learning_rate": 1.0065358605128352e-05,
+      "loss": 0.9561,
+      "step": 1526
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7718025546813219,
+      "learning_rate": 1.0054465622761446e-05,
+      "loss": 0.9468,
+      "step": 1527
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7271893486287399,
+      "learning_rate": 1.0043572575764916e-05,
+      "loss": 0.8929,
+      "step": 1528
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6547655004941632,
+      "learning_rate": 1.0032679477064591e-05,
+      "loss": 0.8797,
+      "step": 1529
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7697916472720387,
+      "learning_rate": 1.0021786339586364e-05,
+      "loss": 0.7607,
+      "step": 1530
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8057002691645274,
+      "learning_rate": 1.0010893176256176e-05,
+      "loss": 1.023,
+      "step": 1531
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6411889097790404,
+      "learning_rate": 1e-05,
+      "loss": 0.9103,
+      "step": 1532
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7251168190645025,
+      "learning_rate": 9.989106823743828e-06,
+      "loss": 0.9058,
+      "step": 1533
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6957753589348845,
+      "learning_rate": 9.97821366041364e-06,
+      "loss": 0.7759,
+      "step": 1534
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5659572790189052,
+      "learning_rate": 9.967320522935412e-06,
+      "loss": 0.5348,
+      "step": 1535
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6908771851992053,
+      "learning_rate": 9.956427424235087e-06,
+      "loss": 0.9767,
+      "step": 1536
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7193814567034799,
+      "learning_rate": 9.945534377238555e-06,
+      "loss": 0.7813,
+      "step": 1537
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7533889828167182,
+      "learning_rate": 9.934641394871651e-06,
+      "loss": 0.9932,
+      "step": 1538
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8000663310826273,
+      "learning_rate": 9.923748490060135e-06,
+      "loss": 0.9375,
+      "step": 1539
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6734807501067632,
+      "learning_rate": 9.91285567572967e-06,
+      "loss": 0.9004,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6860450060393635,
+      "learning_rate": 9.901962964805813e-06,
+      "loss": 0.8324,
+      "step": 1541
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7939912792103084,
+      "learning_rate": 9.891070370214e-06,
+      "loss": 0.8535,
+      "step": 1542
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.714146225664509,
+      "learning_rate": 9.88017790487953e-06,
+      "loss": 0.9147,
+      "step": 1543
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5535567417649389,
+      "learning_rate": 9.869285581727543e-06,
+      "loss": 0.764,
+      "step": 1544
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7101588866899047,
+      "learning_rate": 9.858393413683017e-06,
+      "loss": 1.0679,
+      "step": 1545
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7754206754489977,
+      "learning_rate": 9.847501413670742e-06,
+      "loss": 0.7328,
+      "step": 1546
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7448616079465998,
+      "learning_rate": 9.836609594615309e-06,
+      "loss": 1.0447,
+      "step": 1547
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7557792107516138,
+      "learning_rate": 9.825717969441092e-06,
+      "loss": 0.7826,
+      "step": 1548
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7436472530352235,
+      "learning_rate": 9.81482655107224e-06,
+      "loss": 0.7302,
+      "step": 1549
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8645955659460801,
+      "learning_rate": 9.803935352432657e-06,
+      "loss": 1.0617,
+      "step": 1550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.736948601813171,
+      "learning_rate": 9.79304438644598e-06,
+      "loss": 0.9349,
+      "step": 1551
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5881576264352402,
+      "learning_rate": 9.782153666035575e-06,
+      "loss": 0.6945,
+      "step": 1552
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6814628180452632,
+      "learning_rate": 9.771263204124513e-06,
+      "loss": 0.8398,
+      "step": 1553
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7483613609790991,
+      "learning_rate": 9.760373013635562e-06,
+      "loss": 0.7986,
+      "step": 1554
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7821894753649512,
+      "learning_rate": 9.749483107491167e-06,
+      "loss": 0.9577,
+      "step": 1555
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6688775953578028,
+      "learning_rate": 9.738593498613435e-06,
+      "loss": 0.9382,
+      "step": 1556
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6198596304190267,
+      "learning_rate": 9.727704199924117e-06,
+      "loss": 0.9855,
+      "step": 1557
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7395144863478701,
+      "learning_rate": 9.7168152243446e-06,
+      "loss": 0.8365,
+      "step": 1558
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6453893447751904,
+      "learning_rate": 9.705926584795891e-06,
+      "loss": 0.8,
+      "step": 1559
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8164599592195675,
+      "learning_rate": 9.695038294198588e-06,
+      "loss": 0.9033,
+      "step": 1560
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5564884709761309,
+      "learning_rate": 9.684150365472883e-06,
+      "loss": 0.681,
+      "step": 1561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6837337795815835,
+      "learning_rate": 9.673262811538537e-06,
+      "loss": 0.8749,
+      "step": 1562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8069230320150671,
+      "learning_rate": 9.662375645314867e-06,
+      "loss": 0.8318,
+      "step": 1563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6604308743102597,
+      "learning_rate": 9.651488879720724e-06,
+      "loss": 0.8957,
+      "step": 1564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7023665575973599,
+      "learning_rate": 9.640602527674495e-06,
+      "loss": 0.7538,
+      "step": 1565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7080489607134623,
+      "learning_rate": 9.629716602094062e-06,
+      "loss": 0.8769,
+      "step": 1566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.758472629033268,
+      "learning_rate": 9.618831115896814e-06,
+      "loss": 0.9596,
+      "step": 1567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6730774502225324,
+      "learning_rate": 9.607946081999609e-06,
+      "loss": 0.9961,
+      "step": 1568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6677388187017332,
+      "learning_rate": 9.597061513318778e-06,
+      "loss": 0.8819,
+      "step": 1569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5403823177724486,
+      "learning_rate": 9.586177422770089e-06,
+      "loss": 0.5733,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9635158150361185,
+      "learning_rate": 9.575293823268749e-06,
+      "loss": 0.827,
+      "step": 1571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6060210876901149,
+      "learning_rate": 9.564410727729383e-06,
+      "loss": 0.8822,
+      "step": 1572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8879870416737371,
+      "learning_rate": 9.553528149066016e-06,
+      "loss": 0.8939,
+      "step": 1573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6902052130730717,
+      "learning_rate": 9.542646100192056e-06,
+      "loss": 0.9718,
+      "step": 1574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7754585596272977,
+      "learning_rate": 9.531764594020293e-06,
+      "loss": 0.7437,
+      "step": 1575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7480065497765546,
+      "learning_rate": 9.520883643462862e-06,
+      "loss": 0.8972,
+      "step": 1576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6860397688206373,
+      "learning_rate": 9.510003261431244e-06,
+      "loss": 0.8182,
+      "step": 1577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6835646780753629,
+      "learning_rate": 9.499123460836242e-06,
+      "loss": 0.5364,
+      "step": 1578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7861644819078789,
+      "learning_rate": 9.488244254587977e-06,
+      "loss": 1.0253,
+      "step": 1579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6886066584033688,
+      "learning_rate": 9.477365655595855e-06,
+      "loss": 0.7688,
+      "step": 1580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8038899384816571,
+      "learning_rate": 9.466487676768563e-06,
+      "loss": 0.9226,
+      "step": 1581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6885661854009264,
+      "learning_rate": 9.455610331014064e-06,
+      "loss": 0.9111,
+      "step": 1582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7839179007403801,
+      "learning_rate": 9.444733631239552e-06,
+      "loss": 0.9456,
+      "step": 1583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6887165704766368,
+      "learning_rate": 9.433857590351465e-06,
+      "loss": 0.9816,
+      "step": 1584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6661835923294264,
+      "learning_rate": 9.422982221255458e-06,
+      "loss": 0.7671,
+      "step": 1585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7205260521684148,
+      "learning_rate": 9.41210753685639e-06,
+      "loss": 0.9986,
+      "step": 1586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7009818858080893,
+      "learning_rate": 9.401233550058303e-06,
+      "loss": 0.7358,
+      "step": 1587
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7626831284871393,
+      "learning_rate": 9.390360273764411e-06,
+      "loss": 0.5798,
+      "step": 1588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7771362872820684,
+      "learning_rate": 9.379487720877093e-06,
+      "loss": 0.8387,
+      "step": 1589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6416539038527429,
+      "learning_rate": 9.368615904297861e-06,
+      "loss": 0.9419,
+      "step": 1590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8344879614969797,
+      "learning_rate": 9.357744836927359e-06,
+      "loss": 1.0109,
+      "step": 1591
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7472015344690933,
+      "learning_rate": 9.346874531665336e-06,
+      "loss": 0.8447,
+      "step": 1592
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5813953685982289,
+      "learning_rate": 9.336005001410644e-06,
+      "loss": 0.9816,
+      "step": 1593
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6632612405901143,
+      "learning_rate": 9.325136259061207e-06,
+      "loss": 0.7072,
+      "step": 1594
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7133646392639159,
+      "learning_rate": 9.314268317514023e-06,
+      "loss": 0.9386,
+      "step": 1595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7454272189839326,
+      "learning_rate": 9.303401189665135e-06,
+      "loss": 0.8929,
+      "step": 1596
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7845819415481897,
+      "learning_rate": 9.292534888409618e-06,
+      "loss": 0.5589,
+      "step": 1597
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7654319002816523,
+      "learning_rate": 9.28166942664157e-06,
+      "loss": 1.0066,
+      "step": 1598
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8949218428955654,
+      "learning_rate": 9.270804817254091e-06,
+      "loss": 0.7762,
+      "step": 1599
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6584932916209817,
+      "learning_rate": 9.259941073139272e-06,
+      "loss": 0.9751,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8943124791591465,
+      "learning_rate": 9.249078207188175e-06,
+      "loss": 0.8086,
+      "step": 1601
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7105500770241614,
+      "learning_rate": 9.238216232290821e-06,
+      "loss": 0.9062,
+      "step": 1602
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7445065686814096,
+      "learning_rate": 9.227355161336174e-06,
+      "loss": 0.7864,
+      "step": 1603
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.75597220629178,
+      "learning_rate": 9.216495007212122e-06,
+      "loss": 0.769,
+      "step": 1604
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5800008984013463,
+      "learning_rate": 9.20563578280547e-06,
+      "loss": 0.7649,
+      "step": 1605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5927770821538151,
+      "learning_rate": 9.194777501001919e-06,
+      "loss": 0.8898,
+      "step": 1606
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7654925483781662,
+      "learning_rate": 9.183920174686046e-06,
+      "loss": 0.9538,
+      "step": 1607
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7825237471194422,
+      "learning_rate": 9.173063816741296e-06,
+      "loss": 0.8274,
+      "step": 1608
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7562223396982372,
+      "learning_rate": 9.162208440049976e-06,
+      "loss": 0.9078,
+      "step": 1609
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7195308361860675,
+      "learning_rate": 9.151354057493213e-06,
+      "loss": 0.8585,
+      "step": 1610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6994832491128421,
+      "learning_rate": 9.14050068195096e-06,
+      "loss": 0.8573,
+      "step": 1611
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6885258097251525,
+      "learning_rate": 9.129648326301983e-06,
+      "loss": 1.0766,
+      "step": 1612
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7887053541363115,
+      "learning_rate": 9.118797003423826e-06,
+      "loss": 0.7977,
+      "step": 1613
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.45977616093556306,
+      "learning_rate": 9.10794672619281e-06,
+      "loss": 0.5701,
+      "step": 1614
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7685626339941233,
+      "learning_rate": 9.097097507484023e-06,
+      "loss": 0.8452,
+      "step": 1615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7317262001953612,
+      "learning_rate": 9.086249360171291e-06,
+      "loss": 0.8472,
+      "step": 1616
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7070950225436009,
+      "learning_rate": 9.075402297127167e-06,
+      "loss": 0.9052,
+      "step": 1617
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5796475012230846,
+      "learning_rate": 9.064556331222918e-06,
+      "loss": 0.9475,
+      "step": 1618
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.706707526110707,
+      "learning_rate": 9.053711475328515e-06,
+      "loss": 0.8851,
+      "step": 1619
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8194149879644732,
+      "learning_rate": 9.042867742312605e-06,
+      "loss": 0.9183,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6671813030571792,
+      "learning_rate": 9.032025145042504e-06,
+      "loss": 0.9159,
+      "step": 1621
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6744117006010405,
+      "learning_rate": 9.021183696384186e-06,
+      "loss": 0.9388,
+      "step": 1622
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5808458319917782,
+      "learning_rate": 9.010343409202256e-06,
+      "loss": 0.5519,
+      "step": 1623
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.699637055776632,
+      "learning_rate": 8.99950429635994e-06,
+      "loss": 0.9622,
+      "step": 1624
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.724721038141183,
+      "learning_rate": 8.988666370719071e-06,
+      "loss": 0.7969,
+      "step": 1625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7112913248658582,
+      "learning_rate": 8.977829645140081e-06,
+      "loss": 0.9013,
+      "step": 1626
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6954633329622331,
+      "learning_rate": 8.96699413248197e-06,
+      "loss": 0.7848,
+      "step": 1627
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7729513555516705,
+      "learning_rate": 8.956159845602297e-06,
+      "loss": 0.7595,
+      "step": 1628
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7325174079002889,
+      "learning_rate": 8.945326797357174e-06,
+      "loss": 0.9691,
+      "step": 1629
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5665305855475479,
+      "learning_rate": 8.934495000601241e-06,
+      "loss": 0.9475,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6192625702108923,
+      "learning_rate": 8.923664468187647e-06,
+      "loss": 1.009,
+      "step": 1631
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6485670242016236,
+      "learning_rate": 8.912835212968047e-06,
+      "loss": 0.5876,
+      "step": 1632
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6817935645029677,
+      "learning_rate": 8.902007247792582e-06,
+      "loss": 0.8601,
+      "step": 1633
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7861393495163198,
+      "learning_rate": 8.891180585509854e-06,
+      "loss": 0.9124,
+      "step": 1634
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7979806001561457,
+      "learning_rate": 8.880355238966923e-06,
+      "loss": 0.8043,
+      "step": 1635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6423204868585869,
+      "learning_rate": 8.869531221009294e-06,
+      "loss": 0.7692,
+      "step": 1636
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6384575302147154,
+      "learning_rate": 8.858708544480887e-06,
+      "loss": 0.8278,
+      "step": 1637
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.667638367490074,
+      "learning_rate": 8.847887222224032e-06,
+      "loss": 0.9263,
+      "step": 1638
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7745881470313642,
+      "learning_rate": 8.837067267079456e-06,
+      "loss": 0.8757,
+      "step": 1639
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7021486274555653,
+      "learning_rate": 8.82624869188626e-06,
+      "loss": 0.9287,
+      "step": 1640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5569032900192082,
+      "learning_rate": 8.815431509481909e-06,
+      "loss": 0.5325,
+      "step": 1641
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7479474106829915,
+      "learning_rate": 8.804615732702213e-06,
+      "loss": 0.9146,
+      "step": 1642
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5570253894101501,
+      "learning_rate": 8.793801374381319e-06,
+      "loss": 1.0109,
+      "step": 1643
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6617601527479559,
+      "learning_rate": 8.782988447351685e-06,
+      "loss": 0.86,
+      "step": 1644
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8043112434122923,
+      "learning_rate": 8.772176964444073e-06,
+      "loss": 0.9094,
+      "step": 1645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7151356114474868,
+      "learning_rate": 8.761366938487536e-06,
+      "loss": 0.8469,
+      "step": 1646
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.743487362288343,
+      "learning_rate": 8.75055838230939e-06,
+      "loss": 0.8432,
+      "step": 1647
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6937456257435942,
+      "learning_rate": 8.73975130873521e-06,
+      "loss": 0.9275,
+      "step": 1648
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6831343826481954,
+      "learning_rate": 8.728945730588818e-06,
+      "loss": 0.733,
+      "step": 1649
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.49903859203069006,
+      "learning_rate": 8.718141660692252e-06,
+      "loss": 0.674,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7264336171340051,
+      "learning_rate": 8.707339111865763e-06,
+      "loss": 0.7348,
+      "step": 1651
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6139219746588015,
+      "learning_rate": 8.6965380969278e-06,
+      "loss": 0.7855,
+      "step": 1652
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7349061627206158,
+      "learning_rate": 8.685738628694995e-06,
+      "loss": 0.9281,
+      "step": 1653
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7250524399708869,
+      "learning_rate": 8.674940719982135e-06,
+      "loss": 0.9055,
+      "step": 1654
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2630342376706098,
+      "learning_rate": 8.66414438360216e-06,
+      "loss": 1.0799,
+      "step": 1655
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5618181053377975,
+      "learning_rate": 8.65334963236615e-06,
+      "loss": 0.9192,
+      "step": 1656
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7357296363301792,
+      "learning_rate": 8.642556479083297e-06,
+      "loss": 0.9552,
+      "step": 1657
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5091531858291974,
+      "learning_rate": 8.6317649365609e-06,
+      "loss": 0.5336,
+      "step": 1658
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6829777938484564,
+      "learning_rate": 8.620975017604347e-06,
+      "loss": 0.8537,
+      "step": 1659
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6238660354333496,
+      "learning_rate": 8.610186735017102e-06,
+      "loss": 0.7697,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.649183348491599,
+      "learning_rate": 8.599400101600675e-06,
+      "loss": 0.8085,
+      "step": 1661
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7319356654724375,
+      "learning_rate": 8.588615130154636e-06,
+      "loss": 1.0631,
+      "step": 1662
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7237071696164814,
+      "learning_rate": 8.577831833476574e-06,
+      "loss": 0.798,
+      "step": 1663
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6336043838727821,
+      "learning_rate": 8.567050224362088e-06,
+      "loss": 0.9272,
+      "step": 1664
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7447956103489328,
+      "learning_rate": 8.556270315604779e-06,
+      "loss": 0.7364,
+      "step": 1665
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.727552754441498,
+      "learning_rate": 8.545492119996233e-06,
+      "loss": 1.0757,
+      "step": 1666
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5572696653344834,
+      "learning_rate": 8.534715650325997e-06,
+      "loss": 0.6668,
+      "step": 1667
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6787805060215026,
+      "learning_rate": 8.523940919381571e-06,
+      "loss": 0.7779,
+      "step": 1668
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6746532030506853,
+      "learning_rate": 8.513167939948397e-06,
+      "loss": 0.9754,
+      "step": 1669
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5582564133794369,
+      "learning_rate": 8.502396724809834e-06,
+      "loss": 0.9414,
+      "step": 1670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6869721489956255,
+      "learning_rate": 8.491627286747145e-06,
+      "loss": 0.8754,
+      "step": 1671
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7353818182712629,
+      "learning_rate": 8.480859638539492e-06,
+      "loss": 0.7771,
+      "step": 1672
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6043763179312086,
+      "learning_rate": 8.47009379296391e-06,
+      "loss": 0.6042,
+      "step": 1673
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.713343503229483,
+      "learning_rate": 8.45932976279529e-06,
+      "loss": 0.8738,
+      "step": 1674
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7583117155423813,
+      "learning_rate": 8.448567560806371e-06,
+      "loss": 0.8438,
+      "step": 1675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6821227202521786,
+      "learning_rate": 8.437807199767731e-06,
+      "loss": 1.0269,
+      "step": 1676
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5614515513335271,
+      "learning_rate": 8.427048692447754e-06,
+      "loss": 0.5612,
+      "step": 1677
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7424005388018091,
+      "learning_rate": 8.416292051612625e-06,
+      "loss": 0.9706,
+      "step": 1678
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6928021574760348,
+      "learning_rate": 8.40553729002632e-06,
+      "loss": 1.0278,
+      "step": 1679
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7326716273471002,
+      "learning_rate": 8.394784420450578e-06,
+      "loss": 0.8013,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5222932253072017,
+      "learning_rate": 8.384033455644897e-06,
+      "loss": 0.9373,
+      "step": 1681
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7462394706780808,
+      "learning_rate": 8.373284408366518e-06,
+      "loss": 0.8718,
+      "step": 1682
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7214348604163312,
+      "learning_rate": 8.3625372913704e-06,
+      "loss": 0.9882,
+      "step": 1683
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6963306643563456,
+      "learning_rate": 8.351792117409214e-06,
+      "loss": 0.815,
+      "step": 1684
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6336637184156824,
+      "learning_rate": 8.341048899233324e-06,
+      "loss": 0.5481,
+      "step": 1685
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7125953212510485,
+      "learning_rate": 8.330307649590782e-06,
+      "loss": 0.9519,
+      "step": 1686
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6780154161337357,
+      "learning_rate": 8.319568381227288e-06,
+      "loss": 0.7692,
+      "step": 1687
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7095488782224392,
+      "learning_rate": 8.308831106886206e-06,
+      "loss": 1.1021,
+      "step": 1688
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.737774844587685,
+      "learning_rate": 8.298095839308526e-06,
+      "loss": 0.843,
+      "step": 1689
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6592290013823849,
+      "learning_rate": 8.287362591232859e-06,
+      "loss": 0.929,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6618806474547598,
+      "learning_rate": 8.276631375395417e-06,
+      "loss": 0.9266,
+      "step": 1691
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6085260774769718,
+      "learning_rate": 8.265902204530007e-06,
+      "loss": 0.937,
+      "step": 1692
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6008577188819363,
+      "learning_rate": 8.255175091368004e-06,
+      "loss": 0.6368,
+      "step": 1693
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7390835573395318,
+      "learning_rate": 8.244450048638341e-06,
+      "loss": 0.8543,
+      "step": 1694
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7174036437505333,
+      "learning_rate": 8.233727089067494e-06,
+      "loss": 0.8841,
+      "step": 1695
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5770658044600386,
+      "learning_rate": 8.223006225379474e-06,
+      "loss": 0.6952,
+      "step": 1696
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6878536620656572,
+      "learning_rate": 8.212287470295796e-06,
+      "loss": 0.7937,
+      "step": 1697
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7092593201184705,
+      "learning_rate": 8.201570836535477e-06,
+      "loss": 0.959,
+      "step": 1698
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7054100262208285,
+      "learning_rate": 8.190856336815019e-06,
+      "loss": 0.925,
+      "step": 1699
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6972895502300941,
+      "learning_rate": 8.180143983848388e-06,
+      "loss": 0.8519,
+      "step": 1700
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7220934785711492,
+      "learning_rate": 8.169433790347002e-06,
+      "loss": 0.9029,
+      "step": 1701
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6368858073326332,
+      "learning_rate": 8.158725769019719e-06,
+      "loss": 0.9277,
+      "step": 1702
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.611436888147847,
+      "learning_rate": 8.14801993257282e-06,
+      "loss": 0.5446,
+      "step": 1703
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7317820783051292,
+      "learning_rate": 8.137316293709994e-06,
+      "loss": 0.7679,
+      "step": 1704
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6198833054207816,
+      "learning_rate": 8.126614865132314e-06,
+      "loss": 0.9966,
+      "step": 1705
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.713774683858911,
+      "learning_rate": 8.115915659538246e-06,
+      "loss": 0.6831,
+      "step": 1706
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6881292006514677,
+      "learning_rate": 8.105218689623605e-06,
+      "loss": 0.9993,
+      "step": 1707
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7418483398655865,
+      "learning_rate": 8.094523968081553e-06,
+      "loss": 0.7525,
+      "step": 1708
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7579700865135881,
+      "learning_rate": 8.083831507602593e-06,
+      "loss": 0.7328,
+      "step": 1709
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6538269778594099,
+      "learning_rate": 8.07314132087454e-06,
+      "loss": 0.9702,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.549125555005927,
+      "learning_rate": 8.06245342058251e-06,
+      "loss": 0.5667,
+      "step": 1711
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7006407750052608,
+      "learning_rate": 8.051767819408907e-06,
+      "loss": 0.9514,
+      "step": 1712
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6509866127973132,
+      "learning_rate": 8.041084530033406e-06,
+      "loss": 0.9537,
+      "step": 1713
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7906197409152267,
+      "learning_rate": 8.030403565132942e-06,
+      "loss": 0.9711,
+      "step": 1714
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7310934734624227,
+      "learning_rate": 8.019724937381687e-06,
+      "loss": 0.7729,
+      "step": 1715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7135028261309075,
+      "learning_rate": 8.009048659451044e-06,
+      "loss": 0.8201,
+      "step": 1716
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7452827612655907,
+      "learning_rate": 7.998374744009624e-06,
+      "loss": 1.0026,
+      "step": 1717
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5525833345803298,
+      "learning_rate": 7.987703203723236e-06,
+      "loss": 0.8853,
+      "step": 1718
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6409580334481615,
+      "learning_rate": 7.977034051254875e-06,
+      "loss": 0.8732,
+      "step": 1719
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6461889845417218,
+      "learning_rate": 7.966367299264696e-06,
+      "loss": 0.7652,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5259890205712489,
+      "learning_rate": 7.955702960410006e-06,
+      "loss": 0.6195,
+      "step": 1721
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7436022397239227,
+      "learning_rate": 7.945041047345254e-06,
+      "loss": 0.9833,
+      "step": 1722
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6578725892247318,
+      "learning_rate": 7.934381572722006e-06,
+      "loss": 0.7309,
+      "step": 1723
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7494020583723983,
+      "learning_rate": 7.923724549188935e-06,
+      "loss": 0.9509,
+      "step": 1724
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7589395365353236,
+      "learning_rate": 7.913069989391805e-06,
+      "loss": 0.9283,
+      "step": 1725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6798376547639001,
+      "learning_rate": 7.902417905973463e-06,
+      "loss": 0.9476,
+      "step": 1726
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7387876324598346,
+      "learning_rate": 7.89176831157381e-06,
+      "loss": 0.8884,
+      "step": 1727
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7240804958153796,
+      "learning_rate": 7.881121218829789e-06,
+      "loss": 0.8712,
+      "step": 1728
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6490997649942696,
+      "learning_rate": 7.87047664037539e-06,
+      "loss": 0.9079,
+      "step": 1729
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4718932087671903,
+      "learning_rate": 7.859834588841608e-06,
+      "loss": 0.7653,
+      "step": 1730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7423334030113727,
+      "learning_rate": 7.849195076856443e-06,
+      "loss": 1.049,
+      "step": 1731
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8544209269840272,
+      "learning_rate": 7.838558117044876e-06,
+      "loss": 0.7821,
+      "step": 1732
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6526523877275728,
+      "learning_rate": 7.827923722028871e-06,
+      "loss": 1.0044,
+      "step": 1733
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6697865746348456,
+      "learning_rate": 7.81729190442734e-06,
+      "loss": 0.7286,
+      "step": 1734
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.890889672440269,
+      "learning_rate": 7.806662676856134e-06,
+      "loss": 0.79,
+      "step": 1735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7309962696412541,
+      "learning_rate": 7.79603605192804e-06,
+      "loss": 0.983,
+      "step": 1736
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6854807838947442,
+      "learning_rate": 7.785412042252752e-06,
+      "loss": 0.8916,
+      "step": 1737
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.524545879334837,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.6399,
+      "step": 1738
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6766029477209267,
+      "learning_rate": 7.764171919083831e-06,
+      "loss": 0.8571,
+      "step": 1739
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6792128414615759,
+      "learning_rate": 7.753555830794013e-06,
+      "loss": 0.7404,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7461842077024413,
+      "learning_rate": 7.742942408164591e-06,
+      "loss": 0.9542,
+      "step": 1741
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6218539071122476,
+      "learning_rate": 7.732331663789592e-06,
+      "loss": 0.9507,
+      "step": 1742
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6033012527497507,
+      "learning_rate": 7.72172361025987e-06,
+      "loss": 0.9891,
+      "step": 1743
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7067904057537774,
+      "learning_rate": 7.71111826016308e-06,
+      "loss": 0.8004,
+      "step": 1744
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5989177394101183,
+      "learning_rate": 7.700515626083669e-06,
+      "loss": 0.7989,
+      "step": 1745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6846845927852715,
+      "learning_rate": 7.689915720602868e-06,
+      "loss": 0.7491,
+      "step": 1746
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5491725111716264,
+      "learning_rate": 7.67931855629866e-06,
+      "loss": 0.6684,
+      "step": 1747
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7244342939256788,
+      "learning_rate": 7.66872414574578e-06,
+      "loss": 0.8743,
+      "step": 1748
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6518813214313978,
+      "learning_rate": 7.658132501515702e-06,
+      "loss": 0.7853,
+      "step": 1749
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7020949188848223,
+      "learning_rate": 7.647543636176606e-06,
+      "loss": 0.9545,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7224547778898189,
+      "learning_rate": 7.636957562293383e-06,
+      "loss": 0.7461,
+      "step": 1751
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6786548662992827,
+      "learning_rate": 7.626374292427604e-06,
+      "loss": 0.8463,
+      "step": 1752
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7416767537425248,
+      "learning_rate": 7.615793839137522e-06,
+      "loss": 0.9304,
+      "step": 1753
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5244848870218479,
+      "learning_rate": 7.60521621497804e-06,
+      "loss": 0.9877,
+      "step": 1754
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6820429367683478,
+      "learning_rate": 7.594641432500707e-06,
+      "loss": 0.8155,
+      "step": 1755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5221056987746426,
+      "learning_rate": 7.584069504253703e-06,
+      "loss": 0.5735,
+      "step": 1756
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7345694016822417,
+      "learning_rate": 7.573500442781815e-06,
+      "loss": 1.0419,
+      "step": 1757
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5971974841469628,
+      "learning_rate": 7.56293426062643e-06,
+      "loss": 0.7871,
+      "step": 1758
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6571344382310238,
+      "learning_rate": 7.552370970325524e-06,
+      "loss": 0.877,
+      "step": 1759
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6594029578512945,
+      "learning_rate": 7.541810584413635e-06,
+      "loss": 0.973,
+      "step": 1760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.735429488675651,
+      "learning_rate": 7.531253115421856e-06,
+      "loss": 0.7033,
+      "step": 1761
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6692503266185699,
+      "learning_rate": 7.520698575877819e-06,
+      "loss": 0.8739,
+      "step": 1762
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6456269146183282,
+      "learning_rate": 7.510146978305683e-06,
+      "loss": 0.8732,
+      "step": 1763
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5781461909143998,
+      "learning_rate": 7.499598335226113e-06,
+      "loss": 0.6494,
+      "step": 1764
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7365092965335938,
+      "learning_rate": 7.489052659156264e-06,
+      "loss": 0.9526,
+      "step": 1765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7006953378732733,
+      "learning_rate": 7.47850996260978e-06,
+      "loss": 0.7306,
+      "step": 1766
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7401274449319911,
+      "learning_rate": 7.467970258096764e-06,
+      "loss": 0.93,
+      "step": 1767
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5102109383826746,
+      "learning_rate": 7.457433558123764e-06,
+      "loss": 0.8923,
+      "step": 1768
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.702221327173387,
+      "learning_rate": 7.446899875193775e-06,
+      "loss": 0.9525,
+      "step": 1769
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6668200128444906,
+      "learning_rate": 7.436369221806201e-06,
+      "loss": 0.9013,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6584035385550988,
+      "learning_rate": 7.425841610456854e-06,
+      "loss": 0.721,
+      "step": 1771
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6714518136229897,
+      "learning_rate": 7.415317053637936e-06,
+      "loss": 0.9708,
+      "step": 1772
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7033814278197627,
+      "learning_rate": 7.404795563838027e-06,
+      "loss": 0.7189,
+      "step": 1773
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5646674209427941,
+      "learning_rate": 7.394277153542069e-06,
+      "loss": 0.5799,
+      "step": 1774
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6312305542565991,
+      "learning_rate": 7.3837618352313426e-06,
+      "loss": 0.9254,
+      "step": 1775
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7547641291777699,
+      "learning_rate": 7.373249621383466e-06,
+      "loss": 0.9676,
+      "step": 1776
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7581804857977283,
+      "learning_rate": 7.3627405244723715e-06,
+      "loss": 0.8712,
+      "step": 1777
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6655251283073568,
+      "learning_rate": 7.352234556968289e-06,
+      "loss": 0.7973,
+      "step": 1778
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5292540227119652,
+      "learning_rate": 7.341731731337745e-06,
+      "loss": 0.9549,
+      "step": 1779
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.690255296323784,
+      "learning_rate": 7.331232060043527e-06,
+      "loss": 0.8119,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7340097191722108,
+      "learning_rate": 7.320735555544685e-06,
+      "loss": 0.9735,
+      "step": 1781
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6541694072437196,
+      "learning_rate": 7.310242230296508e-06,
+      "loss": 0.8638,
+      "step": 1782
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6227449889660871,
+      "learning_rate": 7.29975209675052e-06,
+      "loss": 0.6589,
+      "step": 1783
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6829850391267809,
+      "learning_rate": 7.289265167354449e-06,
+      "loss": 0.893,
+      "step": 1784
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7130371626759777,
+      "learning_rate": 7.278781454552222e-06,
+      "loss": 0.7887,
+      "step": 1785
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6350493879669773,
+      "learning_rate": 7.2683009707839555e-06,
+      "loss": 1.0084,
+      "step": 1786
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6914509230623667,
+      "learning_rate": 7.257823728485931e-06,
+      "loss": 0.783,
+      "step": 1787
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6964375459215134,
+      "learning_rate": 7.247349740090579e-06,
+      "loss": 0.8849,
+      "step": 1788
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6981859450581894,
+      "learning_rate": 7.236879018026472e-06,
+      "loss": 0.76,
+      "step": 1789
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.726301347822432,
+      "learning_rate": 7.226411574718314e-06,
+      "loss": 0.9033,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5575118126296639,
+      "learning_rate": 7.215947422586907e-06,
+      "loss": 0.6589,
+      "step": 1791
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5416511710221537,
+      "learning_rate": 7.205486574049152e-06,
+      "loss": 0.8804,
+      "step": 1792
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.717765902726211,
+      "learning_rate": 7.195029041518032e-06,
+      "loss": 0.885,
+      "step": 1793
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7928615194161007,
+      "learning_rate": 7.1845748374025955e-06,
+      "loss": 0.9305,
+      "step": 1794
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5723239149515668,
+      "learning_rate": 7.174123974107934e-06,
+      "loss": 0.935,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7176608750282097,
+      "learning_rate": 7.163676464035186e-06,
+      "loss": 0.7652,
+      "step": 1796
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6533407998441993,
+      "learning_rate": 7.153232319581502e-06,
+      "loss": 0.8703,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7132211218942227,
+      "learning_rate": 7.142791553140045e-06,
+      "loss": 1.0472,
+      "step": 1798
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.803907255327582,
+      "learning_rate": 7.1323541770999615e-06,
+      "loss": 0.762,
+      "step": 1799
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.48191084273380824,
+      "learning_rate": 7.121920203846387e-06,
+      "loss": 0.5713,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7652628604329168,
+      "learning_rate": 7.111489645760405e-06,
+      "loss": 0.8743,
+      "step": 1801
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.679035170158573,
+      "learning_rate": 7.101062515219059e-06,
+      "loss": 0.9603,
+      "step": 1802
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5379510819446139,
+      "learning_rate": 7.09063882459532e-06,
+      "loss": 0.9405,
+      "step": 1803
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6503392571693488,
+      "learning_rate": 7.0802185862580756e-06,
+      "loss": 0.8077,
+      "step": 1804
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7456117265475023,
+      "learning_rate": 7.069801812572117e-06,
+      "loss": 0.8762,
+      "step": 1805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6826856178546811,
+      "learning_rate": 7.059388515898129e-06,
+      "loss": 0.8736,
+      "step": 1806
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6544765753261702,
+      "learning_rate": 7.0489787085926664e-06,
+      "loss": 0.9306,
+      "step": 1807
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7014183954782715,
+      "learning_rate": 7.038572403008146e-06,
+      "loss": 0.8909,
+      "step": 1808
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5156031501254631,
+      "learning_rate": 7.0281696114928214e-06,
+      "loss": 0.6386,
+      "step": 1809
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6999139562388769,
+      "learning_rate": 7.017770346390791e-06,
+      "loss": 0.9535,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7814202522773929,
+      "learning_rate": 7.007374620041956e-06,
+      "loss": 0.8042,
+      "step": 1811
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6408981295926104,
+      "learning_rate": 6.996982444782021e-06,
+      "loss": 0.8549,
+      "step": 1812
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6566254488686086,
+      "learning_rate": 6.986593832942486e-06,
+      "loss": 0.7591,
+      "step": 1813
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7777589352103744,
+      "learning_rate": 6.976208796850608e-06,
+      "loss": 0.818,
+      "step": 1814
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7290632317708274,
+      "learning_rate": 6.96582734882941e-06,
+      "loss": 0.9267,
+      "step": 1815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5997751277370135,
+      "learning_rate": 6.95544950119766e-06,
+      "loss": 0.9453,
+      "step": 1816
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6021263211980341,
+      "learning_rate": 6.945075266269845e-06,
+      "loss": 1.005,
+      "step": 1817
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.48948058636528724,
+      "learning_rate": 6.934704656356171e-06,
+      "loss": 0.5517,
+      "step": 1818
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6406333141539708,
+      "learning_rate": 6.92433768376254e-06,
+      "loss": 0.8464,
+      "step": 1819
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7577023694386048,
+      "learning_rate": 6.9139743607905405e-06,
+      "loss": 0.9124,
+      "step": 1820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7414753654347307,
+      "learning_rate": 6.903614699737429e-06,
+      "loss": 0.9706,
+      "step": 1821
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5786551878407901,
+      "learning_rate": 6.8932587128961124e-06,
+      "loss": 0.6606,
+      "step": 1822
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.739178729904386,
+      "learning_rate": 6.8829064125551485e-06,
+      "loss": 0.8864,
+      "step": 1823
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6705237809553937,
+      "learning_rate": 6.872557810998709e-06,
+      "loss": 0.8968,
+      "step": 1824
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7902663455634784,
+      "learning_rate": 6.862212920506581e-06,
+      "loss": 0.8189,
+      "step": 1825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.563613182303183,
+      "learning_rate": 6.851871753354154e-06,
+      "loss": 0.6528,
+      "step": 1826
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8032225802947186,
+      "learning_rate": 6.841534321812388e-06,
+      "loss": 0.804,
+      "step": 1827
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5299920777440559,
+      "learning_rate": 6.8312006381478216e-06,
+      "loss": 0.9662,
+      "step": 1828
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7439166741948375,
+      "learning_rate": 6.8208707146225324e-06,
+      "loss": 1.027,
+      "step": 1829
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6415203779067836,
+      "learning_rate": 6.810544563494153e-06,
+      "loss": 0.8332,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6797766926665287,
+      "learning_rate": 6.800222197015829e-06,
+      "loss": 0.9202,
+      "step": 1831
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7683402530301165,
+      "learning_rate": 6.7899036274362115e-06,
+      "loss": 0.7977,
+      "step": 1832
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7780389432260649,
+      "learning_rate": 6.77958886699946e-06,
+      "loss": 0.9044,
+      "step": 1833
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7055308686761134,
+      "learning_rate": 6.769277927945201e-06,
+      "loss": 0.8481,
+      "step": 1834
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6819113716404949,
+      "learning_rate": 6.758970822508532e-06,
+      "loss": 0.7465,
+      "step": 1835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6842643634847965,
+      "learning_rate": 6.748667562920002e-06,
+      "loss": 0.624,
+      "step": 1836
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7419362158503421,
+      "learning_rate": 6.738368161405593e-06,
+      "loss": 0.7705,
+      "step": 1837
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7220117216166924,
+      "learning_rate": 6.728072630186715e-06,
+      "loss": 0.8799,
+      "step": 1838
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6317338225728113,
+      "learning_rate": 6.717780981480175e-06,
+      "loss": 0.7979,
+      "step": 1839
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7143836351477585,
+      "learning_rate": 6.707493227498187e-06,
+      "loss": 0.9105,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7469130681687856,
+      "learning_rate": 6.697209380448333e-06,
+      "loss": 1.0402,
+      "step": 1841
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5736051684163728,
+      "learning_rate": 6.68692945253356e-06,
+      "loss": 0.9254,
+      "step": 1842
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.778114029225764,
+      "learning_rate": 6.6766534559521686e-06,
+      "loss": 0.9418,
+      "step": 1843
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6366002898578709,
+      "learning_rate": 6.666381402897794e-06,
+      "loss": 0.5658,
+      "step": 1844
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7038053217767299,
+      "learning_rate": 6.656113305559388e-06,
+      "loss": 0.9414,
+      "step": 1845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5909416148698952,
+      "learning_rate": 6.645849176121213e-06,
+      "loss": 0.5667,
+      "step": 1846
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7277073393476866,
+      "learning_rate": 6.635589026762819e-06,
+      "loss": 0.8758,
+      "step": 1847
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.733852740228684,
+      "learning_rate": 6.625332869659036e-06,
+      "loss": 0.9987,
+      "step": 1848
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8361375313214474,
+      "learning_rate": 6.615080716979954e-06,
+      "loss": 0.7879,
+      "step": 1849
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7219324493615907,
+      "learning_rate": 6.604832580890917e-06,
+      "loss": 0.9013,
+      "step": 1850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6448874276896279,
+      "learning_rate": 6.5945884735524945e-06,
+      "loss": 0.9218,
+      "step": 1851
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7154280450535363,
+      "learning_rate": 6.584348407120481e-06,
+      "loss": 0.9208,
+      "step": 1852
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5697322433429735,
+      "learning_rate": 6.5741123937458776e-06,
+      "loss": 0.6625,
+      "step": 1853
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6478994281941471,
+      "learning_rate": 6.563880445574873e-06,
+      "loss": 0.8465,
+      "step": 1854
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7122707162733919,
+      "learning_rate": 6.553652574748828e-06,
+      "loss": 0.9419,
+      "step": 1855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5431514699709773,
+      "learning_rate": 6.543428793404274e-06,
+      "loss": 0.9083,
+      "step": 1856
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7369849104462219,
+      "learning_rate": 6.5332091136728846e-06,
+      "loss": 0.8548,
+      "step": 1857
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5861574877747664,
+      "learning_rate": 6.5229935476814664e-06,
+      "loss": 0.6323,
+      "step": 1858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6467854294371839,
+      "learning_rate": 6.512782107551945e-06,
+      "loss": 0.6902,
+      "step": 1859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6808397440938877,
+      "learning_rate": 6.502574805401351e-06,
+      "loss": 0.8667,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7845767479336205,
+      "learning_rate": 6.4923716533418035e-06,
+      "loss": 0.9203,
+      "step": 1861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6732526856200135,
+      "learning_rate": 6.482172663480497e-06,
+      "loss": 0.9114,
+      "step": 1862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5101012664189575,
+      "learning_rate": 6.471977847919691e-06,
+      "loss": 0.5093,
+      "step": 1863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7397957447070114,
+      "learning_rate": 6.461787218756687e-06,
+      "loss": 0.9285,
+      "step": 1864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7806167604968243,
+      "learning_rate": 6.451600788083822e-06,
+      "loss": 1.0008,
+      "step": 1865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6685792184445112,
+      "learning_rate": 6.441418567988444e-06,
+      "loss": 0.8106,
+      "step": 1866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5166700292548878,
+      "learning_rate": 6.431240570552917e-06,
+      "loss": 0.9568,
+      "step": 1867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8039732488589699,
+      "learning_rate": 6.421066807854585e-06,
+      "loss": 0.7941,
+      "step": 1868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6467712385545956,
+      "learning_rate": 6.410897291965766e-06,
+      "loss": 0.8814,
+      "step": 1869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5155224252374998,
+      "learning_rate": 6.400732034953749e-06,
+      "loss": 0.526,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7287313889756828,
+      "learning_rate": 6.390571048880758e-06,
+      "loss": 0.8462,
+      "step": 1871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7342693790176126,
+      "learning_rate": 6.3804143458039535e-06,
+      "loss": 0.8513,
+      "step": 1872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7767987433392833,
+      "learning_rate": 6.370261937775415e-06,
+      "loss": 0.7843,
+      "step": 1873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6277486902301519,
+      "learning_rate": 6.360113836842125e-06,
+      "loss": 1.0842,
+      "step": 1874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6824685066115305,
+      "learning_rate": 6.3499700550459554e-06,
+      "loss": 0.9306,
+      "step": 1875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7226235269484088,
+      "learning_rate": 6.339830604423647e-06,
+      "loss": 0.8153,
+      "step": 1876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6870243672699314,
+      "learning_rate": 6.3296954970068125e-06,
+      "loss": 0.9767,
+      "step": 1877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49890836778340475,
+      "learning_rate": 6.3195647448219e-06,
+      "loss": 0.9214,
+      "step": 1878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5556079965428196,
+      "learning_rate": 6.309438359890195e-06,
+      "loss": 0.5856,
+      "step": 1879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7073759252259366,
+      "learning_rate": 6.2993163542278025e-06,
+      "loss": 0.811,
+      "step": 1880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7461615238865188,
+      "learning_rate": 6.2891987398456265e-06,
+      "loss": 0.9747,
+      "step": 1881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5941628617805149,
+      "learning_rate": 6.2790855287493605e-06,
+      "loss": 0.6692,
+      "step": 1882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7441794493684063,
+      "learning_rate": 6.268976732939478e-06,
+      "loss": 0.9659,
+      "step": 1883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.820610272156693,
+      "learning_rate": 6.2588723644112095e-06,
+      "loss": 0.8297,
+      "step": 1884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7095475192396932,
+      "learning_rate": 6.248772435154531e-06,
+      "loss": 0.8812,
+      "step": 1885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6567277905478895,
+      "learning_rate": 6.238676957154151e-06,
+      "loss": 0.8363,
+      "step": 1886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8321269164762839,
+      "learning_rate": 6.2285859423895005e-06,
+      "loss": 0.8984,
+      "step": 1887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6758988271843699,
+      "learning_rate": 6.21849940283471e-06,
+      "loss": 0.9023,
+      "step": 1888
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8154862191681895,
+      "learning_rate": 6.208417350458599e-06,
+      "loss": 0.5274,
+      "step": 1889
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6162553745747915,
+      "learning_rate": 6.1983397972246686e-06,
+      "loss": 0.9364,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6386831980089631,
+      "learning_rate": 6.188266755091073e-06,
+      "loss": 0.8478,
+      "step": 1891
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.732500127601793,
+      "learning_rate": 6.178198236010619e-06,
+      "loss": 0.7329,
+      "step": 1892
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7133543768150488,
+      "learning_rate": 6.168134251930743e-06,
+      "loss": 0.9876,
+      "step": 1893
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8058168896489069,
+      "learning_rate": 6.1580748147935046e-06,
+      "loss": 0.7225,
+      "step": 1894
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.736058051397082,
+      "learning_rate": 6.148019936535564e-06,
+      "loss": 0.9903,
+      "step": 1895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.617312955090941,
+      "learning_rate": 6.137969629088174e-06,
+      "loss": 0.8418,
+      "step": 1896
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7783149761637823,
+      "learning_rate": 6.127923904377159e-06,
+      "loss": 0.524,
+      "step": 1897
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7803368211129369,
+      "learning_rate": 6.117882774322912e-06,
+      "loss": 1.0169,
+      "step": 1898
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7559542415838236,
+      "learning_rate": 6.107846250840371e-06,
+      "loss": 0.8233,
+      "step": 1899
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7613901456347975,
+      "learning_rate": 6.097814345839005e-06,
+      "loss": 0.9657,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7028245289906917,
+      "learning_rate": 6.087787071222808e-06,
+      "loss": 0.7773,
+      "step": 1901
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8329974878396011,
+      "learning_rate": 6.077764438890277e-06,
+      "loss": 0.7934,
+      "step": 1902
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8741037266044405,
+      "learning_rate": 6.067746460734399e-06,
+      "loss": 0.957,
+      "step": 1903
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6758337184253174,
+      "learning_rate": 6.057733148642637e-06,
+      "loss": 0.8936,
+      "step": 1904
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7145732480295636,
+      "learning_rate": 6.047724514496924e-06,
+      "loss": 0.8825,
+      "step": 1905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.559340430938733,
+      "learning_rate": 6.037720570173637e-06,
+      "loss": 0.7317,
+      "step": 1906
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5261420367428784,
+      "learning_rate": 6.027721327543585e-06,
+      "loss": 0.7228,
+      "step": 1907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8552300336561074,
+      "learning_rate": 6.0177267984720074e-06,
+      "loss": 0.8659,
+      "step": 1908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8062384685605656,
+      "learning_rate": 6.007736994818542e-06,
+      "loss": 0.7065,
+      "step": 1909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6245954320974394,
+      "learning_rate": 5.99775192843722e-06,
+      "loss": 1.014,
+      "step": 1910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7332265371513851,
+      "learning_rate": 5.987771611176457e-06,
+      "loss": 0.8224,
+      "step": 1911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7028542707626835,
+      "learning_rate": 5.977796054879028e-06,
+      "loss": 0.8938,
+      "step": 1912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8535166353587976,
+      "learning_rate": 5.9678252713820616e-06,
+      "loss": 0.945,
+      "step": 1913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6926714198078066,
+      "learning_rate": 5.957859272517017e-06,
+      "loss": 0.9518,
+      "step": 1914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8907693880261743,
+      "learning_rate": 5.947898070109686e-06,
+      "loss": 0.7707,
+      "step": 1915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.48271854670940223,
+      "learning_rate": 5.93794167598016e-06,
+      "loss": 0.7595,
+      "step": 1916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7644158636622143,
+      "learning_rate": 5.9279901019428275e-06,
+      "loss": 1.0613,
+      "step": 1917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7203291587345553,
+      "learning_rate": 5.918043359806363e-06,
+      "loss": 0.8277,
+      "step": 1918
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6186197955931014,
+      "learning_rate": 5.908101461373696e-06,
+      "loss": 0.9661,
+      "step": 1919
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6750760361225336,
+      "learning_rate": 5.898164418442018e-06,
+      "loss": 0.7459,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8203768297472994,
+      "learning_rate": 5.888232242802757e-06,
+      "loss": 0.8318,
+      "step": 1921
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7740181164062684,
+      "learning_rate": 5.878304946241562e-06,
+      "loss": 0.9896,
+      "step": 1922
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49370859396196226,
+      "learning_rate": 5.868382540538296e-06,
+      "loss": 0.6946,
+      "step": 1923
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7058535877004728,
+      "learning_rate": 5.858465037467014e-06,
+      "loss": 0.8976,
+      "step": 1924
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7374983588809172,
+      "learning_rate": 5.848552448795961e-06,
+      "loss": 0.8753,
+      "step": 1925
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6908752037162598,
+      "learning_rate": 5.838644786287544e-06,
+      "loss": 0.9083,
+      "step": 1926
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7437962433380004,
+      "learning_rate": 5.828742061698325e-06,
+      "loss": 0.7315,
+      "step": 1927
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5754990157061716,
+      "learning_rate": 5.818844286779008e-06,
+      "loss": 0.9153,
+      "step": 1928
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6736824048401506,
+      "learning_rate": 5.8089514732744245e-06,
+      "loss": 1.0192,
+      "step": 1929
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7467831647234724,
+      "learning_rate": 5.799063632923523e-06,
+      "loss": 0.7103,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6315405088141953,
+      "learning_rate": 5.789180777459337e-06,
+      "loss": 0.8182,
+      "step": 1931
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8394582862526561,
+      "learning_rate": 5.779302918608998e-06,
+      "loss": 0.7882,
+      "step": 1932
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4793029362903519,
+      "learning_rate": 5.769430068093708e-06,
+      "loss": 0.7297,
+      "step": 1933
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7202795140459363,
+      "learning_rate": 5.7595622376287144e-06,
+      "loss": 0.7322,
+      "step": 1934
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.685769120763145,
+      "learning_rate": 5.749699438923324e-06,
+      "loss": 0.809,
+      "step": 1935
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6497119071528586,
+      "learning_rate": 5.739841683680855e-06,
+      "loss": 0.8469,
+      "step": 1936
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.833151062847387,
+      "learning_rate": 5.7299889835986555e-06,
+      "loss": 0.8055,
+      "step": 1937
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6687224532793165,
+      "learning_rate": 5.720141350368072e-06,
+      "loss": 0.8623,
+      "step": 1938
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7023804578510442,
+      "learning_rate": 5.7102987956744315e-06,
+      "loss": 0.8719,
+      "step": 1939
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5477786243782716,
+      "learning_rate": 5.700461331197041e-06,
+      "loss": 0.986,
+      "step": 1940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5917885349328286,
+      "learning_rate": 5.690628968609172e-06,
+      "loss": 0.5775,
+      "step": 1941
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7074830591891349,
+      "learning_rate": 5.680801719578026e-06,
+      "loss": 0.8476,
+      "step": 1942
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7384912077038218,
+      "learning_rate": 5.670979595764754e-06,
+      "loss": 0.9227,
+      "step": 1943
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6198908899347969,
+      "learning_rate": 5.66116260882442e-06,
+      "loss": 0.7642,
+      "step": 1944
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6961705307897714,
+      "learning_rate": 5.6513507704059835e-06,
+      "loss": 0.881,
+      "step": 1945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6553553597058192,
+      "learning_rate": 5.6415440921523045e-06,
+      "loss": 0.9459,
+      "step": 1946
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.780243105783307,
+      "learning_rate": 5.6317425857001214e-06,
+      "loss": 0.7317,
+      "step": 1947
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6693427552081439,
+      "learning_rate": 5.621946262680026e-06,
+      "loss": 0.8171,
+      "step": 1948
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6305508063438907,
+      "learning_rate": 5.612155134716466e-06,
+      "loss": 0.9075,
+      "step": 1949
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5330698574573791,
+      "learning_rate": 5.6023692134277296e-06,
+      "loss": 0.6898,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6946489423911643,
+      "learning_rate": 5.5925885104259115e-06,
+      "loss": 0.8167,
+      "step": 1951
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7552970824896482,
+      "learning_rate": 5.5828130373169274e-06,
+      "loss": 0.7868,
+      "step": 1952
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.658354268238233,
+      "learning_rate": 5.573042805700488e-06,
+      "loss": 0.873,
+      "step": 1953
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5848924967229643,
+      "learning_rate": 5.5632778271700705e-06,
+      "loss": 0.8803,
+      "step": 1954
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7006280143290678,
+      "learning_rate": 5.553518113312936e-06,
+      "loss": 0.9788,
+      "step": 1955
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6713812395447044,
+      "learning_rate": 5.543763675710082e-06,
+      "loss": 0.8967,
+      "step": 1956
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7368829505643745,
+      "learning_rate": 5.534014525936258e-06,
+      "loss": 0.9929,
+      "step": 1957
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.691005877677479,
+      "learning_rate": 5.524270675559938e-06,
+      "loss": 0.7751,
+      "step": 1958
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6528082423370446,
+      "learning_rate": 5.514532136143296e-06,
+      "loss": 0.6984,
+      "step": 1959
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.46410123340100645,
+      "learning_rate": 5.504798919242216e-06,
+      "loss": 0.5528,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7488982521871377,
+      "learning_rate": 5.495071036406263e-06,
+      "loss": 0.9469,
+      "step": 1961
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7156944808127667,
+      "learning_rate": 5.4853484991786686e-06,
+      "loss": 0.9526,
+      "step": 1962
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7894903189675024,
+      "learning_rate": 5.475631319096325e-06,
+      "loss": 0.904,
+      "step": 1963
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6666653422870564,
+      "learning_rate": 5.465919507689772e-06,
+      "loss": 0.7562,
+      "step": 1964
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6508653900515908,
+      "learning_rate": 5.4562130764831645e-06,
+      "loss": 0.9573,
+      "step": 1965
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6556445557989542,
+      "learning_rate": 5.446512036994287e-06,
+      "loss": 0.7676,
+      "step": 1966
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6952938661268032,
+      "learning_rate": 5.436816400734526e-06,
+      "loss": 0.9461,
+      "step": 1967
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6038143420968791,
+      "learning_rate": 5.427126179208844e-06,
+      "loss": 0.9062,
+      "step": 1968
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5684285085497045,
+      "learning_rate": 5.4174413839157865e-06,
+      "loss": 0.6419,
+      "step": 1969
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7241739208289807,
+      "learning_rate": 5.407762026347469e-06,
+      "loss": 0.8734,
+      "step": 1970
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6851604414338658,
+      "learning_rate": 5.398088117989535e-06,
+      "loss": 0.7705,
+      "step": 1971
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.642845005523124,
+      "learning_rate": 5.388419670321177e-06,
+      "loss": 1.0581,
+      "step": 1972
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7108464215582592,
+      "learning_rate": 5.378756694815106e-06,
+      "loss": 0.7354,
+      "step": 1973
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.667604796610558,
+      "learning_rate": 5.369099202937528e-06,
+      "loss": 0.8854,
+      "step": 1974
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6477364327830916,
+      "learning_rate": 5.359447206148161e-06,
+      "loss": 0.742,
+      "step": 1975
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.793590233795198,
+      "learning_rate": 5.349800715900182e-06,
+      "loss": 1.0257,
+      "step": 1976
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5000296458737314,
+      "learning_rate": 5.34015974364025e-06,
+      "loss": 0.5661,
+      "step": 1977
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5620530930695591,
+      "learning_rate": 5.330524300808473e-06,
+      "loss": 0.86,
+      "step": 1978
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6840110748599355,
+      "learning_rate": 5.320894398838388e-06,
+      "loss": 0.9265,
+      "step": 1979
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6477356983444397,
+      "learning_rate": 5.311270049156967e-06,
+      "loss": 0.9949,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5963546689571246,
+      "learning_rate": 5.301651263184595e-06,
+      "loss": 0.8457,
+      "step": 1981
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.712106047097693,
+      "learning_rate": 5.292038052335045e-06,
+      "loss": 0.8548,
+      "step": 1982
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6194956719616822,
+      "learning_rate": 5.2824304280154815e-06,
+      "loss": 0.8106,
+      "step": 1983
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7018004881801182,
+      "learning_rate": 5.272828401626443e-06,
+      "loss": 1.0364,
+      "step": 1984
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.740983854196439,
+      "learning_rate": 5.263231984561814e-06,
+      "loss": 0.8368,
+      "step": 1985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.45833165637798273,
+      "learning_rate": 5.253641188208836e-06,
+      "loss": 0.5851,
+      "step": 1986
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7082430555673556,
+      "learning_rate": 5.244056023948075e-06,
+      "loss": 0.779,
+      "step": 1987
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7062332980446742,
+      "learning_rate": 5.234476503153409e-06,
+      "loss": 0.9842,
+      "step": 1988
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4845284248772645,
+      "learning_rate": 5.2249026371920265e-06,
+      "loss": 0.9223,
+      "step": 1989
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6703953758737417,
+      "learning_rate": 5.2153344374244086e-06,
+      "loss": 0.8267,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6101581194628595,
+      "learning_rate": 5.2057719152042985e-06,
+      "loss": 0.7591,
+      "step": 1991
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6678990459135342,
+      "learning_rate": 5.196215081878717e-06,
+      "loss": 0.9404,
+      "step": 1992
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6661000921536138,
+      "learning_rate": 5.186663948787932e-06,
+      "loss": 0.9227,
+      "step": 1993
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5199754332091556,
+      "learning_rate": 5.177118527265438e-06,
+      "loss": 0.6107,
+      "step": 1994
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6518024597101755,
+      "learning_rate": 5.167578828637965e-06,
+      "loss": 0.9518,
+      "step": 1995
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7447617466663471,
+      "learning_rate": 5.15804486422544e-06,
+      "loss": 0.8307,
+      "step": 1996
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6946246956265928,
+      "learning_rate": 5.148516645340992e-06,
+      "loss": 0.7909,
+      "step": 1997
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9873510094844609,
+      "learning_rate": 5.138994183290937e-06,
+      "loss": 0.8439,
+      "step": 1998
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.629779265478488,
+      "learning_rate": 5.129477489374749e-06,
+      "loss": 0.7958,
+      "step": 1999
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8019616812619768,
+      "learning_rate": 5.119966574885065e-06,
+      "loss": 1.0014,
+      "step": 2000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.653044346107504,
+      "learning_rate": 5.110461451107664e-06,
+      "loss": 0.7617,
+      "step": 2001
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.49363758581020745,
+      "learning_rate": 5.100962129321448e-06,
+      "loss": 0.927,
+      "step": 2002
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6234428082257368,
+      "learning_rate": 5.09146862079844e-06,
+      "loss": 0.9914,
+      "step": 2003
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5223048144222575,
+      "learning_rate": 5.081980936803766e-06,
+      "loss": 0.4986,
+      "step": 2004
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7147030168165954,
+      "learning_rate": 5.07249908859563e-06,
+      "loss": 0.9234,
+      "step": 2005
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6723492244397234,
+      "learning_rate": 5.063023087425324e-06,
+      "loss": 0.9379,
+      "step": 2006
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9337146686685436,
+      "learning_rate": 5.053552944537199e-06,
+      "loss": 0.9277,
+      "step": 2007
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5956786039128061,
+      "learning_rate": 5.044088671168644e-06,
+      "loss": 0.7089,
+      "step": 2008
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6450952030518344,
+      "learning_rate": 5.034630278550095e-06,
+      "loss": 0.8549,
+      "step": 2009
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7105843458832765,
+      "learning_rate": 5.0251777779050114e-06,
+      "loss": 0.9117,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6343926196337059,
+      "learning_rate": 5.015731180449847e-06,
+      "loss": 0.7546,
+      "step": 2011
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5735339095324778,
+      "learning_rate": 5.006290497394066e-06,
+      "loss": 0.6554,
+      "step": 2012
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6862709494330962,
+      "learning_rate": 4.996855739940101e-06,
+      "loss": 0.7394,
+      "step": 2013
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4869900532390591,
+      "learning_rate": 4.9874269192833645e-06,
+      "loss": 1.0238,
+      "step": 2014
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7326228321568051,
+      "learning_rate": 4.978004046612224e-06,
+      "loss": 0.8394,
+      "step": 2015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6272948446255527,
+      "learning_rate": 4.968587133107975e-06,
+      "loss": 0.8381,
+      "step": 2016
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7278007810656959,
+      "learning_rate": 4.959176189944859e-06,
+      "loss": 0.9378,
+      "step": 2017
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5996800384112759,
+      "learning_rate": 4.9497712282900265e-06,
+      "loss": 0.7245,
+      "step": 2018
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7300303756626207,
+      "learning_rate": 4.9403722593035245e-06,
+      "loss": 0.9821,
+      "step": 2019
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6510684301956882,
+      "learning_rate": 4.930979294138296e-06,
+      "loss": 0.8091,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7139733255049603,
+      "learning_rate": 4.921592343940161e-06,
+      "loss": 0.852,
+      "step": 2021
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5758718083183753,
+      "learning_rate": 4.912211419847795e-06,
+      "loss": 0.6004,
+      "step": 2022
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6805778652732652,
+      "learning_rate": 4.902836532992727e-06,
+      "loss": 0.7244,
+      "step": 2023
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6014869888986808,
+      "learning_rate": 4.8934676944993265e-06,
+      "loss": 0.86,
+      "step": 2024
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6368759542036896,
+      "learning_rate": 4.8841049154847744e-06,
+      "loss": 0.7935,
+      "step": 2025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6663186539764517,
+      "learning_rate": 4.874748207059071e-06,
+      "loss": 0.9046,
+      "step": 2026
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.46347563332290626,
+      "learning_rate": 4.865397580325016e-06,
+      "loss": 1.0149,
+      "step": 2027
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7271736400568178,
+      "learning_rate": 4.8560530463781775e-06,
+      "loss": 0.763,
+      "step": 2028
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5572983272424973,
+      "learning_rate": 4.846714616306908e-06,
+      "loss": 0.6668,
+      "step": 2029
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6660525186723519,
+      "learning_rate": 4.837382301192316e-06,
+      "loss": 0.7517,
+      "step": 2030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6274274688235412,
+      "learning_rate": 4.828056112108243e-06,
+      "loss": 0.9124,
+      "step": 2031
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6257627609197232,
+      "learning_rate": 4.818736060121274e-06,
+      "loss": 0.616,
+      "step": 2032
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6330063679096679,
+      "learning_rate": 4.8094221562906996e-06,
+      "loss": 0.7945,
+      "step": 2033
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7087995674909017,
+      "learning_rate": 4.800114411668527e-06,
+      "loss": 0.9971,
+      "step": 2034
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6828047630138379,
+      "learning_rate": 4.790812837299451e-06,
+      "loss": 0.7718,
+      "step": 2035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6623312064777013,
+      "learning_rate": 4.781517444220836e-06,
+      "loss": 0.8395,
+      "step": 2036
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6711862380232978,
+      "learning_rate": 4.772228243462722e-06,
+      "loss": 1.0344,
+      "step": 2037
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7310879337054969,
+      "learning_rate": 4.762945246047801e-06,
+      "loss": 0.7776,
+      "step": 2038
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.48231550770091364,
+      "learning_rate": 4.7536684629913945e-06,
+      "loss": 0.6475,
+      "step": 2039
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6918798273640775,
+      "learning_rate": 4.744397905301459e-06,
+      "loss": 0.8604,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.526475992975871,
+      "learning_rate": 4.7351335839785655e-06,
+      "loss": 0.9511,
+      "step": 2041
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6582490594742586,
+      "learning_rate": 4.72587551001587e-06,
+      "loss": 0.6998,
+      "step": 2042
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7100416254099833,
+      "learning_rate": 4.716623694399134e-06,
+      "loss": 0.9475,
+      "step": 2043
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6256623386810976,
+      "learning_rate": 4.707378148106684e-06,
+      "loss": 0.6206,
+      "step": 2044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6703618479961849,
+      "learning_rate": 4.698138882109403e-06,
+      "loss": 0.6511,
+      "step": 2045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6716810014716759,
+      "learning_rate": 4.688905907370729e-06,
+      "loss": 0.897,
+      "step": 2046
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6800056439515755,
+      "learning_rate": 4.679679234846636e-06,
+      "loss": 0.8753,
+      "step": 2047
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4855346611511633,
+      "learning_rate": 4.670458875485607e-06,
+      "loss": 0.6467,
+      "step": 2048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7562399130858586,
+      "learning_rate": 4.6612448402286485e-06,
+      "loss": 0.8022,
+      "step": 2049
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6544470261913151,
+      "learning_rate": 4.652037140009259e-06,
+      "loss": 1.014,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6791973020129337,
+      "learning_rate": 4.642835785753412e-06,
+      "loss": 0.8976,
+      "step": 2051
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5081823508216157,
+      "learning_rate": 4.63364078837956e-06,
+      "loss": 0.9511,
+      "step": 2052
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6316076268300626,
+      "learning_rate": 4.624452158798604e-06,
+      "loss": 0.8896,
+      "step": 2053
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.658162899619152,
+      "learning_rate": 4.615269907913896e-06,
+      "loss": 0.7839,
+      "step": 2054
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6820770616009004,
+      "learning_rate": 4.606094046621222e-06,
+      "loss": 0.8855,
+      "step": 2055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4836889352525524,
+      "learning_rate": 4.596924585808772e-06,
+      "loss": 0.5188,
+      "step": 2056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7148590050920972,
+      "learning_rate": 4.587761536357153e-06,
+      "loss": 0.823,
+      "step": 2057
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7525853834043489,
+      "learning_rate": 4.5786049091393645e-06,
+      "loss": 0.9136,
+      "step": 2058
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.707165159899719,
+      "learning_rate": 4.5694547150207745e-06,
+      "loss": 0.8098,
+      "step": 2059
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7122027557713732,
+      "learning_rate": 4.560310964859129e-06,
+      "loss": 1.1463,
+      "step": 2060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6955379227597899,
+      "learning_rate": 4.551173669504525e-06,
+      "loss": 0.9017,
+      "step": 2061
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6328817805740354,
+      "learning_rate": 4.542042839799392e-06,
+      "loss": 0.9542,
+      "step": 2062
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7360331325665587,
+      "learning_rate": 4.532918486578497e-06,
+      "loss": 0.7739,
+      "step": 2063
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5684283281911148,
+      "learning_rate": 4.523800620668921e-06,
+      "loss": 0.8931,
+      "step": 2064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5270517275938974,
+      "learning_rate": 4.514689252890039e-06,
+      "loss": 0.576,
+      "step": 2065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6802336609437861,
+      "learning_rate": 4.505584394053523e-06,
+      "loss": 0.8627,
+      "step": 2066
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6210446532228715,
+      "learning_rate": 4.496486054963323e-06,
+      "loss": 0.7818,
+      "step": 2067
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.655113538252824,
+      "learning_rate": 4.487394246415641e-06,
+      "loss": 0.8072,
+      "step": 2068
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7452060221591561,
+      "learning_rate": 4.47830897919894e-06,
+      "loss": 0.9891,
+      "step": 2069
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6475896165176398,
+      "learning_rate": 4.469230264093924e-06,
+      "loss": 0.7906,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6531260555337133,
+      "learning_rate": 4.460158111873511e-06,
+      "loss": 0.865,
+      "step": 2071
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6744042382423473,
+      "learning_rate": 4.45109253330284e-06,
+      "loss": 0.8588,
+      "step": 2072
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6171899507606846,
+      "learning_rate": 4.442033539139241e-06,
+      "loss": 0.8482,
+      "step": 2073
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6052381126136799,
+      "learning_rate": 4.43298114013224e-06,
+      "loss": 0.9135,
+      "step": 2074
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5119848229479084,
+      "learning_rate": 4.423935347023538e-06,
+      "loss": 0.5184,
+      "step": 2075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4971017777848442,
+      "learning_rate": 4.414896170546985e-06,
+      "loss": 0.9294,
+      "step": 2076
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6176372031560139,
+      "learning_rate": 4.405863621428591e-06,
+      "loss": 0.8086,
+      "step": 2077
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7233086247699272,
+      "learning_rate": 4.396837710386503e-06,
+      "loss": 0.769,
+      "step": 2078
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6185436529343598,
+      "learning_rate": 4.387818448130977e-06,
+      "loss": 0.9589,
+      "step": 2079
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7136783732093929,
+      "learning_rate": 4.378805845364395e-06,
+      "loss": 0.7482,
+      "step": 2080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6687231015350262,
+      "learning_rate": 4.369799912781233e-06,
+      "loss": 0.8475,
+      "step": 2081
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7084756715036554,
+      "learning_rate": 4.360800661068043e-06,
+      "loss": 0.9631,
+      "step": 2082
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5843399695848432,
+      "learning_rate": 4.35180810090346e-06,
+      "loss": 0.5323,
+      "step": 2083
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5982175947125257,
+      "learning_rate": 4.34282224295818e-06,
+      "loss": 0.9772,
+      "step": 2084
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6483897652492514,
+      "learning_rate": 4.333843097894932e-06,
+      "loss": 0.8416,
+      "step": 2085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7049017140129,
+      "learning_rate": 4.324870676368491e-06,
+      "loss": 1.0235,
+      "step": 2086
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6854641279781812,
+      "learning_rate": 4.315904989025657e-06,
+      "loss": 0.7763,
+      "step": 2087
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7243251065115472,
+      "learning_rate": 4.306946046505226e-06,
+      "loss": 0.7634,
+      "step": 2088
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6924150507104198,
+      "learning_rate": 4.297993859438003e-06,
+      "loss": 0.9463,
+      "step": 2089
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48286004180473213,
+      "learning_rate": 4.289048438446766e-06,
+      "loss": 0.8876,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5756625786395444,
+      "learning_rate": 4.2801097941462755e-06,
+      "loss": 0.721,
+      "step": 2091
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6253148471452301,
+      "learning_rate": 4.271177937143245e-06,
+      "loss": 0.836,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6316329833701521,
+      "learning_rate": 4.2622528780363295e-06,
+      "loss": 0.7289,
+      "step": 2093
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6081056317732925,
+      "learning_rate": 4.253334627416124e-06,
+      "loss": 0.7589,
+      "step": 2094
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6818812176727539,
+      "learning_rate": 4.244423195865149e-06,
+      "loss": 0.6996,
+      "step": 2095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6867089162199181,
+      "learning_rate": 4.2355185939578184e-06,
+      "loss": 1.0411,
+      "step": 2096
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7242858113576849,
+      "learning_rate": 4.226620832260454e-06,
+      "loss": 0.7963,
+      "step": 2097
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6665186470997914,
+      "learning_rate": 4.217729921331263e-06,
+      "loss": 0.9227,
+      "step": 2098
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6826009982779955,
+      "learning_rate": 4.208845871720309e-06,
+      "loss": 0.9363,
+      "step": 2099
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6854052447047603,
+      "learning_rate": 4.199968693969526e-06,
+      "loss": 0.8954,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5212829080644962,
+      "learning_rate": 4.191098398612696e-06,
+      "loss": 0.5296,
+      "step": 2101
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5010290515853325,
+      "learning_rate": 4.182234996175421e-06,
+      "loss": 0.878,
+      "step": 2102
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.695115778237856,
+      "learning_rate": 4.173378497175137e-06,
+      "loss": 1.0642,
+      "step": 2103
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.684438796216104,
+      "learning_rate": 4.164528912121084e-06,
+      "loss": 0.6874,
+      "step": 2104
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7000724746335677,
+      "learning_rate": 4.1556862515142905e-06,
+      "loss": 0.9866,
+      "step": 2105
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6438350386984236,
+      "learning_rate": 4.1468505258475785e-06,
+      "loss": 0.7243,
+      "step": 2106
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6732539655535382,
+      "learning_rate": 4.13802174560554e-06,
+      "loss": 0.7994,
+      "step": 2107
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7756622938119393,
+      "learning_rate": 4.129199921264518e-06,
+      "loss": 0.9683,
+      "step": 2108
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4665229545643634,
+      "learning_rate": 4.120385063292609e-06,
+      "loss": 0.6674,
+      "step": 2109
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6709382286404953,
+      "learning_rate": 4.111577182149635e-06,
+      "loss": 0.8965,
+      "step": 2110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8522334736304162,
+      "learning_rate": 4.1027762882871496e-06,
+      "loss": 0.875,
+      "step": 2111
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7066278136465135,
+      "learning_rate": 4.093982392148411e-06,
+      "loss": 0.911,
+      "step": 2112
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6379177785612096,
+      "learning_rate": 4.085195504168367e-06,
+      "loss": 0.7538,
+      "step": 2113
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4623075292950711,
+      "learning_rate": 4.07641563477366e-06,
+      "loss": 0.9652,
+      "step": 2114
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.639070921541774,
+      "learning_rate": 4.0676427943826e-06,
+      "loss": 0.9069,
+      "step": 2115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7115630246249623,
+      "learning_rate": 4.0588769934051495e-06,
+      "loss": 0.6797,
+      "step": 2116
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6018900442354636,
+      "learning_rate": 4.050118242242929e-06,
+      "loss": 0.8455,
+      "step": 2117
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7213572690063461,
+      "learning_rate": 4.041366551289191e-06,
+      "loss": 0.8108,
+      "step": 2118
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5128568889077801,
+      "learning_rate": 4.0326219309288e-06,
+      "loss": 0.7676,
+      "step": 2119
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6886126815733765,
+      "learning_rate": 4.023884391538244e-06,
+      "loss": 0.7418,
+      "step": 2120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6567287041255676,
+      "learning_rate": 4.015153943485606e-06,
+      "loss": 0.7759,
+      "step": 2121
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6121590993169667,
+      "learning_rate": 4.0064305971305416e-06,
+      "loss": 0.8362,
+      "step": 2122
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7325777734200493,
+      "learning_rate": 3.997714362824295e-06,
+      "loss": 0.8155,
+      "step": 2123
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.648816732096085,
+      "learning_rate": 3.989005250909668e-06,
+      "loss": 0.9117,
+      "step": 2124
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7217629576191555,
+      "learning_rate": 3.9803032717210024e-06,
+      "loss": 0.7784,
+      "step": 2125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4926781101285906,
+      "learning_rate": 3.971608435584183e-06,
+      "loss": 0.9723,
+      "step": 2126
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.48880897031729353,
+      "learning_rate": 3.9629207528166224e-06,
+      "loss": 0.5767,
+      "step": 2127
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.71389639413657,
+      "learning_rate": 3.954240233727234e-06,
+      "loss": 0.8071,
+      "step": 2128
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6502999780011932,
+      "learning_rate": 3.945566888616442e-06,
+      "loss": 0.8565,
+      "step": 2129
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6451114662351605,
+      "learning_rate": 3.936900727776146e-06,
+      "loss": 0.839,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6102558177605083,
+      "learning_rate": 3.928241761489732e-06,
+      "loss": 0.8785,
+      "step": 2131
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6368872525564815,
+      "learning_rate": 3.919590000032046e-06,
+      "loss": 0.9419,
+      "step": 2132
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6809484248577519,
+      "learning_rate": 3.910945453669375e-06,
+      "loss": 0.6921,
+      "step": 2133
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6748752459977817,
+      "learning_rate": 3.902308132659457e-06,
+      "loss": 0.8426,
+      "step": 2134
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6573201648696158,
+      "learning_rate": 3.8936780472514555e-06,
+      "loss": 0.863,
+      "step": 2135
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5411463555701974,
+      "learning_rate": 3.885055207685937e-06,
+      "loss": 0.698,
+      "step": 2136
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6604401221677799,
+      "learning_rate": 3.876439624194879e-06,
+      "loss": 0.8318,
+      "step": 2137
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7068035567195362,
+      "learning_rate": 3.867831307001652e-06,
+      "loss": 0.7751,
+      "step": 2138
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5094975210860478,
+      "learning_rate": 3.859230266320992e-06,
+      "loss": 0.9586,
+      "step": 2139
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.696986555172945,
+      "learning_rate": 3.85063651235901e-06,
+      "loss": 0.667,
+      "step": 2140
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7322255287513454,
+      "learning_rate": 3.842050055313174e-06,
+      "loss": 0.9578,
+      "step": 2141
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6677947500564159,
+      "learning_rate": 3.833470905372278e-06,
+      "loss": 0.8978,
+      "step": 2142
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6600489018588734,
+      "learning_rate": 3.8248990727164616e-06,
+      "loss": 0.9568,
+      "step": 2143
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6796514088746362,
+      "learning_rate": 3.8163345675171745e-06,
+      "loss": 0.7626,
+      "step": 2144
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7574423123520847,
+      "learning_rate": 3.807777399937168e-06,
+      "loss": 0.7807,
+      "step": 2145
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4906309876763768,
+      "learning_rate": 3.7992275801304935e-06,
+      "loss": 0.5593,
+      "step": 2146
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6481078990454756,
+      "learning_rate": 3.790685118242482e-06,
+      "loss": 0.9823,
+      "step": 2147
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7442055344713733,
+      "learning_rate": 3.7821500244097276e-06,
+      "loss": 0.908,
+      "step": 2148
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6859620567346446,
+      "learning_rate": 3.773622308760091e-06,
+      "loss": 0.9002,
+      "step": 2149
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7082586246231567,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 0.8045,
+      "step": 2150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.549667530958227,
+      "learning_rate": 3.7565890524777883e-06,
+      "loss": 0.9667,
+      "step": 2151
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6173869382054518,
+      "learning_rate": 3.7480835320570173e-06,
+      "loss": 0.8056,
+      "step": 2152
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.747266699837184,
+      "learning_rate": 3.739585430243109e-06,
+      "loss": 0.8788,
+      "step": 2153
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49258938137660363,
+      "learning_rate": 3.731094757120025e-06,
+      "loss": 0.6737,
+      "step": 2154
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7437051890126254,
+      "learning_rate": 3.722611522762917e-06,
+      "loss": 0.8773,
+      "step": 2155
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7093134775163158,
+      "learning_rate": 3.7141357372380926e-06,
+      "loss": 0.8602,
+      "step": 2156
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6595628830953958,
+      "learning_rate": 3.7056674106030376e-06,
+      "loss": 0.7299,
+      "step": 2157
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6468173406051053,
+      "learning_rate": 3.6972065529063837e-06,
+      "loss": 1.046,
+      "step": 2158
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7241272311263748,
+      "learning_rate": 3.688753174187888e-06,
+      "loss": 0.7334,
+      "step": 2159
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.699506937664231,
+      "learning_rate": 3.6803072844784473e-06,
+      "loss": 0.8894,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.73507016031194,
+      "learning_rate": 3.6718688938000667e-06,
+      "loss": 0.7123,
+      "step": 2161
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7260891167523732,
+      "learning_rate": 3.6634380121658484e-06,
+      "loss": 0.9817,
+      "step": 2162
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.48814200465662283,
+      "learning_rate": 3.65501464957999e-06,
+      "loss": 0.5404,
+      "step": 2163
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.45152267328783935,
+      "learning_rate": 3.6465988160377695e-06,
+      "loss": 0.8797,
+      "step": 2164
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6459777709152673,
+      "learning_rate": 3.6381905215255187e-06,
+      "loss": 0.9098,
+      "step": 2165
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6351937664313682,
+      "learning_rate": 3.6297897760206347e-06,
+      "loss": 0.9483,
+      "step": 2166
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.730293505933139,
+      "learning_rate": 3.621396589491557e-06,
+      "loss": 0.8657,
+      "step": 2167
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6898914342997126,
+      "learning_rate": 3.6130109718977457e-06,
+      "loss": 0.8788,
+      "step": 2168
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.620344817907706,
+      "learning_rate": 3.604632933189691e-06,
+      "loss": 0.6979,
+      "step": 2169
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7111876732458482,
+      "learning_rate": 3.5962624833088788e-06,
+      "loss": 1.0629,
+      "step": 2170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.825342605273937,
+      "learning_rate": 3.587899632187799e-06,
+      "loss": 0.7464,
+      "step": 2171
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4316846376685312,
+      "learning_rate": 3.579544389749925e-06,
+      "loss": 0.5815,
+      "step": 2172
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6786019766099279,
+      "learning_rate": 3.571196765909691e-06,
+      "loss": 0.7577,
+      "step": 2173
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6633687460292526,
+      "learning_rate": 3.5628567705725014e-06,
+      "loss": 1.0085,
+      "step": 2174
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4845795895552016,
+      "learning_rate": 3.554524413634709e-06,
+      "loss": 0.9092,
+      "step": 2175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7057881390359655,
+      "learning_rate": 3.546199704983592e-06,
+      "loss": 0.7575,
+      "step": 2176
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3115435233045634,
+      "learning_rate": 3.537882654497362e-06,
+      "loss": 0.8518,
+      "step": 2177
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6595145262927179,
+      "learning_rate": 3.529573272045146e-06,
+      "loss": 0.8939,
+      "step": 2178
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6818628511847635,
+      "learning_rate": 3.52127156748696e-06,
+      "loss": 0.9293,
+      "step": 2179
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4748843498023325,
+      "learning_rate": 3.512977550673721e-06,
+      "loss": 0.604,
+      "step": 2180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6457239256430855,
+      "learning_rate": 3.5046912314472204e-06,
+      "loss": 0.9724,
+      "step": 2181
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.746376561859907,
+      "learning_rate": 3.4964126196401094e-06,
+      "loss": 0.7702,
+      "step": 2182
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7116567767641869,
+      "learning_rate": 3.4881417250759007e-06,
+      "loss": 0.7709,
+      "step": 2183
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6855893072848194,
+      "learning_rate": 3.479878557568952e-06,
+      "loss": 0.9105,
+      "step": 2184
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6650212241246617,
+      "learning_rate": 3.4716231269244384e-06,
+      "loss": 0.827,
+      "step": 2185
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1964713973194412,
+      "learning_rate": 3.4633754429383726e-06,
+      "loss": 0.9281,
+      "step": 2186
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7032457512011129,
+      "learning_rate": 3.455135515397556e-06,
+      "loss": 0.7866,
+      "step": 2187
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.47826055749492785,
+      "learning_rate": 3.4469033540796017e-06,
+      "loss": 0.9734,
+      "step": 2188
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.44060528944927774,
+      "learning_rate": 3.438678968752903e-06,
+      "loss": 0.659,
+      "step": 2189
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6355871092119993,
+      "learning_rate": 3.4304623691766193e-06,
+      "loss": 0.7194,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7184090574147229,
+      "learning_rate": 3.422253565100678e-06,
+      "loss": 0.9655,
+      "step": 2191
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7265012216525767,
+      "learning_rate": 3.4140525662657597e-06,
+      "loss": 0.9198,
+      "step": 2192
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7634622103164013,
+      "learning_rate": 3.405859382403269e-06,
+      "loss": 0.8733,
+      "step": 2193
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6337616799758639,
+      "learning_rate": 3.397674023235353e-06,
+      "loss": 0.7369,
+      "step": 2194
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.647398951782728,
+      "learning_rate": 3.3894964984748692e-06,
+      "loss": 0.8111,
+      "step": 2195
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7663698492557278,
+      "learning_rate": 3.3813268178253687e-06,
+      "loss": 0.93,
+      "step": 2196
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6669264671514514,
+      "learning_rate": 3.373164990981108e-06,
+      "loss": 0.8326,
+      "step": 2197
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4879843977485444,
+      "learning_rate": 3.365011027627021e-06,
+      "loss": 0.6671,
+      "step": 2198
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.800000764481683,
+      "learning_rate": 3.356864937438703e-06,
+      "loss": 0.7396,
+      "step": 2199
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4681725761030607,
+      "learning_rate": 3.3487267300824155e-06,
+      "loss": 0.9982,
+      "step": 2200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6751505913158792,
+      "learning_rate": 3.340596415215065e-06,
+      "loss": 0.8186,
+      "step": 2201
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5984760098984053,
+      "learning_rate": 3.3324740024841863e-06,
+      "loss": 0.7917,
+      "step": 2202
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7212863882177397,
+      "learning_rate": 3.324359501527944e-06,
+      "loss": 0.8981,
+      "step": 2203
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6390226236306796,
+      "learning_rate": 3.316252921975116e-06,
+      "loss": 0.786,
+      "step": 2204
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8289559093913698,
+      "learning_rate": 3.30815427344507e-06,
+      "loss": 0.9368,
+      "step": 2205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6806867523411009,
+      "learning_rate": 3.3000635655477773e-06,
+      "loss": 0.7671,
+      "step": 2206
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7642217699523067,
+      "learning_rate": 3.2919808078837712e-06,
+      "loss": 0.8741,
+      "step": 2207
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4904443579867889,
+      "learning_rate": 3.283906010044162e-06,
+      "loss": 0.5856,
+      "step": 2208
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6751480382154583,
+      "learning_rate": 3.275839181610616e-06,
+      "loss": 0.6937,
+      "step": 2209
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7012710795783327,
+      "learning_rate": 3.2677803321553324e-06,
+      "loss": 0.9398,
+      "step": 2210
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6197041539727123,
+      "learning_rate": 3.259729471241051e-06,
+      "loss": 0.8116,
+      "step": 2211
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6222006562382175,
+      "learning_rate": 3.251686608421033e-06,
+      "loss": 0.8809,
+      "step": 2212
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.43501261131824215,
+      "learning_rate": 3.24365175323904e-06,
+      "loss": 1.0226,
+      "step": 2213
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6683841851708803,
+      "learning_rate": 3.2356249152293416e-06,
+      "loss": 0.6998,
+      "step": 2214
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5458624893895695,
+      "learning_rate": 3.2276061039166917e-06,
+      "loss": 0.6387,
+      "step": 2215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6846314393029048,
+      "learning_rate": 3.219595328816313e-06,
+      "loss": 0.781,
+      "step": 2216
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7030793239340819,
+      "learning_rate": 3.211592599433898e-06,
+      "loss": 0.901,
+      "step": 2217
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6011025083941539,
+      "learning_rate": 3.203597925265598e-06,
+      "loss": 0.6647,
+      "step": 2218
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7180259369316908,
+      "learning_rate": 3.1956113157979906e-06,
+      "loss": 0.8098,
+      "step": 2219
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6965785349123059,
+      "learning_rate": 3.1876327805080966e-06,
+      "loss": 1.0048,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6615156393600395,
+      "learning_rate": 3.1796623288633544e-06,
+      "loss": 0.8097,
+      "step": 2221
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6412857817599577,
+      "learning_rate": 3.171699970321601e-06,
+      "loss": 0.8264,
+      "step": 2222
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6719834978382391,
+      "learning_rate": 3.163745714331079e-06,
+      "loss": 1.0237,
+      "step": 2223
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6754913141815491,
+      "learning_rate": 3.155799570330419e-06,
+      "loss": 0.9267,
+      "step": 2224
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5188903859227881,
+      "learning_rate": 3.147861547748612e-06,
+      "loss": 0.5477,
+      "step": 2225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6217561349132759,
+      "learning_rate": 3.1399316560050287e-06,
+      "loss": 0.7996,
+      "step": 2226
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5026233410247226,
+      "learning_rate": 3.1320099045093766e-06,
+      "loss": 0.9809,
+      "step": 2227
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6383363551156956,
+      "learning_rate": 3.1240963026617142e-06,
+      "loss": 0.7461,
+      "step": 2228
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6992005884620875,
+      "learning_rate": 3.1161908598524293e-06,
+      "loss": 0.8848,
+      "step": 2229
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.612524945486159,
+      "learning_rate": 3.1082935854622186e-06,
+      "loss": 0.6485,
+      "step": 2230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6990137636677829,
+      "learning_rate": 3.1004044888620963e-06,
+      "loss": 0.6871,
+      "step": 2231
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6846668181193133,
+      "learning_rate": 3.0925235794133725e-06,
+      "loss": 0.8927,
+      "step": 2232
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6460005196786736,
+      "learning_rate": 3.08465086646763e-06,
+      "loss": 0.8557,
+      "step": 2233
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5773220967659701,
+      "learning_rate": 3.0767863593667404e-06,
+      "loss": 0.6508,
+      "step": 2234
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7102600306768844,
+      "learning_rate": 3.0689300674428325e-06,
+      "loss": 0.796,
+      "step": 2235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6342956903858432,
+      "learning_rate": 3.0610820000182817e-06,
+      "loss": 1.014,
+      "step": 2236
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6310723377848577,
+      "learning_rate": 3.053242166405711e-06,
+      "loss": 0.9121,
+      "step": 2237
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.44585632400321135,
+      "learning_rate": 3.0454105759079745e-06,
+      "loss": 0.9307,
+      "step": 2238
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6187725081371377,
+      "learning_rate": 3.0375872378181336e-06,
+      "loss": 0.8616,
+      "step": 2239
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7579313604337706,
+      "learning_rate": 3.0297721614194697e-06,
+      "loss": 0.7385,
+      "step": 2240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6297661342739483,
+      "learning_rate": 3.021965355985459e-06,
+      "loss": 0.9036,
+      "step": 2241
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5501186734989928,
+      "learning_rate": 3.014166830779753e-06,
+      "loss": 0.518,
+      "step": 2242
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6808262460651279,
+      "learning_rate": 3.0063765950561884e-06,
+      "loss": 0.9035,
+      "step": 2243
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6803505377341778,
+      "learning_rate": 2.9985946580587653e-06,
+      "loss": 0.7757,
+      "step": 2244
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7654929355265322,
+      "learning_rate": 2.9908210290216266e-06,
+      "loss": 0.7831,
+      "step": 2245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6528190666117311,
+      "learning_rate": 2.98305571716907e-06,
+      "loss": 1.1268,
+      "step": 2246
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5878374994538296,
+      "learning_rate": 2.9752987317155115e-06,
+      "loss": 0.9039,
+      "step": 2247
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6202457903161799,
+      "learning_rate": 2.9675500818654956e-06,
+      "loss": 0.9235,
+      "step": 2248
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6370717420393878,
+      "learning_rate": 2.959809776813676e-06,
+      "loss": 0.6837,
+      "step": 2249
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.44142247220885505,
+      "learning_rate": 2.9520778257447935e-06,
+      "loss": 0.898,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.48648231892951466,
+      "learning_rate": 2.944354237833689e-06,
+      "loss": 0.5927,
+      "step": 2251
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6571394040950647,
+      "learning_rate": 2.9366390222452756e-06,
+      "loss": 0.7951,
+      "step": 2252
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6269322370285243,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.8131,
+      "step": 2253
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6324308196369749,
+      "learning_rate": 2.921233744646472e-06,
+      "loss": 0.7785,
+      "step": 2254
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7486330481506791,
+      "learning_rate": 2.913543700916194e-06,
+      "loss": 0.9402,
+      "step": 2255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.694604073663237,
+      "learning_rate": 2.905862066068793e-06,
+      "loss": 0.7893,
+      "step": 2256
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6226625135485533,
+      "learning_rate": 2.8981888492194e-06,
+      "loss": 0.8616,
+      "step": 2257
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.694043947518676,
+      "learning_rate": 2.890524059473161e-06,
+      "loss": 0.9104,
+      "step": 2258
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6224187710448047,
+      "learning_rate": 2.8828677059252107e-06,
+      "loss": 0.9327,
+      "step": 2259
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.44583332342896886,
+      "learning_rate": 2.875219797660681e-06,
+      "loss": 0.574,
+      "step": 2260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6695013369717018,
+      "learning_rate": 2.8675803437546854e-06,
+      "loss": 0.7713,
+      "step": 2261
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.46980359828962437,
+      "learning_rate": 2.8599493532722954e-06,
+      "loss": 0.9249,
+      "step": 2262
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6655748246102849,
+      "learning_rate": 2.8523268352685495e-06,
+      "loss": 0.8649,
+      "step": 2263
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8085166543184366,
+      "learning_rate": 2.8447127987884247e-06,
+      "loss": 0.8195,
+      "step": 2264
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6041253557479749,
+      "learning_rate": 2.83710725286684e-06,
+      "loss": 0.9241,
+      "step": 2265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7130735169443682,
+      "learning_rate": 2.8295102065286385e-06,
+      "loss": 0.7007,
+      "step": 2266
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.660565884300248,
+      "learning_rate": 2.821921668788571e-06,
+      "loss": 0.9672,
+      "step": 2267
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5138063402857037,
+      "learning_rate": 2.8143416486513e-06,
+      "loss": 0.5345,
+      "step": 2268
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.659316296906705,
+      "learning_rate": 2.80677015511138e-06,
+      "loss": 0.7344,
+      "step": 2269
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.567937831507043,
+      "learning_rate": 2.799207197153241e-06,
+      "loss": 1.006,
+      "step": 2270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6405433148808177,
+      "learning_rate": 2.791652783751191e-06,
+      "loss": 0.9154,
+      "step": 2271
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7136747015590078,
+      "learning_rate": 2.7841069238693997e-06,
+      "loss": 0.9021,
+      "step": 2272
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.684889879908983,
+      "learning_rate": 2.776569626461879e-06,
+      "loss": 0.7727,
+      "step": 2273
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7180704685438809,
+      "learning_rate": 2.7690409004724883e-06,
+      "loss": 0.93,
+      "step": 2274
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6688826781996293,
+      "learning_rate": 2.7615207548349164e-06,
+      "loss": 0.8017,
+      "step": 2275
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4965900157830675,
+      "learning_rate": 2.7540091984726625e-06,
+      "loss": 0.8803,
+      "step": 2276
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6756762100763423,
+      "learning_rate": 2.7465062402990417e-06,
+      "loss": 0.8292,
+      "step": 2277
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6832053174331756,
+      "learning_rate": 2.739011889217166e-06,
+      "loss": 0.9006,
+      "step": 2278
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5053634093759354,
+      "learning_rate": 2.7315261541199266e-06,
+      "loss": 0.7041,
+      "step": 2279
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6246195049130531,
+      "learning_rate": 2.724049043889997e-06,
+      "loss": 0.791,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7238314152730244,
+      "learning_rate": 2.71658056739982e-06,
+      "loss": 0.7548,
+      "step": 2281
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6326707305237464,
+      "learning_rate": 2.7091207335115822e-06,
+      "loss": 0.983,
+      "step": 2282
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7631849138120923,
+      "learning_rate": 2.701669551077226e-06,
+      "loss": 0.8086,
+      "step": 2283
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6651874843291395,
+      "learning_rate": 2.69422702893842e-06,
+      "loss": 0.9527,
+      "step": 2284
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9457605793419165,
+      "learning_rate": 2.6867931759265598e-06,
+      "loss": 0.874,
+      "step": 2285
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6661092656269938,
+      "learning_rate": 2.6793680008627578e-06,
+      "loss": 0.8651,
+      "step": 2286
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5399752304511284,
+      "learning_rate": 2.6719515125578198e-06,
+      "loss": 0.5013,
+      "step": 2287
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4650887185753994,
+      "learning_rate": 2.6645437198122503e-06,
+      "loss": 0.9044,
+      "step": 2288
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6813372345051815,
+      "learning_rate": 2.6571446314162385e-06,
+      "loss": 1.0441,
+      "step": 2289
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7658722128360373,
+      "learning_rate": 2.649754256149635e-06,
+      "loss": 0.8458,
+      "step": 2290
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6253141620038408,
+      "learning_rate": 2.6423726027819595e-06,
+      "loss": 0.9011,
+      "step": 2291
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6490374331143542,
+      "learning_rate": 2.6349996800723822e-06,
+      "loss": 0.7301,
+      "step": 2292
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6287020160857213,
+      "learning_rate": 2.627635496769705e-06,
+      "loss": 0.9146,
+      "step": 2293
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.631413281068037,
+      "learning_rate": 2.620280061612368e-06,
+      "loss": 0.8968,
+      "step": 2294
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5043351240494259,
+      "learning_rate": 2.612933383328432e-06,
+      "loss": 0.514,
+      "step": 2295
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6480978976688068,
+      "learning_rate": 2.605595470635556e-06,
+      "loss": 0.9268,
+      "step": 2296
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6761775869280762,
+      "learning_rate": 2.598266332241007e-06,
+      "loss": 0.8157,
+      "step": 2297
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.741370891261188,
+      "learning_rate": 2.5909459768416424e-06,
+      "loss": 0.9668,
+      "step": 2298
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6896871316021411,
+      "learning_rate": 2.5836344131238844e-06,
+      "loss": 0.7695,
+      "step": 2299
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.47461527209796806,
+      "learning_rate": 2.576331649763736e-06,
+      "loss": 0.9521,
+      "step": 2300
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7000812361038975,
+      "learning_rate": 2.569037695426756e-06,
+      "loss": 0.8212,
+      "step": 2301
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6829845439039016,
+      "learning_rate": 2.56175255876804e-06,
+      "loss": 0.7279,
+      "step": 2302
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6421268098444571,
+      "learning_rate": 2.554476248432234e-06,
+      "loss": 0.8645,
+      "step": 2303
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7354103026471493,
+      "learning_rate": 2.547208773053499e-06,
+      "loss": 0.7226,
+      "step": 2304
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4559044175170171,
+      "learning_rate": 2.5399501412555207e-06,
+      "loss": 0.7223,
+      "step": 2305
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6889391624566933,
+      "learning_rate": 2.532700361651489e-06,
+      "loss": 0.751,
+      "step": 2306
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.684039239305929,
+      "learning_rate": 2.5254594428440826e-06,
+      "loss": 0.7784,
+      "step": 2307
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6426796367936908,
+      "learning_rate": 2.5182273934254765e-06,
+      "loss": 0.8654,
+      "step": 2308
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.700397573967982,
+      "learning_rate": 2.5110042219773178e-06,
+      "loss": 0.7331,
+      "step": 2309
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7272492571384055,
+      "learning_rate": 2.503789937070712e-06,
+      "loss": 0.9939,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6910214137972508,
+      "learning_rate": 2.4965845472662287e-06,
+      "loss": 0.8931,
+      "step": 2311
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5348895982248338,
+      "learning_rate": 2.48938806111388e-06,
+      "loss": 0.9321,
+      "step": 2312
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.48526558546427595,
+      "learning_rate": 2.482200487153107e-06,
+      "loss": 0.5937,
+      "step": 2313
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7056069342644873,
+      "learning_rate": 2.475021833912783e-06,
+      "loss": 0.8118,
+      "step": 2314
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6215963854397342,
+      "learning_rate": 2.467852109911195e-06,
+      "loss": 0.8771,
+      "step": 2315
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6384175315743763,
+      "learning_rate": 2.4606913236560283e-06,
+      "loss": 0.7849,
+      "step": 2316
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5965071531491869,
+      "learning_rate": 2.453539483644368e-06,
+      "loss": 0.9846,
+      "step": 2317
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6762462726283567,
+      "learning_rate": 2.4463965983626857e-06,
+      "loss": 0.8614,
+      "step": 2318
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6793141992714566,
+      "learning_rate": 2.4392626762868187e-06,
+      "loss": 0.693,
+      "step": 2319
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6926289571443951,
+      "learning_rate": 2.4321377258819747e-06,
+      "loss": 0.8132,
+      "step": 2320
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4797655320856672,
+      "learning_rate": 2.4250217556027176e-06,
+      "loss": 0.6328,
+      "step": 2321
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7060132199778675,
+      "learning_rate": 2.4179147738929475e-06,
+      "loss": 1.0049,
+      "step": 2322
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5993195892065991,
+      "learning_rate": 2.410816789185907e-06,
+      "loss": 0.7902,
+      "step": 2323
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7390690709643216,
+      "learning_rate": 2.403727809904154e-06,
+      "loss": 0.7484,
+      "step": 2324
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.45783200627629966,
+      "learning_rate": 2.3966478444595674e-06,
+      "loss": 0.97,
+      "step": 2325
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7064477457411197,
+      "learning_rate": 2.389576901253331e-06,
+      "loss": 0.6561,
+      "step": 2326
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7231992767066588,
+      "learning_rate": 2.3825149886759147e-06,
+      "loss": 0.9678,
+      "step": 2327
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5729242958480179,
+      "learning_rate": 2.375462115107079e-06,
+      "loss": 0.8304,
+      "step": 2328
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.655240343042949,
+      "learning_rate": 2.3684182889158615e-06,
+      "loss": 1.0021,
+      "step": 2329
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7088003947118239,
+      "learning_rate": 2.3613835184605527e-06,
+      "loss": 0.7388,
+      "step": 2330
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6733152470005171,
+      "learning_rate": 2.354357812088707e-06,
+      "loss": 0.7546,
+      "step": 2331
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.43950080576686085,
+      "learning_rate": 2.3473411781371245e-06,
+      "loss": 0.5762,
+      "step": 2332
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6068185102651897,
+      "learning_rate": 2.3403336249318297e-06,
+      "loss": 0.8921,
+      "step": 2333
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7558327517878356,
+      "learning_rate": 2.333335160788082e-06,
+      "loss": 1.0185,
+      "step": 2334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7054577967735393,
+      "learning_rate": 2.326345794010352e-06,
+      "loss": 0.8256,
+      "step": 2335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7010617177752255,
+      "learning_rate": 2.3193655328923116e-06,
+      "loss": 0.7715,
+      "step": 2336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.46630239774772386,
+      "learning_rate": 2.3123943857168316e-06,
+      "loss": 0.9462,
+      "step": 2337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.608365917305324,
+      "learning_rate": 2.3054323607559736e-06,
+      "loss": 0.7645,
+      "step": 2338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6816760030772154,
+      "learning_rate": 2.29847946627096e-06,
+      "loss": 0.9513,
+      "step": 2339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4664464654614688,
+      "learning_rate": 2.2915357105121938e-06,
+      "loss": 0.6496,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6931218413750474,
+      "learning_rate": 2.284601101719224e-06,
+      "loss": 0.8775,
+      "step": 2341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6925101922763913,
+      "learning_rate": 2.2776756481207483e-06,
+      "loss": 0.8649,
+      "step": 2342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6525881770572631,
+      "learning_rate": 2.2707593579346075e-06,
+      "loss": 0.846,
+      "step": 2343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6964890800597346,
+      "learning_rate": 2.2638522393677564e-06,
+      "loss": 0.9899,
+      "step": 2344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6822913931237695,
+      "learning_rate": 2.2569543006162753e-06,
+      "loss": 0.7769,
+      "step": 2345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7949456774870411,
+      "learning_rate": 2.2500655498653544e-06,
+      "loss": 0.9682,
+      "step": 2346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7062517720554031,
+      "learning_rate": 2.24318599528927e-06,
+      "loss": 0.7186,
+      "step": 2347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6860741518930461,
+      "learning_rate": 2.2363156450513946e-06,
+      "loss": 0.9642,
+      "step": 2348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5025472667110379,
+      "learning_rate": 2.229454507304181e-06,
+      "loss": 0.614,
+      "step": 2349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.48876426236600445,
+      "learning_rate": 2.222602590189139e-06,
+      "loss": 0.8824,
+      "step": 2350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6918246287093016,
+      "learning_rate": 2.2157599018368492e-06,
+      "loss": 0.9194,
+      "step": 2351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5451777618058643,
+      "learning_rate": 2.2089264503669374e-06,
+      "loss": 0.9251,
+      "step": 2352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6561109752137455,
+      "learning_rate": 2.2021022438880614e-06,
+      "loss": 0.8179,
+      "step": 2353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.62999269486907,
+      "learning_rate": 2.1952872904979184e-06,
+      "loss": 0.8517,
+      "step": 2354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6300353612367606,
+      "learning_rate": 2.1884815982832265e-06,
+      "loss": 0.7009,
+      "step": 2355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6862400033521712,
+      "learning_rate": 2.1816851753197023e-06,
+      "loss": 1.047,
+      "step": 2356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.598315665533078,
+      "learning_rate": 2.1748980296720767e-06,
+      "loss": 0.5749,
+      "step": 2357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.45667376159434303,
+      "learning_rate": 2.1681201693940667e-06,
+      "loss": 0.6382,
+      "step": 2358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6461620465229674,
+      "learning_rate": 2.1613516025283666e-06,
+      "loss": 0.7895,
+      "step": 2359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6902250015913135,
+      "learning_rate": 2.1545923371066534e-06,
+      "loss": 0.9519,
+      "step": 2360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.41704080911808306,
+      "learning_rate": 2.1478423811495554e-06,
+      "loss": 0.9359,
+      "step": 2361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6379733087004128,
+      "learning_rate": 2.1411017426666627e-06,
+      "loss": 0.6978,
+      "step": 2362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6181812371398,
+      "learning_rate": 2.134370429656507e-06,
+      "loss": 0.996,
+      "step": 2363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7460803893143625,
+      "learning_rate": 2.1276484501065508e-06,
+      "loss": 0.7581,
+      "step": 2364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6259765542755062,
+      "learning_rate": 2.1209358119931845e-06,
+      "loss": 0.8939,
+      "step": 2365
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.46877365727238474,
+      "learning_rate": 2.114232523281717e-06,
+      "loss": 0.6205,
+      "step": 2366
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6612713392060227,
+      "learning_rate": 2.1075385919263547e-06,
+      "loss": 0.9642,
+      "step": 2367
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6766337448738902,
+      "learning_rate": 2.100854025870207e-06,
+      "loss": 0.7158,
+      "step": 2368
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6387813786379641,
+      "learning_rate": 2.0941788330452718e-06,
+      "loss": 0.7722,
+      "step": 2369
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.722671561708832,
+      "learning_rate": 2.087513021372415e-06,
+      "loss": 0.853,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7245500267454573,
+      "learning_rate": 2.0808565987613794e-06,
+      "loss": 0.7791,
+      "step": 2371
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7340079239929398,
+      "learning_rate": 2.0742095731107693e-06,
+      "loss": 0.966,
+      "step": 2372
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.45986974251029644,
+      "learning_rate": 2.067571952308025e-06,
+      "loss": 0.9104,
+      "step": 2373
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7421481708481573,
+      "learning_rate": 2.06094374422944e-06,
+      "loss": 1.0144,
+      "step": 2374
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4625089354386122,
+      "learning_rate": 2.0543249567401356e-06,
+      "loss": 0.7233,
+      "step": 2375
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6167591736798371,
+      "learning_rate": 2.0477155976940466e-06,
+      "loss": 0.6674,
+      "step": 2376
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6806018349294172,
+      "learning_rate": 2.041115674933931e-06,
+      "loss": 0.9076,
+      "step": 2377
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7617711146398369,
+      "learning_rate": 2.0345251962913447e-06,
+      "loss": 0.8823,
+      "step": 2378
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7212913635169435,
+      "learning_rate": 2.027944169586633e-06,
+      "loss": 0.934,
+      "step": 2379
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5903120590090827,
+      "learning_rate": 2.0213726026289336e-06,
+      "loss": 0.68,
+      "step": 2380
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5880318842377736,
+      "learning_rate": 2.0148105032161512e-06,
+      "loss": 0.7903,
+      "step": 2381
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6841632079977498,
+      "learning_rate": 2.0082578791349604e-06,
+      "loss": 0.9095,
+      "step": 2382
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7312807338236267,
+      "learning_rate": 2.001714738160796e-06,
+      "loss": 0.8014,
+      "step": 2383
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.47207236443546174,
+      "learning_rate": 1.9951810880578303e-06,
+      "loss": 0.662,
+      "step": 2384
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7145726395418779,
+      "learning_rate": 1.988656936578981e-06,
+      "loss": 0.7417,
+      "step": 2385
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.44734450728611447,
+      "learning_rate": 1.982142291465896e-06,
+      "loss": 0.9931,
+      "step": 2386
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6970032805678085,
+      "learning_rate": 1.9756371604489333e-06,
+      "loss": 0.8376,
+      "step": 2387
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5826690801712714,
+      "learning_rate": 1.96914155124717e-06,
+      "loss": 0.8332,
+      "step": 2388
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6828878486135549,
+      "learning_rate": 1.9626554715683856e-06,
+      "loss": 0.9223,
+      "step": 2389
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6805565183128092,
+      "learning_rate": 1.956178929109043e-06,
+      "loss": 0.8237,
+      "step": 2390
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7575757792577441,
+      "learning_rate": 1.9497119315542945e-06,
+      "loss": 0.8985,
+      "step": 2391
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6183788842322672,
+      "learning_rate": 1.943254486577969e-06,
+      "loss": 0.8021,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6878312447061539,
+      "learning_rate": 1.9368066018425504e-06,
+      "loss": 0.8148,
+      "step": 2393
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4801897254178189,
+      "learning_rate": 1.9303682849991867e-06,
+      "loss": 0.5809,
+      "step": 2394
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.602527964719601,
+      "learning_rate": 1.9239395436876716e-06,
+      "loss": 0.6572,
+      "step": 2395
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7004177470704818,
+      "learning_rate": 1.9175203855364323e-06,
+      "loss": 1.0056,
+      "step": 2396
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6343273490331917,
+      "learning_rate": 1.911110818162526e-06,
+      "loss": 0.8017,
+      "step": 2397
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6621445570658075,
+      "learning_rate": 1.904710849171636e-06,
+      "loss": 0.9891,
+      "step": 2398
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.44779046573141196,
+      "learning_rate": 1.8983204861580418e-06,
+      "loss": 0.9718,
+      "step": 2399
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7177169162760839,
+      "learning_rate": 1.891939736704641e-06,
+      "loss": 0.6968,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4753823712720629,
+      "learning_rate": 1.8855686083829083e-06,
+      "loss": 0.6082,
+      "step": 2401
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7586530457035919,
+      "learning_rate": 1.8792071087529118e-06,
+      "loss": 0.8595,
+      "step": 2402
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6454921146113552,
+      "learning_rate": 1.8728552453632952e-06,
+      "loss": 0.8872,
+      "step": 2403
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5694415594828693,
+      "learning_rate": 1.8665130257512566e-06,
+      "loss": 0.6569,
+      "step": 2404
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6876184121588244,
+      "learning_rate": 1.8601804574425609e-06,
+      "loss": 0.946,
+      "step": 2405
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7286473185566592,
+      "learning_rate": 1.8538575479515208e-06,
+      "loss": 0.8506,
+      "step": 2406
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6334637748908755,
+      "learning_rate": 1.8475443047809782e-06,
+      "loss": 0.7776,
+      "step": 2407
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6581285412531745,
+      "learning_rate": 1.8412407354223149e-06,
+      "loss": 0.8336,
+      "step": 2408
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6118430118157284,
+      "learning_rate": 1.8349468473554311e-06,
+      "loss": 1.0555,
+      "step": 2409
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6915443293659932,
+      "learning_rate": 1.8286626480487336e-06,
+      "loss": 0.9331,
+      "step": 2410
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4709370131229521,
+      "learning_rate": 1.8223881449591385e-06,
+      "loss": 0.5738,
+      "step": 2411
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.739810366039666,
+      "learning_rate": 1.8161233455320571e-06,
+      "loss": 0.8294,
+      "step": 2412
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.48770358979014605,
+      "learning_rate": 1.8098682572013782e-06,
+      "loss": 0.948,
+      "step": 2413
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.624740822469974,
+      "learning_rate": 1.8036228873894745e-06,
+      "loss": 0.6813,
+      "step": 2414
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7120682965325161,
+      "learning_rate": 1.7973872435071882e-06,
+      "loss": 0.9736,
+      "step": 2415
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6138818307429381,
+      "learning_rate": 1.7911613329538102e-06,
+      "loss": 0.68,
+      "step": 2416
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6805840485751142,
+      "learning_rate": 1.7849451631170922e-06,
+      "loss": 0.649,
+      "step": 2417
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.738594571631631,
+      "learning_rate": 1.7787387413732248e-06,
+      "loss": 0.9586,
+      "step": 2418
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6521860095187768,
+      "learning_rate": 1.7725420750868273e-06,
+      "loss": 0.8763,
+      "step": 2419
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4737532500127817,
+      "learning_rate": 1.7663551716109495e-06,
+      "loss": 0.6252,
+      "step": 2420
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7065701556434947,
+      "learning_rate": 1.760178038287048e-06,
+      "loss": 0.7616,
+      "step": 2421
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6024649593234435,
+      "learning_rate": 1.7540106824449943e-06,
+      "loss": 1.1151,
+      "step": 2422
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6878674403158926,
+      "learning_rate": 1.7478531114030573e-06,
+      "loss": 0.7513,
+      "step": 2423
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4044486883847999,
+      "learning_rate": 1.7417053324678856e-06,
+      "loss": 0.9176,
+      "step": 2424
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6462804210926201,
+      "learning_rate": 1.7355673529345184e-06,
+      "loss": 0.8422,
+      "step": 2425
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7044557897322851,
+      "learning_rate": 1.7294391800863664e-06,
+      "loss": 0.776,
+      "step": 2426
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6303590461380676,
+      "learning_rate": 1.7233208211951946e-06,
+      "loss": 0.8748,
+      "step": 2427
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4936892435144729,
+      "learning_rate": 1.7172122835211337e-06,
+      "loss": 0.5005,
+      "step": 2428
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6049832433978877,
+      "learning_rate": 1.711113574312655e-06,
+      "loss": 0.8729,
+      "step": 2429
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7296739103318604,
+      "learning_rate": 1.7050247008065634e-06,
+      "loss": 0.7349,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6827525487577913,
+      "learning_rate": 1.6989456702280005e-06,
+      "loss": 0.844,
+      "step": 2431
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6939918356745617,
+      "learning_rate": 1.692876489790426e-06,
+      "loss": 1.1199,
+      "step": 2432
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5732560903242717,
+      "learning_rate": 1.6868171666956045e-06,
+      "loss": 0.8947,
+      "step": 2433
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5543355876125785,
+      "learning_rate": 1.6807677081336116e-06,
+      "loss": 0.8604,
+      "step": 2434
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7529276226662013,
+      "learning_rate": 1.6747281212828193e-06,
+      "loss": 0.7265,
+      "step": 2435
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4334411205267366,
+      "learning_rate": 1.6686984133098737e-06,
+      "loss": 0.9085,
+      "step": 2436
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.485553510697186,
+      "learning_rate": 1.6626785913697118e-06,
+      "loss": 0.5898,
+      "step": 2437
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6625121327986695,
+      "learning_rate": 1.6566686626055307e-06,
+      "loss": 0.7915,
+      "step": 2438
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6223489156943912,
+      "learning_rate": 1.650668634148792e-06,
+      "loss": 0.8333,
+      "step": 2439
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6304573505600192,
+      "learning_rate": 1.6446785131192123e-06,
+      "loss": 0.7515,
+      "step": 2440
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7334223834286984,
+      "learning_rate": 1.638698306624743e-06,
+      "loss": 0.9483,
+      "step": 2441
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.639234607811471,
+      "learning_rate": 1.6327280217615793e-06,
+      "loss": 0.9245,
+      "step": 2442
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6176694575582399,
+      "learning_rate": 1.6267676656141429e-06,
+      "loss": 0.7226,
+      "step": 2443
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7144494376727545,
+      "learning_rate": 1.6208172452550664e-06,
+      "loss": 0.9819,
+      "step": 2444
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.590328388015457,
+      "learning_rate": 1.6148767677451994e-06,
+      "loss": 0.9275,
+      "step": 2445
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5069526833964784,
+      "learning_rate": 1.6089462401335932e-06,
+      "loss": 0.5973,
+      "step": 2446
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6501240219044018,
+      "learning_rate": 1.6030256694574876e-06,
+      "loss": 0.7229,
+      "step": 2447
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4609773334190982,
+      "learning_rate": 1.5971150627423093e-06,
+      "loss": 0.965,
+      "step": 2448
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6020417774490008,
+      "learning_rate": 1.5912144270016672e-06,
+      "loss": 0.6741,
+      "step": 2449
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.695590918605766,
+      "learning_rate": 1.5853237692373291e-06,
+      "loss": 0.8052,
+      "step": 2450
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6322607150904076,
+      "learning_rate": 1.5794430964392293e-06,
+      "loss": 0.9464,
+      "step": 2451
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7012249313902517,
+      "learning_rate": 1.5735724155854538e-06,
+      "loss": 0.6926,
+      "step": 2452
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6005208424950059,
+      "learning_rate": 1.5677117336422265e-06,
+      "loss": 0.951,
+      "step": 2453
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5030648780659306,
+      "learning_rate": 1.5618610575639115e-06,
+      "loss": 0.4858,
+      "step": 2454
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6686602327406325,
+      "learning_rate": 1.556020394293003e-06,
+      "loss": 0.7309,
+      "step": 2455
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5560488612157071,
+      "learning_rate": 1.5501897507601016e-06,
+      "loss": 0.9812,
+      "step": 2456
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7786432430100668,
+      "learning_rate": 1.5443691338839328e-06,
+      "loss": 0.8893,
+      "step": 2457
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7625111560555672,
+      "learning_rate": 1.5385585505713129e-06,
+      "loss": 0.9471,
+      "step": 2458
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6806509218480257,
+      "learning_rate": 1.5327580077171589e-06,
+      "loss": 0.7564,
+      "step": 2459
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6977232330940281,
+      "learning_rate": 1.5269675122044746e-06,
+      "loss": 0.9112,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6854572849865737,
+      "learning_rate": 1.521187070904333e-06,
+      "loss": 0.7816,
+      "step": 2461
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4477199143884538,
+      "learning_rate": 1.515416690675885e-06,
+      "loss": 0.9232,
+      "step": 2462
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6011365492213298,
+      "learning_rate": 1.509656378366343e-06,
+      "loss": 0.7006,
+      "step": 2463
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.48371493499999635,
+      "learning_rate": 1.5039061408109656e-06,
+      "loss": 0.5854,
+      "step": 2464
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7106604685156193,
+      "learning_rate": 1.4981659848330622e-06,
+      "loss": 1.0159,
+      "step": 2465
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.586961870199062,
+      "learning_rate": 1.4924359172439794e-06,
+      "loss": 0.8035,
+      "step": 2466
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7187859192590325,
+      "learning_rate": 1.486715944843089e-06,
+      "loss": 0.8736,
+      "step": 2467
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6270501067911285,
+      "learning_rate": 1.4810060744177868e-06,
+      "loss": 0.9399,
+      "step": 2468
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7489911102259356,
+      "learning_rate": 1.475306312743483e-06,
+      "loss": 0.7997,
+      "step": 2469
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6737878281644231,
+      "learning_rate": 1.4696166665835853e-06,
+      "loss": 0.945,
+      "step": 2470
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7144855464650665,
+      "learning_rate": 1.4639371426895066e-06,
+      "loss": 0.8853,
+      "step": 2471
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.629123064187094,
+      "learning_rate": 1.4582677478006468e-06,
+      "loss": 0.8558,
+      "step": 2472
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.36849670595997985,
+      "learning_rate": 1.4526084886443813e-06,
+      "loss": 0.7663,
+      "step": 2473
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6736841129797656,
+      "learning_rate": 1.446959371936062e-06,
+      "loss": 0.8426,
+      "step": 2474
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7234802975837086,
+      "learning_rate": 1.4413204043790108e-06,
+      "loss": 0.9207,
+      "step": 2475
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.726119138529403,
+      "learning_rate": 1.4356915926644942e-06,
+      "loss": 0.8744,
+      "step": 2476
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6407665783607683,
+      "learning_rate": 1.4300729434717398e-06,
+      "loss": 0.9344,
+      "step": 2477
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.714420822852437,
+      "learning_rate": 1.4244644634679084e-06,
+      "loss": 0.7813,
+      "step": 2478
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8442384878747011,
+      "learning_rate": 1.4188661593080966e-06,
+      "loss": 0.9407,
+      "step": 2479
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6437450171132162,
+      "learning_rate": 1.4132780376353295e-06,
+      "loss": 0.9348,
+      "step": 2480
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.482221540458382,
+      "learning_rate": 1.4077001050805418e-06,
+      "loss": 0.5337,
+      "step": 2481
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6745655616034317,
+      "learning_rate": 1.4021323682625842e-06,
+      "loss": 0.9021,
+      "step": 2482
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7101464233213702,
+      "learning_rate": 1.39657483378821e-06,
+      "loss": 0.8871,
+      "step": 2483
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7063003919391933,
+      "learning_rate": 1.3910275082520575e-06,
+      "loss": 0.8783,
+      "step": 2484
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7911058982173373,
+      "learning_rate": 1.3854903982366585e-06,
+      "loss": 0.7487,
+      "step": 2485
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38883075419563773,
+      "learning_rate": 1.3799635103124255e-06,
+      "loss": 0.9494,
+      "step": 2486
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7047471686562171,
+      "learning_rate": 1.3744468510376309e-06,
+      "loss": 0.8472,
+      "step": 2487
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6779570763616944,
+      "learning_rate": 1.3689404269584172e-06,
+      "loss": 0.7393,
+      "step": 2488
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.621051286750564,
+      "learning_rate": 1.3634442446087837e-06,
+      "loss": 0.8261,
+      "step": 2489
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4706605675860966,
+      "learning_rate": 1.3579583105105677e-06,
+      "loss": 0.5179,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.604460896494059,
+      "learning_rate": 1.3524826311734552e-06,
+      "loss": 0.9336,
+      "step": 2491
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6853162246017263,
+      "learning_rate": 1.3470172130949599e-06,
+      "loss": 0.7718,
+      "step": 2492
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6610857866301763,
+      "learning_rate": 1.3415620627604154e-06,
+      "loss": 0.8203,
+      "step": 2493
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5666859249248799,
+      "learning_rate": 1.336117186642979e-06,
+      "loss": 0.7993,
+      "step": 2494
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7018753043271011,
+      "learning_rate": 1.3306825912036137e-06,
+      "loss": 0.8005,
+      "step": 2495
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6657790603023427,
+      "learning_rate": 1.3252582828910775e-06,
+      "loss": 0.971,
+      "step": 2496
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.475277983186321,
+      "learning_rate": 1.3198442681419321e-06,
+      "loss": 0.9673,
+      "step": 2497
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6550981966633919,
+      "learning_rate": 1.3144405533805138e-06,
+      "loss": 0.9576,
+      "step": 2498
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.47365978512867046,
+      "learning_rate": 1.3090471450189446e-06,
+      "loss": 0.4447,
+      "step": 2499
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6419575399936105,
+      "learning_rate": 1.3036640494571162e-06,
+      "loss": 0.8176,
+      "step": 2500
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6498734801621247,
+      "learning_rate": 1.298291273082678e-06,
+      "loss": 0.8485,
+      "step": 2501
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6718908331238651,
+      "learning_rate": 1.292928822271038e-06,
+      "loss": 0.7774,
+      "step": 2502
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5272489985338807,
+      "learning_rate": 1.2875767033853547e-06,
+      "loss": 1.0023,
+      "step": 2503
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7148452374990993,
+      "learning_rate": 1.2822349227765185e-06,
+      "loss": 0.8221,
+      "step": 2504
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8465184309105639,
+      "learning_rate": 1.2769034867831588e-06,
+      "loss": 0.6748,
+      "step": 2505
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6511865269340571,
+      "learning_rate": 1.2715824017316302e-06,
+      "loss": 0.7741,
+      "step": 2506
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.46699369991959105,
+      "learning_rate": 1.2662716739360003e-06,
+      "loss": 0.6498,
+      "step": 2507
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7073692355412297,
+      "learning_rate": 1.260971309698048e-06,
+      "loss": 1.0129,
+      "step": 2508
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6402773262981848,
+      "learning_rate": 1.2556813153072612e-06,
+      "loss": 0.8254,
+      "step": 2509
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.720119693510644,
+      "learning_rate": 1.2504016970408105e-06,
+      "loss": 0.9171,
+      "step": 2510
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.48698364646802106,
+      "learning_rate": 1.245132461163564e-06,
+      "loss": 0.9143,
+      "step": 2511
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6692332345986577,
+      "learning_rate": 1.2398736139280687e-06,
+      "loss": 0.6676,
+      "step": 2512
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6273059102743634,
+      "learning_rate": 1.2346251615745375e-06,
+      "loss": 1.0095,
+      "step": 2513
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6421773694968591,
+      "learning_rate": 1.2293871103308596e-06,
+      "loss": 0.7586,
+      "step": 2514
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6911184856857433,
+      "learning_rate": 1.22415946641257e-06,
+      "loss": 1.0077,
+      "step": 2515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6828811679112302,
+      "learning_rate": 1.218942236022862e-06,
+      "loss": 0.7836,
+      "step": 2516
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7133617088589479,
+      "learning_rate": 1.2137354253525734e-06,
+      "loss": 0.907,
+      "step": 2517
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4351218206981329,
+      "learning_rate": 1.208539040580171e-06,
+      "loss": 0.4643,
+      "step": 2518
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6554675972206198,
+      "learning_rate": 1.2033530878717548e-06,
+      "loss": 0.9103,
+      "step": 2519
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7709587251596351,
+      "learning_rate": 1.1981775733810474e-06,
+      "loss": 0.9991,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7245716337638513,
+      "learning_rate": 1.193012503249379e-06,
+      "loss": 0.7971,
+      "step": 2521
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6915732567954039,
+      "learning_rate": 1.187857883605691e-06,
+      "loss": 0.9189,
+      "step": 2522
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.428955604032153,
+      "learning_rate": 1.1827137205665273e-06,
+      "loss": 0.8921,
+      "step": 2523
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6186074262002763,
+      "learning_rate": 1.1775800202360133e-06,
+      "loss": 0.8065,
+      "step": 2524
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7548294658744028,
+      "learning_rate": 1.1724567887058702e-06,
+      "loss": 0.9315,
+      "step": 2525
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.40945493033181046,
+      "learning_rate": 1.1673440320553941e-06,
+      "loss": 0.6199,
+      "step": 2526
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.83037580980046,
+      "learning_rate": 1.1622417563514443e-06,
+      "loss": 0.9364,
+      "step": 2527
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7193459984781028,
+      "learning_rate": 1.1571499676484522e-06,
+      "loss": 0.8827,
+      "step": 2528
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6470171650285496,
+      "learning_rate": 1.1520686719884022e-06,
+      "loss": 0.9611,
+      "step": 2529
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6567163908596944,
+      "learning_rate": 1.1469978754008238e-06,
+      "loss": 0.8542,
+      "step": 2530
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6623125169833212,
+      "learning_rate": 1.1419375839027947e-06,
+      "loss": 0.786,
+      "step": 2531
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6765344588269222,
+      "learning_rate": 1.1368878034989227e-06,
+      "loss": 0.8535,
+      "step": 2532
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6856659780359411,
+      "learning_rate": 1.1318485401813438e-06,
+      "loss": 0.6714,
+      "step": 2533
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6899772273082865,
+      "learning_rate": 1.126819799929716e-06,
+      "loss": 1.0153,
+      "step": 2534
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3771021795620555,
+      "learning_rate": 1.1218015887112055e-06,
+      "loss": 0.766,
+      "step": 2535
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6653135953677525,
+      "learning_rate": 1.1167939124804894e-06,
+      "loss": 0.8309,
+      "step": 2536
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6829822411992534,
+      "learning_rate": 1.1117967771797477e-06,
+      "loss": 0.7525,
+      "step": 2537
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5996740397530873,
+      "learning_rate": 1.1068101887386395e-06,
+      "loss": 0.9177,
+      "step": 2538
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6984486512506664,
+      "learning_rate": 1.1018341530743215e-06,
+      "loss": 0.8059,
+      "step": 2539
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5907703350982964,
+      "learning_rate": 1.0968686760914248e-06,
+      "loss": 0.8501,
+      "step": 2540
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7244274067815717,
+      "learning_rate": 1.0919137636820442e-06,
+      "loss": 0.9009,
+      "step": 2541
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6910463730943475,
+      "learning_rate": 1.0869694217257498e-06,
+      "loss": 0.876,
+      "step": 2542
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.46734495406054366,
+      "learning_rate": 1.0820356560895629e-06,
+      "loss": 0.4524,
+      "step": 2543
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6313040170149081,
+      "learning_rate": 1.0771124726279514e-06,
+      "loss": 0.8883,
+      "step": 2544
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6863727197694797,
+      "learning_rate": 1.0721998771828312e-06,
+      "loss": 0.7994,
+      "step": 2545
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6870879154674185,
+      "learning_rate": 1.0672978755835572e-06,
+      "loss": 0.9128,
+      "step": 2546
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.43278519260298814,
+      "learning_rate": 1.0624064736469054e-06,
+      "loss": 0.9217,
+      "step": 2547
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6179123947231864,
+      "learning_rate": 1.057525677177077e-06,
+      "loss": 0.705,
+      "step": 2548
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6889881702767485,
+      "learning_rate": 1.052655491965695e-06,
+      "loss": 1.0283,
+      "step": 2549
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7000614845558379,
+      "learning_rate": 1.04779592379178e-06,
+      "loss": 0.7907,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5862568414989496,
+      "learning_rate": 1.0429469784217626e-06,
+      "loss": 0.8593,
+      "step": 2551
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4786334474705862,
+      "learning_rate": 1.0381086616094671e-06,
+      "loss": 0.599,
+      "step": 2552
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6371885939645642,
+      "learning_rate": 1.0332809790961019e-06,
+      "loss": 0.9436,
+      "step": 2553
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6547887439492089,
+      "learning_rate": 1.02846393661026e-06,
+      "loss": 0.7411,
+      "step": 2554
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.648099008156304,
+      "learning_rate": 1.023657539867906e-06,
+      "loss": 0.7492,
+      "step": 2555
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6593242935192011,
+      "learning_rate": 1.0188617945723755e-06,
+      "loss": 0.927,
+      "step": 2556
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7679612528540752,
+      "learning_rate": 1.0140767064143653e-06,
+      "loss": 0.7506,
+      "step": 2557
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6985502611272685,
+      "learning_rate": 1.00930228107192e-06,
+      "loss": 0.913,
+      "step": 2558
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.43014680885375506,
+      "learning_rate": 1.0045385242104378e-06,
+      "loss": 0.874,
+      "step": 2559
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6957873713745786,
+      "learning_rate": 9.99785441482659e-07,
+      "loss": 1.0005,
+      "step": 2560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.43108009554773324,
+      "learning_rate": 9.950430385286492e-07,
+      "loss": 0.7088,
+      "step": 2561
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6339218246007119,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 0.6783,
+      "step": 2562
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6298991406678517,
+      "learning_rate": 9.855902944388595e-07,
+      "loss": 0.8697,
+      "step": 2563
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6932603532094853,
+      "learning_rate": 9.808799645198285e-07,
+      "loss": 0.9011,
+      "step": 2564
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6352640178022347,
+      "learning_rate": 9.76180336808058e-07,
+      "loss": 0.7651,
+      "step": 2565
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6276955614937457,
+      "learning_rate": 9.714914168801915e-07,
+      "loss": 0.8008,
+      "step": 2566
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.677285056842483,
+      "learning_rate": 9.66813210300157e-07,
+      "loss": 0.8144,
+      "step": 2567
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6040019654449765,
+      "learning_rate": 9.62145722619182e-07,
+      "loss": 0.8538,
+      "step": 2568
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7652997259729792,
+      "learning_rate": 9.574889593757686e-07,
+      "loss": 0.7687,
+      "step": 2569
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5172432145658996,
+      "learning_rate": 9.528429260956895e-07,
+      "loss": 0.6501,
+      "step": 2570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7459093058973122,
+      "learning_rate": 9.482076282919927e-07,
+      "loss": 0.7704,
+      "step": 2571
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40344375573647456,
+      "learning_rate": 9.435830714649851e-07,
+      "loss": 0.9971,
+      "step": 2572
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6193236268517582,
+      "learning_rate": 9.389692611022239e-07,
+      "loss": 0.6892,
+      "step": 2573
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6196235721116143,
+      "learning_rate": 9.343662026785217e-07,
+      "loss": 0.8336,
+      "step": 2574
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6923942591498149,
+      "learning_rate": 9.297739016559226e-07,
+      "loss": 0.9193,
+      "step": 2575
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6630678517481088,
+      "learning_rate": 9.251923634837156e-07,
+      "loss": 0.798,
+      "step": 2576
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7109948965588807,
+      "learning_rate": 9.206215935984164e-07,
+      "loss": 0.895,
+      "step": 2577
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6485119497240734,
+      "learning_rate": 9.160615974237563e-07,
+      "loss": 0.8024,
+      "step": 2578
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6903284032847088,
+      "learning_rate": 9.115123803706904e-07,
+      "loss": 0.9434,
+      "step": 2579
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.47193262872625286,
+      "learning_rate": 9.069739478373807e-07,
+      "loss": 0.4579,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6284133955726031,
+      "learning_rate": 9.024463052091869e-07,
+      "loss": 0.6909,
+      "step": 2581
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6094641756793487,
+      "learning_rate": 8.979294578586739e-07,
+      "loss": 0.9226,
+      "step": 2582
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6687106366119285,
+      "learning_rate": 8.934234111455919e-07,
+      "loss": 0.8381,
+      "step": 2583
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6336429856429724,
+      "learning_rate": 8.889281704168739e-07,
+      "loss": 0.972,
+      "step": 2584
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.41188309917760585,
+      "learning_rate": 8.844437410066331e-07,
+      "loss": 0.9734,
+      "step": 2585
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6466265315113616,
+      "learning_rate": 8.79970128236154e-07,
+      "loss": 0.8204,
+      "step": 2586
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.48111192273233044,
+      "learning_rate": 8.755073374138823e-07,
+      "loss": 0.4854,
+      "step": 2587
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7241612935162342,
+      "learning_rate": 8.710553738354254e-07,
+      "loss": 0.8533,
+      "step": 2588
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6169772713641836,
+      "learning_rate": 8.666142427835444e-07,
+      "loss": 0.8032,
+      "step": 2589
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5868191599420277,
+      "learning_rate": 8.621839495281414e-07,
+      "loss": 0.7027,
+      "step": 2590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6469800758190206,
+      "learning_rate": 8.577644993262646e-07,
+      "loss": 1.0295,
+      "step": 2591
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7488361630742034,
+      "learning_rate": 8.533558974220868e-07,
+      "loss": 0.755,
+      "step": 2592
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6489730690709755,
+      "learning_rate": 8.489581490469168e-07,
+      "loss": 0.795,
+      "step": 2593
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7386607188913797,
+      "learning_rate": 8.445712594191824e-07,
+      "loss": 0.8593,
+      "step": 2594
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5856504359590724,
+      "learning_rate": 8.401952337444208e-07,
+      "loss": 1.0131,
+      "step": 2595
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.456446922082125,
+      "learning_rate": 8.35830077215285e-07,
+      "loss": 0.605,
+      "step": 2596
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6762591196457043,
+      "learning_rate": 8.314757950115294e-07,
+      "loss": 0.7753,
+      "step": 2597
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6908941495221185,
+      "learning_rate": 8.271323922999974e-07,
+      "loss": 1.0139,
+      "step": 2598
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4163705664394419,
+      "learning_rate": 8.22799874234631e-07,
+      "loss": 0.9073,
+      "step": 2599
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6002436894324834,
+      "learning_rate": 8.184782459564556e-07,
+      "loss": 0.7012,
+      "step": 2600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7281923547731014,
+      "learning_rate": 8.1416751259357e-07,
+      "loss": 0.9567,
+      "step": 2601
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.599644391989456,
+      "learning_rate": 8.098676792611471e-07,
+      "loss": 0.6456,
+      "step": 2602
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6879680508682496,
+      "learning_rate": 8.055787510614288e-07,
+      "loss": 0.825,
+      "step": 2603
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7115138651287,
+      "learning_rate": 8.013007330837109e-07,
+      "loss": 0.8202,
+      "step": 2604
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6856148583801249,
+      "learning_rate": 7.970336304043469e-07,
+      "loss": 0.8818,
+      "step": 2605
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4665135895641313,
+      "learning_rate": 7.927774480867411e-07,
+      "loss": 0.6055,
+      "step": 2606
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6177371216488928,
+      "learning_rate": 7.885321911813304e-07,
+      "loss": 0.9121,
+      "step": 2607
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6051604201252075,
+      "learning_rate": 7.84297864725595e-07,
+      "loss": 1.0635,
+      "step": 2608
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7708323226373489,
+      "learning_rate": 7.800744737440446e-07,
+      "loss": 0.732,
+      "step": 2609
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.40094805127856065,
+      "learning_rate": 7.758620232482083e-07,
+      "loss": 0.9272,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6604330606486384,
+      "learning_rate": 7.716605182366377e-07,
+      "loss": 0.8129,
+      "step": 2611
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6623499841011313,
+      "learning_rate": 7.674699636948913e-07,
+      "loss": 0.7321,
+      "step": 2612
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6188745239157449,
+      "learning_rate": 7.632903645955392e-07,
+      "loss": 0.8369,
+      "step": 2613
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49747168858154356,
+      "learning_rate": 7.591217258981476e-07,
+      "loss": 0.5201,
+      "step": 2614
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5917269682813063,
+      "learning_rate": 7.549640525492796e-07,
+      "loss": 0.9214,
+      "step": 2615
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7587815087987402,
+      "learning_rate": 7.508173494824844e-07,
+      "loss": 0.7453,
+      "step": 2616
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6742818856394366,
+      "learning_rate": 7.46681621618297e-07,
+      "loss": 0.99,
+      "step": 2617
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.69162987829424,
+      "learning_rate": 7.425568738642242e-07,
+      "loss": 0.9668,
+      "step": 2618
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5939054205615947,
+      "learning_rate": 7.384431111147483e-07,
+      "loss": 0.8891,
+      "step": 2619
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5728714742904193,
+      "learning_rate": 7.343403382513115e-07,
+      "loss": 0.9186,
+      "step": 2620
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7815136443553338,
+      "learning_rate": 7.302485601423215e-07,
+      "loss": 0.7374,
+      "step": 2621
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4592230825783512,
+      "learning_rate": 7.261677816431367e-07,
+      "loss": 0.9459,
+      "step": 2622
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.46255159788534866,
+      "learning_rate": 7.220980075960593e-07,
+      "loss": 0.4556,
+      "step": 2623
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6737877504725929,
+      "learning_rate": 7.180392428303395e-07,
+      "loss": 0.8043,
+      "step": 2624
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6068300567310446,
+      "learning_rate": 7.139914921621638e-07,
+      "loss": 0.8313,
+      "step": 2625
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6182863770591279,
+      "learning_rate": 7.099547603946422e-07,
+      "loss": 0.6882,
+      "step": 2626
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6856497815910317,
+      "learning_rate": 7.059290523178153e-07,
+      "loss": 0.9625,
+      "step": 2627
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6427381719418971,
+      "learning_rate": 7.019143727086441e-07,
+      "loss": 0.9026,
+      "step": 2628
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6830180129844278,
+      "learning_rate": 6.979107263309969e-07,
+      "loss": 0.8769,
+      "step": 2629
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6466622006354869,
+      "learning_rate": 6.939181179356568e-07,
+      "loss": 0.8077,
+      "step": 2630
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5580405039441584,
+      "learning_rate": 6.899365522603041e-07,
+      "loss": 0.901,
+      "step": 2631
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4613523609007502,
+      "learning_rate": 6.859660340295171e-07,
+      "loss": 0.5371,
+      "step": 2632
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6740918119465281,
+      "learning_rate": 6.82006567954766e-07,
+      "loss": 0.7535,
+      "step": 2633
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.41629064814287214,
+      "learning_rate": 6.780581587344071e-07,
+      "loss": 0.9724,
+      "step": 2634
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6999227951541851,
+      "learning_rate": 6.741208110536712e-07,
+      "loss": 0.7347,
+      "step": 2635
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6938310764584681,
+      "learning_rate": 6.701945295846701e-07,
+      "loss": 0.8704,
+      "step": 2636
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6380179624619754,
+      "learning_rate": 6.662793189863825e-07,
+      "loss": 0.8883,
+      "step": 2637
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6835713290351423,
+      "learning_rate": 6.623751839046455e-07,
+      "loss": 0.7488,
+      "step": 2638
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5954127655176479,
+      "learning_rate": 6.584821289721599e-07,
+      "loss": 0.9194,
+      "step": 2639
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.48912303640154275,
+      "learning_rate": 6.54600158808476e-07,
+      "loss": 0.5396,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7665059451817832,
+      "learning_rate": 6.507292780199892e-07,
+      "loss": 0.9738,
+      "step": 2641
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5997193106552118,
+      "learning_rate": 6.468694911999418e-07,
+      "loss": 0.9168,
+      "step": 2642
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6250670137314043,
+      "learning_rate": 6.430208029284035e-07,
+      "loss": 0.8489,
+      "step": 2643
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6744924356737301,
+      "learning_rate": 6.391832177722812e-07,
+      "loss": 0.8806,
+      "step": 2644
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6860584645091814,
+      "learning_rate": 6.353567402853056e-07,
+      "loss": 0.7751,
+      "step": 2645
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7365879196231189,
+      "learning_rate": 6.315413750080224e-07,
+      "loss": 0.9086,
+      "step": 2646
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.43160180413111626,
+      "learning_rate": 6.277371264677956e-07,
+      "loss": 0.9165,
+      "step": 2647
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6360143671658426,
+      "learning_rate": 6.239439991787988e-07,
+      "loss": 0.8676,
+      "step": 2648
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5757052455002667,
+      "learning_rate": 6.201619976420037e-07,
+      "loss": 0.7229,
+      "step": 2649
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4675835507191645,
+      "learning_rate": 6.163911263451861e-07,
+      "loss": 0.5573,
+      "step": 2650
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6528150722500822,
+      "learning_rate": 6.126313897629121e-07,
+      "loss": 0.9644,
+      "step": 2651
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5850278164983124,
+      "learning_rate": 6.088827923565321e-07,
+      "loss": 0.7511,
+      "step": 2652
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7243350865729323,
+      "learning_rate": 6.051453385741835e-07,
+      "loss": 0.9568,
+      "step": 2653
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6012575418008448,
+      "learning_rate": 6.014190328507785e-07,
+      "loss": 0.8491,
+      "step": 2654
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6798690070224261,
+      "learning_rate": 5.977038796080004e-07,
+      "loss": 0.8184,
+      "step": 2655
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6599499632839696,
+      "learning_rate": 5.939998832542992e-07,
+      "loss": 0.9594,
+      "step": 2656
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6778121015703253,
+      "learning_rate": 5.90307048184886e-07,
+      "loss": 0.8397,
+      "step": 2657
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6317479704019133,
+      "learning_rate": 5.866253787817267e-07,
+      "loss": 0.9082,
+      "step": 2658
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3709950953339937,
+      "learning_rate": 5.82954879413542e-07,
+      "loss": 0.7468,
+      "step": 2659
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6862314327249862,
+      "learning_rate": 5.792955544357914e-07,
+      "loss": 0.982,
+      "step": 2660
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7414552335451302,
+      "learning_rate": 5.756474081906804e-07,
+      "loss": 0.7236,
+      "step": 2661
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6452780973632599,
+      "learning_rate": 5.720104450071484e-07,
+      "loss": 0.8526,
+      "step": 2662
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5650823031129877,
+      "learning_rate": 5.683846692008621e-07,
+      "loss": 0.8407,
+      "step": 2663
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7628984378680896,
+      "learning_rate": 5.647700850742166e-07,
+      "loss": 0.8431,
+      "step": 2664
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6024206885794379,
+      "learning_rate": 5.611666969163243e-07,
+      "loss": 0.8918,
+      "step": 2665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5779717746394372,
+      "learning_rate": 5.575745090030138e-07,
+      "loss": 0.8889,
+      "step": 2666
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.45923243099609845,
+      "learning_rate": 5.539935255968221e-07,
+      "loss": 0.625,
+      "step": 2667
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6863383596622997,
+      "learning_rate": 5.504237509469934e-07,
+      "loss": 0.8057,
+      "step": 2668
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7034090685376105,
+      "learning_rate": 5.468651892894672e-07,
+      "loss": 0.8424,
+      "step": 2669
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7329272441717904,
+      "learning_rate": 5.433178448468812e-07,
+      "loss": 0.9096,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6976712834449863,
+      "learning_rate": 5.397817218285628e-07,
+      "loss": 0.7157,
+      "step": 2671
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40739916184991404,
+      "learning_rate": 5.362568244305188e-07,
+      "loss": 0.9597,
+      "step": 2672
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6950303903006086,
+      "learning_rate": 5.327431568354402e-07,
+      "loss": 0.8093,
+      "step": 2673
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5690657474527857,
+      "learning_rate": 5.29240723212695e-07,
+      "loss": 0.6202,
+      "step": 2674
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6851284935439612,
+      "learning_rate": 5.257495277183122e-07,
+      "loss": 0.9943,
+      "step": 2675
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4695854817209914,
+      "learning_rate": 5.222695744949957e-07,
+      "loss": 0.4324,
+      "step": 2676
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6025750858608291,
+      "learning_rate": 5.18800867672099e-07,
+      "loss": 0.9718,
+      "step": 2677
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6373175743935323,
+      "learning_rate": 5.153434113656386e-07,
+      "loss": 0.7343,
+      "step": 2678
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6588407807981604,
+      "learning_rate": 5.118972096782782e-07,
+      "loss": 0.7608,
+      "step": 2679
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6415093873970733,
+      "learning_rate": 5.084622666993244e-07,
+      "loss": 0.8154,
+      "step": 2680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7051391801313881,
+      "learning_rate": 5.05038586504727e-07,
+      "loss": 0.8043,
+      "step": 2681
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6189051513828916,
+      "learning_rate": 5.016261731570715e-07,
+      "loss": 0.9007,
+      "step": 2682
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.41994864222127215,
+      "learning_rate": 4.982250307055703e-07,
+      "loss": 0.9512,
+      "step": 2683
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6554948523346603,
+      "learning_rate": 4.948351631860648e-07,
+      "loss": 0.8905,
+      "step": 2684
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4856523467645659,
+      "learning_rate": 4.914565746210198e-07,
+      "loss": 0.4778,
+      "step": 2685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6486222952802199,
+      "learning_rate": 4.880892690195083e-07,
+      "loss": 0.772,
+      "step": 2686
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5561290244334304,
+      "learning_rate": 4.847332503772229e-07,
+      "loss": 0.8074,
+      "step": 2687
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6705087233674022,
+      "learning_rate": 4.813885226764592e-07,
+      "loss": 0.9739,
+      "step": 2688
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5936530274702606,
+      "learning_rate": 4.780550898861147e-07,
+      "loss": 0.9325,
+      "step": 2689
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7157966416372644,
+      "learning_rate": 4.7473295596168536e-07,
+      "loss": 0.8282,
+      "step": 2690
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7036577790580681,
+      "learning_rate": 4.7142212484525995e-07,
+      "loss": 0.8655,
+      "step": 2691
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6195905036509884,
+      "learning_rate": 4.681226004655126e-07,
+      "loss": 0.6474,
+      "step": 2692
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4566977879893652,
+      "learning_rate": 4.6483438673770364e-07,
+      "loss": 0.6012,
+      "step": 2693
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6636467015531696,
+      "learning_rate": 4.61557487563673e-07,
+      "loss": 0.9851,
+      "step": 2694
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5769228662137443,
+      "learning_rate": 4.5829190683183034e-07,
+      "loss": 0.8399,
+      "step": 2695
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7612907694984018,
+      "learning_rate": 4.55037648417157e-07,
+      "loss": 0.8882,
+      "step": 2696
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4006084428464213,
+      "learning_rate": 4.5179471618119973e-07,
+      "loss": 0.8735,
+      "step": 2697
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6610396794302196,
+      "learning_rate": 4.4856311397206364e-07,
+      "loss": 0.6733,
+      "step": 2698
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5808983681927279,
+      "learning_rate": 4.4534284562441356e-07,
+      "loss": 1.0242,
+      "step": 2699
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6725379046131238,
+      "learning_rate": 4.4213391495946054e-07,
+      "loss": 0.7769,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7069060899074552,
+      "learning_rate": 4.3893632578496327e-07,
+      "loss": 0.994,
+      "step": 2701
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6224271667546739,
+      "learning_rate": 4.3575008189522536e-07,
+      "loss": 0.7446,
+      "step": 2702
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.477504241486592,
+      "learning_rate": 4.325751870710848e-07,
+      "loss": 0.5828,
+      "step": 2703
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6476315741302162,
+      "learning_rate": 4.294116450799146e-07,
+      "loss": 0.6922,
+      "step": 2704
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5911032417397631,
+      "learning_rate": 4.2625945967561646e-07,
+      "loss": 0.892,
+      "step": 2705
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7247356224667234,
+      "learning_rate": 4.231186345986116e-07,
+      "loss": 0.9394,
+      "step": 2706
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6682957828075291,
+      "learning_rate": 4.199891735758477e-07,
+      "loss": 0.8786,
+      "step": 2707
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7569252021549989,
+      "learning_rate": 4.168710803207865e-07,
+      "loss": 0.9126,
+      "step": 2708
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.44051300673993765,
+      "learning_rate": 4.137643585333939e-07,
+      "loss": 0.8925,
+      "step": 2709
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6708427521358937,
+      "learning_rate": 4.1066901190015e-07,
+      "loss": 0.8992,
+      "step": 2710
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6642552571679072,
+      "learning_rate": 4.075850440940343e-07,
+      "loss": 0.7437,
+      "step": 2711
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4574771689993498,
+      "learning_rate": 4.0451245877452193e-07,
+      "loss": 0.6677,
+      "step": 2712
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6774825384858724,
+      "learning_rate": 4.014512595875852e-07,
+      "loss": 0.91,
+      "step": 2713
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7329971588889984,
+      "learning_rate": 3.984014501656841e-07,
+      "loss": 0.7945,
+      "step": 2714
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6068186232890825,
+      "learning_rate": 3.953630341277603e-07,
+      "loss": 0.9852,
+      "step": 2715
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7014587468691776,
+      "learning_rate": 3.92336015079241e-07,
+      "loss": 0.7953,
+      "step": 2716
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6312842827156939,
+      "learning_rate": 3.8932039661202514e-07,
+      "loss": 0.7746,
+      "step": 2717
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7434271596114022,
+      "learning_rate": 3.8631618230448696e-07,
+      "loss": 0.8766,
+      "step": 2718
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6227940275555641,
+      "learning_rate": 3.8332337572146606e-07,
+      "loss": 0.7433,
+      "step": 2719
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5025888073879271,
+      "learning_rate": 3.8034198041426715e-07,
+      "loss": 0.6863,
+      "step": 2720
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4098680365008336,
+      "learning_rate": 3.7737199992065154e-07,
+      "loss": 0.8815,
+      "step": 2721
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6875896329207217,
+      "learning_rate": 3.7441343776484116e-07,
+      "loss": 0.9362,
+      "step": 2722
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6891344252308541,
+      "learning_rate": 3.714662974575012e-07,
+      "loss": 0.7752,
+      "step": 2723
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5751356689197045,
+      "learning_rate": 3.6853058249574745e-07,
+      "loss": 0.9346,
+      "step": 2724
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6495055277606431,
+      "learning_rate": 3.6560629636314236e-07,
+      "loss": 0.8132,
+      "step": 2725
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5767576384401792,
+      "learning_rate": 3.62693442529678e-07,
+      "loss": 0.823,
+      "step": 2726
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6761894180987873,
+      "learning_rate": 3.5979202445178716e-07,
+      "loss": 0.8823,
+      "step": 2727
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7604543614462129,
+      "learning_rate": 3.5690204557233157e-07,
+      "loss": 0.997,
+      "step": 2728
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.430404031431052,
+      "learning_rate": 3.540235093205979e-07,
+      "loss": 0.5524,
+      "step": 2729
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.633306207051208,
+      "learning_rate": 3.51156419112294e-07,
+      "loss": 0.7482,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6435556534436399,
+      "learning_rate": 3.4830077834954956e-07,
+      "loss": 0.7744,
+      "step": 2731
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6403180429916674,
+      "learning_rate": 3.4545659042090284e-07,
+      "loss": 0.9451,
+      "step": 2732
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.43448090381199456,
+      "learning_rate": 3.4262385870130866e-07,
+      "loss": 0.9364,
+      "step": 2733
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6095731334751866,
+      "learning_rate": 3.398025865521204e-07,
+      "loss": 0.84,
+      "step": 2734
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6246534239611152,
+      "learning_rate": 3.369927773210979e-07,
+      "loss": 0.8902,
+      "step": 2735
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6878205597790668,
+      "learning_rate": 3.3419443434240083e-07,
+      "loss": 0.7324,
+      "step": 2736
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6252951022646501,
+      "learning_rate": 3.314075609365752e-07,
+      "loss": 0.8838,
+      "step": 2737
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.478298996087301,
+      "learning_rate": 3.2863216041056465e-07,
+      "loss": 0.5735,
+      "step": 2738
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6116015125579515,
+      "learning_rate": 3.25868236057697e-07,
+      "loss": 0.9461,
+      "step": 2739
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7148592293944601,
+      "learning_rate": 3.2311579115767987e-07,
+      "loss": 0.7446,
+      "step": 2740
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6319344251909502,
+      "learning_rate": 3.2037482897660066e-07,
+      "loss": 0.8716,
+      "step": 2741
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.629911488225966,
+      "learning_rate": 3.1764535276692435e-07,
+      "loss": 0.8093,
+      "step": 2742
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7385834710495195,
+      "learning_rate": 3.14927365767479e-07,
+      "loss": 0.7249,
+      "step": 2743
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7548585491240999,
+      "learning_rate": 3.122208712034669e-07,
+      "loss": 0.9447,
+      "step": 2744
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4218659926487407,
+      "learning_rate": 3.0952587228645024e-07,
+      "loss": 0.8898,
+      "step": 2745
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6950243354609579,
+      "learning_rate": 3.0684237221434877e-07,
+      "loss": 1.0493,
+      "step": 2746
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.42879735119311024,
+      "learning_rate": 3.041703741714408e-07,
+      "loss": 0.6849,
+      "step": 2747
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6638515867211157,
+      "learning_rate": 3.015098813283546e-07,
+      "loss": 0.7959,
+      "step": 2748
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6162084845040902,
+      "learning_rate": 2.9886089684206477e-07,
+      "loss": 0.8821,
+      "step": 2749
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7707559886354782,
+      "learning_rate": 2.9622342385589256e-07,
+      "loss": 0.7284,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5803549283798569,
+      "learning_rate": 2.935974654995e-07,
+      "loss": 0.7371,
+      "step": 2751
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6481130574882834,
+      "learning_rate": 2.9098302488888007e-07,
+      "loss": 0.8122,
+      "step": 2752
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6410827204924112,
+      "learning_rate": 2.883801051263657e-07,
+      "loss": 0.9161,
+      "step": 2753
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6147675292861918,
+      "learning_rate": 2.85788709300614e-07,
+      "loss": 0.7772,
+      "step": 2754
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6925023152233675,
+      "learning_rate": 2.8320884048661067e-07,
+      "loss": 0.8147,
+      "step": 2755
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.46882553197085447,
+      "learning_rate": 2.806405017456626e-07,
+      "loss": 0.6331,
+      "step": 2756
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7250248521097109,
+      "learning_rate": 2.780836961253941e-07,
+      "loss": 0.7501,
+      "step": 2757
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.37047158695324345,
+      "learning_rate": 2.7553842665974385e-07,
+      "loss": 0.9894,
+      "step": 2758
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6722318531267583,
+      "learning_rate": 2.7300469636896365e-07,
+      "loss": 0.6866,
+      "step": 2759
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6371864260732003,
+      "learning_rate": 2.7048250825960963e-07,
+      "loss": 1.0002,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6663524080570897,
+      "learning_rate": 2.6797186532454446e-07,
+      "loss": 0.7988,
+      "step": 2761
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7411925293999105,
+      "learning_rate": 2.6547277054293276e-07,
+      "loss": 0.7927,
+      "step": 2762
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6596386405601605,
+      "learning_rate": 2.629852268802291e-07,
+      "loss": 0.9272,
+      "step": 2763
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7205376916360662,
+      "learning_rate": 2.605092372881879e-07,
+      "loss": 0.7001,
+      "step": 2764
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.47260696194795665,
+      "learning_rate": 2.580448047048523e-07,
+      "loss": 0.6553,
+      "step": 2765
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6964148651376042,
+      "learning_rate": 2.555919320545486e-07,
+      "loss": 0.6888,
+      "step": 2766
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6372132725116776,
+      "learning_rate": 2.531506222478886e-07,
+      "loss": 0.7029,
+      "step": 2767
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5818980582618639,
+      "learning_rate": 2.507208781817638e-07,
+      "loss": 0.9016,
+      "step": 2768
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6837081725933732,
+      "learning_rate": 2.483027027393403e-07,
+      "loss": 0.804,
+      "step": 2769
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6798417220556955,
+      "learning_rate": 2.458960987900549e-07,
+      "loss": 1.0129,
+      "step": 2770
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.42619321870905774,
+      "learning_rate": 2.4350106918962e-07,
+      "loss": 0.9606,
+      "step": 2771
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6774208925531973,
+      "learning_rate": 2.4111761678000555e-07,
+      "loss": 0.8871,
+      "step": 2772
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.45733886668083634,
+      "learning_rate": 2.387457443894503e-07,
+      "loss": 0.5118,
+      "step": 2773
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6375798856389312,
+      "learning_rate": 2.3638545483244735e-07,
+      "loss": 0.7971,
+      "step": 2774
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5929836766164001,
+      "learning_rate": 2.3403675090974852e-07,
+      "loss": 0.7712,
+      "step": 2775
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6341016416640781,
+      "learning_rate": 2.3169963540835783e-07,
+      "loss": 0.7831,
+      "step": 2776
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6625166029359167,
+      "learning_rate": 2.2937411110152574e-07,
+      "loss": 1.0075,
+      "step": 2777
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6789436197454038,
+      "learning_rate": 2.2706018074875046e-07,
+      "loss": 0.7543,
+      "step": 2778
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6662936832304609,
+      "learning_rate": 2.2475784709577453e-07,
+      "loss": 0.7924,
+      "step": 2779
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6807032119412719,
+      "learning_rate": 2.224671128745759e-07,
+      "loss": 0.8617,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5760687202748829,
+      "learning_rate": 2.201879808033691e-07,
+      "loss": 1.0217,
+      "step": 2781
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.45649238831884403,
+      "learning_rate": 2.179204535866053e-07,
+      "loss": 0.6272,
+      "step": 2782
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.648896414472766,
+      "learning_rate": 2.156645339149588e-07,
+      "loss": 0.7717,
+      "step": 2783
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.659331125157595,
+      "learning_rate": 2.1342022446533606e-07,
+      "loss": 0.9296,
+      "step": 2784
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.41714337343577523,
+      "learning_rate": 2.111875279008657e-07,
+      "loss": 0.9223,
+      "step": 2785
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6046816914374384,
+      "learning_rate": 2.0896644687089074e-07,
+      "loss": 0.7071,
+      "step": 2786
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7784452751040165,
+      "learning_rate": 2.0675698401097956e-07,
+      "loss": 0.9674,
+      "step": 2787
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5653341979102041,
+      "learning_rate": 2.0455914194290827e-07,
+      "loss": 0.599,
+      "step": 2788
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6468711211464515,
+      "learning_rate": 2.0237292327466406e-07,
+      "loss": 0.863,
+      "step": 2789
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6294006316916326,
+      "learning_rate": 2.0019833060044512e-07,
+      "loss": 0.8016,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6678494944450495,
+      "learning_rate": 1.980353665006507e-07,
+      "loss": 0.7513,
+      "step": 2791
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4713149662646807,
+      "learning_rate": 1.9588403354188324e-07,
+      "loss": 0.6309,
+      "step": 2792
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.639595793871571,
+      "learning_rate": 1.9374433427694405e-07,
+      "loss": 0.8648,
+      "step": 2793
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6103531351131821,
+      "learning_rate": 1.9161627124482552e-07,
+      "loss": 1.0296,
+      "step": 2794
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7810511579484624,
+      "learning_rate": 1.8949984697071767e-07,
+      "loss": 0.7808,
+      "step": 2795
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.40342822528959404,
+      "learning_rate": 1.8739506396599717e-07,
+      "loss": 0.9239,
+      "step": 2796
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6330021049719263,
+      "learning_rate": 1.8530192472822728e-07,
+      "loss": 0.8269,
+      "step": 2797
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6882870189056222,
+      "learning_rate": 1.8322043174115457e-07,
+      "loss": 0.7346,
+      "step": 2798
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.617287584776734,
+      "learning_rate": 1.811505874747066e-07,
+      "loss": 0.8622,
+      "step": 2799
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4786209902163879,
+      "learning_rate": 1.790923943849876e-07,
+      "loss": 0.5372,
+      "step": 2800
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.676670562167987,
+      "learning_rate": 1.770458549142773e-07,
+      "loss": 0.9448,
+      "step": 2801
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.694784558269377,
+      "learning_rate": 1.750109714910264e-07,
+      "loss": 0.7711,
+      "step": 2802
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6987870555917367,
+      "learning_rate": 1.729877465298535e-07,
+      "loss": 1.0072,
+      "step": 2803
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6729206382041002,
+      "learning_rate": 1.709761824315448e-07,
+      "loss": 0.955,
+      "step": 2804
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5697785083856342,
+      "learning_rate": 1.6897628158304979e-07,
+      "loss": 0.9169,
+      "step": 2805
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5782593258996184,
+      "learning_rate": 1.669880463574758e-07,
+      "loss": 0.9289,
+      "step": 2806
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9472757351124677,
+      "learning_rate": 1.6501147911409e-07,
+      "loss": 0.7126,
+      "step": 2807
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.39445752517142363,
+      "learning_rate": 1.6304658219831292e-07,
+      "loss": 0.9233,
+      "step": 2808
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.481021013939613,
+      "learning_rate": 1.6109335794171733e-07,
+      "loss": 0.5763,
+      "step": 2809
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6346577561107241,
+      "learning_rate": 1.5915180866202585e-07,
+      "loss": 0.7073,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5712109354642245,
+      "learning_rate": 1.572219366631045e-07,
+      "loss": 0.7762,
+      "step": 2811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6206037702703069,
+      "learning_rate": 1.5530374423496696e-07,
+      "loss": 0.7308,
+      "step": 2812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6179239220014808,
+      "learning_rate": 1.533972336537648e-07,
+      "loss": 0.9535,
+      "step": 2813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6886724077818089,
+      "learning_rate": 1.5150240718178833e-07,
+      "loss": 0.9104,
+      "step": 2814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7216330549931201,
+      "learning_rate": 1.4961926706746453e-07,
+      "loss": 0.8367,
+      "step": 2815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6901900086787073,
+      "learning_rate": 1.4774781554535377e-07,
+      "loss": 0.8186,
+      "step": 2816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5648219095739478,
+      "learning_rate": 1.4588805483614188e-07,
+      "loss": 0.8683,
+      "step": 2817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.48329217572473715,
+      "learning_rate": 1.440399871466469e-07,
+      "loss": 0.5894,
+      "step": 2818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6725362306724583,
+      "learning_rate": 1.4220361466981248e-07,
+      "loss": 0.7517,
+      "step": 2819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.39778074752235415,
+      "learning_rate": 1.4037893958469994e-07,
+      "loss": 0.9614,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6183955365493483,
+      "learning_rate": 1.3856596405649292e-07,
+      "loss": 0.6468,
+      "step": 2821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6561915360896654,
+      "learning_rate": 1.3676469023649497e-07,
+      "loss": 0.9663,
+      "step": 2822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6451215819533663,
+      "learning_rate": 1.3497512026211857e-07,
+      "loss": 0.7704,
+      "step": 2823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7160201438506223,
+      "learning_rate": 1.3319725625689173e-07,
+      "loss": 0.678,
+      "step": 2824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.608274205537939,
+      "learning_rate": 1.314311003304525e-07,
+      "loss": 0.9381,
+      "step": 2825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4642376539684525,
+      "learning_rate": 1.2967665457854438e-07,
+      "loss": 0.553,
+      "step": 2826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.703811904490751,
+      "learning_rate": 1.2793392108301438e-07,
+      "loss": 0.9617,
+      "step": 2827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5582949969846273,
+      "learning_rate": 1.2620290191181606e-07,
+      "loss": 0.8956,
+      "step": 2828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7447016414674031,
+      "learning_rate": 1.2448359911899743e-07,
+      "loss": 0.8934,
+      "step": 2829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6644563353862788,
+      "learning_rate": 1.2277601474470767e-07,
+      "loss": 0.9279,
+      "step": 2830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7355063571641668,
+      "learning_rate": 1.2108015081518598e-07,
+      "loss": 0.7521,
+      "step": 2831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7429098210155048,
+      "learning_rate": 1.193960093427704e-07,
+      "loss": 0.9599,
+      "step": 2832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.43235244663314504,
+      "learning_rate": 1.1772359232588459e-07,
+      "loss": 0.8882,
+      "step": 2833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6444246256437847,
+      "learning_rate": 1.160629017490389e-07,
+      "loss": 0.8891,
+      "step": 2834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6113688246556247,
+      "learning_rate": 1.1441393958283253e-07,
+      "loss": 0.702,
+      "step": 2835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.44482500036073336,
+      "learning_rate": 1.127767077839459e-07,
+      "loss": 0.512,
+      "step": 2836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6646643810463975,
+      "learning_rate": 1.1115120829513825e-07,
+      "loss": 0.9843,
+      "step": 2837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6169262570111745,
+      "learning_rate": 1.0953744304525116e-07,
+      "loss": 0.8124,
+      "step": 2838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7189415502258311,
+      "learning_rate": 1.0793541394919838e-07,
+      "loss": 0.9085,
+      "step": 2839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5823189826561498,
+      "learning_rate": 1.0634512290796816e-07,
+      "loss": 0.9009,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7568530673848188,
+      "learning_rate": 1.0476657180862327e-07,
+      "loss": 0.8331,
+      "step": 2841
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6537837942343872,
+      "learning_rate": 1.0319976252429086e-07,
+      "loss": 0.9555,
+      "step": 2842
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6554930513372669,
+      "learning_rate": 1.0164469691416934e-07,
+      "loss": 0.806,
+      "step": 2843
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.618364916481457,
+      "learning_rate": 1.0010137682351928e-07,
+      "loss": 0.878,
+      "step": 2844
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3717327314215282,
+      "learning_rate": 9.856980408366468e-08,
+      "loss": 0.773,
+      "step": 2845
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6268937749894009,
+      "learning_rate": 9.704998051199066e-08,
+      "loss": 0.9424,
+      "step": 2846
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6987570884716062,
+      "learning_rate": 9.554190791193907e-08,
+      "loss": 0.6843,
+      "step": 2847
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6717332881621485,
+      "learning_rate": 9.404558807301067e-08,
+      "loss": 0.8884,
+      "step": 2848
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5860214124360067,
+      "learning_rate": 9.256102277075519e-08,
+      "loss": 0.7992,
+      "step": 2849
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.719981664262449,
+      "learning_rate": 9.108821376678123e-08,
+      "loss": 0.8559,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.678776481891488,
+      "learning_rate": 8.962716280874084e-08,
+      "loss": 0.9224,
+      "step": 2851
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6026831108980989,
+      "learning_rate": 8.817787163033609e-08,
+      "loss": 0.8931,
+      "step": 2852
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4679706761570336,
+      "learning_rate": 8.674034195131798e-08,
+      "loss": 0.6316,
+      "step": 2853
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6101586989811233,
+      "learning_rate": 8.531457547747534e-08,
+      "loss": 0.8413,
+      "step": 2854
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.741399992900386,
+      "learning_rate": 8.390057390064266e-08,
+      "loss": 0.7703,
+      "step": 2855
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6342508002606986,
+      "learning_rate": 8.24983388986933e-08,
+      "loss": 0.8618,
+      "step": 2856
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6573970207295808,
+      "learning_rate": 8.11078721355374e-08,
+      "loss": 0.7099,
+      "step": 2857
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.38208000509839696,
+      "learning_rate": 7.972917526112178e-08,
+      "loss": 0.9723,
+      "step": 2858
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.696584972128348,
+      "learning_rate": 7.83622499114256e-08,
+      "loss": 0.7937,
+      "step": 2859
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5517784559032325,
+      "learning_rate": 7.700709770846138e-08,
+      "loss": 0.7872,
+      "step": 2860
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8223360302889021,
+      "learning_rate": 7.566372026026947e-08,
+      "loss": 0.852,
+      "step": 2861
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.46668488011626924,
+      "learning_rate": 7.433211916092143e-08,
+      "loss": 0.4437,
+      "step": 2862
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.556866963118925,
+      "learning_rate": 7.301229599050997e-08,
+      "loss": 0.9776,
+      "step": 2863
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6194546402262693,
+      "learning_rate": 7.170425231515676e-08,
+      "loss": 0.7389,
+      "step": 2864
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6743468247459843,
+      "learning_rate": 7.040798968700357e-08,
+      "loss": 0.7804,
+      "step": 2865
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6192123029541217,
+      "learning_rate": 6.912350964421221e-08,
+      "loss": 0.7914,
+      "step": 2866
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6820063125296908,
+      "learning_rate": 6.785081371096458e-08,
+      "loss": 0.8819,
+      "step": 2867
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6775703398872709,
+      "learning_rate": 6.658990339745708e-08,
+      "loss": 0.9315,
+      "step": 2868
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.42153719697739683,
+      "learning_rate": 6.534078019990397e-08,
+      "loss": 0.9669,
+      "step": 2869
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5901632525886075,
+      "learning_rate": 6.410344560053072e-08,
+      "loss": 0.8499,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.45504041196806017,
+      "learning_rate": 6.287790106757396e-08,
+      "loss": 0.4683,
+      "step": 2871
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7276613079632562,
+      "learning_rate": 6.166414805528265e-08,
+      "loss": 0.8116,
+      "step": 2872
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5487383989346268,
+      "learning_rate": 6.046218800391245e-08,
+      "loss": 0.8477,
+      "step": 2873
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5799726074211679,
+      "learning_rate": 5.9272022339723624e-08,
+      "loss": 0.852,
+      "step": 2874
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6280937161547762,
+      "learning_rate": 5.809365247498311e-08,
+      "loss": 0.9784,
+      "step": 2875
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6524772458441905,
+      "learning_rate": 5.69270798079613e-08,
+      "loss": 0.661,
+      "step": 2876
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7447718740649082,
+      "learning_rate": 5.5772305722926467e-08,
+      "loss": 0.8684,
+      "step": 2877
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6314891372382665,
+      "learning_rate": 5.462933159015027e-08,
+      "loss": 0.6394,
+      "step": 2878
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4766882096151558,
+      "learning_rate": 5.3498158765901145e-08,
+      "loss": 0.6213,
+      "step": 2879
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6491322450131797,
+      "learning_rate": 5.2378788592443167e-08,
+      "loss": 1.0398,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6472597209421643,
+      "learning_rate": 5.1271222398036056e-08,
+      "loss": 0.7765,
+      "step": 2881
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.754282016569779,
+      "learning_rate": 5.017546149693298e-08,
+      "loss": 0.8842,
+      "step": 2882
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4224556052972046,
+      "learning_rate": 4.9091507189377163e-08,
+      "loss": 0.884,
+      "step": 2883
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7077920528209953,
+      "learning_rate": 4.8019360761604186e-08,
+      "loss": 0.8005,
+      "step": 2884
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6345615170917235,
+      "learning_rate": 4.695902348583636e-08,
+      "loss": 1.0124,
+      "step": 2885
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5912646143736082,
+      "learning_rate": 4.591049662028502e-08,
+      "loss": 0.7869,
+      "step": 2886
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7536096415700662,
+      "learning_rate": 4.487378140914489e-08,
+      "loss": 1.0422,
+      "step": 2887
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6866313647389813,
+      "learning_rate": 4.384887908259638e-08,
+      "loss": 0.736,
+      "step": 2888
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4417799780114287,
+      "learning_rate": 4.283579085680223e-08,
+      "loss": 0.5431,
+      "step": 2889
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7088013995743894,
+      "learning_rate": 4.183451793390747e-08,
+      "loss": 0.8389,
+      "step": 2890
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6257354262628168,
+      "learning_rate": 4.0845061502033935e-08,
+      "loss": 0.9114,
+      "step": 2891
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7006892817767445,
+      "learning_rate": 3.9867422735283544e-08,
+      "loss": 0.971,
+      "step": 2892
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6749074861962011,
+      "learning_rate": 3.8901602793738334e-08,
+      "loss": 0.8201,
+      "step": 2893
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7446497214590111,
+      "learning_rate": 3.7947602823449335e-08,
+      "loss": 0.9682,
+      "step": 2894
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.42089092323034294,
+      "learning_rate": 3.7005423956447686e-08,
+      "loss": 0.8677,
+      "step": 2895
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6327603959636279,
+      "learning_rate": 3.607506731073573e-08,
+      "loss": 0.9201,
+      "step": 2896
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.611026387027496,
+      "learning_rate": 3.515653399028596e-08,
+      "loss": 0.7515,
+      "step": 2897
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4248284427441181,
+      "learning_rate": 3.4249825085040936e-08,
+      "loss": 0.6372,
+      "step": 2898
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.721190650486805,
+      "learning_rate": 3.3354941670916687e-08,
+      "loss": 0.998,
+      "step": 2899
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7800041341338213,
+      "learning_rate": 3.247188480979158e-08,
+      "loss": 0.8065,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5929198922122161,
+      "learning_rate": 3.160065554951186e-08,
+      "loss": 0.9427,
+      "step": 2901
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6944203101581948,
+      "learning_rate": 3.074125492389168e-08,
+      "loss": 0.8301,
+      "step": 2902
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6940141599237891,
+      "learning_rate": 2.989368395270531e-08,
+      "loss": 0.6854,
+      "step": 2903
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7355094977435371,
+      "learning_rate": 2.9057943641693787e-08,
+      "loss": 0.9381,
+      "step": 2904
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6487757131570335,
+      "learning_rate": 2.823403498255606e-08,
+      "loss": 0.7327,
+      "step": 2905
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4753912478971177,
+      "learning_rate": 2.742195895295341e-08,
+      "loss": 0.6633,
+      "step": 2906
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.42836117922759304,
+      "learning_rate": 2.6621716516506137e-08,
+      "loss": 0.8891,
+      "step": 2907
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7187318271297163,
+      "learning_rate": 2.583330862279243e-08,
+      "loss": 0.9345,
+      "step": 2908
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6576347684725014,
+      "learning_rate": 2.5056736207348387e-08,
+      "loss": 0.767,
+      "step": 2909
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5598688050393787,
+      "learning_rate": 2.4292000191663555e-08,
+      "loss": 0.9216,
+      "step": 2910
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6570279630532985,
+      "learning_rate": 2.353910148318428e-08,
+      "loss": 0.8365,
+      "step": 2911
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6232284093114551,
+      "learning_rate": 2.2798040975310353e-08,
+      "loss": 0.8244,
+      "step": 2912
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6915230971607258,
+      "learning_rate": 2.2068819547395038e-08,
+      "loss": 1.0132,
+      "step": 2913
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7401799373175983,
+      "learning_rate": 2.1351438064740605e-08,
+      "loss": 0.7586,
+      "step": 2914
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.41615527883213305,
+      "learning_rate": 2.0645897378601677e-08,
+      "loss": 0.5368,
+      "step": 2915
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6157536625752755,
+      "learning_rate": 1.9952198326183004e-08,
+      "loss": 0.7649,
+      "step": 2916
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6770304592385578,
+      "learning_rate": 1.9270341730636134e-08,
+      "loss": 0.8374,
+      "step": 2917
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6673397187075967,
+      "learning_rate": 1.860032840106163e-08,
+      "loss": 0.9115,
+      "step": 2918
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.42543557916684266,
+      "learning_rate": 1.7942159132505742e-08,
+      "loss": 0.9507,
+      "step": 2919
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6151540077702607,
+      "learning_rate": 1.7295834705961524e-08,
+      "loss": 0.7829,
+      "step": 2920
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5927525302980846,
+      "learning_rate": 1.6661355888366593e-08,
+      "loss": 0.9172,
+      "step": 2921
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.665321262016628,
+      "learning_rate": 1.6038723432599823e-08,
+      "loss": 0.8663,
+      "step": 2922
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6057651992239808,
+      "learning_rate": 1.5427938077486883e-08,
+      "loss": 0.8994,
+      "step": 2923
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.46511756762607387,
+      "learning_rate": 1.4829000547793571e-08,
+      "loss": 0.5187,
+      "step": 2924
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5781429719084783,
+      "learning_rate": 1.424191155422583e-08,
+      "loss": 0.9358,
+      "step": 2925
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6979772848321701,
+      "learning_rate": 1.3666671793433062e-08,
+      "loss": 0.7484,
+      "step": 2926
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6647028081138089,
+      "learning_rate": 1.3103281948001478e-08,
+      "loss": 0.8805,
+      "step": 2927
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6132293172138021,
+      "learning_rate": 1.2551742686456313e-08,
+      "loss": 0.7759,
+      "step": 2928
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7409289833279389,
+      "learning_rate": 1.201205466326294e-08,
+      "loss": 0.7509,
+      "step": 2929
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.679398218790593,
+      "learning_rate": 1.1484218518820201e-08,
+      "loss": 0.9519,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4003966888472103,
+      "learning_rate": 1.0968234879468187e-08,
+      "loss": 0.9304,
+      "step": 2931
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6280169608524709,
+      "learning_rate": 1.0464104357477133e-08,
+      "loss": 0.9087,
+      "step": 2932
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.41158555869789276,
+      "learning_rate": 9.971827551056302e-09,
+      "loss": 0.6626,
+      "step": 2933
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6556659368295924,
+      "learning_rate": 9.491405044348422e-09,
+      "loss": 0.8303,
+      "step": 2934
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6365756710336434,
+      "learning_rate": 9.022837407427487e-09,
+      "loss": 0.914,
+      "step": 2935
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6915204267577063,
+      "learning_rate": 8.566125196304287e-09,
+      "loss": 0.6983,
+      "step": 2936
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6084125342418436,
+      "learning_rate": 8.121268952917538e-09,
+      "loss": 0.8037,
+      "step": 2937
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.678406503470794,
+      "learning_rate": 7.688269205141653e-09,
+      "loss": 0.854,
+      "step": 2938
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6628968947602799,
+      "learning_rate": 7.267126466777852e-09,
+      "loss": 0.9249,
+      "step": 2939
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6191879046390241,
+      "learning_rate": 6.857841237560836e-09,
+      "loss": 0.7979,
+      "step": 2940
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6214357408687838,
+      "learning_rate": 6.460414003153226e-09,
+      "loss": 0.7615,
+      "step": 2941
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.475865401506941,
+      "learning_rate": 6.074845235147786e-09,
+      "loss": 0.6003,
+      "step": 2942
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7259863482080148,
+      "learning_rate": 5.701135391065205e-09,
+      "loss": 0.7326,
+      "step": 2943
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.383050825624809,
+      "learning_rate": 5.339284914355203e-09,
+      "loss": 1.0064,
+      "step": 2944
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6445123080978313,
+      "learning_rate": 4.989294234393205e-09,
+      "loss": 0.6662,
+      "step": 2945
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6302355566312514,
+      "learning_rate": 4.651163766484779e-09,
+      "loss": 0.9903,
+      "step": 2946
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.627181791202635,
+      "learning_rate": 4.324893911857864e-09,
+      "loss": 0.7568,
+      "step": 2947
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7096901414988926,
+      "learning_rate": 4.010485057669433e-09,
+      "loss": 0.8181,
+      "step": 2948
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.616762602191851,
+      "learning_rate": 3.707937577002163e-09,
+      "loss": 0.8852,
+      "step": 2949
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6934966504021275,
+      "learning_rate": 3.4172518288611014e-09,
+      "loss": 0.6968,
+      "step": 2950
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4762016413627739,
+      "learning_rate": 3.1384281581792187e-09,
+      "loss": 0.6945,
+      "step": 2951
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6577078074376902,
+      "learning_rate": 2.8714668958118585e-09,
+      "loss": 0.6563,
+      "step": 2952
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6247347077636631,
+      "learning_rate": 2.6163683585389565e-09,
+      "loss": 0.7383,
+      "step": 2953
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5949953013558139,
+      "learning_rate": 2.3731328490628202e-09,
+      "loss": 0.9131,
+      "step": 2954
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6101812883768953,
+      "learning_rate": 2.1417606560114603e-09,
+      "loss": 0.9002,
+      "step": 2955
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.675049506910943,
+      "learning_rate": 1.922252053933038e-09,
+      "loss": 1.046,
+      "step": 2956
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4083524819589728,
+      "learning_rate": 1.7146073033003085e-09,
+      "loss": 0.9161,
+      "step": 2957
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6958884503195137,
+      "learning_rate": 1.5188266505061776e-09,
+      "loss": 0.8736,
+      "step": 2958
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4782538684344312,
+      "learning_rate": 1.3349103278681441e-09,
+      "loss": 0.4921,
+      "step": 2959
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6237705069178578,
+      "learning_rate": 1.1628585536216374e-09,
+      "loss": 0.8007,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5987475966677738,
+      "learning_rate": 1.0026715319277901e-09,
+      "loss": 0.7841,
+      "step": 2961
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6375999675680211,
+      "learning_rate": 8.543494528645557e-10,
+      "loss": 0.7274,
+      "step": 2962
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.699100753592795,
+      "learning_rate": 7.178924924344799e-10,
+      "loss": 1.0252,
+      "step": 2963
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6757832012895416,
+      "learning_rate": 5.933008125580397e-10,
+      "loss": 0.7733,
+      "step": 2964
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.644801693316542,
+      "learning_rate": 4.805745610791945e-10,
+      "loss": 0.8005,
+      "step": 2965
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7130954726842352,
+      "learning_rate": 3.797138717576143e-10,
+      "loss": 0.8401,
+      "step": 2966
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6021958518191793,
+      "learning_rate": 2.907188642786718e-10,
+      "loss": 1.0276,
+      "step": 2967
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4744821165388851,
+      "learning_rate": 2.135896442434504e-10,
+      "loss": 0.6668,
+      "step": 2968
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6168695714316085,
+      "learning_rate": 1.4832630317429542e-10,
+      "loss": 0.7051,
+      "step": 2969
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7209944533988882,
+      "learning_rate": 9.49289185148139e-11,
+      "loss": 0.9535,
+      "step": 2970
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4114004576699493,
+      "learning_rate": 5.339755362543386e-11,
+      "loss": 0.9038,
+      "step": 2971
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6705506754519056,
+      "learning_rate": 2.3732257790065606e-11,
+      "loss": 0.6987,
+      "step": 2972
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7274581937947537,
+      "learning_rate": 5.933066207219895e-12,
+      "loss": 0.92,
+      "step": 2973
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5602166217212989,
+      "learning_rate": 0.0,
+      "loss": 0.5062,
+      "step": 2974
+    },
+    {
+      "epoch": 1.0,
+      "step": 2974,
+      "total_flos": 0.0,
+      "train_loss": 0.06582018055607988,
+      "train_runtime": 4302.496,
+      "train_samples_per_second": 1417.485,
+      "train_steps_per_second": 0.691
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2974,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}