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": 6.020898030295759,
+      "learning_rate": 1.1111111111111112e-07,
+      "loss": 1.7306,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.5093131371851625,
+      "learning_rate": 2.2222222222222224e-07,
+      "loss": 1.5296,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.555254520927707,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 1.6453,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.167517210604746,
+      "learning_rate": 4.444444444444445e-07,
+      "loss": 1.3601,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.337865578358168,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 1.5936,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.197537612384531,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 1.6105,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.7071750793416705,
+      "learning_rate": 7.777777777777779e-07,
+      "loss": 1.5346,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.867444838666045,
+      "learning_rate": 8.88888888888889e-07,
+      "loss": 1.4384,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.024004355148772,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.5733,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.751318881465759,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 1.5568,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.3853621546980115,
+      "learning_rate": 1.2222222222222223e-06,
+      "loss": 1.3708,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.537387205190727,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 1.3072,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.23139313367372,
+      "learning_rate": 1.4444444444444445e-06,
+      "loss": 1.5981,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.4018488285384687,
+      "learning_rate": 1.5555555555555558e-06,
+      "loss": 1.4173,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.684797940945863,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.3716,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.646303872392676,
+      "learning_rate": 1.777777777777778e-06,
+      "loss": 1.4545,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9495758829048704,
+      "learning_rate": 1.888888888888889e-06,
+      "loss": 1.397,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.366263720251264,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.365,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.839976774580383,
+      "learning_rate": 2.1111111111111114e-06,
+      "loss": 1.185,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6342538540467102,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 1.1806,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6599785518630754,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": 1.2685,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8448167155078814,
+      "learning_rate": 2.4444444444444447e-06,
+      "loss": 1.0219,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7512589943688497,
+      "learning_rate": 2.5555555555555557e-06,
+      "loss": 1.2719,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.156554314663689,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.2205,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5304531912685935,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 1.3731,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2500838449139604,
+      "learning_rate": 2.888888888888889e-06,
+      "loss": 1.1428,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5082983802919083,
+      "learning_rate": 3e-06,
+      "loss": 1.3047,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6177261527447628,
+      "learning_rate": 3.1111111111111116e-06,
+      "loss": 1.1996,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8360713710189462,
+      "learning_rate": 3.2222222222222227e-06,
+      "loss": 1.222,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8430975591369023,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.0663,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.046310489866483,
+      "learning_rate": 3.444444444444445e-06,
+      "loss": 1.1678,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.100826125003991,
+      "learning_rate": 3.555555555555556e-06,
+      "loss": 1.2384,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0663202598141646,
+      "learning_rate": 3.6666666666666666e-06,
+      "loss": 1.0626,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.037675341882923,
+      "learning_rate": 3.777777777777778e-06,
+      "loss": 1.3272,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2020884108777876,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 1.0534,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8529417568151625,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.9597,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3783066806716602,
+      "learning_rate": 4.111111111111111e-06,
+      "loss": 1.2806,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.110763970915099,
+      "learning_rate": 4.222222222222223e-06,
+      "loss": 0.9401,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.126179766156456,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 1.2566,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6527927342936612,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 1.2164,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9560977610225554,
+      "learning_rate": 4.555555555555556e-06,
+      "loss": 1.1596,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2602017040522464,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 1.1942,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4946904787680995,
+      "learning_rate": 4.777777777777778e-06,
+      "loss": 1.0445,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6820020626040015,
+      "learning_rate": 4.888888888888889e-06,
+      "loss": 1.4199,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2768139856497662,
+      "learning_rate": 5e-06,
+      "loss": 1.1108,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3685452034608794,
+      "learning_rate": 5.1111111111111115e-06,
+      "loss": 1.1317,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2467870655438416,
+      "learning_rate": 5.2222222222222226e-06,
+      "loss": 1.1229,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8015736433677734,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.8661,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.314129615480521,
+      "learning_rate": 5.444444444444445e-06,
+      "loss": 1.2102,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.376665180773933,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 0.991,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3130460130080732,
+      "learning_rate": 5.666666666666667e-06,
+      "loss": 1.2305,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.248188336808312,
+      "learning_rate": 5.777777777777778e-06,
+      "loss": 1.191,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.752120755859964,
+      "learning_rate": 5.88888888888889e-06,
+      "loss": 1.1772,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4409867828492302,
+      "learning_rate": 6e-06,
+      "loss": 1.0935,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.255521837436117,
+      "learning_rate": 6.111111111111112e-06,
+      "loss": 1.0643,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2462979329005954,
+      "learning_rate": 6.222222222222223e-06,
+      "loss": 1.1121,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1597916496517735,
+      "learning_rate": 6.333333333333333e-06,
+      "loss": 0.9677,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1925051092384764,
+      "learning_rate": 6.444444444444445e-06,
+      "loss": 1.2934,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.45780692652238,
+      "learning_rate": 6.555555555555556e-06,
+      "loss": 0.9344,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3151211868470905,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.2318,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.160434549338062,
+      "learning_rate": 6.777777777777779e-06,
+      "loss": 0.9053,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4310724645048005,
+      "learning_rate": 6.88888888888889e-06,
+      "loss": 0.9287,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5199960623786288,
+      "learning_rate": 7e-06,
+      "loss": 1.2231,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1805906278311316,
+      "learning_rate": 7.111111111111112e-06,
+      "loss": 1.1331,
+      "step": 64
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0271975049666155,
+      "learning_rate": 7.222222222222223e-06,
+      "loss": 0.902,
+      "step": 65
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.819359187837601,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 1.0091,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.56607881386993,
+      "learning_rate": 7.444444444444445e-06,
+      "loss": 0.9476,
+      "step": 67
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2723117056000988,
+      "learning_rate": 7.555555555555556e-06,
+      "loss": 1.1569,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.337812598379636,
+      "learning_rate": 7.666666666666667e-06,
+      "loss": 1.0866,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1037532780574288,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 1.1321,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1769510996262655,
+      "learning_rate": 7.88888888888889e-06,
+      "loss": 1.0507,
+      "step": 71
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1890068962308549,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.1234,
+      "step": 72
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2677899369635037,
+      "learning_rate": 8.111111111111112e-06,
+      "loss": 1.0256,
+      "step": 73
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.356848273679716,
+      "learning_rate": 8.222222222222222e-06,
+      "loss": 0.8915,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7648679960952123,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.9863,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8409602605912934,
+      "learning_rate": 8.444444444444446e-06,
+      "loss": 1.0207,
+      "step": 76
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1684827594433063,
+      "learning_rate": 8.555555555555556e-06,
+      "loss": 1.0476,
+      "step": 77
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1683705523644832,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.8686,
+      "step": 78
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7530105675212482,
+      "learning_rate": 8.777777777777778e-06,
+      "loss": 1.0336,
+      "step": 79
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4445099321125248,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 1.1212,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9730426161163872,
+      "learning_rate": 9e-06,
+      "loss": 1.1248,
+      "step": 81
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.220619936560779,
+      "learning_rate": 9.111111111111112e-06,
+      "loss": 1.0278,
+      "step": 82
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.992570916328277,
+      "learning_rate": 9.222222222222224e-06,
+      "loss": 0.7983,
+      "step": 83
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.721692166786328,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.9634,
+      "step": 84
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.197755462250134,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 1.0935,
+      "step": 85
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3792147570294744,
+      "learning_rate": 9.555555555555556e-06,
+      "loss": 1.0437,
+      "step": 86
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3341305179213792,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 1.1169,
+      "step": 87
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.080347584342399,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 0.8965,
+      "step": 88
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3008940654674894,
+      "learning_rate": 9.88888888888889e-06,
+      "loss": 1.1245,
+      "step": 89
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3413740345111518,
+      "learning_rate": 1e-05,
+      "loss": 0.9169,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.720904819282236,
+      "learning_rate": 9.999997033466898e-06,
+      "loss": 0.8597,
+      "step": 91
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.817693345659933,
+      "learning_rate": 9.999988133871106e-06,
+      "loss": 1.1579,
+      "step": 92
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4862565637632583,
+      "learning_rate": 9.999973301223189e-06,
+      "loss": 0.938,
+      "step": 93
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8793598827617122,
+      "learning_rate": 9.999952535540743e-06,
+      "loss": 1.0742,
+      "step": 94
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9677823570429186,
+      "learning_rate": 9.999925836848413e-06,
+      "loss": 1.0403,
+      "step": 95
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6567718982405544,
+      "learning_rate": 9.99989320517788e-06,
+      "loss": 1.1472,
+      "step": 96
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8356421924262722,
+      "learning_rate": 9.999854640567861e-06,
+      "loss": 1.0842,
+      "step": 97
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3949535007725533,
+      "learning_rate": 9.999810143064122e-06,
+      "loss": 0.9233,
+      "step": 98
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3621910185308956,
+      "learning_rate": 9.999759712719461e-06,
+      "loss": 1.1597,
+      "step": 99
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3165455793194516,
+      "learning_rate": 9.999703349593721e-06,
+      "loss": 0.8999,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.626913349027703,
+      "learning_rate": 9.999641053753784e-06,
+      "loss": 0.9357,
+      "step": 101
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5306786549204892,
+      "learning_rate": 9.999572825273569e-06,
+      "loss": 1.0193,
+      "step": 102
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2216374912201793,
+      "learning_rate": 9.999498664234038e-06,
+      "loss": 1.0678,
+      "step": 103
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5684013544753055,
+      "learning_rate": 9.999418570723189e-06,
+      "loss": 1.1503,
+      "step": 104
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0556370851103432,
+      "learning_rate": 9.999332544836066e-06,
+      "loss": 0.9618,
+      "step": 105
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7871835168527705,
+      "learning_rate": 9.999240586674749e-06,
+      "loss": 1.1076,
+      "step": 106
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0185577076236143,
+      "learning_rate": 9.999142696348351e-06,
+      "loss": 0.8433,
+      "step": 107
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6532580327579758,
+      "learning_rate": 9.999038873973034e-06,
+      "loss": 1.099,
+      "step": 108
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9493570256045315,
+      "learning_rate": 9.998929119671996e-06,
+      "loss": 1.0014,
+      "step": 109
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3968958234582853,
+      "learning_rate": 9.99881343357547e-06,
+      "loss": 0.7398,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2787796104443074,
+      "learning_rate": 9.998691815820732e-06,
+      "loss": 1.1522,
+      "step": 111
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2153670054233487,
+      "learning_rate": 9.998564266552096e-06,
+      "loss": 0.8671,
+      "step": 112
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1083265194148346,
+      "learning_rate": 9.998430785920912e-06,
+      "loss": 1.0972,
+      "step": 113
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6242510943614903,
+      "learning_rate": 9.99829137408557e-06,
+      "loss": 0.9546,
+      "step": 114
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0222089455842167,
+      "learning_rate": 9.9981460312115e-06,
+      "loss": 1.0223,
+      "step": 115
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3186828064280087,
+      "learning_rate": 9.997994757471165e-06,
+      "loss": 0.8844,
+      "step": 116
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7083714252921345,
+      "learning_rate": 9.997837553044072e-06,
+      "loss": 0.8308,
+      "step": 117
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0640704563932113,
+      "learning_rate": 9.997674418116759e-06,
+      "loss": 0.9205,
+      "step": 118
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2234653048015862,
+      "learning_rate": 9.997505352882803e-06,
+      "loss": 0.9909,
+      "step": 119
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.196102935955767,
+      "learning_rate": 9.997330357542823e-06,
+      "loss": 1.0755,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5665366051605698,
+      "learning_rate": 9.997149432304468e-06,
+      "loss": 0.9899,
+      "step": 121
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8863028691261228,
+      "learning_rate": 9.996962577382428e-06,
+      "loss": 1.0132,
+      "step": 122
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3272031835909146,
+      "learning_rate": 9.996769792998424e-06,
+      "loss": 0.9504,
+      "step": 123
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3543972412613536,
+      "learning_rate": 9.996571079381222e-06,
+      "loss": 0.9564,
+      "step": 124
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.959881310588133,
+      "learning_rate": 9.996366436766612e-06,
+      "loss": 1.148,
+      "step": 125
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.767725635901644,
+      "learning_rate": 9.99615586539743e-06,
+      "loss": 0.9383,
+      "step": 126
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3252133962182393,
+      "learning_rate": 9.995939365523542e-06,
+      "loss": 0.7415,
+      "step": 127
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2425723054330782,
+      "learning_rate": 9.99571693740185e-06,
+      "loss": 0.979,
+      "step": 128
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.124642698198707,
+      "learning_rate": 9.995488581296287e-06,
+      "loss": 0.9189,
+      "step": 129
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2777449663358862,
+      "learning_rate": 9.995254297477827e-06,
+      "loss": 1.0288,
+      "step": 130
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7753697791616395,
+      "learning_rate": 9.995014086224473e-06,
+      "loss": 1.0237,
+      "step": 131
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6520124236533547,
+      "learning_rate": 9.994767947821261e-06,
+      "loss": 1.081,
+      "step": 132
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1262038527744473,
+      "learning_rate": 9.994515882560267e-06,
+      "loss": 1.0065,
+      "step": 133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.175415679181585,
+      "learning_rate": 9.99425789074059e-06,
+      "loss": 0.9947,
+      "step": 134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.69246795089213,
+      "learning_rate": 9.99399397266837e-06,
+      "loss": 0.9607,
+      "step": 135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0205633422989124,
+      "learning_rate": 9.993724128656772e-06,
+      "loss": 0.681,
+      "step": 136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.075325446095766,
+      "learning_rate": 9.993448359026e-06,
+      "loss": 1.0778,
+      "step": 137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7501176970520742,
+      "learning_rate": 9.993166664103283e-06,
+      "loss": 0.8985,
+      "step": 138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0342699371811452,
+      "learning_rate": 9.992879044222887e-06,
+      "loss": 0.9397,
+      "step": 139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.778398106495321,
+      "learning_rate": 9.992585499726104e-06,
+      "loss": 0.9627,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7449536230740645,
+      "learning_rate": 9.992286030961257e-06,
+      "loss": 0.8455,
+      "step": 141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.277868994981918,
+      "learning_rate": 9.9919806382837e-06,
+      "loss": 1.0861,
+      "step": 142
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9852189021540414,
+      "learning_rate": 9.991669322055818e-06,
+      "loss": 1.0277,
+      "step": 143
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0610257566280477,
+      "learning_rate": 9.99135208264702e-06,
+      "loss": 1.0823,
+      "step": 144
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.001238969026088,
+      "learning_rate": 9.991028920433747e-06,
+      "loss": 0.746,
+      "step": 145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9329013346269643,
+      "learning_rate": 9.99069983579947e-06,
+      "loss": 0.9651,
+      "step": 146
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3388898507923237,
+      "learning_rate": 9.990364829134682e-06,
+      "loss": 0.9675,
+      "step": 147
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8302543451924096,
+      "learning_rate": 9.990023900836909e-06,
+      "loss": 0.8925,
+      "step": 148
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8805427678747366,
+      "learning_rate": 9.9896770513107e-06,
+      "loss": 0.9304,
+      "step": 149
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7019163874309788,
+      "learning_rate": 9.98932428096763e-06,
+      "loss": 0.957,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4809320058897433,
+      "learning_rate": 9.988965590226303e-06,
+      "loss": 1.0363,
+      "step": 151
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1864570217558925,
+      "learning_rate": 9.988600979512345e-06,
+      "loss": 0.9558,
+      "step": 152
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.134709924917348,
+      "learning_rate": 9.988230449258409e-06,
+      "loss": 0.9881,
+      "step": 153
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0704584402312147,
+      "learning_rate": 9.987853999904169e-06,
+      "loss": 0.6569,
+      "step": 154
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1892081565545338,
+      "learning_rate": 9.987471631896326e-06,
+      "loss": 1.0141,
+      "step": 155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0209922851221758,
+      "learning_rate": 9.987083345688604e-06,
+      "loss": 1.078,
+      "step": 156
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.475176089477989,
+      "learning_rate": 9.986689141741748e-06,
+      "loss": 0.9046,
+      "step": 157
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.558717901289885,
+      "learning_rate": 9.986289020523525e-06,
+      "loss": 1.0373,
+      "step": 158
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.152047193857759,
+      "learning_rate": 9.985882982508723e-06,
+      "loss": 0.9382,
+      "step": 159
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.507976501632076,
+      "learning_rate": 9.985471028179155e-06,
+      "loss": 0.9003,
+      "step": 160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2349219766814303,
+      "learning_rate": 9.985053158023647e-06,
+      "loss": 1.0314,
+      "step": 161
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0134033494622083,
+      "learning_rate": 9.984629372538055e-06,
+      "loss": 0.8469,
+      "step": 162
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2362367211971406,
+      "learning_rate": 9.984199672225246e-06,
+      "loss": 0.7923,
+      "step": 163
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.971708412609744,
+      "learning_rate": 9.983764057595105e-06,
+      "loss": 0.8672,
+      "step": 164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6211625921811355,
+      "learning_rate": 9.983322529164543e-06,
+      "loss": 0.8829,
+      "step": 165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8141433692136149,
+      "learning_rate": 9.98287508745748e-06,
+      "loss": 1.0359,
+      "step": 166
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2789308931823613,
+      "learning_rate": 9.982421733004857e-06,
+      "loss": 0.9886,
+      "step": 167
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.647520427026211,
+      "learning_rate": 9.981962466344633e-06,
+      "loss": 1.1532,
+      "step": 168
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2634153609569436,
+      "learning_rate": 9.981497288021778e-06,
+      "loss": 0.9887,
+      "step": 169
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6324340480418404,
+      "learning_rate": 9.981026198588276e-06,
+      "loss": 1.0668,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0686068956409525,
+      "learning_rate": 9.980549198603132e-06,
+      "loss": 0.6617,
+      "step": 171
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5411742986741408,
+      "learning_rate": 9.98006628863236e-06,
+      "loss": 0.9345,
+      "step": 172
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.236842587604399,
+      "learning_rate": 9.979577469248984e-06,
+      "loss": 0.8768,
+      "step": 173
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3712043970777765,
+      "learning_rate": 9.979082741033047e-06,
+      "loss": 0.8838,
+      "step": 174
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5152752781378522,
+      "learning_rate": 9.9785821045716e-06,
+      "loss": 1.1285,
+      "step": 175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.028389636493706,
+      "learning_rate": 9.978075560458703e-06,
+      "loss": 0.8638,
+      "step": 176
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8745400738197981,
+      "learning_rate": 9.977563109295429e-06,
+      "loss": 1.0348,
+      "step": 177
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6428521355980752,
+      "learning_rate": 9.977044751689858e-06,
+      "loss": 0.7875,
+      "step": 178
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3079870980534036,
+      "learning_rate": 9.976520488257082e-06,
+      "loss": 1.0996,
+      "step": 179
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1281539988864875,
+      "learning_rate": 9.975990319619199e-06,
+      "loss": 0.7946,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3276287051521578,
+      "learning_rate": 9.975454246405312e-06,
+      "loss": 0.8549,
+      "step": 181
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.107249581715156,
+      "learning_rate": 9.974912269251534e-06,
+      "loss": 1.1309,
+      "step": 182
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.427331128768283,
+      "learning_rate": 9.974364388800983e-06,
+      "loss": 1.0148,
+      "step": 183
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.894947622409324,
+      "learning_rate": 9.97381060570378e-06,
+      "loss": 0.9642,
+      "step": 184
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.306718327302065,
+      "learning_rate": 9.97325092061705e-06,
+      "loss": 0.8674,
+      "step": 185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4475386646631423,
+      "learning_rate": 9.972685334204926e-06,
+      "loss": 0.8515,
+      "step": 186
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4958541755642323,
+      "learning_rate": 9.972113847138538e-06,
+      "loss": 0.9773,
+      "step": 187
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.578331568212937,
+      "learning_rate": 9.971536460096021e-06,
+      "loss": 0.946,
+      "step": 188
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.911849955521383,
+      "learning_rate": 9.97095317376251e-06,
+      "loss": 1.1086,
+      "step": 189
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.186402837062041,
+      "learning_rate": 9.970363988830138e-06,
+      "loss": 0.6713,
+      "step": 190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8199481317125474,
+      "learning_rate": 9.969768905998045e-06,
+      "loss": 0.9944,
+      "step": 191
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3772574279376883,
+      "learning_rate": 9.96916792597236e-06,
+      "loss": 1.1249,
+      "step": 192
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6997394329250959,
+      "learning_rate": 9.968561049466214e-06,
+      "loss": 0.8626,
+      "step": 193
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.135885039108074,
+      "learning_rate": 9.967948277199736e-06,
+      "loss": 1.0192,
+      "step": 194
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8654051692311207,
+      "learning_rate": 9.96732960990005e-06,
+      "loss": 0.9099,
+      "step": 195
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.232300589296793,
+      "learning_rate": 9.966705048301272e-06,
+      "loss": 1.0594,
+      "step": 196
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.769839080229836,
+      "learning_rate": 9.96607459314452e-06,
+      "loss": 0.8557,
+      "step": 197
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1819394019822527,
+      "learning_rate": 9.965438245177895e-06,
+      "loss": 0.6591,
+      "step": 198
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6561491251489833,
+      "learning_rate": 9.964796005156498e-06,
+      "loss": 1.0654,
+      "step": 199
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.076566588095923,
+      "learning_rate": 9.964147873842423e-06,
+      "loss": 0.9098,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8274859585049659,
+      "learning_rate": 9.963493852004746e-06,
+      "loss": 1.145,
+      "step": 201
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4743831365425886,
+      "learning_rate": 9.96283394041954e-06,
+      "loss": 0.9402,
+      "step": 202
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2860404212872454,
+      "learning_rate": 9.962168139869866e-06,
+      "loss": 1.0003,
+      "step": 203
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.49537957580469,
+      "learning_rate": 9.96149645114577e-06,
+      "loss": 1.0238,
+      "step": 204
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2340714868980769,
+      "learning_rate": 9.960818875044288e-06,
+      "loss": 1.002,
+      "step": 205
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1284485998257747,
+      "learning_rate": 9.960135412369441e-06,
+      "loss": 0.7313,
+      "step": 206
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5107322710886535,
+      "learning_rate": 9.959446063932232e-06,
+      "loss": 0.906,
+      "step": 207
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1417931537333665,
+      "learning_rate": 9.958750830550654e-06,
+      "loss": 0.9505,
+      "step": 208
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.998294508212294,
+      "learning_rate": 9.95804971304968e-06,
+      "loss": 0.8158,
+      "step": 209
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3031863351309383,
+      "learning_rate": 9.957342712261263e-06,
+      "loss": 0.9033,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.159538509547293,
+      "learning_rate": 9.956629829024342e-06,
+      "loss": 1.025,
+      "step": 211
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1441035816389082,
+      "learning_rate": 9.955911064184834e-06,
+      "loss": 1.0276,
+      "step": 212
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.767131494064827,
+      "learning_rate": 9.95518641859563e-06,
+      "loss": 0.9236,
+      "step": 213
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9853973295182781,
+      "learning_rate": 9.95445589311661e-06,
+      "loss": 0.9335,
+      "step": 214
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9369895446898371,
+      "learning_rate": 9.953719488614625e-06,
+      "loss": 1.0355,
+      "step": 215
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9099470372481967,
+      "learning_rate": 9.952977205963496e-06,
+      "loss": 0.6765,
+      "step": 216
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.381014312258799,
+      "learning_rate": 9.95222904604403e-06,
+      "loss": 0.8395,
+      "step": 217
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8632387427554324,
+      "learning_rate": 9.951475009744006e-06,
+      "loss": 1.0484,
+      "step": 218
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7781729755901865,
+      "learning_rate": 9.950715097958168e-06,
+      "loss": 0.8061,
+      "step": 219
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.96291547717403,
+      "learning_rate": 9.94994931158824e-06,
+      "loss": 1.0707,
+      "step": 220
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.281205157857246,
+      "learning_rate": 9.949177651542917e-06,
+      "loss": 0.8391,
+      "step": 221
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.478232699581741,
+      "learning_rate": 9.948400118737856e-06,
+      "loss": 0.796,
+      "step": 222
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0899418794298212,
+      "learning_rate": 9.94761671409569e-06,
+      "loss": 1.0265,
+      "step": 223
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0519575701736021,
+      "learning_rate": 9.946827438546016e-06,
+      "loss": 0.6633,
+      "step": 224
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8963658903319036,
+      "learning_rate": 9.946032293025402e-06,
+      "loss": 1.0057,
+      "step": 225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1268256551939204,
+      "learning_rate": 9.945231278477374e-06,
+      "loss": 0.9888,
+      "step": 226
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2345005310746198,
+      "learning_rate": 9.944424395852432e-06,
+      "loss": 1.0428,
+      "step": 227
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3889968987892425,
+      "learning_rate": 9.943611646108029e-06,
+      "loss": 0.8652,
+      "step": 228
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7984713289052676,
+      "learning_rate": 9.942793030208585e-06,
+      "loss": 0.8959,
+      "step": 229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6705014873155208,
+      "learning_rate": 9.941968549125481e-06,
+      "loss": 1.0733,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.924048026846623,
+      "learning_rate": 9.94113820383706e-06,
+      "loss": 0.9425,
+      "step": 231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6798412708247372,
+      "learning_rate": 9.940301995328616e-06,
+      "loss": 1.0229,
+      "step": 232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4837906003622596,
+      "learning_rate": 9.939459924592407e-06,
+      "loss": 0.8519,
+      "step": 233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8453585396403996,
+      "learning_rate": 9.938611992627647e-06,
+      "loss": 0.706,
+      "step": 234
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.287422434119563,
+      "learning_rate": 9.937758200440501e-06,
+      "loss": 0.9349,
+      "step": 235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6643902508785695,
+      "learning_rate": 9.936898549044093e-06,
+      "loss": 0.8115,
+      "step": 236
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5928803368755373,
+      "learning_rate": 9.936033039458494e-06,
+      "loss": 1.0133,
+      "step": 237
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3850069675398193,
+      "learning_rate": 9.935161672710728e-06,
+      "loss": 0.9817,
+      "step": 238
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5136027116578243,
+      "learning_rate": 9.934284449834775e-06,
+      "loss": 1.025,
+      "step": 239
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3948037542727116,
+      "learning_rate": 9.933401371871555e-06,
+      "loss": 0.902,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.640860727920044,
+      "learning_rate": 9.932512439868942e-06,
+      "loss": 0.9375,
+      "step": 241
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9323786426178224,
+      "learning_rate": 9.931617654881753e-06,
+      "loss": 0.9494,
+      "step": 242
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7538884342353285,
+      "learning_rate": 9.930717017971753e-06,
+      "loss": 0.8419,
+      "step": 243
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5711238928017217,
+      "learning_rate": 9.929810530207651e-06,
+      "loss": 1.11,
+      "step": 244
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8098648973860392,
+      "learning_rate": 9.928898192665095e-06,
+      "loss": 0.8543,
+      "step": 245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1077144466714448,
+      "learning_rate": 9.927980006426677e-06,
+      "loss": 1.0343,
+      "step": 246
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6357636865901277,
+      "learning_rate": 9.92705597258193e-06,
+      "loss": 0.7851,
+      "step": 247
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1909097105252235,
+      "learning_rate": 9.926126092227323e-06,
+      "loss": 0.8327,
+      "step": 248
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4432654295316647,
+      "learning_rate": 9.925190366466269e-06,
+      "loss": 1.094,
+      "step": 249
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1908048978426817,
+      "learning_rate": 9.924248796409107e-06,
+      "loss": 0.9988,
+      "step": 250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4896957767057013,
+      "learning_rate": 9.923301383173119e-06,
+      "loss": 0.7244,
+      "step": 251
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.699079544790394,
+      "learning_rate": 9.922348127882518e-06,
+      "loss": 0.8914,
+      "step": 252
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.63780151346983,
+      "learning_rate": 9.921389031668449e-06,
+      "loss": 0.8846,
+      "step": 253
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3880782385542485,
+      "learning_rate": 9.920424095668988e-06,
+      "loss": 0.975,
+      "step": 254
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.065532767415578,
+      "learning_rate": 9.919453321029141e-06,
+      "loss": 1.0056,
+      "step": 255
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0707929088580674,
+      "learning_rate": 9.918476708900843e-06,
+      "loss": 0.9698,
+      "step": 256
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.431740771547493,
+      "learning_rate": 9.917494260442956e-06,
+      "loss": 0.8802,
+      "step": 257
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3342704081217773,
+      "learning_rate": 9.916505976821262e-06,
+      "loss": 0.9038,
+      "step": 258
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8467579544129362,
+      "learning_rate": 9.915511859208475e-06,
+      "loss": 0.9163,
+      "step": 259
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7320990520832737,
+      "learning_rate": 9.914511908784227e-06,
+      "loss": 0.7735,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3796335963935416,
+      "learning_rate": 9.913506126735074e-06,
+      "loss": 0.9029,
+      "step": 261
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6835648055465686,
+      "learning_rate": 9.912494514254487e-06,
+      "loss": 0.8729,
+      "step": 262
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1839176968599987,
+      "learning_rate": 9.911477072542861e-06,
+      "loss": 1.0187,
+      "step": 263
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.640338495063523,
+      "learning_rate": 9.910453802807506e-06,
+      "loss": 0.7708,
+      "step": 264
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8871037342726975,
+      "learning_rate": 9.909424706262647e-06,
+      "loss": 0.9192,
+      "step": 265
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8139780230677538,
+      "learning_rate": 9.908389784129424e-06,
+      "loss": 0.9958,
+      "step": 266
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.728058018472273,
+      "learning_rate": 9.907349037635888e-06,
+      "loss": 1.0385,
+      "step": 267
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9086036063238607,
+      "learning_rate": 9.906302468017002e-06,
+      "loss": 0.9161,
+      "step": 268
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6114989125313772,
+      "learning_rate": 9.905250076514642e-06,
+      "loss": 0.6475,
+      "step": 269
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3762741058906207,
+      "learning_rate": 9.904191864377588e-06,
+      "loss": 1.1156,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.086390607075079,
+      "learning_rate": 9.90312783286153e-06,
+      "loss": 0.8221,
+      "step": 271
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5740262812144985,
+      "learning_rate": 9.902057983229059e-06,
+      "loss": 0.9528,
+      "step": 272
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.50666432486656,
+      "learning_rate": 9.900982316749676e-06,
+      "loss": 1.0109,
+      "step": 273
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6309283564461712,
+      "learning_rate": 9.899900834699778e-06,
+      "loss": 0.7297,
+      "step": 274
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6838445889239224,
+      "learning_rate": 9.898813538362669e-06,
+      "loss": 0.9022,
+      "step": 275
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6680955746681156,
+      "learning_rate": 9.897720429028547e-06,
+      "loss": 0.9331,
+      "step": 276
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0756880620121314,
+      "learning_rate": 9.89662150799451e-06,
+      "loss": 0.7475,
+      "step": 277
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.136706594615024,
+      "learning_rate": 9.895516776564555e-06,
+      "loss": 0.9863,
+      "step": 278
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0572881228437248,
+      "learning_rate": 9.894406236049569e-06,
+      "loss": 0.8198,
+      "step": 279
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8180647894989632,
+      "learning_rate": 9.893289887767331e-06,
+      "loss": 1.0312,
+      "step": 280
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1338441221347955,
+      "learning_rate": 9.89216773304252e-06,
+      "loss": 0.9669,
+      "step": 281
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.418049798401187,
+      "learning_rate": 9.891039773206698e-06,
+      "loss": 1.017,
+      "step": 282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.768470259542854,
+      "learning_rate": 9.889906009598316e-06,
+      "loss": 1.0251,
+      "step": 283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.658461872275104,
+      "learning_rate": 9.888766443562712e-06,
+      "loss": 0.8274,
+      "step": 284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7970413119919844,
+      "learning_rate": 9.887621076452114e-06,
+      "loss": 1.0827,
+      "step": 285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3605969891368064,
+      "learning_rate": 9.886469909625624e-06,
+      "loss": 0.7672,
+      "step": 286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1032964256853734,
+      "learning_rate": 9.885312944449239e-06,
+      "loss": 0.6565,
+      "step": 287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3870357241696727,
+      "learning_rate": 9.884150182295821e-06,
+      "loss": 1.0235,
+      "step": 288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.26168882123742,
+      "learning_rate": 9.882981624545127e-06,
+      "loss": 1.0208,
+      "step": 289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.09649325436949,
+      "learning_rate": 9.881807272583776e-06,
+      "loss": 0.9187,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4556902223024557,
+      "learning_rate": 9.880627127805276e-06,
+      "loss": 0.8593,
+      "step": 291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8014947127925711,
+      "learning_rate": 9.879441191609997e-06,
+      "loss": 0.9749,
+      "step": 292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8290301702334593,
+      "learning_rate": 9.87824946540519e-06,
+      "loss": 0.8511,
+      "step": 293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4625527319337002,
+      "learning_rate": 9.877051950604972e-06,
+      "loss": 1.0078,
+      "step": 294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1680384151001364,
+      "learning_rate": 9.87584864863033e-06,
+      "loss": 0.9407,
+      "step": 295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6130682789955968,
+      "learning_rate": 9.874639560909118e-06,
+      "loss": 0.7385,
+      "step": 296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8507225700170138,
+      "learning_rate": 9.873424688876057e-06,
+      "loss": 0.8806,
+      "step": 297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.606513761011154,
+      "learning_rate": 9.872204033972727e-06,
+      "loss": 0.7909,
+      "step": 298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0498835238448727,
+      "learning_rate": 9.870977597647575e-06,
+      "loss": 1.0263,
+      "step": 299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2838274716183784,
+      "learning_rate": 9.869745381355906e-06,
+      "loss": 0.8487,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4135658926148096,
+      "learning_rate": 9.868507386559887e-06,
+      "loss": 0.9428,
+      "step": 301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8282516051562494,
+      "learning_rate": 9.867263614728535e-06,
+      "loss": 0.8022,
+      "step": 302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5671918054431404,
+      "learning_rate": 9.866014067337729e-06,
+      "loss": 0.9707,
+      "step": 303
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5015006133561526,
+      "learning_rate": 9.864758745870196e-06,
+      "loss": 0.7557,
+      "step": 304
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.694032530156948,
+      "learning_rate": 9.863497651815519e-06,
+      "loss": 0.9421,
+      "step": 305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2660115821949285,
+      "learning_rate": 9.862230786670129e-06,
+      "loss": 0.9859,
+      "step": 306
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.776176415672431,
+      "learning_rate": 9.860958151937303e-06,
+      "loss": 0.9903,
+      "step": 307
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9698974117011222,
+      "learning_rate": 9.85967974912717e-06,
+      "loss": 0.9531,
+      "step": 308
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7306546660935704,
+      "learning_rate": 9.858395579756695e-06,
+      "loss": 0.818,
+      "step": 309
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5378493088313332,
+      "learning_rate": 9.857105645349694e-06,
+      "loss": 0.93,
+      "step": 310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9868363561873847,
+      "learning_rate": 9.855809947436819e-06,
+      "loss": 1.1493,
+      "step": 311
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1962522712967694,
+      "learning_rate": 9.854508487555562e-06,
+      "loss": 0.7852,
+      "step": 312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9909144916652012,
+      "learning_rate": 9.853201267250252e-06,
+      "loss": 0.6818,
+      "step": 313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0696694649160716,
+      "learning_rate": 9.851888288072053e-06,
+      "loss": 0.9477,
+      "step": 314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0046874799403294,
+      "learning_rate": 9.850569551578969e-06,
+      "loss": 1.0127,
+      "step": 315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8327101643147752,
+      "learning_rate": 9.849245059335824e-06,
+      "loss": 0.9942,
+      "step": 316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4875784073689209,
+      "learning_rate": 9.847914812914282e-06,
+      "loss": 0.8581,
+      "step": 317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7277337829063115,
+      "learning_rate": 9.846578813892827e-06,
+      "loss": 0.9338,
+      "step": 318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6115308035833913,
+      "learning_rate": 9.845237063856777e-06,
+      "loss": 0.9224,
+      "step": 319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.650709097120753,
+      "learning_rate": 9.843889564398269e-06,
+      "loss": 0.9348,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.031783328039755,
+      "learning_rate": 9.842536317116262e-06,
+      "loss": 0.9041,
+      "step": 321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0241770821372147,
+      "learning_rate": 9.841177323616539e-06,
+      "loss": 0.6927,
+      "step": 322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.154682248097513,
+      "learning_rate": 9.8398125855117e-06,
+      "loss": 1.0406,
+      "step": 323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.288943940967939,
+      "learning_rate": 9.838442104421162e-06,
+      "loss": 0.8838,
+      "step": 324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4263909174801925,
+      "learning_rate": 9.837065881971153e-06,
+      "loss": 0.925,
+      "step": 325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.081150451702571,
+      "learning_rate": 9.835683919794719e-06,
+      "loss": 0.8502,
+      "step": 326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.724855282062143,
+      "learning_rate": 9.834296219531713e-06,
+      "loss": 0.9246,
+      "step": 327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.747634148370311,
+      "learning_rate": 9.832902782828801e-06,
+      "loss": 0.9765,
+      "step": 328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8182794900307626,
+      "learning_rate": 9.831503611339452e-06,
+      "loss": 0.9935,
+      "step": 329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9396062649839036,
+      "learning_rate": 9.83009870672394e-06,
+      "loss": 1.0744,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2542831466335191,
+      "learning_rate": 9.828688070649346e-06,
+      "loss": 0.6068,
+      "step": 331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.919105483365787,
+      "learning_rate": 9.82727170478955e-06,
+      "loss": 0.9518,
+      "step": 332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.148299339216061,
+      "learning_rate": 9.825849610825227e-06,
+      "loss": 0.9806,
+      "step": 333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3008558239223045,
+      "learning_rate": 9.824421790443855e-06,
+      "loss": 1.0473,
+      "step": 334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.528500506176747,
+      "learning_rate": 9.822988245339701e-06,
+      "loss": 0.7412,
+      "step": 335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9746260164403284,
+      "learning_rate": 9.821548977213834e-06,
+      "loss": 0.9016,
+      "step": 336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.396888023011961,
+      "learning_rate": 9.820103987774108e-06,
+      "loss": 0.9455,
+      "step": 337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.441669462873289,
+      "learning_rate": 9.818653278735163e-06,
+      "loss": 0.839,
+      "step": 338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9561609802751962,
+      "learning_rate": 9.81719685181843e-06,
+      "loss": 0.707,
+      "step": 339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4887756664350493,
+      "learning_rate": 9.815734708752126e-06,
+      "loss": 0.8669,
+      "step": 340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9113019813878273,
+      "learning_rate": 9.81426685127125e-06,
+      "loss": 1.0096,
+      "step": 341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7955834893505451,
+      "learning_rate": 9.81279328111758e-06,
+      "loss": 1.0499,
+      "step": 342
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.7312601742716485,
+      "learning_rate": 9.811314000039674e-06,
+      "loss": 0.8725,
+      "step": 343
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0525761578078434,
+      "learning_rate": 9.809829009792868e-06,
+      "loss": 1.0188,
+      "step": 344
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5844767245599929,
+      "learning_rate": 9.808338312139267e-06,
+      "loss": 0.8821,
+      "step": 345
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.129863624516024,
+      "learning_rate": 9.806841908847758e-06,
+      "loss": 1.0471,
+      "step": 346
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8181713596251912,
+      "learning_rate": 9.805339801693988e-06,
+      "loss": 0.9231,
+      "step": 347
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8601507619448951,
+      "learning_rate": 9.80383199246038e-06,
+      "loss": 0.8636,
+      "step": 348
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1432210489943007,
+      "learning_rate": 9.802318482936121e-06,
+      "loss": 0.7001,
+      "step": 349
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.018611922712681,
+      "learning_rate": 9.800799274917159e-06,
+      "loss": 0.8655,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3402611956653754,
+      "learning_rate": 9.799274370206207e-06,
+      "loss": 0.9545,
+      "step": 351
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5563460657938841,
+      "learning_rate": 9.79774377061274e-06,
+      "loss": 0.8218,
+      "step": 352
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.391530342329827,
+      "learning_rate": 9.796207477952984e-06,
+      "loss": 0.9464,
+      "step": 353
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2953341782625847,
+      "learning_rate": 9.794665494049926e-06,
+      "loss": 1.0932,
+      "step": 354
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1845085417724233,
+      "learning_rate": 9.793117820733304e-06,
+      "loss": 0.9574,
+      "step": 355
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.645608598329667,
+      "learning_rate": 9.791564459839609e-06,
+      "loss": 1.014,
+      "step": 356
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4561658177427725,
+      "learning_rate": 9.790005413212075e-06,
+      "loss": 0.6713,
+      "step": 357
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6309249408489093,
+      "learning_rate": 9.788440682700695e-06,
+      "loss": 0.9646,
+      "step": 358
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.68575635715683,
+      "learning_rate": 9.786870270162193e-06,
+      "loss": 0.7148,
+      "step": 359
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6123481246750455,
+      "learning_rate": 9.785294177460043e-06,
+      "loss": 0.9058,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7360508178969465,
+      "learning_rate": 9.783712406464459e-06,
+      "loss": 1.0463,
+      "step": 361
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.067736944526222,
+      "learning_rate": 9.782124959052388e-06,
+      "loss": 0.8575,
+      "step": 362
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5309489406318155,
+      "learning_rate": 9.780531837107519e-06,
+      "loss": 0.9516,
+      "step": 363
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.493444259938855,
+      "learning_rate": 9.77893304252027e-06,
+      "loss": 0.9484,
+      "step": 364
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0759418737704407,
+      "learning_rate": 9.777328577187795e-06,
+      "loss": 0.9026,
+      "step": 365
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8583204411325285,
+      "learning_rate": 9.775718443013969e-06,
+      "loss": 0.7457,
+      "step": 366
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.574646803833619,
+      "learning_rate": 9.774102641909402e-06,
+      "loss": 0.9058,
+      "step": 367
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.320922005721891,
+      "learning_rate": 9.772481175791422e-06,
+      "loss": 1.013,
+      "step": 368
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.481711566653038,
+      "learning_rate": 9.770854046584085e-06,
+      "loss": 0.9677,
+      "step": 369
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8800989233021865,
+      "learning_rate": 9.769221256218165e-06,
+      "loss": 0.929,
+      "step": 370
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8258773251736202,
+      "learning_rate": 9.767582806631148e-06,
+      "loss": 0.7623,
+      "step": 371
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.296101962611138,
+      "learning_rate": 9.765938699767245e-06,
+      "loss": 0.789,
+      "step": 372
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.057979586025814,
+      "learning_rate": 9.76428893757737e-06,
+      "loss": 0.9489,
+      "step": 373
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8367860417723987,
+      "learning_rate": 9.762633522019159e-06,
+      "loss": 0.9672,
+      "step": 374
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1663859183262386,
+      "learning_rate": 9.760972455056943e-06,
+      "loss": 0.9456,
+      "step": 375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.30527024534985,
+      "learning_rate": 9.75930573866177e-06,
+      "loss": 0.5971,
+      "step": 376
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3861028830308717,
+      "learning_rate": 9.75763337481139e-06,
+      "loss": 0.9794,
+      "step": 377
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7344194444541416,
+      "learning_rate": 9.755955365490246e-06,
+      "loss": 1.0426,
+      "step": 378
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6079433181471146,
+      "learning_rate": 9.754271712689491e-06,
+      "loss": 0.8199,
+      "step": 379
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6633791894930898,
+      "learning_rate": 9.752582418406969e-06,
+      "loss": 1.013,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7076892066731,
+      "learning_rate": 9.750887484647215e-06,
+      "loss": 0.8158,
+      "step": 381
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.329265111717286,
+      "learning_rate": 9.749186913421465e-06,
+      "loss": 0.9788,
+      "step": 382
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5735972204730957,
+      "learning_rate": 9.747480706747637e-06,
+      "loss": 0.6201,
+      "step": 383
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.494660920124428,
+      "learning_rate": 9.745768866650339e-06,
+      "loss": 0.9151,
+      "step": 384
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2078144295658277,
+      "learning_rate": 9.744051395160863e-06,
+      "loss": 0.9662,
+      "step": 385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.324400284729826,
+      "learning_rate": 9.742328294317181e-06,
+      "loss": 0.8617,
+      "step": 386
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.722904344979319,
+      "learning_rate": 9.74059956616395e-06,
+      "loss": 1.1377,
+      "step": 387
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8561101869094638,
+      "learning_rate": 9.738865212752505e-06,
+      "loss": 0.9776,
+      "step": 388
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0080385263945146,
+      "learning_rate": 9.737125236140845e-06,
+      "loss": 0.8623,
+      "step": 389
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.347434646805619,
+      "learning_rate": 9.735379638393654e-06,
+      "loss": 0.9776,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4029260064490408,
+      "learning_rate": 9.73362842158228e-06,
+      "loss": 0.9632,
+      "step": 391
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4804049214575348,
+      "learning_rate": 9.73187158778474e-06,
+      "loss": 0.6559,
+      "step": 392
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5548406295290045,
+      "learning_rate": 9.73010913908572e-06,
+      "loss": 0.8577,
+      "step": 393
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7640649519716318,
+      "learning_rate": 9.72834107757656e-06,
+      "loss": 0.958,
+      "step": 394
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3701691748284914,
+      "learning_rate": 9.726567405355268e-06,
+      "loss": 0.7236,
+      "step": 395
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8900703096079545,
+      "learning_rate": 9.724788124526504e-06,
+      "loss": 1.0285,
+      "step": 396
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2808882506780406,
+      "learning_rate": 9.723003237201588e-06,
+      "loss": 0.8586,
+      "step": 397
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.80557161829214,
+      "learning_rate": 9.721212745498493e-06,
+      "loss": 0.9551,
+      "step": 398
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9964719149457895,
+      "learning_rate": 9.719416651541839e-06,
+      "loss": 0.8969,
+      "step": 399
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8315479173208038,
+      "learning_rate": 9.717614957462892e-06,
+      "loss": 0.9313,
+      "step": 400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1464036616731381,
+      "learning_rate": 9.71580766539957e-06,
+      "loss": 0.9788,
+      "step": 401
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2149322509739002,
+      "learning_rate": 9.713994777496427e-06,
+      "loss": 0.6179,
+      "step": 402
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1719801998701387,
+      "learning_rate": 9.712176295904661e-06,
+      "loss": 0.9532,
+      "step": 403
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.498839263376739,
+      "learning_rate": 9.710352222782104e-06,
+      "loss": 0.919,
+      "step": 404
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.884789515670859,
+      "learning_rate": 9.70852256029323e-06,
+      "loss": 0.7941,
+      "step": 405
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5409104114313388,
+      "learning_rate": 9.706687310609137e-06,
+      "loss": 1.0136,
+      "step": 406
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6980505171270477,
+      "learning_rate": 9.704846475907558e-06,
+      "loss": 0.7619,
+      "step": 407
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.859369368639108,
+      "learning_rate": 9.70300005837285e-06,
+      "loss": 1.0314,
+      "step": 408
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5280563044953512,
+      "learning_rate": 9.701148060196001e-06,
+      "loss": 0.9247,
+      "step": 409
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4072925038970214,
+      "learning_rate": 9.699290483574611e-06,
+      "loss": 0.6182,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2163357441401956,
+      "learning_rate": 9.697427330712908e-06,
+      "loss": 1.0617,
+      "step": 411
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4290665980548192,
+      "learning_rate": 9.695558603821735e-06,
+      "loss": 0.8938,
+      "step": 412
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3222848679185883,
+      "learning_rate": 9.693684305118545e-06,
+      "loss": 1.0186,
+      "step": 413
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4926766388462966,
+      "learning_rate": 9.691804436827409e-06,
+      "loss": 0.8521,
+      "step": 414
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.351335649893509,
+      "learning_rate": 9.689919001178998e-06,
+      "loss": 0.8597,
+      "step": 415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6007001191471868,
+      "learning_rate": 9.6880280004106e-06,
+      "loss": 1.0217,
+      "step": 416
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8397365162223847,
+      "learning_rate": 9.686131436766104e-06,
+      "loss": 0.9312,
+      "step": 417
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6323614123234034,
+      "learning_rate": 9.68422931249599e-06,
+      "loss": 0.9664,
+      "step": 418
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0503966006852647,
+      "learning_rate": 9.682321629857348e-06,
+      "loss": 0.7928,
+      "step": 419
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5686737043014571,
+      "learning_rate": 9.68040839111386e-06,
+      "loss": 0.7949,
+      "step": 420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5669615649354094,
+      "learning_rate": 9.678489598535799e-06,
+      "loss": 0.8645,
+      "step": 421
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.420771860807288,
+      "learning_rate": 9.67656525440003e-06,
+      "loss": 0.7538,
+      "step": 422
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2128781893206555,
+      "learning_rate": 9.674635360990005e-06,
+      "loss": 1.0583,
+      "step": 423
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.932790346200411,
+      "learning_rate": 9.672699920595763e-06,
+      "loss": 0.8767,
+      "step": 424
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.143896946306491,
+      "learning_rate": 9.670758935513922e-06,
+      "loss": 0.9357,
+      "step": 425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.903442805092249,
+      "learning_rate": 9.66881240804768e-06,
+      "loss": 0.9596,
+      "step": 426
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.153932480759634,
+      "learning_rate": 9.666860340506809e-06,
+      "loss": 0.9747,
+      "step": 427
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3037176146426408,
+      "learning_rate": 9.664902735207664e-06,
+      "loss": 0.8013,
+      "step": 428
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7915002089490824,
+      "learning_rate": 9.662939594473164e-06,
+      "loss": 0.8294,
+      "step": 429
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9083857748083946,
+      "learning_rate": 9.660970920632798e-06,
+      "loss": 1.0629,
+      "step": 430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.760151448294159,
+      "learning_rate": 9.658996716022618e-06,
+      "loss": 0.91,
+      "step": 431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3396171020788812,
+      "learning_rate": 9.657016982985242e-06,
+      "loss": 0.9304,
+      "step": 432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4447814445151794,
+      "learning_rate": 9.655031723869848e-06,
+      "loss": 0.7932,
+      "step": 433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6325282964164058,
+      "learning_rate": 9.653040941032173e-06,
+      "loss": 0.8823,
+      "step": 434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5779994573964133,
+      "learning_rate": 9.651044636834501e-06,
+      "loss": 1.0506,
+      "step": 435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0692946886960508,
+      "learning_rate": 9.64904281364568e-06,
+      "loss": 0.7855,
+      "step": 436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0139326936705952,
+      "learning_rate": 9.647035473841093e-06,
+      "loss": 0.9462,
+      "step": 437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3609130287166833,
+      "learning_rate": 9.64502261980268e-06,
+      "loss": 0.9459,
+      "step": 438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7957586786398785,
+      "learning_rate": 9.643004253918919e-06,
+      "loss": 0.9851,
+      "step": 439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8260226737266592,
+      "learning_rate": 9.64098037858483e-06,
+      "loss": 0.8245,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8305861785303663,
+      "learning_rate": 9.638950996201972e-06,
+      "loss": 0.9592,
+      "step": 441
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0926487139223684,
+      "learning_rate": 9.636916109178433e-06,
+      "loss": 1.0396,
+      "step": 442
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.454393077284597,
+      "learning_rate": 9.63487571992884e-06,
+      "loss": 0.7575,
+      "step": 443
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2554606450821124,
+      "learning_rate": 9.632829830874345e-06,
+      "loss": 0.8708,
+      "step": 444
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8095281380535917,
+      "learning_rate": 9.630778444442628e-06,
+      "loss": 0.8765,
+      "step": 445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5420482020325055,
+      "learning_rate": 9.628721563067888e-06,
+      "loss": 0.7536,
+      "step": 446
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8577806917692514,
+      "learning_rate": 9.626659189190852e-06,
+      "loss": 0.8307,
+      "step": 447
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.07494269311885,
+      "learning_rate": 9.62459132525876e-06,
+      "loss": 0.8901,
+      "step": 448
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9116144347331061,
+      "learning_rate": 9.62251797372536e-06,
+      "loss": 0.9003,
+      "step": 449
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.141954411374895,
+      "learning_rate": 9.620439137050927e-06,
+      "loss": 0.8517,
+      "step": 450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0991894413797134,
+      "learning_rate": 9.618354817702231e-06,
+      "loss": 0.9166,
+      "step": 451
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0203595460842225,
+      "learning_rate": 9.616265018152555e-06,
+      "loss": 0.9302,
+      "step": 452
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7154060087813795,
+      "learning_rate": 9.614169740881683e-06,
+      "loss": 1.0046,
+      "step": 453
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8167717196305844,
+      "learning_rate": 9.612068988375898e-06,
+      "loss": 0.6463,
+      "step": 454
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.296136973715655,
+      "learning_rate": 9.609962763127978e-06,
+      "loss": 0.9223,
+      "step": 455
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1708130288398655,
+      "learning_rate": 9.607851067637205e-06,
+      "loss": 0.9098,
+      "step": 456
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.062779318950205,
+      "learning_rate": 9.605733904409335e-06,
+      "loss": 0.8124,
+      "step": 457
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9391193243070961,
+      "learning_rate": 9.603611275956632e-06,
+      "loss": 0.9488,
+      "step": 458
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6144470530400434,
+      "learning_rate": 9.601483184797827e-06,
+      "loss": 0.9924,
+      "step": 459
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2145372924230144,
+      "learning_rate": 9.599349633458145e-06,
+      "loss": 0.7585,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0345751473764033,
+      "learning_rate": 9.597210624469288e-06,
+      "loss": 0.926,
+      "step": 461
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0732840522834757,
+      "learning_rate": 9.595066160369428e-06,
+      "loss": 0.9154,
+      "step": 462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7033469178423276,
+      "learning_rate": 9.592916243703217e-06,
+      "loss": 0.7614,
+      "step": 463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1235957715916827,
+      "learning_rate": 9.590760877021775e-06,
+      "loss": 0.8393,
+      "step": 464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5881886760483408,
+      "learning_rate": 9.588600062882686e-06,
+      "loss": 0.8536,
+      "step": 465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1301657579057316,
+      "learning_rate": 9.586433803850002e-06,
+      "loss": 0.9258,
+      "step": 466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2500511975440693,
+      "learning_rate": 9.584262102494238e-06,
+      "loss": 0.9306,
+      "step": 467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2310329672529101,
+      "learning_rate": 9.582084961392358e-06,
+      "loss": 1.022,
+      "step": 468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4426698232908963,
+      "learning_rate": 9.579902383127791e-06,
+      "loss": 0.9068,
+      "step": 469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2235973400657567,
+      "learning_rate": 9.57771437029041e-06,
+      "loss": 1.0053,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2254720777654586,
+      "learning_rate": 9.575520925476543e-06,
+      "loss": 0.8091,
+      "step": 471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.999227783372363,
+      "learning_rate": 9.573322051288958e-06,
+      "loss": 0.762,
+      "step": 472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6228243021504931,
+      "learning_rate": 9.57111775033687e-06,
+      "loss": 0.6266,
+      "step": 473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0892882468931995,
+      "learning_rate": 9.56890802523593e-06,
+      "loss": 0.9988,
+      "step": 474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1065133287443882,
+      "learning_rate": 9.566692878608229e-06,
+      "loss": 0.974,
+      "step": 475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1355852357942586,
+      "learning_rate": 9.564472313082288e-06,
+      "loss": 0.9236,
+      "step": 476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.288887573970032,
+      "learning_rate": 9.56224633129306e-06,
+      "loss": 0.7766,
+      "step": 477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.628628876442106,
+      "learning_rate": 9.560014935881924e-06,
+      "loss": 1.0039,
+      "step": 478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9560744805586425,
+      "learning_rate": 9.557778129496684e-06,
+      "loss": 0.8118,
+      "step": 479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3795656345397382,
+      "learning_rate": 9.555535914791566e-06,
+      "loss": 0.9782,
+      "step": 480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2561230954757427,
+      "learning_rate": 9.553288294427205e-06,
+      "loss": 0.9582,
+      "step": 481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0457951845675297,
+      "learning_rate": 9.551035271070665e-06,
+      "loss": 0.7109,
+      "step": 482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9731405996997973,
+      "learning_rate": 9.548776847395408e-06,
+      "loss": 0.8886,
+      "step": 483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5665377130807763,
+      "learning_rate": 9.546513026081311e-06,
+      "loss": 0.7656,
+      "step": 484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1344864829695132,
+      "learning_rate": 9.544243809814657e-06,
+      "loss": 1.0445,
+      "step": 485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5802330023522864,
+      "learning_rate": 9.541969201288123e-06,
+      "loss": 0.8193,
+      "step": 486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3450065200754924,
+      "learning_rate": 9.539689203200794e-06,
+      "loss": 0.9397,
+      "step": 487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6406332952064706,
+      "learning_rate": 9.537403818258143e-06,
+      "loss": 0.8097,
+      "step": 488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2222115333109085,
+      "learning_rate": 9.53511304917204e-06,
+      "loss": 1.0885,
+      "step": 489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.940353623823162,
+      "learning_rate": 9.532816898660742e-06,
+      "loss": 0.6368,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4326375205582773,
+      "learning_rate": 9.530515369448889e-06,
+      "loss": 0.8942,
+      "step": 491
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.215136960616669,
+      "learning_rate": 9.528208464267508e-06,
+      "loss": 0.9776,
+      "step": 492
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6168651420873563,
+      "learning_rate": 9.525896185854005e-06,
+      "loss": 1.0623,
+      "step": 493
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.478014054178834,
+      "learning_rate": 9.523578536952155e-06,
+      "loss": 0.8342,
+      "step": 494
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4567686232154315,
+      "learning_rate": 9.521255520312117e-06,
+      "loss": 0.8359,
+      "step": 495
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.307014568742023,
+      "learning_rate": 9.51892713869041e-06,
+      "loss": 0.8112,
+      "step": 496
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1288059246290212,
+      "learning_rate": 9.516593394849923e-06,
+      "loss": 1.1127,
+      "step": 497
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1686648397621733,
+      "learning_rate": 9.514254291559905e-06,
+      "loss": 0.8427,
+      "step": 498
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6446512735822925,
+      "learning_rate": 9.51190983159597e-06,
+      "loss": 0.6393,
+      "step": 499
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.427163382087819,
+      "learning_rate": 9.509560017740087e-06,
+      "loss": 0.8235,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.314342747904294,
+      "learning_rate": 9.507204852780571e-06,
+      "loss": 1.0226,
+      "step": 501
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9852323287803421,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.9693,
+      "step": 502
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5507130395076416,
+      "learning_rate": 9.502478480735678e-06,
+      "loss": 0.8802,
+      "step": 503
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6425778492873921,
+      "learning_rate": 9.500107279258672e-06,
+      "loss": 0.8801,
+      "step": 504
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3764444540304093,
+      "learning_rate": 9.49773073789478e-06,
+      "loss": 0.9763,
+      "step": 505
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1007939156349102,
+      "learning_rate": 9.495348859464042e-06,
+      "loss": 0.9332,
+      "step": 506
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8013227309678581,
+      "learning_rate": 9.492961646792819e-06,
+      "loss": 0.7007,
+      "step": 507
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2289869324888618,
+      "learning_rate": 9.490569102713814e-06,
+      "loss": 1.0234,
+      "step": 508
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3453544708371745,
+      "learning_rate": 9.488171230066048e-06,
+      "loss": 0.8388,
+      "step": 509
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4344652018157407,
+      "learning_rate": 9.485768031694872e-06,
+      "loss": 0.8539,
+      "step": 510
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.24320007688377,
+      "learning_rate": 9.48335951045195e-06,
+      "loss": 0.867,
+      "step": 511
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6210681051134521,
+      "learning_rate": 9.480945669195267e-06,
+      "loss": 0.9053,
+      "step": 512
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.60773568927563,
+      "learning_rate": 9.47852651078912e-06,
+      "loss": 1.0244,
+      "step": 513
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7439125920756795,
+      "learning_rate": 9.476102038104112e-06,
+      "loss": 0.8371,
+      "step": 514
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5109553355745977,
+      "learning_rate": 9.473672254017155e-06,
+      "loss": 0.9806,
+      "step": 515
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0968220205248773,
+      "learning_rate": 9.471237161411462e-06,
+      "loss": 1.0234,
+      "step": 516
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8962707950116502,
+      "learning_rate": 9.468796763176549e-06,
+      "loss": 0.5715,
+      "step": 517
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8012406176698051,
+      "learning_rate": 9.466351062208223e-06,
+      "loss": 0.9631,
+      "step": 518
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9680681137853948,
+      "learning_rate": 9.463900061408585e-06,
+      "loss": 0.9166,
+      "step": 519
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0121866140449953,
+      "learning_rate": 9.461443763686027e-06,
+      "loss": 0.9841,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9056496221657028,
+      "learning_rate": 9.45898217195522e-06,
+      "loss": 0.7436,
+      "step": 521
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9813321073208333,
+      "learning_rate": 9.456515289137127e-06,
+      "loss": 0.9048,
+      "step": 522
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8748047606228255,
+      "learning_rate": 9.454043118158979e-06,
+      "loss": 0.974,
+      "step": 523
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0861429024850242,
+      "learning_rate": 9.45156566195429e-06,
+      "loss": 0.8315,
+      "step": 524
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8325103819484405,
+      "learning_rate": 9.44908292346284e-06,
+      "loss": 0.7543,
+      "step": 525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0146920997435467,
+      "learning_rate": 9.446594905630682e-06,
+      "loss": 0.761,
+      "step": 526
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.627442042068908,
+      "learning_rate": 9.444101611410127e-06,
+      "loss": 1.0105,
+      "step": 527
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0079381553671005,
+      "learning_rate": 9.441603043759756e-06,
+      "loss": 0.8585,
+      "step": 528
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8855551400123958,
+      "learning_rate": 9.4390992056444e-06,
+      "loss": 0.9261,
+      "step": 529
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0577641720785922,
+      "learning_rate": 9.436590100035145e-06,
+      "loss": 0.9613,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9576838193012358,
+      "learning_rate": 9.43407572990933e-06,
+      "loss": 0.7968,
+      "step": 531
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.077123047149094,
+      "learning_rate": 9.43155609825054e-06,
+      "loss": 0.9956,
+      "step": 532
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0096188152271284,
+      "learning_rate": 9.429031208048604e-06,
+      "loss": 0.8194,
+      "step": 533
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3378614593659408,
+      "learning_rate": 9.42650106229959e-06,
+      "loss": 0.9074,
+      "step": 534
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7122728646794696,
+      "learning_rate": 9.423965664005801e-06,
+      "loss": 0.6743,
+      "step": 535
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3680058685417984,
+      "learning_rate": 9.421425016175776e-06,
+      "loss": 0.7981,
+      "step": 536
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2838840485947283,
+      "learning_rate": 9.41887912182428e-06,
+      "loss": 0.904,
+      "step": 537
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8840921827519916,
+      "learning_rate": 9.416327983972304e-06,
+      "loss": 0.8125,
+      "step": 538
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.094873044241505,
+      "learning_rate": 9.413771605647067e-06,
+      "loss": 0.9328,
+      "step": 539
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7467876885992745,
+      "learning_rate": 9.411209989881995e-06,
+      "loss": 1.0125,
+      "step": 540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3980481942316298,
+      "learning_rate": 9.408643139716737e-06,
+      "loss": 0.8193,
+      "step": 541
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8420152384548751,
+      "learning_rate": 9.406071058197154e-06,
+      "loss": 0.723,
+      "step": 542
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.4588984234287405,
+      "learning_rate": 9.403493748375311e-06,
+      "loss": 0.8262,
+      "step": 543
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6626698324200608,
+      "learning_rate": 9.400911213309478e-06,
+      "loss": 0.9724,
+      "step": 544
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3027542586093863,
+      "learning_rate": 9.398323456064124e-06,
+      "loss": 0.6904,
+      "step": 545
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2830496489071872,
+      "learning_rate": 9.395730479709916e-06,
+      "loss": 0.8589,
+      "step": 546
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4704335187266249,
+      "learning_rate": 9.393132287323714e-06,
+      "loss": 1.0668,
+      "step": 547
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2205331022939507,
+      "learning_rate": 9.39052888198857e-06,
+      "loss": 0.7981,
+      "step": 548
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0097733191521074,
+      "learning_rate": 9.387920266793718e-06,
+      "loss": 0.8921,
+      "step": 549
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.03750521159318,
+      "learning_rate": 9.385306444834573e-06,
+      "loss": 1.0135,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.289222263655943,
+      "learning_rate": 9.382687419212733e-06,
+      "loss": 0.8318,
+      "step": 551
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8307090240835004,
+      "learning_rate": 9.380063193035968e-06,
+      "loss": 0.7279,
+      "step": 552
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.046765630455092,
+      "learning_rate": 9.377433769418218e-06,
+      "loss": 0.9101,
+      "step": 553
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.336677237457073,
+      "learning_rate": 9.374799151479597e-06,
+      "loss": 1.0168,
+      "step": 554
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1937410837531395,
+      "learning_rate": 9.372159342346371e-06,
+      "loss": 0.7548,
+      "step": 555
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0806306978648021,
+      "learning_rate": 9.369514345150977e-06,
+      "loss": 0.9787,
+      "step": 556
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.642621752210111,
+      "learning_rate": 9.366864163032e-06,
+      "loss": 0.7039,
+      "step": 557
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0973405735956627,
+      "learning_rate": 9.364208799134187e-06,
+      "loss": 0.7039,
+      "step": 558
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9866052510173103,
+      "learning_rate": 9.361548256608421e-06,
+      "loss": 0.9382,
+      "step": 559
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9006535143027823,
+      "learning_rate": 9.358882538611743e-06,
+      "loss": 0.8331,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7218112384508352,
+      "learning_rate": 9.356211648307326e-06,
+      "loss": 0.7227,
+      "step": 561
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1907119099432033,
+      "learning_rate": 9.353535588864481e-06,
+      "loss": 0.8216,
+      "step": 562
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.207443954822768,
+      "learning_rate": 9.350854363458663e-06,
+      "loss": 1.0217,
+      "step": 563
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8749136344935519,
+      "learning_rate": 9.348167975271443e-06,
+      "loss": 0.9209,
+      "step": 564
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7356722382934965,
+      "learning_rate": 9.34547642749053e-06,
+      "loss": 0.9506,
+      "step": 565
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2341019605378554,
+      "learning_rate": 9.342779723309746e-06,
+      "loss": 0.955,
+      "step": 566
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2356001974926119,
+      "learning_rate": 9.340077865929035e-06,
+      "loss": 0.7971,
+      "step": 567
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9549419028423295,
+      "learning_rate": 9.337370858554462e-06,
+      "loss": 0.9289,
+      "step": 568
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2153961220183471,
+      "learning_rate": 9.334658704398193e-06,
+      "loss": 0.5854,
+      "step": 569
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2015088386690136,
+      "learning_rate": 9.33194140667851e-06,
+      "loss": 0.8294,
+      "step": 570
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.244277027442002,
+      "learning_rate": 9.329218968619793e-06,
+      "loss": 0.9575,
+      "step": 571
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7212119972406605,
+      "learning_rate": 9.326491393452521e-06,
+      "loss": 0.8754,
+      "step": 572
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9058761060199059,
+      "learning_rate": 9.323758684413272e-06,
+      "loss": 1.1495,
+      "step": 573
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9029488949041279,
+      "learning_rate": 9.321020844744717e-06,
+      "loss": 0.949,
+      "step": 574
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.390568133990733,
+      "learning_rate": 9.31827787769561e-06,
+      "loss": 0.9594,
+      "step": 575
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3707545249388862,
+      "learning_rate": 9.315529786520792e-06,
+      "loss": 0.8076,
+      "step": 576
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9243052289523535,
+      "learning_rate": 9.312776574481187e-06,
+      "loss": 0.9356,
+      "step": 577
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8914510425130124,
+      "learning_rate": 9.310018244843789e-06,
+      "loss": 0.6447,
+      "step": 578
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5329569898374,
+      "learning_rate": 9.307254800881672e-06,
+      "loss": 0.8768,
+      "step": 579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5607332375358072,
+      "learning_rate": 9.304486245873973e-06,
+      "loss": 0.7949,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4614573898438918,
+      "learning_rate": 9.301712583105898e-06,
+      "loss": 0.8023,
+      "step": 581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4431877372081636,
+      "learning_rate": 9.29893381586871e-06,
+      "loss": 1.0527,
+      "step": 582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8440264236761488,
+      "learning_rate": 9.296149947459731e-06,
+      "loss": 0.8781,
+      "step": 583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2843299453593413,
+      "learning_rate": 9.293360981182338e-06,
+      "loss": 0.8865,
+      "step": 584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0243165718921237,
+      "learning_rate": 9.290566920345954e-06,
+      "loss": 0.9003,
+      "step": 585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8091043703686498,
+      "learning_rate": 9.287767768266046e-06,
+      "loss": 0.8931,
+      "step": 586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1876270537868368,
+      "learning_rate": 9.284963528264133e-06,
+      "loss": 0.9663,
+      "step": 587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2618986594167843,
+      "learning_rate": 9.282154203667754e-06,
+      "loss": 0.5817,
+      "step": 588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.175303294685142,
+      "learning_rate": 9.279339797810497e-06,
+      "loss": 0.9698,
+      "step": 589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5265009051977787,
+      "learning_rate": 9.27652031403197e-06,
+      "loss": 0.857,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1884150336514194,
+      "learning_rate": 9.27369575567781e-06,
+      "loss": 0.85,
+      "step": 591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0085773085873373,
+      "learning_rate": 9.270866126099679e-06,
+      "loss": 0.9564,
+      "step": 592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.48620180213456,
+      "learning_rate": 9.268031428655248e-06,
+      "loss": 0.7692,
+      "step": 593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5210667686386574,
+      "learning_rate": 9.26519166670821e-06,
+      "loss": 0.9032,
+      "step": 594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.296342566756489,
+      "learning_rate": 9.26234684362826e-06,
+      "loss": 0.9831,
+      "step": 595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9216506812060357,
+      "learning_rate": 9.259496962791108e-06,
+      "loss": 0.6179,
+      "step": 596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1667607904364798,
+      "learning_rate": 9.256642027578458e-06,
+      "loss": 0.9614,
+      "step": 597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2136353264659008,
+      "learning_rate": 9.253782041378012e-06,
+      "loss": 0.9046,
+      "step": 598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0807411494760317,
+      "learning_rate": 9.25091700758347e-06,
+      "loss": 1.0713,
+      "step": 599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.643698399414858,
+      "learning_rate": 9.248046929594519e-06,
+      "loss": 0.8228,
+      "step": 600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2221103579270582,
+      "learning_rate": 9.24517181081683e-06,
+      "loss": 0.793,
+      "step": 601
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2966010747488719,
+      "learning_rate": 9.24229165466206e-06,
+      "loss": 0.9598,
+      "step": 602
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9431954874219102,
+      "learning_rate": 9.239406464547835e-06,
+      "loss": 0.8976,
+      "step": 603
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4015088870359067,
+      "learning_rate": 9.236516243897764e-06,
+      "loss": 0.7697,
+      "step": 604
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9381339746883319,
+      "learning_rate": 9.233620996141421e-06,
+      "loss": 0.8818,
+      "step": 605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5400620709431516,
+      "learning_rate": 9.230720724714345e-06,
+      "loss": 0.7534,
+      "step": 606
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0698457031221338,
+      "learning_rate": 9.227815433058037e-06,
+      "loss": 0.7989,
+      "step": 607
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.967038345941606,
+      "learning_rate": 9.22490512461995e-06,
+      "loss": 0.7677,
+      "step": 608
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2863963135005487,
+      "learning_rate": 9.2219898028535e-06,
+      "loss": 1.0701,
+      "step": 609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.684005531137107,
+      "learning_rate": 9.219069471218045e-06,
+      "loss": 0.8121,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.28490250923848,
+      "learning_rate": 9.216144133178889e-06,
+      "loss": 0.9521,
+      "step": 611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.169305378165972,
+      "learning_rate": 9.213213792207275e-06,
+      "loss": 0.9067,
+      "step": 612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.207195100423092,
+      "learning_rate": 9.210278451780386e-06,
+      "loss": 0.9623,
+      "step": 613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8982990163176723,
+      "learning_rate": 9.207338115381337e-06,
+      "loss": 0.5738,
+      "step": 614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7557205400526718,
+      "learning_rate": 9.204392786499168e-06,
+      "loss": 0.9186,
+      "step": 615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1446261331002228,
+      "learning_rate": 9.201442468628847e-06,
+      "loss": 1.0517,
+      "step": 616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1378426254378928,
+      "learning_rate": 9.198487165271259e-06,
+      "loss": 0.7547,
+      "step": 617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0419301776478003,
+      "learning_rate": 9.195526879933206e-06,
+      "loss": 0.9868,
+      "step": 618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4543306844596113,
+      "learning_rate": 9.192561616127403e-06,
+      "loss": 0.7778,
+      "step": 619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2001053531473302,
+      "learning_rate": 9.189591377372468e-06,
+      "loss": 0.8508,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5104289988227275,
+      "learning_rate": 9.18661616719293e-06,
+      "loss": 0.9253,
+      "step": 621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7952005802604254,
+      "learning_rate": 9.183635989119211e-06,
+      "loss": 0.7569,
+      "step": 622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.375065998444244,
+      "learning_rate": 9.18065084668763e-06,
+      "loss": 0.9547,
+      "step": 623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0329085245989464,
+      "learning_rate": 9.177660743440395e-06,
+      "loss": 0.8641,
+      "step": 624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4228520600264396,
+      "learning_rate": 9.174665682925606e-06,
+      "loss": 0.953,
+      "step": 625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.716708873072845,
+      "learning_rate": 9.171665668697236e-06,
+      "loss": 0.8137,
+      "step": 626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9745546547456091,
+      "learning_rate": 9.168660704315146e-06,
+      "loss": 0.9768,
+      "step": 627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0801511859844588,
+      "learning_rate": 9.165650793345064e-06,
+      "loss": 0.9509,
+      "step": 628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.404901227247712,
+      "learning_rate": 9.162635939358593e-06,
+      "loss": 0.7244,
+      "step": 629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.488027180600429,
+      "learning_rate": 9.159616145933194e-06,
+      "loss": 0.8708,
+      "step": 630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1918877529603054,
+      "learning_rate": 9.156591416652198e-06,
+      "loss": 0.8922,
+      "step": 631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6167158512965359,
+      "learning_rate": 9.15356175510479e-06,
+      "loss": 0.7794,
+      "step": 632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5103107976495278,
+      "learning_rate": 9.150527164886e-06,
+      "loss": 0.761,
+      "step": 633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.532004281537576,
+      "learning_rate": 9.14748764959672e-06,
+      "loss": 0.8367,
+      "step": 634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8375335934518672,
+      "learning_rate": 9.144443212843675e-06,
+      "loss": 0.8665,
+      "step": 635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.823326668126576,
+      "learning_rate": 9.141393858239435e-06,
+      "loss": 0.8258,
+      "step": 636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3256439953319359,
+      "learning_rate": 9.138339589402405e-06,
+      "loss": 0.9752,
+      "step": 637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1334967329006307,
+      "learning_rate": 9.135280409956819e-06,
+      "loss": 0.8254,
+      "step": 638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6258745396316695,
+      "learning_rate": 9.132216323532743e-06,
+      "loss": 0.9706,
+      "step": 639
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7234274380340548,
+      "learning_rate": 9.12914733376606e-06,
+      "loss": 0.6766,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.079593009882704,
+      "learning_rate": 9.126073444298473e-06,
+      "loss": 0.8434,
+      "step": 641
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4269894552031246,
+      "learning_rate": 9.122994658777504e-06,
+      "loss": 0.8728,
+      "step": 642
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8980957665579652,
+      "learning_rate": 9.119910980856477e-06,
+      "loss": 0.8777,
+      "step": 643
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1602112046049882,
+      "learning_rate": 9.116822414194528e-06,
+      "loss": 0.8939,
+      "step": 644
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.999773576345978,
+      "learning_rate": 9.113728962456588e-06,
+      "loss": 0.923,
+      "step": 645
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0811971984182707,
+      "learning_rate": 9.110630629313388e-06,
+      "loss": 0.7118,
+      "step": 646
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9340498448077799,
+      "learning_rate": 9.107527418441455e-06,
+      "loss": 0.8993,
+      "step": 647
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1233156389688794,
+      "learning_rate": 9.104419333523096e-06,
+      "loss": 0.9429,
+      "step": 648
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7934575600173343,
+      "learning_rate": 9.101306378246408e-06,
+      "loss": 0.7428,
+      "step": 649
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7945483261174356,
+      "learning_rate": 9.098188556305262e-06,
+      "loss": 0.8848,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9949065504503363,
+      "learning_rate": 9.095065871399312e-06,
+      "loss": 0.8432,
+      "step": 651
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6431092804064721,
+      "learning_rate": 9.091938327233973e-06,
+      "loss": 0.9779,
+      "step": 652
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8477278548571725,
+      "learning_rate": 9.088805927520432e-06,
+      "loss": 0.7433,
+      "step": 653
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0635599333858399,
+      "learning_rate": 9.085668675975634e-06,
+      "loss": 0.972,
+      "step": 654
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9774458808495424,
+      "learning_rate": 9.082526576322286e-06,
+      "loss": 0.9072,
+      "step": 655
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8368942696019095,
+      "learning_rate": 9.079379632288842e-06,
+      "loss": 0.9967,
+      "step": 656
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7435076499052764,
+      "learning_rate": 9.076227847609513e-06,
+      "loss": 0.7968,
+      "step": 657
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4910582753518837,
+      "learning_rate": 9.073071226024242e-06,
+      "loss": 0.7754,
+      "step": 658
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5982721424945875,
+      "learning_rate": 9.06990977127872e-06,
+      "loss": 0.6324,
+      "step": 659
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0951288752239954,
+      "learning_rate": 9.066743487124374e-06,
+      "loss": 1.0456,
+      "step": 660
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.070256365877206,
+      "learning_rate": 9.063572377318356e-06,
+      "loss": 0.8842,
+      "step": 661
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9150302238467513,
+      "learning_rate": 9.060396445623545e-06,
+      "loss": 0.8917,
+      "step": 662
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9313225372542996,
+      "learning_rate": 9.057215695808548e-06,
+      "loss": 0.8655,
+      "step": 663
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5500193571483682,
+      "learning_rate": 9.054030131647682e-06,
+      "loss": 0.9783,
+      "step": 664
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8304394934004465,
+      "learning_rate": 9.05083975692098e-06,
+      "loss": 0.8111,
+      "step": 665
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8592725686364147,
+      "learning_rate": 9.047644575414184e-06,
+      "loss": 0.8922,
+      "step": 666
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6997195398607927,
+      "learning_rate": 9.044444590918737e-06,
+      "loss": 0.726,
+      "step": 667
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1045758216283486,
+      "learning_rate": 9.041239807231785e-06,
+      "loss": 0.9597,
+      "step": 668
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.236932865988219,
+      "learning_rate": 9.038030228156165e-06,
+      "loss": 0.9061,
+      "step": 669
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1191866439006815,
+      "learning_rate": 9.034815857500407e-06,
+      "loss": 0.7433,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.736433022759413,
+      "learning_rate": 9.031596699078727e-06,
+      "loss": 1.0284,
+      "step": 671
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0154876439985687,
+      "learning_rate": 9.028372756711017e-06,
+      "loss": 0.7685,
+      "step": 672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8209672386785943,
+      "learning_rate": 9.025144034222853e-06,
+      "loss": 0.9488,
+      "step": 673
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0343832704513858,
+      "learning_rate": 9.021910535445479e-06,
+      "loss": 0.7681,
+      "step": 674
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9159694622915663,
+      "learning_rate": 9.018672264215808e-06,
+      "loss": 1.0265,
+      "step": 675
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7107034566678979,
+      "learning_rate": 9.015429224376415e-06,
+      "loss": 0.5968,
+      "step": 676
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8342710143728472,
+      "learning_rate": 9.012181419775535e-06,
+      "loss": 0.9074,
+      "step": 677
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.915399387925381,
+      "learning_rate": 9.008928854267054e-06,
+      "loss": 0.9932,
+      "step": 678
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8856582692470547,
+      "learning_rate": 9.005671531710511e-06,
+      "loss": 0.9488,
+      "step": 679
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0502984148044037,
+      "learning_rate": 9.002409455971087e-06,
+      "loss": 0.8899,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9233121574959628,
+      "learning_rate": 8.999142630919604e-06,
+      "loss": 0.8822,
+      "step": 681
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.944667997958301,
+      "learning_rate": 8.99587106043252e-06,
+      "loss": 0.7077,
+      "step": 682
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.790312835987665,
+      "learning_rate": 8.992594748391926e-06,
+      "loss": 1.1275,
+      "step": 683
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.119884662640157,
+      "learning_rate": 8.989313698685536e-06,
+      "loss": 0.7179,
+      "step": 684
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6008560277124985,
+      "learning_rate": 8.986027915206686e-06,
+      "loss": 0.667,
+      "step": 685
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8840673050505539,
+      "learning_rate": 8.982737401854328e-06,
+      "loss": 0.7948,
+      "step": 686
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7928685356672789,
+      "learning_rate": 8.979442162533036e-06,
+      "loss": 0.9866,
+      "step": 687
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5110531435165316,
+      "learning_rate": 8.976142201152978e-06,
+      "loss": 0.9588,
+      "step": 688
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0746660529126315,
+      "learning_rate": 8.972837521629934e-06,
+      "loss": 0.7684,
+      "step": 689
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1763467438514088,
+      "learning_rate": 8.969528127885281e-06,
+      "loss": 0.9092,
+      "step": 690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8589200968277745,
+      "learning_rate": 8.966214023845989e-06,
+      "loss": 0.9324,
+      "step": 691
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8135060504270676,
+      "learning_rate": 8.962895213444618e-06,
+      "loss": 0.9569,
+      "step": 692
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6094295317708318,
+      "learning_rate": 8.959571700619312e-06,
+      "loss": 0.6381,
+      "step": 693
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7365701400364734,
+      "learning_rate": 8.956243489313795e-06,
+      "loss": 0.9837,
+      "step": 694
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8785807812994888,
+      "learning_rate": 8.952910583477366e-06,
+      "loss": 0.7449,
+      "step": 695
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.389569204790213,
+      "learning_rate": 8.949572987064896e-06,
+      "loss": 0.8141,
+      "step": 696
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9126159835606786,
+      "learning_rate": 8.946230704036824e-06,
+      "loss": 0.893,
+      "step": 697
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2556955484027572,
+      "learning_rate": 8.942883738359142e-06,
+      "loss": 0.8644,
+      "step": 698
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7000701714830395,
+      "learning_rate": 8.939532094003409e-06,
+      "loss": 0.9959,
+      "step": 699
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0753417036057253,
+      "learning_rate": 8.936175774946726e-06,
+      "loss": 0.8226,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6935888450521948,
+      "learning_rate": 8.932814785171748e-06,
+      "loss": 1.0221,
+      "step": 701
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.667321218367855,
+      "learning_rate": 8.92944912866667e-06,
+      "loss": 0.7162,
+      "step": 702
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0998587955176187,
+      "learning_rate": 8.926078809425224e-06,
+      "loss": 0.7335,
+      "step": 703
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9200383785746729,
+      "learning_rate": 8.922703831446675e-06,
+      "loss": 0.9732,
+      "step": 704
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2451855853102167,
+      "learning_rate": 8.919324198735817e-06,
+      "loss": 0.874,
+      "step": 705
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0171756210129577,
+      "learning_rate": 8.91593991530297e-06,
+      "loss": 0.9182,
+      "step": 706
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1137365343321568,
+      "learning_rate": 8.912550985163963e-06,
+      "loss": 0.7535,
+      "step": 707
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8789396917517228,
+      "learning_rate": 8.90915741234015e-06,
+      "loss": 0.8429,
+      "step": 708
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.839135585259204,
+      "learning_rate": 8.905759200858388e-06,
+      "loss": 0.9132,
+      "step": 709
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0858023196002093,
+      "learning_rate": 8.902356354751042e-06,
+      "loss": 0.8381,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6501612836734532,
+      "learning_rate": 8.898948878055971e-06,
+      "loss": 0.6908,
+      "step": 711
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.207023654895919,
+      "learning_rate": 8.895536774816533e-06,
+      "loss": 0.8366,
+      "step": 712
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5899827533492924,
+      "learning_rate": 8.892120049081577e-06,
+      "loss": 1.0192,
+      "step": 713
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.200821296536281,
+      "learning_rate": 8.888698704905431e-06,
+      "loss": 0.8596,
+      "step": 714
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.869025744191145,
+      "learning_rate": 8.885272746347912e-06,
+      "loss": 0.8378,
+      "step": 715
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6349647024498257,
+      "learning_rate": 8.881842177474304e-06,
+      "loss": 0.9176,
+      "step": 716
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9794345123278975,
+      "learning_rate": 8.878407002355367e-06,
+      "loss": 0.8512,
+      "step": 717
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9750988804780665,
+      "learning_rate": 8.874967225067325e-06,
+      "loss": 0.9765,
+      "step": 718
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8299589560175472,
+      "learning_rate": 8.871522849691863e-06,
+      "loss": 0.7937,
+      "step": 719
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.813587226289581,
+      "learning_rate": 8.868073880316125e-06,
+      "loss": 0.8879,
+      "step": 720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5927136330017142,
+      "learning_rate": 8.864620321032698e-06,
+      "loss": 0.6243,
+      "step": 721
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9935740595558086,
+      "learning_rate": 8.861162175939626e-06,
+      "loss": 0.731,
+      "step": 722
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8273522716266456,
+      "learning_rate": 8.85769944914039e-06,
+      "loss": 0.9303,
+      "step": 723
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8664560732698762,
+      "learning_rate": 8.854232144743905e-06,
+      "loss": 0.8397,
+      "step": 724
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8795324317171525,
+      "learning_rate": 8.85076026686452e-06,
+      "loss": 0.9418,
+      "step": 725
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.47514978949546083,
+      "learning_rate": 8.847283819622015e-06,
+      "loss": 1.0166,
+      "step": 726
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7980007360325794,
+      "learning_rate": 8.843802807141584e-06,
+      "loss": 0.7881,
+      "step": 727
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8492475749325418,
+      "learning_rate": 8.840317233553845e-06,
+      "loss": 0.7084,
+      "step": 728
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2025963820725836,
+      "learning_rate": 8.836827102994824e-06,
+      "loss": 0.8476,
+      "step": 729
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2621153293965255,
+      "learning_rate": 8.83333241960596e-06,
+      "loss": 0.8681,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8303011684221875,
+      "learning_rate": 8.829833187534086e-06,
+      "loss": 0.7202,
+      "step": 731
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7978675805720584,
+      "learning_rate": 8.82632941093144e-06,
+      "loss": 0.8491,
+      "step": 732
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8056709319578862,
+      "learning_rate": 8.822821093955646e-06,
+      "loss": 1.016,
+      "step": 733
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9194418720649518,
+      "learning_rate": 8.819308240769726e-06,
+      "loss": 0.8422,
+      "step": 734
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8509670955795444,
+      "learning_rate": 8.81579085554207e-06,
+      "loss": 0.8708,
+      "step": 735
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7952817728374606,
+      "learning_rate": 8.812268942446461e-06,
+      "loss": 1.0425,
+      "step": 736
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0351944767698529,
+      "learning_rate": 8.808742505662045e-06,
+      "loss": 0.9979,
+      "step": 737
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9977130186663479,
+      "learning_rate": 8.805211549373335e-06,
+      "loss": 0.6375,
+      "step": 738
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7268650973070016,
+      "learning_rate": 8.801676077770217e-06,
+      "loss": 0.8224,
+      "step": 739
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6389459203077702,
+      "learning_rate": 8.798136095047925e-06,
+      "loss": 0.9964,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.287174593009283,
+      "learning_rate": 8.794591605407047e-06,
+      "loss": 0.7786,
+      "step": 741
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0032073314866823,
+      "learning_rate": 8.791042613053527e-06,
+      "loss": 0.957,
+      "step": 742
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1050781831623215,
+      "learning_rate": 8.787489122198643e-06,
+      "loss": 0.7367,
+      "step": 743
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6596508951811606,
+      "learning_rate": 8.783931137059013e-06,
+      "loss": 0.787,
+      "step": 744
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9627021749638279,
+      "learning_rate": 8.780368661856592e-06,
+      "loss": 0.8925,
+      "step": 745
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.032497189815307,
+      "learning_rate": 8.776801700818658e-06,
+      "loss": 0.8613,
+      "step": 746
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8561846898296334,
+      "learning_rate": 8.773230258177815e-06,
+      "loss": 0.7079,
+      "step": 747
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.562196474179583,
+      "learning_rate": 8.769654338171986e-06,
+      "loss": 0.8124,
+      "step": 748
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8627547167883806,
+      "learning_rate": 8.766073945044404e-06,
+      "loss": 1.0245,
+      "step": 749
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8063526964330535,
+      "learning_rate": 8.76248908304361e-06,
+      "loss": 0.9308,
+      "step": 750
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5679485751014459,
+      "learning_rate": 8.758899756423448e-06,
+      "loss": 0.9334,
+      "step": 751
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8460967681229478,
+      "learning_rate": 8.755305969443062e-06,
+      "loss": 0.8938,
+      "step": 752
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3955745549757175,
+      "learning_rate": 8.751707726366887e-06,
+      "loss": 0.7695,
+      "step": 753
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6976517086084567,
+      "learning_rate": 8.748105031464644e-06,
+      "loss": 0.9488,
+      "step": 754
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8227305608123205,
+      "learning_rate": 8.744497889011344e-06,
+      "loss": 0.5736,
+      "step": 755
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1507977529620217,
+      "learning_rate": 8.740886303287263e-06,
+      "loss": 0.9303,
+      "step": 756
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8405979506816968,
+      "learning_rate": 8.737270278577959e-06,
+      "loss": 0.8085,
+      "step": 757
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1503043032357627,
+      "learning_rate": 8.733649819174257e-06,
+      "loss": 0.7984,
+      "step": 758
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7930591071514794,
+      "learning_rate": 8.730024929372241e-06,
+      "loss": 1.1093,
+      "step": 759
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0042285767543015,
+      "learning_rate": 8.726395613473251e-06,
+      "loss": 0.9278,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8989044875590482,
+      "learning_rate": 8.722761875783885e-06,
+      "loss": 0.9476,
+      "step": 761
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8728438071464727,
+      "learning_rate": 8.71912372061598e-06,
+      "loss": 0.7512,
+      "step": 762
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5251355194097443,
+      "learning_rate": 8.715481152286625e-06,
+      "loss": 0.9197,
+      "step": 763
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7094920054213856,
+      "learning_rate": 8.711834175118132e-06,
+      "loss": 0.665,
+      "step": 764
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7629617798060091,
+      "learning_rate": 8.708182793438059e-06,
+      "loss": 0.7845,
+      "step": 765
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.795087784218763,
+      "learning_rate": 8.704527011579181e-06,
+      "loss": 0.8322,
+      "step": 766
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1909423530557952,
+      "learning_rate": 8.700866833879497e-06,
+      "loss": 0.7995,
+      "step": 767
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0444528224611986,
+      "learning_rate": 8.697202264682223e-06,
+      "loss": 0.9388,
+      "step": 768
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9486689717726335,
+      "learning_rate": 8.693533308335786e-06,
+      "loss": 0.8126,
+      "step": 769
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7980989240587661,
+      "learning_rate": 8.689859969193817e-06,
+      "loss": 0.8441,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7734499415049558,
+      "learning_rate": 8.68618225161515e-06,
+      "loss": 0.9764,
+      "step": 771
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6777349526982211,
+      "learning_rate": 8.682500159963812e-06,
+      "loss": 0.9546,
+      "step": 772
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7953075646069098,
+      "learning_rate": 8.67881369860902e-06,
+      "loss": 0.6741,
+      "step": 773
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0343676348392423,
+      "learning_rate": 8.675122871925183e-06,
+      "loss": 0.829,
+      "step": 774
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.806864851444203,
+      "learning_rate": 8.671427684291882e-06,
+      "loss": 0.9515,
+      "step": 775
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7572416906745798,
+      "learning_rate": 8.667728140093876e-06,
+      "loss": 0.8483,
+      "step": 776
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8380140423011045,
+      "learning_rate": 8.664024243721092e-06,
+      "loss": 0.8425,
+      "step": 777
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8123587951986472,
+      "learning_rate": 8.660315999568623e-06,
+      "loss": 0.9526,
+      "step": 778
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9184399337784447,
+      "learning_rate": 8.656603412036722e-06,
+      "loss": 0.7108,
+      "step": 779
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.831843338935016,
+      "learning_rate": 8.652886485530791e-06,
+      "loss": 1.0123,
+      "step": 780
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8066641019349248,
+      "learning_rate": 8.64916522446139e-06,
+      "loss": 0.5859,
+      "step": 781
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0685522911949512,
+      "learning_rate": 8.64543963324421e-06,
+      "loss": 0.8115,
+      "step": 782
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7367019584287984,
+      "learning_rate": 8.641709716300092e-06,
+      "loss": 1.0002,
+      "step": 783
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1169280348996233,
+      "learning_rate": 8.637975478055002e-06,
+      "loss": 0.9091,
+      "step": 784
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9765798942044751,
+      "learning_rate": 8.634236922940038e-06,
+      "loss": 0.953,
+      "step": 785
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9018057962498353,
+      "learning_rate": 8.630494055391418e-06,
+      "loss": 0.783,
+      "step": 786
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5765299875173697,
+      "learning_rate": 8.626746879850479e-06,
+      "loss": 0.9463,
+      "step": 787
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9907858154987383,
+      "learning_rate": 8.62299540076367e-06,
+      "loss": 0.8049,
+      "step": 788
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7614608864418805,
+      "learning_rate": 8.619239622582542e-06,
+      "loss": 0.9006,
+      "step": 789
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8314799942415151,
+      "learning_rate": 8.615479549763756e-06,
+      "loss": 0.8107,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0173814058372952,
+      "learning_rate": 8.611715186769063e-06,
+      "loss": 0.8898,
+      "step": 791
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6696667729133458,
+      "learning_rate": 8.607946538065301e-06,
+      "loss": 0.7322,
+      "step": 792
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9904979750452643,
+      "learning_rate": 8.604173608124405e-06,
+      "loss": 0.784,
+      "step": 793
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4815666969700425,
+      "learning_rate": 8.600396401423382e-06,
+      "loss": 0.7898,
+      "step": 794
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.79736383984785,
+      "learning_rate": 8.596614922444313e-06,
+      "loss": 1.02,
+      "step": 795
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.235172792133491,
+      "learning_rate": 8.592829175674351e-06,
+      "loss": 0.7608,
+      "step": 796
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8831479685583735,
+      "learning_rate": 8.589039165605716e-06,
+      "loss": 0.9778,
+      "step": 797
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3325843559676318,
+      "learning_rate": 8.585244896735683e-06,
+      "loss": 0.8733,
+      "step": 798
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7415911990332329,
+      "learning_rate": 8.581446373566581e-06,
+      "loss": 0.8813,
+      "step": 799
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6142722155335408,
+      "learning_rate": 8.577643600605789e-06,
+      "loss": 0.5659,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.164915953950598,
+      "learning_rate": 8.573836582365727e-06,
+      "loss": 0.9224,
+      "step": 801
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8887148955452323,
+      "learning_rate": 8.570025323363853e-06,
+      "loss": 0.9773,
+      "step": 802
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1121431206247039,
+      "learning_rate": 8.566209828122658e-06,
+      "loss": 0.9311,
+      "step": 803
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7657005061872006,
+      "learning_rate": 8.56239010116966e-06,
+      "loss": 0.9378,
+      "step": 804
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9184179350866519,
+      "learning_rate": 8.558566147037396e-06,
+      "loss": 0.7559,
+      "step": 805
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6859068986825837,
+      "learning_rate": 8.55473797026342e-06,
+      "loss": 0.9342,
+      "step": 806
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0397352573350611,
+      "learning_rate": 8.550905575390301e-06,
+      "loss": 0.9045,
+      "step": 807
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0139829500896023,
+      "learning_rate": 8.547068966965605e-06,
+      "loss": 0.627,
+      "step": 808
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7665184065352235,
+      "learning_rate": 8.543228149541905e-06,
+      "loss": 0.9516,
+      "step": 809
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.32804703137598,
+      "learning_rate": 8.539383127676764e-06,
+      "loss": 0.882,
+      "step": 810
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9471702430398453,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 1.0203,
+      "step": 811
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0015739911913293,
+      "learning_rate": 8.531680488877363e-06,
+      "loss": 0.7979,
+      "step": 812
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5050714297786606,
+      "learning_rate": 8.527822881083157e-06,
+      "loss": 0.9781,
+      "step": 813
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3518666750574064,
+      "learning_rate": 8.523961087127605e-06,
+      "loss": 0.8297,
+      "step": 814
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.778193580631305,
+      "learning_rate": 8.520095111593163e-06,
+      "loss": 0.796,
+      "step": 815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7695262802798188,
+      "learning_rate": 8.516224959067253e-06,
+      "loss": 0.9157,
+      "step": 816
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.110871835067212,
+      "learning_rate": 8.512350634142246e-06,
+      "loss": 0.7789,
+      "step": 817
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7787376924589497,
+      "learning_rate": 8.508472141415468e-06,
+      "loss": 0.7708,
+      "step": 818
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4361529285393615,
+      "learning_rate": 8.504589485489187e-06,
+      "loss": 0.7806,
+      "step": 819
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.185897245847663,
+      "learning_rate": 8.50070267097062e-06,
+      "loss": 0.8031,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7802504526913002,
+      "learning_rate": 8.496811702471906e-06,
+      "loss": 0.8748,
+      "step": 821
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2989596454546593,
+      "learning_rate": 8.492916584610124e-06,
+      "loss": 0.7786,
+      "step": 822
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9256969514806701,
+      "learning_rate": 8.489017322007272e-06,
+      "loss": 0.9849,
+      "step": 823
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.952553329800984,
+      "learning_rate": 8.485113919290265e-06,
+      "loss": 0.91,
+      "step": 824
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7209085321556576,
+      "learning_rate": 8.481206381090934e-06,
+      "loss": 0.9276,
+      "step": 825
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8291082877728642,
+      "learning_rate": 8.477294712046015e-06,
+      "loss": 0.6321,
+      "step": 826
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0907699956283343,
+      "learning_rate": 8.473378916797146e-06,
+      "loss": 0.853,
+      "step": 827
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2046922736040153,
+      "learning_rate": 8.46945899999086e-06,
+      "loss": 0.8632,
+      "step": 828
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0539160247510744,
+      "learning_rate": 8.465534966278586e-06,
+      "loss": 0.8275,
+      "step": 829
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.266370021064102,
+      "learning_rate": 8.46160682031663e-06,
+      "loss": 1.0156,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3889032566021533,
+      "learning_rate": 8.457674566766185e-06,
+      "loss": 0.8499,
+      "step": 831
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.122553181570519,
+      "learning_rate": 8.453738210293316e-06,
+      "loss": 0.676,
+      "step": 832
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1032953502747713,
+      "learning_rate": 8.449797755568952e-06,
+      "loss": 0.7756,
+      "step": 833
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9110242227527524,
+      "learning_rate": 8.44585320726889e-06,
+      "loss": 0.6729,
+      "step": 834
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9965870395398055,
+      "learning_rate": 8.441904570073787e-06,
+      "loss": 1.0118,
+      "step": 835
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8949529812144408,
+      "learning_rate": 8.437951848669144e-06,
+      "loss": 0.8111,
+      "step": 836
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2287296177699212,
+      "learning_rate": 8.433995047745313e-06,
+      "loss": 0.8355,
+      "step": 837
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6986915942190378,
+      "learning_rate": 8.430034171997487e-06,
+      "loss": 0.9937,
+      "step": 838
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2243048207305436,
+      "learning_rate": 8.426069226125695e-06,
+      "loss": 0.6981,
+      "step": 839
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2890503319160356,
+      "learning_rate": 8.422100214834791e-06,
+      "loss": 0.9871,
+      "step": 840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1405743394590862,
+      "learning_rate": 8.41812714283446e-06,
+      "loss": 0.8869,
+      "step": 841
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4141031891180675,
+      "learning_rate": 8.4141500148392e-06,
+      "loss": 1.0139,
+      "step": 842
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9506695661027441,
+      "learning_rate": 8.410168835568324e-06,
+      "loss": 0.7528,
+      "step": 843
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.048432559091801,
+      "learning_rate": 8.406183609745953e-06,
+      "loss": 0.7277,
+      "step": 844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7491997329514749,
+      "learning_rate": 8.402194342101006e-06,
+      "loss": 0.6636,
+      "step": 845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.249957360002299,
+      "learning_rate": 8.398201037367202e-06,
+      "loss": 0.9759,
+      "step": 846
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2391824248179655,
+      "learning_rate": 8.39420370028305e-06,
+      "loss": 1.0611,
+      "step": 847
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9146882503950583,
+      "learning_rate": 8.390202335591845e-06,
+      "loss": 0.8546,
+      "step": 848
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1189970398897255,
+      "learning_rate": 8.386196948041657e-06,
+      "loss": 0.7866,
+      "step": 849
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0731715413795784,
+      "learning_rate": 8.382187542385329e-06,
+      "loss": 0.9732,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3420075613887534,
+      "learning_rate": 8.37817412338048e-06,
+      "loss": 0.7469,
+      "step": 851
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0545752044494447,
+      "learning_rate": 8.374156695789487e-06,
+      "loss": 0.9445,
+      "step": 852
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.663373084138263,
+      "learning_rate": 8.370135264379475e-06,
+      "loss": 0.6901,
+      "step": 853
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0088637339395825,
+      "learning_rate": 8.366109833922335e-06,
+      "loss": 0.9301,
+      "step": 854
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9511385411457727,
+      "learning_rate": 8.362080409194694e-06,
+      "loss": 0.9078,
+      "step": 855
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2120900437898698,
+      "learning_rate": 8.35804699497792e-06,
+      "loss": 0.8426,
+      "step": 856
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8733276210247882,
+      "learning_rate": 8.354009596058116e-06,
+      "loss": 0.9818,
+      "step": 857
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.474019455428399,
+      "learning_rate": 8.349968217226114e-06,
+      "loss": 0.776,
+      "step": 858
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.439806816557651,
+      "learning_rate": 8.345922863277466e-06,
+      "loss": 0.997,
+      "step": 859
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.263637853376208,
+      "learning_rate": 8.341873539012443e-06,
+      "loss": 0.7339,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.29429948830183,
+      "learning_rate": 8.337820249236028e-06,
+      "loss": 0.9998,
+      "step": 861
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7649891382789833,
+      "learning_rate": 8.333762998757908e-06,
+      "loss": 0.6363,
+      "step": 862
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3762963649493423,
+      "learning_rate": 8.329701792392468e-06,
+      "loss": 0.9192,
+      "step": 863
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2834958508988321,
+      "learning_rate": 8.325636634958793e-06,
+      "loss": 0.9404,
+      "step": 864
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3683699979147976,
+      "learning_rate": 8.32156753128065e-06,
+      "loss": 0.9702,
+      "step": 865
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.020354836961051,
+      "learning_rate": 8.31749448618649e-06,
+      "loss": 0.8147,
+      "step": 866
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.551310838441252,
+      "learning_rate": 8.313417504509446e-06,
+      "loss": 0.8767,
+      "step": 867
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4199355214556608,
+      "learning_rate": 8.309336591087318e-06,
+      "loss": 0.7192,
+      "step": 868
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0390533416614771,
+      "learning_rate": 8.305251750762568e-06,
+      "loss": 1.0747,
+      "step": 869
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3506345301978289,
+      "learning_rate": 8.301162988382325e-06,
+      "loss": 0.6064,
+      "step": 870
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8892017615992809,
+      "learning_rate": 8.297070308798367e-06,
+      "loss": 0.6984,
+      "step": 871
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6237415398598127,
+      "learning_rate": 8.292973716867124e-06,
+      "loss": 0.8098,
+      "step": 872
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6766065073210752,
+      "learning_rate": 8.288873217449662e-06,
+      "loss": 0.978,
+      "step": 873
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7039990899176017,
+      "learning_rate": 8.284768815411693e-06,
+      "loss": 0.9505,
+      "step": 874
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6977261587693264,
+      "learning_rate": 8.28066051562355e-06,
+      "loss": 0.7394,
+      "step": 875
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9880844135399935,
+      "learning_rate": 8.2765483229602e-06,
+      "loss": 1.0306,
+      "step": 876
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5385173129182235,
+      "learning_rate": 8.272432242301223e-06,
+      "loss": 0.8292,
+      "step": 877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.368301871987185,
+      "learning_rate": 8.268312278530816e-06,
+      "loss": 0.9065,
+      "step": 878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7259198105838697,
+      "learning_rate": 8.26418843653778e-06,
+      "loss": 0.6417,
+      "step": 879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0440203258360932,
+      "learning_rate": 8.260060721215527e-06,
+      "loss": 0.9819,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0604976561483186,
+      "learning_rate": 8.255929137462049e-06,
+      "loss": 0.7589,
+      "step": 881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0293501019486986,
+      "learning_rate": 8.251793690179947e-06,
+      "loss": 0.8365,
+      "step": 882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9732039627933925,
+      "learning_rate": 8.247654384276391e-06,
+      "loss": 0.8773,
+      "step": 883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.177655649900707,
+      "learning_rate": 8.243511224663141e-06,
+      "loss": 0.8219,
+      "step": 884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1250324463737151,
+      "learning_rate": 8.239364216256521e-06,
+      "loss": 0.985,
+      "step": 885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.884342702861109,
+      "learning_rate": 8.23521336397743e-06,
+      "loss": 0.9346,
+      "step": 886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8367264851250735,
+      "learning_rate": 8.231058672751319e-06,
+      "loss": 1.0037,
+      "step": 887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6719823684712402,
+      "learning_rate": 8.226900147508205e-06,
+      "loss": 0.7054,
+      "step": 888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.241775236733515,
+      "learning_rate": 8.222737793182648e-06,
+      "loss": 0.704,
+      "step": 889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.950632996277853,
+      "learning_rate": 8.21857161471375e-06,
+      "loss": 0.9298,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9645520876525666,
+      "learning_rate": 8.214401617045155e-06,
+      "loss": 0.9243,
+      "step": 891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9164952823331624,
+      "learning_rate": 8.210227805125039e-06,
+      "loss": 0.9277,
+      "step": 892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0293442009991178,
+      "learning_rate": 8.206050183906102e-06,
+      "loss": 0.7351,
+      "step": 893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.003659171802231,
+      "learning_rate": 8.201868758345561e-06,
+      "loss": 0.8095,
+      "step": 894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.477532499836891,
+      "learning_rate": 8.197683533405156e-06,
+      "loss": 0.9701,
+      "step": 895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1091540769626786,
+      "learning_rate": 8.193494514051128e-06,
+      "loss": 0.8069,
+      "step": 896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7067479529711058,
+      "learning_rate": 8.189301705254222e-06,
+      "loss": 0.721,
+      "step": 897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0787158066908829,
+      "learning_rate": 8.185105111989683e-06,
+      "loss": 0.7952,
+      "step": 898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5695368855950586,
+      "learning_rate": 8.180904739237241e-06,
+      "loss": 0.999,
+      "step": 899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9408099207033626,
+      "learning_rate": 8.176700591981118e-06,
+      "loss": 0.8486,
+      "step": 900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0142747022418177,
+      "learning_rate": 8.172492675210005e-06,
+      "loss": 0.801,
+      "step": 901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8872842597495222,
+      "learning_rate": 8.168280993917078e-06,
+      "loss": 0.9517,
+      "step": 902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3496550092319346,
+      "learning_rate": 8.164065553099968e-06,
+      "loss": 0.8605,
+      "step": 903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9900621854726843,
+      "learning_rate": 8.159846357760776e-06,
+      "loss": 0.9396,
+      "step": 904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.029177293236724,
+      "learning_rate": 8.155623412906058e-06,
+      "loss": 0.8247,
+      "step": 905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9660397779772322,
+      "learning_rate": 8.15139672354681e-06,
+      "loss": 0.8172,
+      "step": 906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6538798360233261,
+      "learning_rate": 8.147166294698482e-06,
+      "loss": 0.6002,
+      "step": 907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8931919375433652,
+      "learning_rate": 8.142932131380955e-06,
+      "loss": 0.7126,
+      "step": 908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5600772809628976,
+      "learning_rate": 8.138694238618543e-06,
+      "loss": 1.0385,
+      "step": 909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.174765635692957,
+      "learning_rate": 8.134452621439988e-06,
+      "loss": 0.7862,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9626358330113229,
+      "learning_rate": 8.130207284878447e-06,
+      "loss": 1.0167,
+      "step": 911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5470569074959682,
+      "learning_rate": 8.125958233971494e-06,
+      "loss": 0.9686,
+      "step": 912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.228029459482363,
+      "learning_rate": 8.121705473761107e-06,
+      "loss": 0.7442,
+      "step": 913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7684762075524795,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.6393,
+      "step": 914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4536118945044414,
+      "learning_rate": 8.113188845619957e-06,
+      "loss": 0.8389,
+      "step": 915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8036503965901991,
+      "learning_rate": 8.108924987795137e-06,
+      "loss": 0.9026,
+      "step": 916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9550100302168791,
+      "learning_rate": 8.10465744087876e-06,
+      "loss": 0.7133,
+      "step": 917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9654381779268255,
+      "learning_rate": 8.100386209934754e-06,
+      "loss": 0.9829,
+      "step": 918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8594717377715599,
+      "learning_rate": 8.096111300031417e-06,
+      "loss": 0.8682,
+      "step": 919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.159642140574289,
+      "learning_rate": 8.091832716241414e-06,
+      "loss": 0.7898,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8386443781708032,
+      "learning_rate": 8.08755046364177e-06,
+      "loss": 0.8149,
+      "step": 921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.219553550779851,
+      "learning_rate": 8.083264547313863e-06,
+      "loss": 1.0387,
+      "step": 922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9836705709688978,
+      "learning_rate": 8.078974972343414e-06,
+      "loss": 0.9657,
+      "step": 923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0659421126791389,
+      "learning_rate": 8.074681743820496e-06,
+      "loss": 0.6063,
+      "step": 924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0247730742944847,
+      "learning_rate": 8.070384866839504e-06,
+      "loss": 0.8527,
+      "step": 925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6234497868297239,
+      "learning_rate": 8.066084346499176e-06,
+      "loss": 0.9743,
+      "step": 926
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1216442255869334,
+      "learning_rate": 8.061780187902562e-06,
+      "loss": 0.7287,
+      "step": 927
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3805264935364454,
+      "learning_rate": 8.057472396157034e-06,
+      "loss": 0.9549,
+      "step": 928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1340563073007235,
+      "learning_rate": 8.053160976374275e-06,
+      "loss": 0.7431,
+      "step": 929
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9679458482571414,
+      "learning_rate": 8.048845933670274e-06,
+      "loss": 0.6525,
+      "step": 930
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0351998469283725,
+      "learning_rate": 8.044527273165315e-06,
+      "loss": 0.9526,
+      "step": 931
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9352437525153571,
+      "learning_rate": 8.040204999983981e-06,
+      "loss": 0.8644,
+      "step": 932
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6980281236378553,
+      "learning_rate": 8.035879119255134e-06,
+      "loss": 0.6839,
+      "step": 933
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9393989164819734,
+      "learning_rate": 8.031549636111928e-06,
+      "loss": 0.7585,
+      "step": 934
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7472837028782404,
+      "learning_rate": 8.027216555691781e-06,
+      "loss": 1.0852,
+      "step": 935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0365880465923214,
+      "learning_rate": 8.022879883136385e-06,
+      "loss": 0.7707,
+      "step": 936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6927075998732038,
+      "learning_rate": 8.01853962359169e-06,
+      "loss": 0.9542,
+      "step": 937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9355394779424284,
+      "learning_rate": 8.01419578220791e-06,
+      "loss": 0.8241,
+      "step": 938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3353418899611251,
+      "learning_rate": 8.0098483641395e-06,
+      "loss": 0.7849,
+      "step": 939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9741583911800749,
+      "learning_rate": 8.005497374545169e-06,
+      "loss": 0.8547,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.89894580627274,
+      "learning_rate": 8.001142818587853e-06,
+      "loss": 0.5285,
+      "step": 941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2472422404982646,
+      "learning_rate": 7.99678470143473e-06,
+      "loss": 0.9272,
+      "step": 942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.241543087398163,
+      "learning_rate": 7.9924230282572e-06,
+      "loss": 0.7559,
+      "step": 943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.314974118228036,
+      "learning_rate": 7.988057804230878e-06,
+      "loss": 0.8753,
+      "step": 944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1985171601422697,
+      "learning_rate": 7.9836890345356e-06,
+      "loss": 1.1384,
+      "step": 945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0958702457674099,
+      "learning_rate": 7.979316724355406e-06,
+      "loss": 0.9082,
+      "step": 946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8882825697595312,
+      "learning_rate": 7.974940878878537e-06,
+      "loss": 0.9196,
+      "step": 947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9518474852338964,
+      "learning_rate": 7.970561503297426e-06,
+      "loss": 0.7575,
+      "step": 948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6773057064066956,
+      "learning_rate": 7.966178602808703e-06,
+      "loss": 0.9431,
+      "step": 949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1017440570642512,
+      "learning_rate": 7.96179218261317e-06,
+      "loss": 0.643,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3057245928497756,
+      "learning_rate": 7.957402247915817e-06,
+      "loss": 0.7813,
+      "step": 951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9167034007934262,
+      "learning_rate": 7.953008803925797e-06,
+      "loss": 0.8323,
+      "step": 952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0521269352875444,
+      "learning_rate": 7.948611855856426e-06,
+      "loss": 0.7614,
+      "step": 953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.457919053386127,
+      "learning_rate": 7.944211408925184e-06,
+      "loss": 0.9283,
+      "step": 954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0502752886821072,
+      "learning_rate": 7.939807468353697e-06,
+      "loss": 0.8974,
+      "step": 955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0848599132562,
+      "learning_rate": 7.935400039367742e-06,
+      "loss": 0.7772,
+      "step": 956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8868108145589925,
+      "learning_rate": 7.930989127197232e-06,
+      "loss": 0.9553,
+      "step": 957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8113631071749853,
+      "learning_rate": 7.92657473707621e-06,
+      "loss": 0.9223,
+      "step": 958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8413830436372951,
+      "learning_rate": 7.922156874242857e-06,
+      "loss": 0.6574,
+      "step": 959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9959831332838394,
+      "learning_rate": 7.91773554393946e-06,
+      "loss": 0.7391,
+      "step": 960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6483827089867299,
+      "learning_rate": 7.913310751412431e-06,
+      "loss": 1.0088,
+      "step": 961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0605312504405813,
+      "learning_rate": 7.908882501912289e-06,
+      "loss": 0.7376,
+      "step": 962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8879775799602105,
+      "learning_rate": 7.904450800693652e-06,
+      "loss": 0.8324,
+      "step": 963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2068817906406053,
+      "learning_rate": 7.900015653015238e-06,
+      "loss": 0.9902,
+      "step": 964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3953758747807117,
+      "learning_rate": 7.895577064139847e-06,
+      "loss": 0.6958,
+      "step": 965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.079921701725994,
+      "learning_rate": 7.89113503933437e-06,
+      "loss": 0.9397,
+      "step": 966
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9938460852116291,
+      "learning_rate": 7.886689583869773e-06,
+      "loss": 0.5547,
+      "step": 967
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0392974773320938,
+      "learning_rate": 7.882240703021092e-06,
+      "loss": 0.7879,
+      "step": 968
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6087681622176686,
+      "learning_rate": 7.877788402067427e-06,
+      "loss": 1.01,
+      "step": 969
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.042478789369615,
+      "learning_rate": 7.873332686291939e-06,
+      "loss": 0.8875,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.063784870023408,
+      "learning_rate": 7.868873560981837e-06,
+      "loss": 0.9814,
+      "step": 971
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1871192000802355,
+      "learning_rate": 7.864411031428379e-06,
+      "loss": 0.7879,
+      "step": 972
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0970062901383153,
+      "learning_rate": 7.859945102926862e-06,
+      "loss": 0.9866,
+      "step": 973
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2217542614365495,
+      "learning_rate": 7.855475780776618e-06,
+      "loss": 0.7673,
+      "step": 974
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8519255201515229,
+      "learning_rate": 7.851003070281e-06,
+      "loss": 0.9271,
+      "step": 975
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7786337728022726,
+      "learning_rate": 7.846526976747388e-06,
+      "loss": 0.7,
+      "step": 976
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0446667487299308,
+      "learning_rate": 7.842047505487173e-06,
+      "loss": 0.6139,
+      "step": 977
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7541250080728523,
+      "learning_rate": 7.837564661815755e-06,
+      "loss": 0.9584,
+      "step": 978
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0487353131458552,
+      "learning_rate": 7.833078451052537e-06,
+      "loss": 0.7919,
+      "step": 979
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.00793885814896,
+      "learning_rate": 7.828588878520911e-06,
+      "loss": 0.8981,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8993617441212048,
+      "learning_rate": 7.824095949548269e-06,
+      "loss": 0.9346,
+      "step": 981
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.891832722128992,
+      "learning_rate": 7.819599669465979e-06,
+      "loss": 0.8465,
+      "step": 982
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0376568369011432,
+      "learning_rate": 7.815100043609385e-06,
+      "loss": 1.0096,
+      "step": 983
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8628317165219468,
+      "learning_rate": 7.810597077317802e-06,
+      "loss": 0.861,
+      "step": 984
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0006142889443894,
+      "learning_rate": 7.806090775934513e-06,
+      "loss": 0.8869,
+      "step": 985
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5926355919148432,
+      "learning_rate": 7.801581144806752e-06,
+      "loss": 0.7927,
+      "step": 986
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1183269004770366,
+      "learning_rate": 7.797068189285705e-06,
+      "loss": 0.8013,
+      "step": 987
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0511218621060956,
+      "learning_rate": 7.792551914726509e-06,
+      "loss": 0.9205,
+      "step": 988
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9978706075358954,
+      "learning_rate": 7.788032326488233e-06,
+      "loss": 0.8705,
+      "step": 989
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.888064083510313,
+      "learning_rate": 7.78350942993388e-06,
+      "loss": 0.9114,
+      "step": 990
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3819226863788523,
+      "learning_rate": 7.778983230430382e-06,
+      "loss": 0.7916,
+      "step": 991
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0919204192742067,
+      "learning_rate": 7.77445373334858e-06,
+      "loss": 0.9728,
+      "step": 992
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8122749559057681,
+      "learning_rate": 7.769920944063244e-06,
+      "loss": 0.9395,
+      "step": 993
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8996396462506161,
+      "learning_rate": 7.765384867953038e-06,
+      "loss": 0.6014,
+      "step": 994
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2796795028402865,
+      "learning_rate": 7.76084551040053e-06,
+      "loss": 0.8527,
+      "step": 995
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.473970126543929,
+      "learning_rate": 7.75630287679218e-06,
+      "loss": 0.8912,
+      "step": 996
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8388641269549169,
+      "learning_rate": 7.751756972518341e-06,
+      "loss": 0.9719,
+      "step": 997
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2621661445664245,
+      "learning_rate": 7.74720780297324e-06,
+      "loss": 0.7821,
+      "step": 998
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6088860116137997,
+      "learning_rate": 7.742655373554982e-06,
+      "loss": 0.9552,
+      "step": 999
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3069504177458988,
+      "learning_rate": 7.73809968966554e-06,
+      "loss": 0.885,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2262048413044229,
+      "learning_rate": 7.733540756710752e-06,
+      "loss": 0.7823,
+      "step": 1001
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0786921811512518,
+      "learning_rate": 7.728978580100304e-06,
+      "loss": 0.8818,
+      "step": 1002
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8208923407742453,
+      "learning_rate": 7.72441316524774e-06,
+      "loss": 0.5364,
+      "step": 1003
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.797117469608383,
+      "learning_rate": 7.719844517570437e-06,
+      "loss": 0.9051,
+      "step": 1004
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0421093238214658,
+      "learning_rate": 7.715272642489612e-06,
+      "loss": 0.8128,
+      "step": 1005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9270379316265225,
+      "learning_rate": 7.71069754543032e-06,
+      "loss": 0.8863,
+      "step": 1006
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6581336546787458,
+      "learning_rate": 7.706119231821423e-06,
+      "loss": 0.7783,
+      "step": 1007
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.631583279800118,
+      "learning_rate": 7.701537707095615e-06,
+      "loss": 0.7855,
+      "step": 1008
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9341468116931029,
+      "learning_rate": 7.69695297668939e-06,
+      "loss": 0.9793,
+      "step": 1009
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6204932481526119,
+      "learning_rate": 7.692365046043053e-06,
+      "loss": 0.9782,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3022135661992034,
+      "learning_rate": 7.687773920600699e-06,
+      "loss": 0.9242,
+      "step": 1011
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9205848580181927,
+      "learning_rate": 7.683179605810221e-06,
+      "loss": 0.4714,
+      "step": 1012
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3151060793455438,
+      "learning_rate": 7.678582107123295e-06,
+      "loss": 0.8683,
+      "step": 1013
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8128950648003733,
+      "learning_rate": 7.673981429995372e-06,
+      "loss": 0.8802,
+      "step": 1014
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0289533886881097,
+      "learning_rate": 7.669377579885674e-06,
+      "loss": 0.8198,
+      "step": 1015
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8825901538983683,
+      "learning_rate": 7.664770562257197e-06,
+      "loss": 0.997,
+      "step": 1016
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0629027489913274,
+      "learning_rate": 7.660160382576683e-06,
+      "loss": 0.8576,
+      "step": 1017
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0352060704859878,
+      "learning_rate": 7.655547046314635e-06,
+      "loss": 0.6718,
+      "step": 1018
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7732580323618514,
+      "learning_rate": 7.650930558945299e-06,
+      "loss": 0.7735,
+      "step": 1019
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7735686642664795,
+      "learning_rate": 7.646310925946659e-06,
+      "loss": 0.6913,
+      "step": 1020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.040953763589302,
+      "learning_rate": 7.641688152800433e-06,
+      "loss": 1.0254,
+      "step": 1021
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8996151733026105,
+      "learning_rate": 7.637062244992065e-06,
+      "loss": 0.8196,
+      "step": 1022
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9054617915608173,
+      "learning_rate": 7.63243320801072e-06,
+      "loss": 0.9015,
+      "step": 1023
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7529356003796459,
+      "learning_rate": 7.62780104734927e-06,
+      "loss": 0.9139,
+      "step": 1024
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9080598466185776,
+      "learning_rate": 7.6231657685043036e-06,
+      "loss": 0.6567,
+      "step": 1025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9374441711647022,
+      "learning_rate": 7.6185273769761015e-06,
+      "loss": 0.9889,
+      "step": 1026
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.259815335347003,
+      "learning_rate": 7.61388587826864e-06,
+      "loss": 0.8032,
+      "step": 1027
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8551565681696168,
+      "learning_rate": 7.609241277889583e-06,
+      "loss": 0.9947,
+      "step": 1028
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9405166729236499,
+      "learning_rate": 7.6045935813502756e-06,
+      "loss": 0.8011,
+      "step": 1029
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.768129290504335,
+      "learning_rate": 7.599942794165738e-06,
+      "loss": 0.8805,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6458916343463573,
+      "learning_rate": 7.595288921854651e-06,
+      "loss": 0.5078,
+      "step": 1031
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9940623690490097,
+      "learning_rate": 7.590631969939364e-06,
+      "loss": 0.8916,
+      "step": 1032
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0424585837470128,
+      "learning_rate": 7.58597194394588e-06,
+      "loss": 1.0165,
+      "step": 1033
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8638826914417733,
+      "learning_rate": 7.581308849403843e-06,
+      "loss": 0.7855,
+      "step": 1034
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.903662391147223,
+      "learning_rate": 7.5766426918465455e-06,
+      "loss": 0.9338,
+      "step": 1035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6829896535882392,
+      "learning_rate": 7.5719734768109125e-06,
+      "loss": 0.9159,
+      "step": 1036
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9196836801632842,
+      "learning_rate": 7.567301209837493e-06,
+      "loss": 0.8095,
+      "step": 1037
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7862694732455047,
+      "learning_rate": 7.5626258964704634e-06,
+      "loss": 0.9286,
+      "step": 1038
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6271390146320459,
+      "learning_rate": 7.557947542257613e-06,
+      "loss": 0.6664,
+      "step": 1039
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9458259813128382,
+      "learning_rate": 7.553266152750338e-06,
+      "loss": 0.9078,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.175479381283007,
+      "learning_rate": 7.548581733503636e-06,
+      "loss": 0.9285,
+      "step": 1041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6400693854181663,
+      "learning_rate": 7.5438942900761035e-06,
+      "loss": 0.9623,
+      "step": 1042
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0117328666631071,
+      "learning_rate": 7.539203828029919e-06,
+      "loss": 0.8453,
+      "step": 1043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8033644203072535,
+      "learning_rate": 7.534510352930853e-06,
+      "loss": 0.7603,
+      "step": 1044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2838525899558926,
+      "learning_rate": 7.529813870348238e-06,
+      "loss": 0.9572,
+      "step": 1045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0939580166021312,
+      "learning_rate": 7.525114385854988e-06,
+      "loss": 0.7115,
+      "step": 1046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9050193991046418,
+      "learning_rate": 7.520411905027571e-06,
+      "loss": 1.0383,
+      "step": 1047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5960555823237125,
+      "learning_rate": 7.515706433446013e-06,
+      "loss": 0.782,
+      "step": 1048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1944213358772537,
+      "learning_rate": 7.51099797669389e-06,
+      "loss": 0.8776,
+      "step": 1049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.102035561083673,
+      "learning_rate": 7.506286540358318e-06,
+      "loss": 0.7766,
+      "step": 1050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6980703146938201,
+      "learning_rate": 7.50157213002995e-06,
+      "loss": 0.9222,
+      "step": 1051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9081866793707962,
+      "learning_rate": 7.4968547513029685e-06,
+      "loss": 0.7944,
+      "step": 1052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7704346559365427,
+      "learning_rate": 7.492134409775077e-06,
+      "loss": 0.8285,
+      "step": 1053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1969236541113868,
+      "learning_rate": 7.4874111110474955e-06,
+      "loss": 0.9177,
+      "step": 1054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.894577700145311,
+      "learning_rate": 7.482684860724951e-06,
+      "loss": 0.9185,
+      "step": 1055
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6731811928748919,
+      "learning_rate": 7.477955664415678e-06,
+      "loss": 0.4605,
+      "step": 1056
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8875398867394554,
+      "learning_rate": 7.473223527731401e-06,
+      "loss": 0.9302,
+      "step": 1057
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9676287958627418,
+      "learning_rate": 7.4684884562873375e-06,
+      "loss": 0.8202,
+      "step": 1058
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7102875419028689,
+      "learning_rate": 7.463750455702185e-06,
+      "loss": 0.9034,
+      "step": 1059
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.512004082576946,
+      "learning_rate": 7.459009531598118e-06,
+      "loss": 0.912,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8133400663918698,
+      "learning_rate": 7.4542656896007804e-06,
+      "loss": 0.7414,
+      "step": 1061
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8022659280832216,
+      "learning_rate": 7.449518935339276e-06,
+      "loss": 1.0322,
+      "step": 1062
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.082594196687515,
+      "learning_rate": 7.444769274446168e-06,
+      "loss": 0.7803,
+      "step": 1063
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7689284494063036,
+      "learning_rate": 7.440016712557468e-06,
+      "loss": 0.8766,
+      "step": 1064
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6805491251563369,
+      "learning_rate": 7.435261255312626e-06,
+      "loss": 0.5801,
+      "step": 1065
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9129187107443577,
+      "learning_rate": 7.430502908354532e-06,
+      "loss": 0.9552,
+      "step": 1066
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7904575636987102,
+      "learning_rate": 7.4257416773295045e-06,
+      "loss": 0.7357,
+      "step": 1067
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.298875254806093,
+      "learning_rate": 7.420977567887281e-06,
+      "loss": 0.7654,
+      "step": 1068
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7881390775452288,
+      "learning_rate": 7.416210585681019e-06,
+      "loss": 0.9732,
+      "step": 1069
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8198614323062058,
+      "learning_rate": 7.411440736367281e-06,
+      "loss": 0.8276,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7100340322246931,
+      "learning_rate": 7.406668025606035e-06,
+      "loss": 0.8978,
+      "step": 1071
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6890110368813483,
+      "learning_rate": 7.401892459060641e-06,
+      "loss": 0.8738,
+      "step": 1072
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7700836609357556,
+      "learning_rate": 7.397114042397852e-06,
+      "loss": 0.9646,
+      "step": 1073
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6227961423142602,
+      "learning_rate": 7.392332781287798e-06,
+      "loss": 0.7505,
+      "step": 1074
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9003469763301504,
+      "learning_rate": 7.387548681403986e-06,
+      "loss": 0.6569,
+      "step": 1075
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2059568201062187,
+      "learning_rate": 7.382761748423296e-06,
+      "loss": 0.9523,
+      "step": 1076
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7036934257688677,
+      "learning_rate": 7.377971988025964e-06,
+      "loss": 0.9798,
+      "step": 1077
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7520443645098317,
+      "learning_rate": 7.373179405895582e-06,
+      "loss": 0.7982,
+      "step": 1078
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8234806449758567,
+      "learning_rate": 7.368384007719093e-06,
+      "loss": 0.7965,
+      "step": 1079
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.874987261796812,
+      "learning_rate": 7.36358579918678e-06,
+      "loss": 0.8427,
+      "step": 1080
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8234891392269017,
+      "learning_rate": 7.35878478599226e-06,
+      "loss": 0.9125,
+      "step": 1081
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3257139357330774,
+      "learning_rate": 7.353980973832479e-06,
+      "loss": 0.7566,
+      "step": 1082
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7059380048199736,
+      "learning_rate": 7.349174368407703e-06,
+      "loss": 0.7143,
+      "step": 1083
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6366955095510047,
+      "learning_rate": 7.3443649754215175e-06,
+      "loss": 0.8213,
+      "step": 1084
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6190274249130907,
+      "learning_rate": 7.339552800580807e-06,
+      "loss": 0.995,
+      "step": 1085
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.040582093631083,
+      "learning_rate": 7.3347378495957655e-06,
+      "loss": 0.7801,
+      "step": 1086
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9303557411019188,
+      "learning_rate": 7.329920128179875e-06,
+      "loss": 0.8328,
+      "step": 1087
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0167090611855578,
+      "learning_rate": 7.325099642049909e-06,
+      "loss": 0.9605,
+      "step": 1088
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8067185705552913,
+      "learning_rate": 7.320276396925921e-06,
+      "loss": 0.844,
+      "step": 1089
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1029800521628075,
+      "learning_rate": 7.3154503985312366e-06,
+      "loss": 0.9206,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9151217724788081,
+      "learning_rate": 7.310621652592449e-06,
+      "loss": 0.8348,
+      "step": 1091
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0167189970304824,
+      "learning_rate": 7.305790164839411e-06,
+      "loss": 0.9326,
+      "step": 1092
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6265654668741276,
+      "learning_rate": 7.300955941005232e-06,
+      "loss": 0.4529,
+      "step": 1093
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8457298669244813,
+      "learning_rate": 7.296118986826266e-06,
+      "loss": 0.7684,
+      "step": 1094
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7576145484819609,
+      "learning_rate": 7.291279308042105e-06,
+      "loss": 0.9262,
+      "step": 1095
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0794333223803838,
+      "learning_rate": 7.286436910395579e-06,
+      "loss": 0.8142,
+      "step": 1096
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9907106463557742,
+      "learning_rate": 7.281591799632738e-06,
+      "loss": 1.0324,
+      "step": 1097
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.47553919472812695,
+      "learning_rate": 7.276743981502856e-06,
+      "loss": 0.9581,
+      "step": 1098
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9514807517236695,
+      "learning_rate": 7.271893461758418e-06,
+      "loss": 0.8591,
+      "step": 1099
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.839383923383816,
+      "learning_rate": 7.267040246155115e-06,
+      "loss": 0.5166,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5724760268701603,
+      "learning_rate": 7.262184340451838e-06,
+      "loss": 0.8515,
+      "step": 1101
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8376379077103544,
+      "learning_rate": 7.2573257504106665e-06,
+      "loss": 0.832,
+      "step": 1102
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.776764717900711,
+      "learning_rate": 7.252464481796869e-06,
+      "loss": 0.7515,
+      "step": 1103
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8600893844940484,
+      "learning_rate": 7.247600540378893e-06,
+      "loss": 1.0153,
+      "step": 1104
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.554942102850519,
+      "learning_rate": 7.242733931928352e-06,
+      "loss": 0.8014,
+      "step": 1105
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9374932086891602,
+      "learning_rate": 7.237864662220032e-06,
+      "loss": 0.772,
+      "step": 1106
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9072647644515902,
+      "learning_rate": 7.232992737031871e-06,
+      "loss": 0.8342,
+      "step": 1107
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7101727659196688,
+      "learning_rate": 7.2281181621449594e-06,
+      "loss": 1.0232,
+      "step": 1108
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9096215418254705,
+      "learning_rate": 7.223240943343533e-06,
+      "loss": 0.6449,
+      "step": 1109
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3182757291386697,
+      "learning_rate": 7.2183610864149655e-06,
+      "loss": 0.7756,
+      "step": 1110
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9319457016068583,
+      "learning_rate": 7.213478597149758e-06,
+      "loss": 0.9944,
+      "step": 1111
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5654342812883145,
+      "learning_rate": 7.208593481341536e-06,
+      "loss": 0.8903,
+      "step": 1112
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1402051089444134,
+      "learning_rate": 7.203705744787044e-06,
+      "loss": 0.7282,
+      "step": 1113
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9294127465484998,
+      "learning_rate": 7.198815393286136e-06,
+      "loss": 1.0281,
+      "step": 1114
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3011185839280694,
+      "learning_rate": 7.193922432641767e-06,
+      "loss": 0.6644,
+      "step": 1115
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4905861135450518,
+      "learning_rate": 7.1890268686599875e-06,
+      "loss": 0.8269,
+      "step": 1116
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8994124165040088,
+      "learning_rate": 7.18412870714994e-06,
+      "loss": 0.8006,
+      "step": 1117
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.094986216393371,
+      "learning_rate": 7.179227953923848e-06,
+      "loss": 0.885,
+      "step": 1118
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8352196698352151,
+      "learning_rate": 7.1743246147970095e-06,
+      "loss": 0.6045,
+      "step": 1119
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9685109753804898,
+      "learning_rate": 7.169418695587791e-06,
+      "loss": 0.8697,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8447602163358183,
+      "learning_rate": 7.1645102021176225e-06,
+      "loss": 1.0596,
+      "step": 1121
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3558033074183098,
+      "learning_rate": 7.1595991402109865e-06,
+      "loss": 0.7786,
+      "step": 1122
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6263239367320915,
+      "learning_rate": 7.154685515695415e-06,
+      "loss": 0.9268,
+      "step": 1123
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4185427905522858,
+      "learning_rate": 7.14976933440148e-06,
+      "loss": 0.8544,
+      "step": 1124
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9059748529608416,
+      "learning_rate": 7.144850602162785e-06,
+      "loss": 0.7074,
+      "step": 1125
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7788570988921534,
+      "learning_rate": 7.139929324815965e-06,
+      "loss": 0.8635,
+      "step": 1126
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.652269845843319,
+      "learning_rate": 7.135005508200673e-06,
+      "loss": 0.5325,
+      "step": 1127
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6478304505640102,
+      "learning_rate": 7.130079158159573e-06,
+      "loss": 0.929,
+      "step": 1128
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.143261818168066,
+      "learning_rate": 7.12515028053834e-06,
+      "loss": 0.775,
+      "step": 1129
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.245328994889597,
+      "learning_rate": 7.120218881185644e-06,
+      "loss": 0.9895,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5608399467409726,
+      "learning_rate": 7.115284965953147e-06,
+      "loss": 0.96,
+      "step": 1131
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9166365396285775,
+      "learning_rate": 7.110348540695501e-06,
+      "loss": 0.8682,
+      "step": 1132
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5495250629909987,
+      "learning_rate": 7.105409611270332e-06,
+      "loss": 0.9158,
+      "step": 1133
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1295779304628286,
+      "learning_rate": 7.100468183538241e-06,
+      "loss": 0.7402,
+      "step": 1134
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5756273751705822,
+      "learning_rate": 7.095524263362787e-06,
+      "loss": 0.9576,
+      "step": 1135
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.140812510716515,
+      "learning_rate": 7.090577856610497e-06,
+      "loss": 0.4864,
+      "step": 1136
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8635566832549894,
+      "learning_rate": 7.0856289691508395e-06,
+      "loss": 0.8098,
+      "step": 1137
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8654882837475819,
+      "learning_rate": 7.08067760685623e-06,
+      "loss": 0.8161,
+      "step": 1138
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.204875110884719,
+      "learning_rate": 7.07572377560202e-06,
+      "loss": 0.7172,
+      "step": 1139
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8659094371151999,
+      "learning_rate": 7.070767481266493e-06,
+      "loss": 1.0083,
+      "step": 1140
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.816214305729195,
+      "learning_rate": 7.065808729730853e-06,
+      "loss": 0.8717,
+      "step": 1141
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8750464057168982,
+      "learning_rate": 7.0608475268792186e-06,
+      "loss": 0.8559,
+      "step": 1142
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9498912654406104,
+      "learning_rate": 7.055883878598622e-06,
+      "loss": 0.8086,
+      "step": 1143
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9475559728912553,
+      "learning_rate": 7.050917790778992e-06,
+      "loss": 0.9414,
+      "step": 1144
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6320197659236363,
+      "learning_rate": 7.045949269313153e-06,
+      "loss": 0.5524,
+      "step": 1145
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0490216689168523,
+      "learning_rate": 7.040978320096819e-06,
+      "loss": 0.7967,
+      "step": 1146
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6389003018602956,
+      "learning_rate": 7.036004949028587e-06,
+      "loss": 0.945,
+      "step": 1147
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.773991348507658,
+      "learning_rate": 7.031029162009921e-06,
+      "loss": 0.749,
+      "step": 1148
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7407889086611532,
+      "learning_rate": 7.026050964945158e-06,
+      "loss": 0.866,
+      "step": 1149
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8971331293535137,
+      "learning_rate": 7.021070363741492e-06,
+      "loss": 0.9104,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0318500147803904,
+      "learning_rate": 7.016087364308971e-06,
+      "loss": 0.7299,
+      "step": 1151
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3420002733497536,
+      "learning_rate": 7.0111019725604856e-06,
+      "loss": 0.9241,
+      "step": 1152
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9664437130941114,
+      "learning_rate": 7.006114194411772e-06,
+      "loss": 0.5706,
+      "step": 1153
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.872068997093268,
+      "learning_rate": 7.00112403578139e-06,
+      "loss": 0.967,
+      "step": 1154
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9154252229033849,
+      "learning_rate": 6.996131502590729e-06,
+      "loss": 0.9138,
+      "step": 1155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9416640254354681,
+      "learning_rate": 6.991136600763997e-06,
+      "loss": 0.8749,
+      "step": 1156
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9761275314905827,
+      "learning_rate": 6.986139336228208e-06,
+      "loss": 0.9622,
+      "step": 1157
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1517864710795096,
+      "learning_rate": 6.9811397149131835e-06,
+      "loss": 0.8269,
+      "step": 1158
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8208612517714133,
+      "learning_rate": 6.976137742751538e-06,
+      "loss": 0.9305,
+      "step": 1159
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6633806773579906,
+      "learning_rate": 6.971133425678682e-06,
+      "loss": 0.92,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0133215801975257,
+      "learning_rate": 6.9661267696328015e-06,
+      "loss": 0.9023,
+      "step": 1161
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7582735822314888,
+      "learning_rate": 6.961117780554862e-06,
+      "loss": 0.7123,
+      "step": 1162
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7068760809126409,
+      "learning_rate": 6.956106464388596e-06,
+      "loss": 0.5383,
+      "step": 1163
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9511756028367826,
+      "learning_rate": 6.9510928270804975e-06,
+      "loss": 0.9171,
+      "step": 1164
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0368816213638288,
+      "learning_rate": 6.9460768745798144e-06,
+      "loss": 0.785,
+      "step": 1165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9395465616103404,
+      "learning_rate": 6.941058612838544e-06,
+      "loss": 0.9192,
+      "step": 1166
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7396632467423322,
+      "learning_rate": 6.936038047811421e-06,
+      "loss": 0.9164,
+      "step": 1167
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4342150260299302,
+      "learning_rate": 6.931015185455915e-06,
+      "loss": 0.8633,
+      "step": 1168
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8151253045604222,
+      "learning_rate": 6.925990031732218e-06,
+      "loss": 0.9538,
+      "step": 1169
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5477399483692502,
+      "learning_rate": 6.9209625926032485e-06,
+      "loss": 0.8573,
+      "step": 1170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2848084739575163,
+      "learning_rate": 6.91593287403463e-06,
+      "loss": 0.9098,
+      "step": 1171
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5717768372623621,
+      "learning_rate": 6.910900881994692e-06,
+      "loss": 0.765,
+      "step": 1172
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9214920115632345,
+      "learning_rate": 6.905866622454466e-06,
+      "loss": 0.9794,
+      "step": 1173
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.123514711377056,
+      "learning_rate": 6.900830101387667e-06,
+      "loss": 0.6874,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9591195194171481,
+      "learning_rate": 6.895791324770702e-06,
+      "loss": 0.8602,
+      "step": 1175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8915865200699761,
+      "learning_rate": 6.890750298582647e-06,
+      "loss": 0.8345,
+      "step": 1176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.771458480866771,
+      "learning_rate": 6.885707028805253e-06,
+      "loss": 0.8788,
+      "step": 1177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3014312886989445,
+      "learning_rate": 6.8806615214229275e-06,
+      "loss": 0.9111,
+      "step": 1178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.932918971467558,
+      "learning_rate": 6.875613782422737e-06,
+      "loss": 0.9338,
+      "step": 1179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6074042917276643,
+      "learning_rate": 6.870563817794398e-06,
+      "loss": 0.6185,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4815773993633037,
+      "learning_rate": 6.865511633530263e-06,
+      "loss": 0.8302,
+      "step": 1181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0998175505162082,
+      "learning_rate": 6.860457235625322e-06,
+      "loss": 0.8268,
+      "step": 1182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9414302580918265,
+      "learning_rate": 6.855400630077189e-06,
+      "loss": 0.9085,
+      "step": 1183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9947700987454516,
+      "learning_rate": 6.8503418228861e-06,
+      "loss": 0.7305,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5035024463953894,
+      "learning_rate": 6.845280820054904e-06,
+      "loss": 0.9642,
+      "step": 1185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.236289102289844,
+      "learning_rate": 6.840217627589052e-06,
+      "loss": 0.7863,
+      "step": 1186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1245541265531942,
+      "learning_rate": 6.835152251496596e-06,
+      "loss": 0.6703,
+      "step": 1187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0468009209534792,
+      "learning_rate": 6.8300846977881785e-06,
+      "loss": 0.9715,
+      "step": 1188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6579768411419812,
+      "learning_rate": 6.825014972477024e-06,
+      "loss": 0.4529,
+      "step": 1189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1616103355880574,
+      "learning_rate": 6.819943081578939e-06,
+      "loss": 0.916,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9162206110142164,
+      "learning_rate": 6.814869031112294e-06,
+      "loss": 0.8111,
+      "step": 1191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9201501123930744,
+      "learning_rate": 6.809792827098025e-06,
+      "loss": 0.7749,
+      "step": 1192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.347090566131551,
+      "learning_rate": 6.804714475559624e-06,
+      "loss": 0.797,
+      "step": 1193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2317816077767876,
+      "learning_rate": 6.799633982523128e-06,
+      "loss": 0.8818,
+      "step": 1194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8109912669434763,
+      "learning_rate": 6.794551354017118e-06,
+      "loss": 0.9477,
+      "step": 1195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5859769714742059,
+      "learning_rate": 6.7894665960727105e-06,
+      "loss": 0.9453,
+      "step": 1196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.160912134061732,
+      "learning_rate": 6.784379714723544e-06,
+      "loss": 0.9048,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8911929718745089,
+      "learning_rate": 6.7792907160057796e-06,
+      "loss": 0.4981,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1101579093639147,
+      "learning_rate": 6.774199605958091e-06,
+      "loss": 0.8212,
+      "step": 1199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7237091409676937,
+      "learning_rate": 6.7691063906216585e-06,
+      "loss": 0.8031,
+      "step": 1200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8221819413739208,
+      "learning_rate": 6.764011076040157e-06,
+      "loss": 0.9612,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.109349045030578,
+      "learning_rate": 6.758913668259753e-06,
+      "loss": 0.9263,
+      "step": 1202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3256621856344526,
+      "learning_rate": 6.7538141733291e-06,
+      "loss": 0.8534,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.924135039999999,
+      "learning_rate": 6.748712597299327e-06,
+      "loss": 0.7765,
+      "step": 1204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8672294207900395,
+      "learning_rate": 6.743608946224029e-06,
+      "loss": 0.6714,
+      "step": 1205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7630615404014595,
+      "learning_rate": 6.738503226159269e-06,
+      "loss": 0.6067,
+      "step": 1206
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7571332916783056,
+      "learning_rate": 6.73339544316356e-06,
+      "loss": 0.9593,
+      "step": 1207
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1768656530600423,
+      "learning_rate": 6.728285603297865e-06,
+      "loss": 0.8576,
+      "step": 1208
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9288949371765874,
+      "learning_rate": 6.723173712625588e-06,
+      "loss": 0.9242,
+      "step": 1209
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6905065700968571,
+      "learning_rate": 6.7180597772125665e-06,
+      "loss": 0.882,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0334662774676926,
+      "learning_rate": 6.712943803127063e-06,
+      "loss": 0.6132,
+      "step": 1211
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1866744077921096,
+      "learning_rate": 6.707825796439761e-06,
+      "loss": 1.0319,
+      "step": 1212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.009122571248358,
+      "learning_rate": 6.702705763223756e-06,
+      "loss": 0.7896,
+      "step": 1213
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8752140980641042,
+      "learning_rate": 6.697583709554545e-06,
+      "loss": 1.0161,
+      "step": 1214
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7204635647834511,
+      "learning_rate": 6.692459641510025e-06,
+      "loss": 0.772,
+      "step": 1215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9343040082102373,
+      "learning_rate": 6.687333565170484e-06,
+      "loss": 0.5986,
+      "step": 1216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.050286626503953,
+      "learning_rate": 6.682205486618592e-06,
+      "loss": 0.6702,
+      "step": 1217
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9106619291902527,
+      "learning_rate": 6.677075411939396e-06,
+      "loss": 0.913,
+      "step": 1218
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.990423024447723,
+      "learning_rate": 6.671943347220308e-06,
+      "loss": 0.9285,
+      "step": 1219
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9550923395204381,
+      "learning_rate": 6.666809298551105e-06,
+      "loss": 0.8438,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.206941123993105,
+      "learning_rate": 6.661673272023917e-06,
+      "loss": 0.9146,
+      "step": 1221
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8043618317396037,
+      "learning_rate": 6.656535273733222e-06,
+      "loss": 0.8915,
+      "step": 1222
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7401394946968677,
+      "learning_rate": 6.651395309775837e-06,
+      "loss": 0.9065,
+      "step": 1223
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2326666708989282,
+      "learning_rate": 6.646253386250909e-06,
+      "loss": 0.7758,
+      "step": 1224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6483102573736766,
+      "learning_rate": 6.641109509259916e-06,
+      "loss": 0.6544,
+      "step": 1225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.73083690280401,
+      "learning_rate": 6.635963684906646e-06,
+      "loss": 0.9472,
+      "step": 1226
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3252531041755309,
+      "learning_rate": 6.630815919297205e-06,
+      "loss": 0.8088,
+      "step": 1227
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.799567750318313,
+      "learning_rate": 6.625666218540001e-06,
+      "loss": 0.9934,
+      "step": 1228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2633893422871425,
+      "learning_rate": 6.620514588745736e-06,
+      "loss": 0.817,
+      "step": 1229
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.926083824961928,
+      "learning_rate": 6.6153610360274014e-06,
+      "loss": 0.7814,
+      "step": 1230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1851272252625251,
+      "learning_rate": 6.610205566500272e-06,
+      "loss": 0.887,
+      "step": 1231
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.018703208671171,
+      "learning_rate": 6.605048186281895e-06,
+      "loss": 0.7741,
+      "step": 1232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.850075662767509,
+      "learning_rate": 6.5998889014920884e-06,
+      "loss": 0.685,
+      "step": 1233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.581681750012993,
+      "learning_rate": 6.594727718252925e-06,
+      "loss": 0.8827,
+      "step": 1234
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9115517978596477,
+      "learning_rate": 6.589564642688736e-06,
+      "loss": 0.9661,
+      "step": 1235
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3705587145064546,
+      "learning_rate": 6.584399680926092e-06,
+      "loss": 0.792,
+      "step": 1236
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1733456859937748,
+      "learning_rate": 6.579232839093808e-06,
+      "loss": 0.9095,
+      "step": 1237
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1852387752153666,
+      "learning_rate": 6.574064123322925e-06,
+      "loss": 0.8079,
+      "step": 1238
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2799427782250876,
+      "learning_rate": 6.56889353974671e-06,
+      "loss": 0.833,
+      "step": 1239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3095490548667523,
+      "learning_rate": 6.563721094500648e-06,
+      "loss": 0.9104,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.544899181673468,
+      "learning_rate": 6.558546793722429e-06,
+      "loss": 0.9487,
+      "step": 1241
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.984036327228971,
+      "learning_rate": 6.5533706435519454e-06,
+      "loss": 0.5467,
+      "step": 1242
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2997321719593171,
+      "learning_rate": 6.548192650131289e-06,
+      "loss": 0.7659,
+      "step": 1243
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.409932866569754,
+      "learning_rate": 6.543012819604732e-06,
+      "loss": 0.8077,
+      "step": 1244
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.007245901045933,
+      "learning_rate": 6.537831158118733e-06,
+      "loss": 0.9302,
+      "step": 1245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6658316716038252,
+      "learning_rate": 6.5326476718219165e-06,
+      "loss": 0.9236,
+      "step": 1246
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0756842513095213,
+      "learning_rate": 6.5274623668650805e-06,
+      "loss": 0.8308,
+      "step": 1247
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9575725637165821,
+      "learning_rate": 6.522275249401173e-06,
+      "loss": 0.8838,
+      "step": 1248
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4255529129061295,
+      "learning_rate": 6.517086325585296e-06,
+      "loss": 0.7466,
+      "step": 1249
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9315449349246908,
+      "learning_rate": 6.511895601574698e-06,
+      "loss": 0.907,
+      "step": 1250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9076752386859259,
+      "learning_rate": 6.5067030835287605e-06,
+      "loss": 0.5541,
+      "step": 1251
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9463305630214507,
+      "learning_rate": 6.50150877760899e-06,
+      "loss": 0.9675,
+      "step": 1252
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5608346404590374,
+      "learning_rate": 6.496312689979024e-06,
+      "loss": 0.7445,
+      "step": 1253
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1279755334813248,
+      "learning_rate": 6.491114826804607e-06,
+      "loss": 0.8409,
+      "step": 1254
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7886495916999157,
+      "learning_rate": 6.485915194253591e-06,
+      "loss": 0.8201,
+      "step": 1255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.356201034088837,
+      "learning_rate": 6.48071379849593e-06,
+      "loss": 0.7505,
+      "step": 1256
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6887002042513384,
+      "learning_rate": 6.475510645703669e-06,
+      "loss": 0.9495,
+      "step": 1257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6257779787963043,
+      "learning_rate": 6.470305742050938e-06,
+      "loss": 0.8955,
+      "step": 1258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8345860783742126,
+      "learning_rate": 6.465099093713944e-06,
+      "loss": 1.0711,
+      "step": 1259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6542028472803113,
+      "learning_rate": 6.459890706870965e-06,
+      "loss": 0.6768,
+      "step": 1260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.289979058754802,
+      "learning_rate": 6.454680587702342e-06,
+      "loss": 0.8318,
+      "step": 1261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9351106119600974,
+      "learning_rate": 6.449468742390472e-06,
+      "loss": 0.8914,
+      "step": 1262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.428922291446092,
+      "learning_rate": 6.444255177119799e-06,
+      "loss": 0.7606,
+      "step": 1263
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9610584278985153,
+      "learning_rate": 6.43903989807681e-06,
+      "loss": 0.7497,
+      "step": 1264
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.417101003170088,
+      "learning_rate": 6.433822911450022e-06,
+      "loss": 0.8393,
+      "step": 1265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1555990066362924,
+      "learning_rate": 6.42860422342998e-06,
+      "loss": 0.922,
+      "step": 1266
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3740612319040875,
+      "learning_rate": 6.423383840209251e-06,
+      "loss": 0.7712,
+      "step": 1267
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.746440067069165,
+      "learning_rate": 6.41816176798241e-06,
+      "loss": 0.8066,
+      "step": 1268
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7477247382345116,
+      "learning_rate": 6.412938012946035e-06,
+      "loss": 0.6282,
+      "step": 1269
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4323695407523391,
+      "learning_rate": 6.407712581298705e-06,
+      "loss": 0.7759,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7399212107866275,
+      "learning_rate": 6.4024854792409864e-06,
+      "loss": 0.9947,
+      "step": 1271
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4045256440565486,
+      "learning_rate": 6.397256712975426e-06,
+      "loss": 0.7085,
+      "step": 1272
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1050456407771019,
+      "learning_rate": 6.392026288706549e-06,
+      "loss": 0.9781,
+      "step": 1273
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2165767848841602,
+      "learning_rate": 6.386794212640846e-06,
+      "loss": 0.767,
+      "step": 1274
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.52490279742241,
+      "learning_rate": 6.381560490986766e-06,
+      "loss": 0.8293,
+      "step": 1275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.126856488991011,
+      "learning_rate": 6.376325129954712e-06,
+      "loss": 0.9152,
+      "step": 1276
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5111803694182788,
+      "learning_rate": 6.371088135757038e-06,
+      "loss": 0.6841,
+      "step": 1277
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9457998875553426,
+      "learning_rate": 6.365849514608025e-06,
+      "loss": 0.6474,
+      "step": 1278
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1024531118220533,
+      "learning_rate": 6.36060927272389e-06,
+      "loss": 0.683,
+      "step": 1279
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2187559556905074,
+      "learning_rate": 6.3553674163227786e-06,
+      "loss": 0.6838,
+      "step": 1280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7733187073494139,
+      "learning_rate": 6.3501239516247424e-06,
+      "loss": 0.912,
+      "step": 1281
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0877060660824964,
+      "learning_rate": 6.344878884851748e-06,
+      "loss": 0.7554,
+      "step": 1282
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9414545018093784,
+      "learning_rate": 6.339632222227661e-06,
+      "loss": 1.0241,
+      "step": 1283
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7321885427661501,
+      "learning_rate": 6.33438396997824e-06,
+      "loss": 0.9533,
+      "step": 1284
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.857702227152555,
+      "learning_rate": 6.329134134331129e-06,
+      "loss": 0.879,
+      "step": 1285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8486763929584309,
+      "learning_rate": 6.3238827215158575e-06,
+      "loss": 0.5252,
+      "step": 1286
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3858337954485223,
+      "learning_rate": 6.318629737763818e-06,
+      "loss": 0.8192,
+      "step": 1287
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9438469485874288,
+      "learning_rate": 6.313375189308269e-06,
+      "loss": 0.7766,
+      "step": 1288
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9508423947680671,
+      "learning_rate": 6.3081190823843295e-06,
+      "loss": 0.8073,
+      "step": 1289
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4652354509789247,
+      "learning_rate": 6.302861423228967e-06,
+      "loss": 0.9896,
+      "step": 1290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4402240406545395,
+      "learning_rate": 6.297602218080988e-06,
+      "loss": 0.7602,
+      "step": 1291
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5733341976153,
+      "learning_rate": 6.292341473181034e-06,
+      "loss": 0.7596,
+      "step": 1292
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0492939317251995,
+      "learning_rate": 6.287079194771576e-06,
+      "loss": 0.8163,
+      "step": 1293
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8456451429638168,
+      "learning_rate": 6.281815389096903e-06,
+      "loss": 1.0001,
+      "step": 1294
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9212509390016326,
+      "learning_rate": 6.276550062403115e-06,
+      "loss": 0.621,
+      "step": 1295
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0768576312432112,
+      "learning_rate": 6.271283220938119e-06,
+      "loss": 0.7721,
+      "step": 1296
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0125761358475456,
+      "learning_rate": 6.26601487095162e-06,
+      "loss": 0.9497,
+      "step": 1297
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6011264115744833,
+      "learning_rate": 6.260745018695112e-06,
+      "loss": 0.9213,
+      "step": 1298
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9959624161099369,
+      "learning_rate": 6.25547367042187e-06,
+      "loss": 0.7281,
+      "step": 1299
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9587986673955071,
+      "learning_rate": 6.250200832386947e-06,
+      "loss": 0.9736,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3352778981052014,
+      "learning_rate": 6.244926510847162e-06,
+      "loss": 0.672,
+      "step": 1301
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1172366365813378,
+      "learning_rate": 6.239650712061093e-06,
+      "loss": 0.8747,
+      "step": 1302
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2764381971678307,
+      "learning_rate": 6.2343734422890745e-06,
+      "loss": 0.8447,
+      "step": 1303
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2379752116455003,
+      "learning_rate": 6.229094707793185e-06,
+      "loss": 0.8231,
+      "step": 1304
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9707107961079281,
+      "learning_rate": 6.223814514837241e-06,
+      "loss": 0.617,
+      "step": 1305
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.085701721974298,
+      "learning_rate": 6.2185328696867866e-06,
+      "loss": 0.8789,
+      "step": 1306
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7744852960099238,
+      "learning_rate": 6.2132497786090955e-06,
+      "loss": 1.024,
+      "step": 1307
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2999336228859184,
+      "learning_rate": 6.207965247873151e-06,
+      "loss": 0.7818,
+      "step": 1308
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.651658027288747,
+      "learning_rate": 6.202679283749647e-06,
+      "loss": 0.9575,
+      "step": 1309
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4059413100540181,
+      "learning_rate": 6.197391892510982e-06,
+      "loss": 0.868,
+      "step": 1310
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1747589150430753,
+      "learning_rate": 6.192103080431242e-06,
+      "loss": 0.6845,
+      "step": 1311
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0618020328513644,
+      "learning_rate": 6.1868128537862005e-06,
+      "loss": 0.8496,
+      "step": 1312
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3891138188435146,
+      "learning_rate": 6.181521218853312e-06,
+      "loss": 0.5271,
+      "step": 1313
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.256412819614988,
+      "learning_rate": 6.176228181911699e-06,
+      "loss": 0.9478,
+      "step": 1314
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.334912310039142,
+      "learning_rate": 6.1709337492421515e-06,
+      "loss": 0.7765,
+      "step": 1315
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1590860603075632,
+      "learning_rate": 6.165637927127111e-06,
+      "loss": 1.0136,
+      "step": 1316
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0885973573127805,
+      "learning_rate": 6.1603407218506726e-06,
+      "loss": 0.9118,
+      "step": 1317
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1295328307865153,
+      "learning_rate": 6.155042139698568e-06,
+      "loss": 0.9429,
+      "step": 1318
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0802471058737697,
+      "learning_rate": 6.149742186958166e-06,
+      "loss": 0.9084,
+      "step": 1319
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2274670765856943,
+      "learning_rate": 6.144440869918462e-06,
+      "loss": 0.7233,
+      "step": 1320
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7878180237826575,
+      "learning_rate": 6.139138194870067e-06,
+      "loss": 0.9406,
+      "step": 1321
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0755452494242859,
+      "learning_rate": 6.133834168105206e-06,
+      "loss": 0.5382,
+      "step": 1322
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9989535456675933,
+      "learning_rate": 6.128528795917708e-06,
+      "loss": 0.7308,
+      "step": 1323
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3930980271191187,
+      "learning_rate": 6.123222084602997e-06,
+      "loss": 0.7778,
+      "step": 1324
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8026029148545433,
+      "learning_rate": 6.117914040458087e-06,
+      "loss": 0.7984,
+      "step": 1325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.231249997273463,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.918,
+      "step": 1326
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7387057931635226,
+      "learning_rate": 6.107293978873626e-06,
+      "loss": 0.888,
+      "step": 1327
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.425250256693401,
+      "learning_rate": 6.101981974035982e-06,
+      "loss": 0.8477,
+      "step": 1328
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0995945158258784,
+      "learning_rate": 6.096668661571934e-06,
+      "loss": 0.8374,
+      "step": 1329
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7749026719794037,
+      "learning_rate": 6.091354047786333e-06,
+      "loss": 0.9029,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1896211180652976,
+      "learning_rate": 6.086038138985566e-06,
+      "loss": 0.5726,
+      "step": 1331
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7339269684043337,
+      "learning_rate": 6.080720941477564e-06,
+      "loss": 0.7994,
+      "step": 1332
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7110393754265275,
+      "learning_rate": 6.075402461571781e-06,
+      "loss": 0.9504,
+      "step": 1333
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.359466589924437,
+      "learning_rate": 6.070082705579198e-06,
+      "loss": 0.6854,
+      "step": 1334
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.361677611543213,
+      "learning_rate": 6.0647616798123075e-06,
+      "loss": 0.9855,
+      "step": 1335
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3761732374581144,
+      "learning_rate": 6.0594393905851065e-06,
+      "loss": 0.7617,
+      "step": 1336
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.534559697693697,
+      "learning_rate": 6.054115844213099e-06,
+      "loss": 0.6984,
+      "step": 1337
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.12766780188686,
+      "learning_rate": 6.048791047013272e-06,
+      "loss": 0.9301,
+      "step": 1338
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.322423975304647,
+      "learning_rate": 6.043465005304099e-06,
+      "loss": 0.56,
+      "step": 1339
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2206113648194485,
+      "learning_rate": 6.038137725405535e-06,
+      "loss": 0.9508,
+      "step": 1340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3909528892252907,
+      "learning_rate": 6.032809213639e-06,
+      "loss": 0.9,
+      "step": 1341
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4173161441704973,
+      "learning_rate": 6.027479476327376e-06,
+      "loss": 0.8805,
+      "step": 1342
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3403062490849877,
+      "learning_rate": 6.0221485197949995e-06,
+      "loss": 0.9217,
+      "step": 1343
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.538685832741167,
+      "learning_rate": 6.016816350367655e-06,
+      "loss": 0.7135,
+      "step": 1344
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4166107201187692,
+      "learning_rate": 6.011482974372564e-06,
+      "loss": 0.9492,
+      "step": 1345
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7444391025702888,
+      "learning_rate": 6.006148398138383e-06,
+      "loss": 0.888,
+      "step": 1346
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6793441092684698,
+      "learning_rate": 6.00081262799519e-06,
+      "loss": 0.9188,
+      "step": 1347
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4535667346082388,
+      "learning_rate": 5.995475670274481e-06,
+      "loss": 0.7256,
+      "step": 1348
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9610341512407925,
+      "learning_rate": 5.990137531309158e-06,
+      "loss": 0.5137,
+      "step": 1349
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7995138587439042,
+      "learning_rate": 5.9847982174335314e-06,
+      "loss": 0.9156,
+      "step": 1350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8618615693530038,
+      "learning_rate": 5.9794577349833e-06,
+      "loss": 0.7894,
+      "step": 1351
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1564411798381742,
+      "learning_rate": 5.974116090295548e-06,
+      "loss": 0.9656,
+      "step": 1352
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9024751184478257,
+      "learning_rate": 5.968773289708746e-06,
+      "loss": 0.8981,
+      "step": 1353
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.97768948772318,
+      "learning_rate": 5.963429339562731e-06,
+      "loss": 0.7964,
+      "step": 1354
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6261717204154365,
+      "learning_rate": 5.958084246198704e-06,
+      "loss": 0.9648,
+      "step": 1355
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8843456329512893,
+      "learning_rate": 5.952738015959225e-06,
+      "loss": 0.8282,
+      "step": 1356
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.048278380999577,
+      "learning_rate": 5.9473906551881985e-06,
+      "loss": 0.8319,
+      "step": 1357
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.650466221339909,
+      "learning_rate": 5.942042170230879e-06,
+      "loss": 0.7619,
+      "step": 1358
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1982729367807619,
+      "learning_rate": 5.936692567433844e-06,
+      "loss": 0.9448,
+      "step": 1359
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.727822282639748,
+      "learning_rate": 5.931341853145005e-06,
+      "loss": 0.8029,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.50999872800413,
+      "learning_rate": 5.925990033713591e-06,
+      "loss": 0.9091,
+      "step": 1361
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0214045873640203,
+      "learning_rate": 5.920637115490142e-06,
+      "loss": 0.7939,
+      "step": 1362
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.251073332874601,
+      "learning_rate": 5.9152831048265005e-06,
+      "loss": 0.8331,
+      "step": 1363
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.607071036730256,
+      "learning_rate": 5.9099280080758085e-06,
+      "loss": 0.9181,
+      "step": 1364
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.250464872786283,
+      "learning_rate": 5.904571831592491e-06,
+      "loss": 0.8959,
+      "step": 1365
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7887879755870933,
+      "learning_rate": 5.899214581732262e-06,
+      "loss": 0.6251,
+      "step": 1366
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3900114918346054,
+      "learning_rate": 5.893856264852103e-06,
+      "loss": 0.8249,
+      "step": 1367
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6110441027876499,
+      "learning_rate": 5.888496887310265e-06,
+      "loss": 0.7651,
+      "step": 1368
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0768368408429463,
+      "learning_rate": 5.883136455466254e-06,
+      "loss": 0.8879,
+      "step": 1369
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6844277679245316,
+      "learning_rate": 5.877774975680831e-06,
+      "loss": 0.7489,
+      "step": 1370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5634697061659643,
+      "learning_rate": 5.872412454315999e-06,
+      "loss": 0.953,
+      "step": 1371
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9655709964804797,
+      "learning_rate": 5.867048897734997e-06,
+      "loss": 0.7612,
+      "step": 1372
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0450384611910801,
+      "learning_rate": 5.8616843123022925e-06,
+      "loss": 0.7982,
+      "step": 1373
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6097976628334678,
+      "learning_rate": 5.856318704383572e-06,
+      "loss": 0.8672,
+      "step": 1374
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8434915364190128,
+      "learning_rate": 5.850952080345739e-06,
+      "loss": 0.4481,
+      "step": 1375
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1475614382432056,
+      "learning_rate": 5.845584446556899e-06,
+      "loss": 0.9809,
+      "step": 1376
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0757039270463167,
+      "learning_rate": 5.840215809386357e-06,
+      "loss": 0.7501,
+      "step": 1377
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6159555237660301,
+      "learning_rate": 5.834846175204612e-06,
+      "loss": 0.7847,
+      "step": 1378
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2297666251100745,
+      "learning_rate": 5.829475550383339e-06,
+      "loss": 0.7928,
+      "step": 1379
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3775546761125024,
+      "learning_rate": 5.824103941295394e-06,
+      "loss": 0.8716,
+      "step": 1380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9777125769598205,
+      "learning_rate": 5.818731354314801e-06,
+      "loss": 0.9623,
+      "step": 1381
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6771628303032851,
+      "learning_rate": 5.813357795816742e-06,
+      "loss": 0.9677,
+      "step": 1382
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5787705262051395,
+      "learning_rate": 5.807983272177553e-06,
+      "loss": 0.885,
+      "step": 1383
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8476714212819934,
+      "learning_rate": 5.802607789774713e-06,
+      "loss": 0.5111,
+      "step": 1384
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5559952400268289,
+      "learning_rate": 5.797231354986842e-06,
+      "loss": 0.8368,
+      "step": 1385
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1184057749718035,
+      "learning_rate": 5.791853974193688e-06,
+      "loss": 0.8436,
+      "step": 1386
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9323492710153323,
+      "learning_rate": 5.7864756537761245e-06,
+      "loss": 0.9014,
+      "step": 1387
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9521046881382375,
+      "learning_rate": 5.781096400116135e-06,
+      "loss": 0.975,
+      "step": 1388
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4074576716938811,
+      "learning_rate": 5.775716219596815e-06,
+      "loss": 0.7334,
+      "step": 1389
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8557326905905628,
+      "learning_rate": 5.7703351186023575e-06,
+      "loss": 0.8407,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8320263875185963,
+      "learning_rate": 5.7649531035180476e-06,
+      "loss": 0.6704,
+      "step": 1391
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9264963717741087,
+      "learning_rate": 5.759570180730255e-06,
+      "loss": 0.5923,
+      "step": 1392
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6839184021403117,
+      "learning_rate": 5.754186356626429e-06,
+      "loss": 0.9807,
+      "step": 1393
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1217211338341593,
+      "learning_rate": 5.748801637595085e-06,
+      "loss": 0.778,
+      "step": 1394
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2649891237679443,
+      "learning_rate": 5.743416030025804e-06,
+      "loss": 0.909,
+      "step": 1395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8299235627232515,
+      "learning_rate": 5.738029540309215e-06,
+      "loss": 0.903,
+      "step": 1396
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2069254706225743,
+      "learning_rate": 5.732642174837003e-06,
+      "loss": 0.8127,
+      "step": 1397
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.823357087809232,
+      "learning_rate": 5.727253940001884e-06,
+      "loss": 0.9976,
+      "step": 1398
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2415895015488119,
+      "learning_rate": 5.721864842197612e-06,
+      "loss": 0.7725,
+      "step": 1399
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0593107005156956,
+      "learning_rate": 5.716474887818957e-06,
+      "loss": 1.0465,
+      "step": 1400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.24440938310932,
+      "learning_rate": 5.711084083261716e-06,
+      "loss": 0.7271,
+      "step": 1401
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7695182027737566,
+      "learning_rate": 5.705692434922684e-06,
+      "loss": 0.5349,
+      "step": 1402
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2096987389049612,
+      "learning_rate": 5.700299949199664e-06,
+      "loss": 0.8126,
+      "step": 1403
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4974012313116924,
+      "learning_rate": 5.694906632491452e-06,
+      "loss": 0.8731,
+      "step": 1404
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5597630413497399,
+      "learning_rate": 5.689512491197827e-06,
+      "loss": 0.9875,
+      "step": 1405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1550836293112545,
+      "learning_rate": 5.684117531719552e-06,
+      "loss": 0.8011,
+      "step": 1406
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.240063029233875,
+      "learning_rate": 5.678721760458352e-06,
+      "loss": 0.9672,
+      "step": 1407
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7224219994924103,
+      "learning_rate": 5.673325183816927e-06,
+      "loss": 0.8681,
+      "step": 1408
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0155375914626805,
+      "learning_rate": 5.667927808198922e-06,
+      "loss": 0.9099,
+      "step": 1409
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6994101786369953,
+      "learning_rate": 5.662529640008933e-06,
+      "loss": 0.7807,
+      "step": 1410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7407554961030611,
+      "learning_rate": 5.657130685652503e-06,
+      "loss": 0.6453,
+      "step": 1411
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.282694155555053,
+      "learning_rate": 5.6517309515361005e-06,
+      "loss": 0.9494,
+      "step": 1412
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4126242941256903,
+      "learning_rate": 5.646330444067121e-06,
+      "loss": 0.8299,
+      "step": 1413
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8964144993539503,
+      "learning_rate": 5.640929169653876e-06,
+      "loss": 0.9589,
+      "step": 1414
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7135912087848546,
+      "learning_rate": 5.635527134705593e-06,
+      "loss": 0.8007,
+      "step": 1415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.87627715873684,
+      "learning_rate": 5.630124345632396e-06,
+      "loss": 0.6767,
+      "step": 1416
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3964569409167993,
+      "learning_rate": 5.6247208088453065e-06,
+      "loss": 0.9147,
+      "step": 1417
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.204387504715069,
+      "learning_rate": 5.619316530756234e-06,
+      "loss": 0.7678,
+      "step": 1418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8135518503850294,
+      "learning_rate": 5.613911517777965e-06,
+      "loss": 0.658,
+      "step": 1419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7930697184938414,
+      "learning_rate": 5.608505776324158e-06,
+      "loss": 0.8915,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4069318788987446,
+      "learning_rate": 5.6030993128093415e-06,
+      "loss": 0.9395,
+      "step": 1421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9503243898218101,
+      "learning_rate": 5.597692133648894e-06,
+      "loss": 0.7598,
+      "step": 1422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7340782067712569,
+      "learning_rate": 5.592284245259047e-06,
+      "loss": 0.9026,
+      "step": 1423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1181795784382287,
+      "learning_rate": 5.5868756540568704e-06,
+      "loss": 0.8399,
+      "step": 1424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0550817795880885,
+      "learning_rate": 5.5814663664602735e-06,
+      "loss": 0.79,
+      "step": 1425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3120495891856772,
+      "learning_rate": 5.576056388887985e-06,
+      "loss": 0.9909,
+      "step": 1426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8934060649887934,
+      "learning_rate": 5.570645727759558e-06,
+      "loss": 0.8146,
+      "step": 1427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.920684607554885,
+      "learning_rate": 5.565234389495354e-06,
+      "loss": 0.5183,
+      "step": 1428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.153206698770488,
+      "learning_rate": 5.559822380516539e-06,
+      "loss": 0.8107,
+      "step": 1429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2455180987840444,
+      "learning_rate": 5.554409707245076e-06,
+      "loss": 0.81,
+      "step": 1430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.48173237168619,
+      "learning_rate": 5.548996376103712e-06,
+      "loss": 0.8798,
+      "step": 1431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6985747450085716,
+      "learning_rate": 5.543582393515978e-06,
+      "loss": 0.929,
+      "step": 1432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.278396424196154,
+      "learning_rate": 5.538167765906178e-06,
+      "loss": 0.8309,
+      "step": 1433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4199499410743877,
+      "learning_rate": 5.532752499699381e-06,
+      "loss": 0.894,
+      "step": 1434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2066317465558989,
+      "learning_rate": 5.527336601321414e-06,
+      "loss": 0.8874,
+      "step": 1435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9917347231328922,
+      "learning_rate": 5.521920077198853e-06,
+      "loss": 0.8792,
+      "step": 1436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9622909466291227,
+      "learning_rate": 5.516502933759017e-06,
+      "loss": 0.4659,
+      "step": 1437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2462695929473209,
+      "learning_rate": 5.511085177429961e-06,
+      "loss": 0.9711,
+      "step": 1438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.585717748916246,
+      "learning_rate": 5.505666814640466e-06,
+      "loss": 0.7825,
+      "step": 1439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1696951552266461,
+      "learning_rate": 5.5002478518200316e-06,
+      "loss": 0.8092,
+      "step": 1440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6823823954777597,
+      "learning_rate": 5.494828295398874e-06,
+      "loss": 0.8161,
+      "step": 1441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0945013610938656,
+      "learning_rate": 5.489408151807908e-06,
+      "loss": 0.6995,
+      "step": 1442
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.476666013321854,
+      "learning_rate": 5.483987427478749e-06,
+      "loss": 0.9307,
+      "step": 1443
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6355741684115963,
+      "learning_rate": 5.4785661288436985e-06,
+      "loss": 0.9404,
+      "step": 1444
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8806929293384578,
+      "learning_rate": 5.473144262335744e-06,
+      "loss": 0.8946,
+      "step": 1445
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.873040471686788,
+      "learning_rate": 5.467721834388543e-06,
+      "loss": 0.634,
+      "step": 1446
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1977758930684796,
+      "learning_rate": 5.462298851436418e-06,
+      "loss": 0.8687,
+      "step": 1447
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0313327700504056,
+      "learning_rate": 5.456875319914355e-06,
+      "loss": 0.8376,
+      "step": 1448
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4205522176232346,
+      "learning_rate": 5.451451246257989e-06,
+      "loss": 0.6908,
+      "step": 1449
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0283692673375737,
+      "learning_rate": 5.446026636903597e-06,
+      "loss": 0.8025,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9538982374836444,
+      "learning_rate": 5.440601498288088e-06,
+      "loss": 0.8634,
+      "step": 1451
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3506867677336547,
+      "learning_rate": 5.43517583684901e-06,
+      "loss": 0.9185,
+      "step": 1452
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5651409304249373,
+      "learning_rate": 5.429749659024521e-06,
+      "loss": 0.7897,
+      "step": 1453
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2266115232709012,
+      "learning_rate": 5.424322971253395e-06,
+      "loss": 0.7612,
+      "step": 1454
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8323033492887951,
+      "learning_rate": 5.4188957799750145e-06,
+      "loss": 0.6213,
+      "step": 1455
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9965590817667975,
+      "learning_rate": 5.413468091629353e-06,
+      "loss": 0.7432,
+      "step": 1456
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6336978892297456,
+      "learning_rate": 5.40803991265698e-06,
+      "loss": 1.0066,
+      "step": 1457
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3906217417748743,
+      "learning_rate": 5.402611249499042e-06,
+      "loss": 0.6617,
+      "step": 1458
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.112214616402955,
+      "learning_rate": 5.397182108597266e-06,
+      "loss": 1.0069,
+      "step": 1459
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0474791498171876,
+      "learning_rate": 5.391752496393941e-06,
+      "loss": 0.7956,
+      "step": 1460
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3134000695516703,
+      "learning_rate": 5.386322419331914e-06,
+      "loss": 0.8127,
+      "step": 1461
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0982396313256826,
+      "learning_rate": 5.380891883854591e-06,
+      "loss": 0.9252,
+      "step": 1462
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5405390293411296,
+      "learning_rate": 5.375460896405914e-06,
+      "loss": 0.7162,
+      "step": 1463
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.921991294834801,
+      "learning_rate": 5.370029463430365e-06,
+      "loss": 0.6743,
+      "step": 1464
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5252436777217095,
+      "learning_rate": 5.364597591372957e-06,
+      "loss": 0.6596,
+      "step": 1465
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.73641034149806,
+      "learning_rate": 5.359165286679218e-06,
+      "loss": 0.7065,
+      "step": 1466
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0007106940214074,
+      "learning_rate": 5.3537325557951925e-06,
+      "loss": 0.9059,
+      "step": 1467
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8957692792620751,
+      "learning_rate": 5.348299405167434e-06,
+      "loss": 0.9342,
+      "step": 1468
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.156413148137658,
+      "learning_rate": 5.34286584124299e-06,
+      "loss": 1.0154,
+      "step": 1469
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7293081164919134,
+      "learning_rate": 5.337431870469398e-06,
+      "loss": 0.9107,
+      "step": 1470
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0183732758427553,
+      "learning_rate": 5.33199749929468e-06,
+      "loss": 0.908,
+      "step": 1471
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4597843930047711,
+      "learning_rate": 5.3265627341673336e-06,
+      "loss": 0.4835,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.768915889791925,
+      "learning_rate": 5.321127581536323e-06,
+      "loss": 0.7993,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0154072962721385,
+      "learning_rate": 5.31569204785107e-06,
+      "loss": 0.7998,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9193634815006646,
+      "learning_rate": 5.310256139561455e-06,
+      "loss": 0.7326,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.144858461443132,
+      "learning_rate": 5.304819863117796e-06,
+      "loss": 1.0072,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 48.39733757267215,
+      "learning_rate": 5.29938322497085e-06,
+      "loss": 0.7892,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3506851810751819,
+      "learning_rate": 5.293946231571806e-06,
+      "loss": 0.7995,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5367310892227926,
+      "learning_rate": 5.2885088893722725e-06,
+      "loss": 0.8136,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2239651267229594,
+      "learning_rate": 5.283071204824269e-06,
+      "loss": 0.9965,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1214503434849266,
+      "learning_rate": 5.277633184380226e-06,
+      "loss": 0.661,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7280744700963329,
+      "learning_rate": 5.27219483449297e-06,
+      "loss": 0.7117,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.347921731507998,
+      "learning_rate": 5.266756161615719e-06,
+      "loss": 1.0434,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8043938133184256,
+      "learning_rate": 5.261317172202074e-06,
+      "loss": 0.9156,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7307739860048534,
+      "learning_rate": 5.255877872706013e-06,
+      "loss": 0.7135,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4868514291346695,
+      "learning_rate": 5.25043826958188e-06,
+      "loss": 0.9253,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.520095142918424,
+      "learning_rate": 5.244998369284379e-06,
+      "loss": 0.5692,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.04397526549605,
+      "learning_rate": 5.23955817826857e-06,
+      "loss": 0.9303,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6056604010938338,
+      "learning_rate": 5.234117702989855e-06,
+      "loss": 0.8121,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.230624896594264,
+      "learning_rate": 5.228676949903974e-06,
+      "loss": 0.922,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3141040904254504,
+      "learning_rate": 5.223235925466994e-06,
+      "loss": 0.5237,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0729018400029147,
+      "learning_rate": 5.217794636135311e-06,
+      "loss": 0.9214,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.97247371107511,
+      "learning_rate": 5.212353088365627e-06,
+      "loss": 1.0189,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.428822602035591,
+      "learning_rate": 5.2069112886149564e-06,
+      "loss": 0.7454,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.005145488936022,
+      "learning_rate": 5.201469243340612e-06,
+      "loss": 0.926,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1798059810159578,
+      "learning_rate": 5.1960269590001965e-06,
+      "loss": 0.8118,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1831107616555165,
+      "learning_rate": 5.190584442051594e-06,
+      "loss": 0.866,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4288799286802236,
+      "learning_rate": 5.1851416989529705e-06,
+      "loss": 0.7623,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3005982611699751,
+      "learning_rate": 5.179698736162755e-06,
+      "loss": 0.4809,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5433104779737237,
+      "learning_rate": 5.17425556013964e-06,
+      "loss": 0.903,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9795477959088617,
+      "learning_rate": 5.168812177342568e-06,
+      "loss": 0.8004,
+      "step": 1501
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5413444775631016,
+      "learning_rate": 5.163368594230732e-06,
+      "loss": 0.9352,
+      "step": 1502
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1797675590776637,
+      "learning_rate": 5.15792481726356e-06,
+      "loss": 0.8767,
+      "step": 1503
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.982571836839815,
+      "learning_rate": 5.1524808529007075e-06,
+      "loss": 0.9031,
+      "step": 1504
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3469319223681064,
+      "learning_rate": 5.147036707602057e-06,
+      "loss": 0.8414,
+      "step": 1505
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7767685686999175,
+      "learning_rate": 5.141592387827701e-06,
+      "loss": 0.7317,
+      "step": 1506
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8653858094494804,
+      "learning_rate": 5.136147900037942e-06,
+      "loss": 0.7882,
+      "step": 1507
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1438157947702285,
+      "learning_rate": 5.130703250693284e-06,
+      "loss": 0.9046,
+      "step": 1508
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.335634748726229,
+      "learning_rate": 5.1252584462544175e-06,
+      "loss": 0.6892,
+      "step": 1509
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2908491496763213,
+      "learning_rate": 5.119813493182221e-06,
+      "loss": 0.7917,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6303483092102882,
+      "learning_rate": 5.114368397937744e-06,
+      "loss": 0.7988,
+      "step": 1511
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4004197466139023,
+      "learning_rate": 5.108923166982215e-06,
+      "loss": 0.9391,
+      "step": 1512
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.481704756079357,
+      "learning_rate": 5.103477806777012e-06,
+      "loss": 0.8592,
+      "step": 1513
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0861459925141546,
+      "learning_rate": 5.098032323783673e-06,
+      "loss": 0.8417,
+      "step": 1514
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4535271304013029,
+      "learning_rate": 5.092586724463881e-06,
+      "loss": 0.8532,
+      "step": 1515
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3266410240082385,
+      "learning_rate": 5.087141015279455e-06,
+      "loss": 0.8912,
+      "step": 1516
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8948163815546327,
+      "learning_rate": 5.081695202692347e-06,
+      "loss": 0.5654,
+      "step": 1517
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.495510702363034,
+      "learning_rate": 5.07624929316463e-06,
+      "loss": 0.7857,
+      "step": 1518
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8438391616148538,
+      "learning_rate": 5.070803293158492e-06,
+      "loss": 0.9633,
+      "step": 1519
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5912093727028693,
+      "learning_rate": 5.06535720913623e-06,
+      "loss": 0.6717,
+      "step": 1520
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3014066077246338,
+      "learning_rate": 5.059911047560236e-06,
+      "loss": 0.9342,
+      "step": 1521
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.121448156462971,
+      "learning_rate": 5.054464814893001e-06,
+      "loss": 0.7667,
+      "step": 1522
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6461943791365161,
+      "learning_rate": 5.049018517597095e-06,
+      "loss": 0.6683,
+      "step": 1523
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0211149761618439,
+      "learning_rate": 5.043572162135166e-06,
+      "loss": 0.8978,
+      "step": 1524
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8768649129847246,
+      "learning_rate": 5.038125754969933e-06,
+      "loss": 0.5339,
+      "step": 1525
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7391230144409158,
+      "learning_rate": 5.032679302564176e-06,
+      "loss": 0.904,
+      "step": 1526
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2845930750441346,
+      "learning_rate": 5.027232811380723e-06,
+      "loss": 0.9181,
+      "step": 1527
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2250507947320406,
+      "learning_rate": 5.021786287882458e-06,
+      "loss": 0.8529,
+      "step": 1528
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0117412469967957,
+      "learning_rate": 5.0163397385322955e-06,
+      "loss": 0.8549,
+      "step": 1529
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2687879856448727,
+      "learning_rate": 5.010893169793182e-06,
+      "loss": 0.7316,
+      "step": 1530
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.168060832379654,
+      "learning_rate": 5.005446588128088e-06,
+      "loss": 1.0047,
+      "step": 1531
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8733292750470187,
+      "learning_rate": 5e-06,
+      "loss": 0.8819,
+      "step": 1532
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4049350258524866,
+      "learning_rate": 4.994553411871914e-06,
+      "loss": 0.8775,
+      "step": 1533
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0952591031193046,
+      "learning_rate": 4.98910683020682e-06,
+      "loss": 0.745,
+      "step": 1534
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8588639622821992,
+      "learning_rate": 4.983660261467706e-06,
+      "loss": 0.4801,
+      "step": 1535
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.046256805668577,
+      "learning_rate": 4.978213712117543e-06,
+      "loss": 0.8805,
+      "step": 1536
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.291638607679887,
+      "learning_rate": 4.972767188619278e-06,
+      "loss": 0.7446,
+      "step": 1537
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.202945138286398,
+      "learning_rate": 4.9673206974358254e-06,
+      "loss": 0.9649,
+      "step": 1538
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5198696677746755,
+      "learning_rate": 4.9618742450300675e-06,
+      "loss": 0.9119,
+      "step": 1539
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6894009311956983,
+      "learning_rate": 4.956427837864835e-06,
+      "loss": 0.8668,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0627929288126472,
+      "learning_rate": 4.9509814824029065e-06,
+      "loss": 0.8091,
+      "step": 1541
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2378287402494215,
+      "learning_rate": 4.945535185107e-06,
+      "loss": 0.8149,
+      "step": 1542
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9697965444187244,
+      "learning_rate": 4.940088952439765e-06,
+      "loss": 0.8804,
+      "step": 1543
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5908115887007626,
+      "learning_rate": 4.934642790863772e-06,
+      "loss": 0.724,
+      "step": 1544
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1265739812737676,
+      "learning_rate": 4.9291967068415085e-06,
+      "loss": 1.0216,
+      "step": 1545
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9546473776051394,
+      "learning_rate": 4.923750706835371e-06,
+      "loss": 0.7134,
+      "step": 1546
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4236180415436994,
+      "learning_rate": 4.9183047973076546e-06,
+      "loss": 0.9875,
+      "step": 1547
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9444922532219139,
+      "learning_rate": 4.912858984720546e-06,
+      "loss": 0.746,
+      "step": 1548
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0008149705315714,
+      "learning_rate": 4.90741327553612e-06,
+      "loss": 0.6975,
+      "step": 1549
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3707108290038612,
+      "learning_rate": 4.901967676216329e-06,
+      "loss": 1.0188,
+      "step": 1550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1176007791942846,
+      "learning_rate": 4.89652219322299e-06,
+      "loss": 0.9374,
+      "step": 1551
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8259341787327605,
+      "learning_rate": 4.8910768330177875e-06,
+      "loss": 0.6279,
+      "step": 1552
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4883238054331664,
+      "learning_rate": 4.8856316020622564e-06,
+      "loss": 0.798,
+      "step": 1553
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3591652312461857,
+      "learning_rate": 4.880186506817781e-06,
+      "loss": 0.7811,
+      "step": 1554
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.071891064080181,
+      "learning_rate": 4.874741553745583e-06,
+      "loss": 0.9436,
+      "step": 1555
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.457918862155709,
+      "learning_rate": 4.8692967493067175e-06,
+      "loss": 0.9126,
+      "step": 1556
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7794227766680848,
+      "learning_rate": 4.8638520999620585e-06,
+      "loss": 0.9263,
+      "step": 1557
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.283598593634706,
+      "learning_rate": 4.8584076121723e-06,
+      "loss": 0.7945,
+      "step": 1558
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0527723644876628,
+      "learning_rate": 4.852963292397946e-06,
+      "loss": 0.7762,
+      "step": 1559
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2632617507480353,
+      "learning_rate": 4.847519147099294e-06,
+      "loss": 0.867,
+      "step": 1560
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7252715988079687,
+      "learning_rate": 4.8420751827364415e-06,
+      "loss": 0.6314,
+      "step": 1561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.060295660839889,
+      "learning_rate": 4.8366314057692684e-06,
+      "loss": 0.8165,
+      "step": 1562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5088617457824505,
+      "learning_rate": 4.831187822657434e-06,
+      "loss": 0.7961,
+      "step": 1563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3495854823363904,
+      "learning_rate": 4.825744439860362e-06,
+      "loss": 0.859,
+      "step": 1564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.548457815815163,
+      "learning_rate": 4.8203012638372475e-06,
+      "loss": 0.6944,
+      "step": 1565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.524203609025586,
+      "learning_rate": 4.814858301047031e-06,
+      "loss": 0.8518,
+      "step": 1566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.846046636555615,
+      "learning_rate": 4.809415557948407e-06,
+      "loss": 0.9278,
+      "step": 1567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3196028545678415,
+      "learning_rate": 4.803973040999804e-06,
+      "loss": 0.9613,
+      "step": 1568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6167644809968278,
+      "learning_rate": 4.798530756659389e-06,
+      "loss": 0.8358,
+      "step": 1569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9241210146245262,
+      "learning_rate": 4.793088711385044e-06,
+      "loss": 0.5133,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7669861475368007,
+      "learning_rate": 4.7876469116343745e-06,
+      "loss": 0.8021,
+      "step": 1571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1009157913468193,
+      "learning_rate": 4.782205363864692e-06,
+      "loss": 0.8494,
+      "step": 1572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.089512269838726,
+      "learning_rate": 4.776764074533008e-06,
+      "loss": 0.8616,
+      "step": 1573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5511471636690424,
+      "learning_rate": 4.771323050096028e-06,
+      "loss": 0.9278,
+      "step": 1574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4425789952754333,
+      "learning_rate": 4.765882297010147e-06,
+      "loss": 0.6989,
+      "step": 1575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1252742904898303,
+      "learning_rate": 4.760441821731431e-06,
+      "loss": 0.8717,
+      "step": 1576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.074922222954093,
+      "learning_rate": 4.755001630715622e-06,
+      "loss": 0.7696,
+      "step": 1577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2099823439018484,
+      "learning_rate": 4.749561730418121e-06,
+      "loss": 0.4762,
+      "step": 1578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.131131717466025,
+      "learning_rate": 4.7441221272939886e-06,
+      "loss": 0.9728,
+      "step": 1579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3629250838421492,
+      "learning_rate": 4.738682827797928e-06,
+      "loss": 0.7444,
+      "step": 1580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9191929637847713,
+      "learning_rate": 4.733243838384282e-06,
+      "loss": 0.8972,
+      "step": 1581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6715266984759934,
+      "learning_rate": 4.727805165507032e-06,
+      "loss": 0.8941,
+      "step": 1582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7675285702447916,
+      "learning_rate": 4.722366815619776e-06,
+      "loss": 0.9254,
+      "step": 1583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.542761054392612,
+      "learning_rate": 4.716928795175732e-06,
+      "loss": 0.9442,
+      "step": 1584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.647602289463234,
+      "learning_rate": 4.711491110627729e-06,
+      "loss": 0.7544,
+      "step": 1585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.426344091545656,
+      "learning_rate": 4.706053768428195e-06,
+      "loss": 0.9579,
+      "step": 1586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.288847432808909,
+      "learning_rate": 4.700616775029151e-06,
+      "loss": 0.6951,
+      "step": 1587
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8459104988374413,
+      "learning_rate": 4.6951801368822055e-06,
+      "loss": 0.5111,
+      "step": 1588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.192731551276934,
+      "learning_rate": 4.689743860438547e-06,
+      "loss": 0.8584,
+      "step": 1589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1490872506746799,
+      "learning_rate": 4.684307952148931e-06,
+      "loss": 0.8527,
+      "step": 1590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.9226257909840445,
+      "learning_rate": 4.678872418463679e-06,
+      "loss": 0.9861,
+      "step": 1591
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.387333096986182,
+      "learning_rate": 4.673437265832668e-06,
+      "loss": 0.824,
+      "step": 1592
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8846581991357186,
+      "learning_rate": 4.668002500705322e-06,
+      "loss": 0.9517,
+      "step": 1593
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.242830662798849,
+      "learning_rate": 4.662568129530603e-06,
+      "loss": 0.6822,
+      "step": 1594
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6414621713108402,
+      "learning_rate": 4.6571341587570114e-06,
+      "loss": 0.896,
+      "step": 1595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9912927046865645,
+      "learning_rate": 4.6517005948325674e-06,
+      "loss": 0.8787,
+      "step": 1596
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8018100658075413,
+      "learning_rate": 4.646267444204809e-06,
+      "loss": 0.4883,
+      "step": 1597
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7765747096413576,
+      "learning_rate": 4.640834713320785e-06,
+      "loss": 0.9673,
+      "step": 1598
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8448031443321673,
+      "learning_rate": 4.635402408627046e-06,
+      "loss": 0.7233,
+      "step": 1599
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.654480844297652,
+      "learning_rate": 4.629970536569636e-06,
+      "loss": 0.928,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3525299874027032,
+      "learning_rate": 4.624539103594087e-06,
+      "loss": 0.7701,
+      "step": 1601
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0255905168379158,
+      "learning_rate": 4.619108116145411e-06,
+      "loss": 0.8782,
+      "step": 1602
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8611557003932115,
+      "learning_rate": 4.613677580668087e-06,
+      "loss": 0.7678,
+      "step": 1603
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.95382535165409,
+      "learning_rate": 4.608247503606061e-06,
+      "loss": 0.7565,
+      "step": 1604
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7634388929075117,
+      "learning_rate": 4.602817891402735e-06,
+      "loss": 0.6973,
+      "step": 1605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8100622519669148,
+      "learning_rate": 4.597388750500959e-06,
+      "loss": 0.8579,
+      "step": 1606
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4998445800889175,
+      "learning_rate": 4.591960087343023e-06,
+      "loss": 0.9334,
+      "step": 1607
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3730716388935327,
+      "learning_rate": 4.586531908370648e-06,
+      "loss": 0.8052,
+      "step": 1608
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9496378435130904,
+      "learning_rate": 4.581104220024988e-06,
+      "loss": 0.8832,
+      "step": 1609
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.440290310087043,
+      "learning_rate": 4.575677028746606e-06,
+      "loss": 0.8012,
+      "step": 1610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3275328405060844,
+      "learning_rate": 4.57025034097548e-06,
+      "loss": 0.8214,
+      "step": 1611
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0434205849984983,
+      "learning_rate": 4.564824163150992e-06,
+      "loss": 1.0111,
+      "step": 1612
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6805536987833303,
+      "learning_rate": 4.559398501711913e-06,
+      "loss": 0.7747,
+      "step": 1613
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7017629106777692,
+      "learning_rate": 4.553973363096405e-06,
+      "loss": 0.5091,
+      "step": 1614
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7521231010187839,
+      "learning_rate": 4.548548753742012e-06,
+      "loss": 0.8396,
+      "step": 1615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.123578249868797,
+      "learning_rate": 4.5431246800856455e-06,
+      "loss": 0.8013,
+      "step": 1616
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3134319499254585,
+      "learning_rate": 4.537701148563584e-06,
+      "loss": 0.8826,
+      "step": 1617
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7244515345944992,
+      "learning_rate": 4.532278165611459e-06,
+      "loss": 0.9266,
+      "step": 1618
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5035356520428338,
+      "learning_rate": 4.526855737664257e-06,
+      "loss": 0.8738,
+      "step": 1619
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3694637412976869,
+      "learning_rate": 4.521433871156302e-06,
+      "loss": 0.8967,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6201910247283535,
+      "learning_rate": 4.516012572521252e-06,
+      "loss": 0.8754,
+      "step": 1621
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.859713647226442,
+      "learning_rate": 4.510591848192093e-06,
+      "loss": 0.9056,
+      "step": 1622
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7601441931151104,
+      "learning_rate": 4.505171704601128e-06,
+      "loss": 0.4849,
+      "step": 1623
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5320844623845529,
+      "learning_rate": 4.49975214817997e-06,
+      "loss": 0.9196,
+      "step": 1624
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2652773584289472,
+      "learning_rate": 4.494333185359536e-06,
+      "loss": 0.7785,
+      "step": 1625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0753726969041137,
+      "learning_rate": 4.4889148225700406e-06,
+      "loss": 0.8633,
+      "step": 1626
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1904667316515922,
+      "learning_rate": 4.483497066240985e-06,
+      "loss": 0.7795,
+      "step": 1627
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.066707688077118,
+      "learning_rate": 4.4780799228011485e-06,
+      "loss": 0.7543,
+      "step": 1628
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.306247038032537,
+      "learning_rate": 4.472663398678587e-06,
+      "loss": 0.946,
+      "step": 1629
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.549657505884265,
+      "learning_rate": 4.467247500300621e-06,
+      "loss": 0.9225,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6023804335213845,
+      "learning_rate": 4.461832234093823e-06,
+      "loss": 0.9712,
+      "step": 1631
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9281768150078866,
+      "learning_rate": 4.456417606484023e-06,
+      "loss": 0.5186,
+      "step": 1632
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3712213984432633,
+      "learning_rate": 4.451003623896291e-06,
+      "loss": 0.8357,
+      "step": 1633
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8627929980353939,
+      "learning_rate": 4.445590292754927e-06,
+      "loss": 0.8689,
+      "step": 1634
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2121727909233175,
+      "learning_rate": 4.4401776194834615e-06,
+      "loss": 0.7658,
+      "step": 1635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7619931755184829,
+      "learning_rate": 4.434765610504647e-06,
+      "loss": 0.7479,
+      "step": 1636
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1692894814943924,
+      "learning_rate": 4.4293542722404435e-06,
+      "loss": 0.8188,
+      "step": 1637
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.976670123754082,
+      "learning_rate": 4.423943611112016e-06,
+      "loss": 0.8789,
+      "step": 1638
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6955102158914201,
+      "learning_rate": 4.418533633539728e-06,
+      "loss": 0.8534,
+      "step": 1639
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0390280290560177,
+      "learning_rate": 4.41312434594313e-06,
+      "loss": 0.8558,
+      "step": 1640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6562128341921831,
+      "learning_rate": 4.407715754740954e-06,
+      "loss": 0.4739,
+      "step": 1641
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8823526227154683,
+      "learning_rate": 4.402307866351107e-06,
+      "loss": 0.8986,
+      "step": 1642
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.606533922728858,
+      "learning_rate": 4.396900687190659e-06,
+      "loss": 0.9721,
+      "step": 1643
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9004175513604052,
+      "learning_rate": 4.391494223675843e-06,
+      "loss": 0.8045,
+      "step": 1644
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.178252751626005,
+      "learning_rate": 4.386088482222037e-06,
+      "loss": 0.8768,
+      "step": 1645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9704803686213136,
+      "learning_rate": 4.380683469243768e-06,
+      "loss": 0.829,
+      "step": 1646
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.473573506644909,
+      "learning_rate": 4.375279191154695e-06,
+      "loss": 0.8206,
+      "step": 1647
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.30094436952466,
+      "learning_rate": 4.369875654367605e-06,
+      "loss": 0.8902,
+      "step": 1648
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0190244955597008,
+      "learning_rate": 4.364472865294409e-06,
+      "loss": 0.7133,
+      "step": 1649
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6156156504950216,
+      "learning_rate": 4.359070830346126e-06,
+      "loss": 0.5891,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.44135175816039,
+      "learning_rate": 4.3536695559328816e-06,
+      "loss": 0.7164,
+      "step": 1651
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1450597277589063,
+      "learning_rate": 4.3482690484639e-06,
+      "loss": 0.7422,
+      "step": 1652
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8384762145293162,
+      "learning_rate": 4.342869314347498e-06,
+      "loss": 0.883,
+      "step": 1653
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0326618237384986,
+      "learning_rate": 4.337470359991068e-06,
+      "loss": 0.8808,
+      "step": 1654
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9841879016537473,
+      "learning_rate": 4.33207219180108e-06,
+      "loss": 1.0278,
+      "step": 1655
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7752242933860405,
+      "learning_rate": 4.326674816183075e-06,
+      "loss": 0.9001,
+      "step": 1656
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.088763667151357,
+      "learning_rate": 4.321278239541649e-06,
+      "loss": 0.9263,
+      "step": 1657
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7354535227515983,
+      "learning_rate": 4.31588246828045e-06,
+      "loss": 0.4634,
+      "step": 1658
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.106877547680572,
+      "learning_rate": 4.310487508802174e-06,
+      "loss": 0.8272,
+      "step": 1659
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0499956906778987,
+      "learning_rate": 4.305093367508551e-06,
+      "loss": 0.7512,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0556738665815522,
+      "learning_rate": 4.299700050800337e-06,
+      "loss": 0.7865,
+      "step": 1661
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9559763187829152,
+      "learning_rate": 4.294307565077318e-06,
+      "loss": 1.0337,
+      "step": 1662
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2101660291676115,
+      "learning_rate": 4.288915916738287e-06,
+      "loss": 0.791,
+      "step": 1663
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2239888045422613,
+      "learning_rate": 4.283525112181044e-06,
+      "loss": 0.873,
+      "step": 1664
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.361263585283182,
+      "learning_rate": 4.278135157802389e-06,
+      "loss": 0.6851,
+      "step": 1665
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.739855839915373,
+      "learning_rate": 4.272746059998117e-06,
+      "loss": 1.0263,
+      "step": 1666
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1467692435083565,
+      "learning_rate": 4.267357825162999e-06,
+      "loss": 0.61,
+      "step": 1667
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9497868330416064,
+      "learning_rate": 4.2619704596907856e-06,
+      "loss": 0.7483,
+      "step": 1668
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8442115903128304,
+      "learning_rate": 4.256583969974199e-06,
+      "loss": 0.9425,
+      "step": 1669
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6951783844588145,
+      "learning_rate": 4.251198362404917e-06,
+      "loss": 0.9112,
+      "step": 1670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0944091711488397,
+      "learning_rate": 4.2458136433735726e-06,
+      "loss": 0.8521,
+      "step": 1671
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1835787021540454,
+      "learning_rate": 4.240429819269746e-06,
+      "loss": 0.7578,
+      "step": 1672
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9958367874778015,
+      "learning_rate": 4.235046896481955e-06,
+      "loss": 0.6159,
+      "step": 1673
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.013069089000961,
+      "learning_rate": 4.229664881397645e-06,
+      "loss": 0.8525,
+      "step": 1674
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1101683861005187,
+      "learning_rate": 4.2242837804031855e-06,
+      "loss": 0.8081,
+      "step": 1675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9538053882135212,
+      "learning_rate": 4.218903599883866e-06,
+      "loss": 0.9961,
+      "step": 1676
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7889129486209985,
+      "learning_rate": 4.213524346223877e-06,
+      "loss": 0.4985,
+      "step": 1677
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5734467065744162,
+      "learning_rate": 4.208146025806313e-06,
+      "loss": 0.9256,
+      "step": 1678
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7807400372819789,
+      "learning_rate": 4.20276864501316e-06,
+      "loss": 0.9973,
+      "step": 1679
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.196857182271887,
+      "learning_rate": 4.197392210225289e-06,
+      "loss": 0.7604,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3829898878749505,
+      "learning_rate": 4.192016727822449e-06,
+      "loss": 0.9187,
+      "step": 1681
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.716998424206162,
+      "learning_rate": 4.186642204183259e-06,
+      "loss": 0.8153,
+      "step": 1682
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1741483407518944,
+      "learning_rate": 4.1812686456852e-06,
+      "loss": 0.9597,
+      "step": 1683
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0467623995603748,
+      "learning_rate": 4.175896058704607e-06,
+      "loss": 0.7831,
+      "step": 1684
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.821981634826077,
+      "learning_rate": 4.170524449616662e-06,
+      "loss": 0.4841,
+      "step": 1685
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1789021283576218,
+      "learning_rate": 4.165153824795391e-06,
+      "loss": 0.9215,
+      "step": 1686
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3743165133279927,
+      "learning_rate": 4.159784190613644e-06,
+      "loss": 0.747,
+      "step": 1687
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9734074687822725,
+      "learning_rate": 4.154415553443103e-06,
+      "loss": 1.0422,
+      "step": 1688
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0216668962263555,
+      "learning_rate": 4.149047919654263e-06,
+      "loss": 0.8325,
+      "step": 1689
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9928411580612923,
+      "learning_rate": 4.143681295616429e-06,
+      "loss": 0.9002,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1060942845201984,
+      "learning_rate": 4.138315687697708e-06,
+      "loss": 0.8821,
+      "step": 1691
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6312913528320311,
+      "learning_rate": 4.1329511022650035e-06,
+      "loss": 0.9152,
+      "step": 1692
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8109249184213578,
+      "learning_rate": 4.127587545684002e-06,
+      "loss": 0.5775,
+      "step": 1693
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0762123613855916,
+      "learning_rate": 4.122225024319171e-06,
+      "loss": 0.8316,
+      "step": 1694
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2022227237208092,
+      "learning_rate": 4.116863544533747e-06,
+      "loss": 0.8302,
+      "step": 1695
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1366764496497024,
+      "learning_rate": 4.111503112689737e-06,
+      "loss": 0.6708,
+      "step": 1696
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.354530598445704,
+      "learning_rate": 4.106143735147898e-06,
+      "loss": 0.7803,
+      "step": 1697
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6507867419304685,
+      "learning_rate": 4.100785418267739e-06,
+      "loss": 0.9314,
+      "step": 1698
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.051025142192599,
+      "learning_rate": 4.0954281684075096e-06,
+      "loss": 0.8723,
+      "step": 1699
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2188368554794997,
+      "learning_rate": 4.090071991924194e-06,
+      "loss": 0.8261,
+      "step": 1700
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2077313407849142,
+      "learning_rate": 4.084716895173501e-06,
+      "loss": 0.862,
+      "step": 1701
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9631642475460992,
+      "learning_rate": 4.0793628845098595e-06,
+      "loss": 0.9033,
+      "step": 1702
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8494031076475121,
+      "learning_rate": 4.07400996628641e-06,
+      "loss": 0.479,
+      "step": 1703
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9766750402117619,
+      "learning_rate": 4.068658146854997e-06,
+      "loss": 0.7543,
+      "step": 1704
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.656326475343753,
+      "learning_rate": 4.063307432566157e-06,
+      "loss": 0.9684,
+      "step": 1705
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6215748040225293,
+      "learning_rate": 4.057957829769123e-06,
+      "loss": 0.6804,
+      "step": 1706
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3653019737070682,
+      "learning_rate": 4.052609344811802e-06,
+      "loss": 0.9464,
+      "step": 1707
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.16155910845944,
+      "learning_rate": 4.0472619840407765e-06,
+      "loss": 0.7329,
+      "step": 1708
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2963438572827695,
+      "learning_rate": 4.041915753801297e-06,
+      "loss": 0.6782,
+      "step": 1709
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3685549222952307,
+      "learning_rate": 4.03657066043727e-06,
+      "loss": 0.9331,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9966723017514155,
+      "learning_rate": 4.031226710291255e-06,
+      "loss": 0.5091,
+      "step": 1711
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0188874937600332,
+      "learning_rate": 4.025883909704453e-06,
+      "loss": 0.8949,
+      "step": 1712
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8542664968816456,
+      "learning_rate": 4.020542265016703e-06,
+      "loss": 0.9265,
+      "step": 1713
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0939717086597422,
+      "learning_rate": 4.015201782566471e-06,
+      "loss": 0.9368,
+      "step": 1714
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9266349752380412,
+      "learning_rate": 4.009862468690844e-06,
+      "loss": 0.761,
+      "step": 1715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2965490010088125,
+      "learning_rate": 4.004524329725522e-06,
+      "loss": 0.7992,
+      "step": 1716
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4211650055257108,
+      "learning_rate": 3.999187372004812e-06,
+      "loss": 0.9554,
+      "step": 1717
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2771418603377567,
+      "learning_rate": 3.993851601861618e-06,
+      "loss": 0.8699,
+      "step": 1718
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.56947610455339,
+      "learning_rate": 3.988517025627437e-06,
+      "loss": 0.8431,
+      "step": 1719
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.741394522012726,
+      "learning_rate": 3.983183649632348e-06,
+      "loss": 0.7406,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6925069465705461,
+      "learning_rate": 3.977851480205003e-06,
+      "loss": 0.5458,
+      "step": 1721
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.089933631273524,
+      "learning_rate": 3.972520523672627e-06,
+      "loss": 0.8813,
+      "step": 1722
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1951737553056558,
+      "learning_rate": 3.967190786361003e-06,
+      "loss": 0.7159,
+      "step": 1723
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.125878781921648,
+      "learning_rate": 3.9618622745944675e-06,
+      "loss": 0.9124,
+      "step": 1724
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6057667625455498,
+      "learning_rate": 3.956534994695903e-06,
+      "loss": 0.9101,
+      "step": 1725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7831277577841966,
+      "learning_rate": 3.951208952986731e-06,
+      "loss": 0.9351,
+      "step": 1726
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2099054061854602,
+      "learning_rate": 3.945884155786905e-06,
+      "loss": 0.8546,
+      "step": 1727
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4436715605065216,
+      "learning_rate": 3.940560609414894e-06,
+      "loss": 0.8377,
+      "step": 1728
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0348277611887815,
+      "learning_rate": 3.935238320187695e-06,
+      "loss": 0.8734,
+      "step": 1729
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7923533313293999,
+      "learning_rate": 3.929917294420804e-06,
+      "loss": 0.7293,
+      "step": 1730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4250332608525142,
+      "learning_rate": 3.9245975384282216e-06,
+      "loss": 0.9946,
+      "step": 1731
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7432438684267635,
+      "learning_rate": 3.919279058522438e-06,
+      "loss": 0.7491,
+      "step": 1732
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.194413982906636,
+      "learning_rate": 3.913961861014436e-06,
+      "loss": 0.9549,
+      "step": 1733
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0255930356243597,
+      "learning_rate": 3.90864595221367e-06,
+      "loss": 0.7088,
+      "step": 1734
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6650478478610728,
+      "learning_rate": 3.903331338428067e-06,
+      "loss": 0.7633,
+      "step": 1735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6364094400500908,
+      "learning_rate": 3.89801802596402e-06,
+      "loss": 0.9581,
+      "step": 1736
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0504207322702752,
+      "learning_rate": 3.892706021126376e-06,
+      "loss": 0.8926,
+      "step": 1737
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7600968707768407,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.571,
+      "step": 1738
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2477541566612533,
+      "learning_rate": 3.882085959541916e-06,
+      "loss": 0.8146,
+      "step": 1739
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1606552245938127,
+      "learning_rate": 3.876777915397006e-06,
+      "loss": 0.7229,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1535568152069566,
+      "learning_rate": 3.8714712040822954e-06,
+      "loss": 0.928,
+      "step": 1741
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7032524697564326,
+      "learning_rate": 3.866165831894796e-06,
+      "loss": 0.9289,
+      "step": 1742
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1283013999449054,
+      "learning_rate": 3.860861805129935e-06,
+      "loss": 0.9355,
+      "step": 1743
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8281699267270377,
+      "learning_rate": 3.85555913008154e-06,
+      "loss": 0.7639,
+      "step": 1744
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0319724140093811,
+      "learning_rate": 3.8502578130418345e-06,
+      "loss": 0.7764,
+      "step": 1745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1311134161926293,
+      "learning_rate": 3.844957860301434e-06,
+      "loss": 0.7332,
+      "step": 1746
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6230143361415247,
+      "learning_rate": 3.83965927814933e-06,
+      "loss": 0.6129,
+      "step": 1747
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2759460661835056,
+      "learning_rate": 3.83436207287289e-06,
+      "loss": 0.8232,
+      "step": 1748
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3805582604464486,
+      "learning_rate": 3.829066250757851e-06,
+      "loss": 0.7526,
+      "step": 1749
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7493027866518287,
+      "learning_rate": 3.823771818088303e-06,
+      "loss": 0.9199,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4950066911242845,
+      "learning_rate": 3.818478781146692e-06,
+      "loss": 0.6925,
+      "step": 1751
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4324734361036784,
+      "learning_rate": 3.813187146213802e-06,
+      "loss": 0.8257,
+      "step": 1752
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8951406557658588,
+      "learning_rate": 3.807896919568761e-06,
+      "loss": 0.9121,
+      "step": 1753
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7327704994788523,
+      "learning_rate": 3.80260810748902e-06,
+      "loss": 0.9408,
+      "step": 1754
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4427010050826792,
+      "learning_rate": 3.7973207162503535e-06,
+      "loss": 0.7915,
+      "step": 1755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7711455750010803,
+      "learning_rate": 3.7920347521268514e-06,
+      "loss": 0.512,
+      "step": 1756
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0566293894549703,
+      "learning_rate": 3.7867502213909075e-06,
+      "loss": 1.0165,
+      "step": 1757
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9768256762700678,
+      "learning_rate": 3.781467130313215e-06,
+      "loss": 0.765,
+      "step": 1758
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3481937532706605,
+      "learning_rate": 3.776185485162762e-06,
+      "loss": 0.8547,
+      "step": 1759
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.138134756700102,
+      "learning_rate": 3.7709052922068174e-06,
+      "loss": 0.9369,
+      "step": 1760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6214419649870244,
+      "learning_rate": 3.765626557710928e-06,
+      "loss": 0.6738,
+      "step": 1761
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1192642545750209,
+      "learning_rate": 3.7603492879389093e-06,
+      "loss": 0.8478,
+      "step": 1762
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9609317323338377,
+      "learning_rate": 3.7550734891528413e-06,
+      "loss": 0.84,
+      "step": 1763
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8685069696170549,
+      "learning_rate": 3.7497991676130563e-06,
+      "loss": 0.5869,
+      "step": 1764
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3032358855717079,
+      "learning_rate": 3.744526329578132e-06,
+      "loss": 0.9088,
+      "step": 1765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8895599076298986,
+      "learning_rate": 3.73925498130489e-06,
+      "loss": 0.707,
+      "step": 1766
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6307544217788137,
+      "learning_rate": 3.733985129048382e-06,
+      "loss": 0.9057,
+      "step": 1767
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7798975275179516,
+      "learning_rate": 3.728716779061882e-06,
+      "loss": 0.8735,
+      "step": 1768
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.098106490779679,
+      "learning_rate": 3.7234499375968876e-06,
+      "loss": 0.9235,
+      "step": 1769
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0191306094757255,
+      "learning_rate": 3.7181846109031007e-06,
+      "loss": 0.8756,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9436618712853083,
+      "learning_rate": 3.712920805228427e-06,
+      "loss": 0.7091,
+      "step": 1771
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9821158137496327,
+      "learning_rate": 3.707658526818968e-06,
+      "loss": 0.9396,
+      "step": 1772
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7562072892546529,
+      "learning_rate": 3.7023977819190136e-06,
+      "loss": 0.6887,
+      "step": 1773
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6152398868495805,
+      "learning_rate": 3.6971385767710345e-06,
+      "loss": 0.5127,
+      "step": 1774
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.945946987458467,
+      "learning_rate": 3.6918809176156713e-06,
+      "loss": 0.9103,
+      "step": 1775
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9795610976975223,
+      "learning_rate": 3.686624810691733e-06,
+      "loss": 0.9186,
+      "step": 1776
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2221468175595303,
+      "learning_rate": 3.6813702622361858e-06,
+      "loss": 0.8633,
+      "step": 1777
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8757942514094176,
+      "learning_rate": 3.6761172784841446e-06,
+      "loss": 0.7696,
+      "step": 1778
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6362980540753906,
+      "learning_rate": 3.6708658656688723e-06,
+      "loss": 0.929,
+      "step": 1779
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8198264318166325,
+      "learning_rate": 3.6656160300217637e-06,
+      "loss": 0.7876,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0478137369531435,
+      "learning_rate": 3.6603677777723425e-06,
+      "loss": 0.9223,
+      "step": 1781
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7433633034278525,
+      "learning_rate": 3.655121115148254e-06,
+      "loss": 0.8396,
+      "step": 1782
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9943643255290605,
+      "learning_rate": 3.64987604837526e-06,
+      "loss": 0.5895,
+      "step": 1783
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2940123747331547,
+      "learning_rate": 3.6446325836772244e-06,
+      "loss": 0.8676,
+      "step": 1784
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.442188388447257,
+      "learning_rate": 3.639390727276111e-06,
+      "loss": 0.7628,
+      "step": 1785
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9475160149731682,
+      "learning_rate": 3.6341504853919778e-06,
+      "loss": 0.9332,
+      "step": 1786
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2222394161713523,
+      "learning_rate": 3.6289118642429654e-06,
+      "loss": 0.7778,
+      "step": 1787
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8137100185613839,
+      "learning_rate": 3.6236748700452894e-06,
+      "loss": 0.8544,
+      "step": 1788
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1692218282636948,
+      "learning_rate": 3.618439509013236e-06,
+      "loss": 0.7399,
+      "step": 1789
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.246012677919529,
+      "learning_rate": 3.613205787359157e-06,
+      "loss": 0.8872,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7445038830454733,
+      "learning_rate": 3.6079737112934533e-06,
+      "loss": 0.5929,
+      "step": 1791
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6448989276657444,
+      "learning_rate": 3.602743287024576e-06,
+      "loss": 0.8537,
+      "step": 1792
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9042320440550045,
+      "learning_rate": 3.597514520759016e-06,
+      "loss": 0.8705,
+      "step": 1793
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.529008588991439,
+      "learning_rate": 3.5922874187012977e-06,
+      "loss": 0.9153,
+      "step": 1794
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1550256358657647,
+      "learning_rate": 3.587061987053967e-06,
+      "loss": 0.8874,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5667157215884582,
+      "learning_rate": 3.581838232017593e-06,
+      "loss": 0.7437,
+      "step": 1796
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1244821154176883,
+      "learning_rate": 3.576616159790751e-06,
+      "loss": 0.8528,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0647487269478164,
+      "learning_rate": 3.5713957765700224e-06,
+      "loss": 0.9904,
+      "step": 1798
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7794296909290523,
+      "learning_rate": 3.5661770885499808e-06,
+      "loss": 0.744,
+      "step": 1799
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.176803610464933,
+      "learning_rate": 3.5609601019231934e-06,
+      "loss": 0.5268,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.365169940496684,
+      "learning_rate": 3.5557448228802026e-06,
+      "loss": 0.8704,
+      "step": 1801
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0775241681631589,
+      "learning_rate": 3.5505312576095295e-06,
+      "loss": 0.9141,
+      "step": 1802
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6721126997918822,
+      "learning_rate": 3.54531941229766e-06,
+      "loss": 0.9214,
+      "step": 1803
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2045268207298043,
+      "learning_rate": 3.5401092931290378e-06,
+      "loss": 0.799,
+      "step": 1804
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1580048316121614,
+      "learning_rate": 3.5349009062860586e-06,
+      "loss": 0.8544,
+      "step": 1805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1421362522062475,
+      "learning_rate": 3.5296942579490645e-06,
+      "loss": 0.8559,
+      "step": 1806
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9551805783620073,
+      "learning_rate": 3.5244893542963332e-06,
+      "loss": 0.8869,
+      "step": 1807
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9730665537314701,
+      "learning_rate": 3.519286201504073e-06,
+      "loss": 0.8436,
+      "step": 1808
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8360284293032966,
+      "learning_rate": 3.5140848057464107e-06,
+      "loss": 0.5689,
+      "step": 1809
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9751618193995529,
+      "learning_rate": 3.5088851731953956e-06,
+      "loss": 0.9335,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.562111171496152,
+      "learning_rate": 3.503687310020978e-06,
+      "loss": 0.7894,
+      "step": 1811
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6619389822211792,
+      "learning_rate": 3.4984912223910105e-06,
+      "loss": 0.8045,
+      "step": 1812
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.237300338968151,
+      "learning_rate": 3.493296916471243e-06,
+      "loss": 0.7404,
+      "step": 1813
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.286794832362563,
+      "learning_rate": 3.488104398425304e-06,
+      "loss": 0.8008,
+      "step": 1814
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7347011696788135,
+      "learning_rate": 3.482913674414705e-06,
+      "loss": 0.9009,
+      "step": 1815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7554668182065728,
+      "learning_rate": 3.47772475059883e-06,
+      "loss": 0.9212,
+      "step": 1816
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9453254719467715,
+      "learning_rate": 3.4725376331349224e-06,
+      "loss": 0.9776,
+      "step": 1817
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8562269989168798,
+      "learning_rate": 3.4673523281780856e-06,
+      "loss": 0.487,
+      "step": 1818
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0268574946415083,
+      "learning_rate": 3.46216884188127e-06,
+      "loss": 0.8236,
+      "step": 1819
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6167545336858147,
+      "learning_rate": 3.4569871803952702e-06,
+      "loss": 0.8682,
+      "step": 1820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4858345142491183,
+      "learning_rate": 3.4518073498687145e-06,
+      "loss": 0.9433,
+      "step": 1821
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8876572923975693,
+      "learning_rate": 3.4466293564480562e-06,
+      "loss": 0.6327,
+      "step": 1822
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3602809063372636,
+      "learning_rate": 3.4414532062775743e-06,
+      "loss": 0.8461,
+      "step": 1823
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0993352964353744,
+      "learning_rate": 3.4362789054993544e-06,
+      "loss": 0.877,
+      "step": 1824
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5080931259288262,
+      "learning_rate": 3.4311064602532906e-06,
+      "loss": 0.8012,
+      "step": 1825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6967216483492461,
+      "learning_rate": 3.425935876677077e-06,
+      "loss": 0.5776,
+      "step": 1826
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9699113572616324,
+      "learning_rate": 3.420767160906194e-06,
+      "loss": 0.7931,
+      "step": 1827
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7397001046623557,
+      "learning_rate": 3.4156003190739108e-06,
+      "loss": 0.943,
+      "step": 1828
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.365028433442601,
+      "learning_rate": 3.4104353573112662e-06,
+      "loss": 0.9779,
+      "step": 1829
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0271756547526254,
+      "learning_rate": 3.4052722817470767e-06,
+      "loss": 0.8018,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6766742641424515,
+      "learning_rate": 3.4001110985079145e-06,
+      "loss": 0.8974,
+      "step": 1831
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.478892603023643,
+      "learning_rate": 3.3949518137181058e-06,
+      "loss": 0.7743,
+      "step": 1832
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.184545946632175,
+      "learning_rate": 3.38979443349973e-06,
+      "loss": 0.8825,
+      "step": 1833
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8215017839350283,
+      "learning_rate": 3.3846389639726007e-06,
+      "loss": 0.8189,
+      "step": 1834
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9335042599702176,
+      "learning_rate": 3.379485411254266e-06,
+      "loss": 0.7172,
+      "step": 1835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7157829725784235,
+      "learning_rate": 3.374333781460001e-06,
+      "loss": 0.5454,
+      "step": 1836
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2412635833364452,
+      "learning_rate": 3.3691840807027965e-06,
+      "loss": 0.7555,
+      "step": 1837
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3469432041947713,
+      "learning_rate": 3.3640363150933574e-06,
+      "loss": 0.8323,
+      "step": 1838
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.074259286706321,
+      "learning_rate": 3.3588904907400876e-06,
+      "loss": 0.7576,
+      "step": 1839
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9721966333579976,
+      "learning_rate": 3.3537466137490937e-06,
+      "loss": 0.8861,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9533874799813001,
+      "learning_rate": 3.3486046902241663e-06,
+      "loss": 0.998,
+      "step": 1841
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.677390049716744,
+      "learning_rate": 3.34346472626678e-06,
+      "loss": 0.9028,
+      "step": 1842
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6618244707699643,
+      "learning_rate": 3.3383267279760843e-06,
+      "loss": 0.9197,
+      "step": 1843
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9021495028984114,
+      "learning_rate": 3.333190701448897e-06,
+      "loss": 0.5027,
+      "step": 1844
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4662181864213324,
+      "learning_rate": 3.328056652779694e-06,
+      "loss": 0.9072,
+      "step": 1845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2514189974830376,
+      "learning_rate": 3.3229245880606063e-06,
+      "loss": 0.5718,
+      "step": 1846
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1432160947439578,
+      "learning_rate": 3.3177945133814093e-06,
+      "loss": 0.8566,
+      "step": 1847
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1413590470842585,
+      "learning_rate": 3.312666434829518e-06,
+      "loss": 0.9598,
+      "step": 1848
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8500546431237077,
+      "learning_rate": 3.307540358489977e-06,
+      "loss": 0.7856,
+      "step": 1849
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0316231539706027,
+      "learning_rate": 3.3024162904454584e-06,
+      "loss": 0.8624,
+      "step": 1850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0361121631440025,
+      "learning_rate": 3.2972942367762473e-06,
+      "loss": 0.8297,
+      "step": 1851
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.855199153280017,
+      "learning_rate": 3.2921742035602404e-06,
+      "loss": 0.8937,
+      "step": 1852
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6268178254953594,
+      "learning_rate": 3.2870561968729388e-06,
+      "loss": 0.6025,
+      "step": 1853
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0151428666972433,
+      "learning_rate": 3.2819402227874364e-06,
+      "loss": 0.8228,
+      "step": 1854
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2294820312971966,
+      "learning_rate": 3.276826287374414e-06,
+      "loss": 0.8958,
+      "step": 1855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8679894914755037,
+      "learning_rate": 3.271714396702137e-06,
+      "loss": 0.8838,
+      "step": 1856
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2561112593307222,
+      "learning_rate": 3.2666045568364423e-06,
+      "loss": 0.8341,
+      "step": 1857
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.018077566751702,
+      "learning_rate": 3.2614967738407332e-06,
+      "loss": 0.623,
+      "step": 1858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2109109138028347,
+      "learning_rate": 3.2563910537759723e-06,
+      "loss": 0.7002,
+      "step": 1859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4320156314163426,
+      "learning_rate": 3.2512874027006753e-06,
+      "loss": 0.8445,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0329544787489917,
+      "learning_rate": 3.2461858266709017e-06,
+      "loss": 0.8735,
+      "step": 1861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9984863093594726,
+      "learning_rate": 3.2410863317402486e-06,
+      "loss": 0.8821,
+      "step": 1862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.947382081593161,
+      "learning_rate": 3.2359889239598456e-06,
+      "loss": 0.4531,
+      "step": 1863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.797570404260827,
+      "learning_rate": 3.2308936093783436e-06,
+      "loss": 0.8917,
+      "step": 1864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7846002280653357,
+      "learning_rate": 3.225800394041911e-06,
+      "loss": 0.9698,
+      "step": 1865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.371136543215513,
+      "learning_rate": 3.220709283994222e-06,
+      "loss": 0.7689,
+      "step": 1866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3112106472443754,
+      "learning_rate": 3.2156202852764584e-06,
+      "loss": 0.9321,
+      "step": 1867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1147641583086627,
+      "learning_rate": 3.2105334039272924e-06,
+      "loss": 0.7254,
+      "step": 1868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0736848268363257,
+      "learning_rate": 3.205448645982883e-06,
+      "loss": 0.8504,
+      "step": 1869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4378396771998043,
+      "learning_rate": 3.2003660174768746e-06,
+      "loss": 0.4646,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.07727015136679,
+      "learning_rate": 3.195285524440379e-06,
+      "loss": 0.82,
+      "step": 1871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1775394380919233,
+      "learning_rate": 3.1902071729019767e-06,
+      "loss": 0.832,
+      "step": 1872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.307036238036918,
+      "learning_rate": 3.1851309688877074e-06,
+      "loss": 0.7695,
+      "step": 1873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9312624399342299,
+      "learning_rate": 3.1800569184210627e-06,
+      "loss": 1.0246,
+      "step": 1874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9041755825406368,
+      "learning_rate": 3.1749850275229777e-06,
+      "loss": 0.9141,
+      "step": 1875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.224193339333444,
+      "learning_rate": 3.1699153022118236e-06,
+      "loss": 0.7876,
+      "step": 1876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3052274218787692,
+      "learning_rate": 3.1648477485034062e-06,
+      "loss": 0.9262,
+      "step": 1877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.023954407049467,
+      "learning_rate": 3.15978237241095e-06,
+      "loss": 0.9061,
+      "step": 1878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7805774284463238,
+      "learning_rate": 3.1547191799450976e-06,
+      "loss": 0.5237,
+      "step": 1879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1740378668405023,
+      "learning_rate": 3.1496581771139013e-06,
+      "loss": 0.7852,
+      "step": 1880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1175281041949063,
+      "learning_rate": 3.1445993699228132e-06,
+      "loss": 0.9243,
+      "step": 1881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2755442944882132,
+      "learning_rate": 3.1395427643746802e-06,
+      "loss": 0.6282,
+      "step": 1882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.498024663936799,
+      "learning_rate": 3.134488366469739e-06,
+      "loss": 0.944,
+      "step": 1883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.006933399046772,
+      "learning_rate": 3.1294361822056048e-06,
+      "loss": 0.8043,
+      "step": 1884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0090687859418357,
+      "learning_rate": 3.1243862175772656e-06,
+      "loss": 0.8403,
+      "step": 1885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8823022491814754,
+      "learning_rate": 3.1193384785770755e-06,
+      "loss": 0.8125,
+      "step": 1886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9967397312708116,
+      "learning_rate": 3.1142929711947502e-06,
+      "loss": 0.866,
+      "step": 1887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9257595299879717,
+      "learning_rate": 3.109249701417355e-06,
+      "loss": 0.883,
+      "step": 1888
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.883517971842169,
+      "learning_rate": 3.1042086752292995e-06,
+      "loss": 0.4565,
+      "step": 1889
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7391787478526181,
+      "learning_rate": 3.0991698986123343e-06,
+      "loss": 0.918,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2288338324735373,
+      "learning_rate": 3.0941333775455367e-06,
+      "loss": 0.8212,
+      "step": 1891
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6256256556957354,
+      "learning_rate": 3.0890991180053094e-06,
+      "loss": 0.7121,
+      "step": 1892
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9653149782155358,
+      "learning_rate": 3.0840671259653714e-06,
+      "loss": 0.9479,
+      "step": 1893
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.754654089825459,
+      "learning_rate": 3.0790374073967523e-06,
+      "loss": 0.6859,
+      "step": 1894
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6921570371012875,
+      "learning_rate": 3.074009968267782e-06,
+      "loss": 0.9587,
+      "step": 1895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.125897978181646,
+      "learning_rate": 3.068984814544087e-06,
+      "loss": 0.8131,
+      "step": 1896
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8067118595052013,
+      "learning_rate": 3.0639619521885795e-06,
+      "loss": 0.4694,
+      "step": 1897
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9091776926579623,
+      "learning_rate": 3.058941387161456e-06,
+      "loss": 0.9819,
+      "step": 1898
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1754807142009933,
+      "learning_rate": 3.0539231254201855e-06,
+      "loss": 0.7959,
+      "step": 1899
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0459605800208533,
+      "learning_rate": 3.0489071729195025e-06,
+      "loss": 0.9322,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6493301805983154,
+      "learning_rate": 3.043893535611404e-06,
+      "loss": 0.7793,
+      "step": 1901
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4738027904025492,
+      "learning_rate": 3.0388822194451385e-06,
+      "loss": 0.7773,
+      "step": 1902
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.240447137280675,
+      "learning_rate": 3.0338732303671993e-06,
+      "loss": 0.9162,
+      "step": 1903
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6639066137246032,
+      "learning_rate": 3.0288665743213186e-06,
+      "loss": 0.8798,
+      "step": 1904
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.063022926759733,
+      "learning_rate": 3.023862257248462e-06,
+      "loss": 0.8756,
+      "step": 1905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7328469535612668,
+      "learning_rate": 3.0188602850868186e-06,
+      "loss": 0.7051,
+      "step": 1906
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6892447487894612,
+      "learning_rate": 3.0138606637717925e-06,
+      "loss": 0.6388,
+      "step": 1907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0655609185590362,
+      "learning_rate": 3.0088633992360037e-06,
+      "loss": 0.8215,
+      "step": 1908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6907783789557127,
+      "learning_rate": 3.003868497409271e-06,
+      "loss": 0.7035,
+      "step": 1909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.358122114377092,
+      "learning_rate": 2.99887596421861e-06,
+      "loss": 0.9796,
+      "step": 1910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.37191162376772,
+      "learning_rate": 2.9938858055882287e-06,
+      "loss": 0.8124,
+      "step": 1911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4924508603057514,
+      "learning_rate": 2.988898027439514e-06,
+      "loss": 0.875,
+      "step": 1912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0846326837330404,
+      "learning_rate": 2.9839126356910308e-06,
+      "loss": 0.901,
+      "step": 1913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0702514173849522,
+      "learning_rate": 2.9789296362585084e-06,
+      "loss": 0.9229,
+      "step": 1914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8422699658039259,
+      "learning_rate": 2.973949035054843e-06,
+      "loss": 0.7518,
+      "step": 1915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5881199844453088,
+      "learning_rate": 2.96897083799008e-06,
+      "loss": 0.7213,
+      "step": 1916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3890820375384423,
+      "learning_rate": 2.9639950509714138e-06,
+      "loss": 1.0035,
+      "step": 1917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9923798192601897,
+      "learning_rate": 2.9590216799031814e-06,
+      "loss": 0.8212,
+      "step": 1918
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.13029863252314,
+      "learning_rate": 2.954050730686848e-06,
+      "loss": 0.9201,
+      "step": 1919
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.239945897039648,
+      "learning_rate": 2.949082209221009e-06,
+      "loss": 0.7144,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5761533391615632,
+      "learning_rate": 2.9441161214013787e-06,
+      "loss": 0.8108,
+      "step": 1921
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1255151895180664,
+      "learning_rate": 2.939152473120781e-06,
+      "loss": 0.9607,
+      "step": 1922
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8113776382939397,
+      "learning_rate": 2.934191270269148e-06,
+      "loss": 0.6585,
+      "step": 1923
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4187338109415966,
+      "learning_rate": 2.929232518733507e-06,
+      "loss": 0.8609,
+      "step": 1924
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5506853845108772,
+      "learning_rate": 2.9242762243979804e-06,
+      "loss": 0.8427,
+      "step": 1925
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.136869471582903,
+      "learning_rate": 2.919322393143772e-06,
+      "loss": 0.8773,
+      "step": 1926
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2351465713343792,
+      "learning_rate": 2.9143710308491625e-06,
+      "loss": 0.7255,
+      "step": 1927
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5182923564730045,
+      "learning_rate": 2.909422143389504e-06,
+      "loss": 0.8922,
+      "step": 1928
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1548808335549023,
+      "learning_rate": 2.9044757366372123e-06,
+      "loss": 0.9608,
+      "step": 1929
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1730406667698414,
+      "learning_rate": 2.8995318164617614e-06,
+      "loss": 0.6996,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0573552684064784,
+      "learning_rate": 2.8945903887296686e-06,
+      "loss": 0.7994,
+      "step": 1931
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9235412589414317,
+      "learning_rate": 2.889651459304499e-06,
+      "loss": 0.7663,
+      "step": 1932
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6004894378705573,
+      "learning_rate": 2.884715034046854e-06,
+      "loss": 0.6623,
+      "step": 1933
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.328600857471778,
+      "learning_rate": 2.8797811188143572e-06,
+      "loss": 0.7102,
+      "step": 1934
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8464323146432448,
+      "learning_rate": 2.874849719461662e-06,
+      "loss": 0.7846,
+      "step": 1935
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8528350014222902,
+      "learning_rate": 2.8699208418404273e-06,
+      "loss": 0.8003,
+      "step": 1936
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9945318822218923,
+      "learning_rate": 2.8649944917993278e-06,
+      "loss": 0.7911,
+      "step": 1937
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1683491104592674,
+      "learning_rate": 2.860070675184036e-06,
+      "loss": 0.8419,
+      "step": 1938
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0480237192568085,
+      "learning_rate": 2.8551493978372158e-06,
+      "loss": 0.8495,
+      "step": 1939
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6968489450806228,
+      "learning_rate": 2.8502306655985207e-06,
+      "loss": 0.9498,
+      "step": 1940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7260581293937647,
+      "learning_rate": 2.845314484304586e-06,
+      "loss": 0.5207,
+      "step": 1941
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4095074574463686,
+      "learning_rate": 2.840400859789013e-06,
+      "loss": 0.8355,
+      "step": 1942
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.487605664149953,
+      "learning_rate": 2.835489797882377e-06,
+      "loss": 0.8978,
+      "step": 1943
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0271678266529929,
+      "learning_rate": 2.83058130441221e-06,
+      "loss": 0.7307,
+      "step": 1944
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0638515883833133,
+      "learning_rate": 2.8256753852029917e-06,
+      "loss": 0.8666,
+      "step": 1945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0017478836103846,
+      "learning_rate": 2.8207720460761523e-06,
+      "loss": 0.9163,
+      "step": 1946
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9503949918034911,
+      "learning_rate": 2.8158712928500607e-06,
+      "loss": 0.6821,
+      "step": 1947
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1191057389343484,
+      "learning_rate": 2.810973131340013e-06,
+      "loss": 0.799,
+      "step": 1948
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9332408795255149,
+      "learning_rate": 2.806077567358233e-06,
+      "loss": 0.8636,
+      "step": 1949
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8040096719282792,
+      "learning_rate": 2.8011846067138648e-06,
+      "loss": 0.6217,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.526462114283599,
+      "learning_rate": 2.7962942552129558e-06,
+      "loss": 0.7807,
+      "step": 1951
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0758445707853983,
+      "learning_rate": 2.7914065186584637e-06,
+      "loss": 0.7769,
+      "step": 1952
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3273730998669404,
+      "learning_rate": 2.786521402850244e-06,
+      "loss": 0.8464,
+      "step": 1953
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7077143443626882,
+      "learning_rate": 2.7816389135850353e-06,
+      "loss": 0.8623,
+      "step": 1954
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.565599934111027,
+      "learning_rate": 2.776759056656468e-06,
+      "loss": 0.9636,
+      "step": 1955
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.073496490310275,
+      "learning_rate": 2.771881837855041e-06,
+      "loss": 0.8663,
+      "step": 1956
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9891234962671787,
+      "learning_rate": 2.767007262968129e-06,
+      "loss": 0.9634,
+      "step": 1957
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2637347247787116,
+      "learning_rate": 2.762135337779969e-06,
+      "loss": 0.7669,
+      "step": 1958
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8618682925588199,
+      "learning_rate": 2.757266068071648e-06,
+      "loss": 0.6582,
+      "step": 1959
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6319585048724882,
+      "learning_rate": 2.752399459621108e-06,
+      "loss": 0.4955,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.32878086826792,
+      "learning_rate": 2.7475355182031315e-06,
+      "loss": 0.9226,
+      "step": 1961
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9006164755818888,
+      "learning_rate": 2.7426742495893343e-06,
+      "loss": 0.9144,
+      "step": 1962
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3932404460116814,
+      "learning_rate": 2.7378156595481625e-06,
+      "loss": 0.8821,
+      "step": 1963
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3587817060046663,
+      "learning_rate": 2.732959753844886e-06,
+      "loss": 0.7464,
+      "step": 1964
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.853393559543957,
+      "learning_rate": 2.7281065382415822e-06,
+      "loss": 0.9381,
+      "step": 1965
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9936965844511633,
+      "learning_rate": 2.7232560184971437e-06,
+      "loss": 0.7461,
+      "step": 1966
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9981127133008043,
+      "learning_rate": 2.718408200367263e-06,
+      "loss": 0.8982,
+      "step": 1967
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1895179252054031,
+      "learning_rate": 2.713563089604422e-06,
+      "loss": 0.8917,
+      "step": 1968
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.156848474133586,
+      "learning_rate": 2.7087206919578932e-06,
+      "loss": 0.584,
+      "step": 1969
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5252815529486679,
+      "learning_rate": 2.7038810131737346e-06,
+      "loss": 0.8527,
+      "step": 1970
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0257215506573054,
+      "learning_rate": 2.6990440589947676e-06,
+      "loss": 0.7357,
+      "step": 1971
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3155864765833492,
+      "learning_rate": 2.6942098351605883e-06,
+      "loss": 0.996,
+      "step": 1972
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.54147789199519,
+      "learning_rate": 2.689378347407553e-06,
+      "loss": 0.7204,
+      "step": 1973
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9836387407001115,
+      "learning_rate": 2.684549601468764e-06,
+      "loss": 0.8579,
+      "step": 1974
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0424257461347062,
+      "learning_rate": 2.6797236030740803e-06,
+      "loss": 0.7278,
+      "step": 1975
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1853349398798707,
+      "learning_rate": 2.674900357950091e-06,
+      "loss": 0.9864,
+      "step": 1976
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7556662197643502,
+      "learning_rate": 2.670079871820125e-06,
+      "loss": 0.5067,
+      "step": 1977
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7981325804839247,
+      "learning_rate": 2.6652621504042366e-06,
+      "loss": 0.8426,
+      "step": 1978
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 24.275275655508104,
+      "learning_rate": 2.660447199419194e-06,
+      "loss": 0.9051,
+      "step": 1979
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1021108249315148,
+      "learning_rate": 2.6556350245784833e-06,
+      "loss": 0.9799,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8462434083370993,
+      "learning_rate": 2.6508256315922975e-06,
+      "loss": 0.7994,
+      "step": 1981
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0159623684049648,
+      "learning_rate": 2.6460190261675223e-06,
+      "loss": 0.827,
+      "step": 1982
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1259654266492716,
+      "learning_rate": 2.6412152140077408e-06,
+      "loss": 0.7657,
+      "step": 1983
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4829185334449881,
+      "learning_rate": 2.6364142008132216e-06,
+      "loss": 0.9873,
+      "step": 1984
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2646298269331766,
+      "learning_rate": 2.631615992280907e-06,
+      "loss": 0.8197,
+      "step": 1985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7863583977784274,
+      "learning_rate": 2.626820594104418e-06,
+      "loss": 0.5339,
+      "step": 1986
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1691825412539085,
+      "learning_rate": 2.6220280119740376e-06,
+      "loss": 0.7619,
+      "step": 1987
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9464063243079284,
+      "learning_rate": 2.6172382515767046e-06,
+      "loss": 0.9381,
+      "step": 1988
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6001557266079228,
+      "learning_rate": 2.6124513185960132e-06,
+      "loss": 0.9029,
+      "step": 1989
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3315053183565022,
+      "learning_rate": 2.6076672187122043e-06,
+      "loss": 0.8151,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3883512839097574,
+      "learning_rate": 2.6028859576021492e-06,
+      "loss": 0.7267,
+      "step": 1991
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0105735158578668,
+      "learning_rate": 2.5981075409393587e-06,
+      "loss": 0.9239,
+      "step": 1992
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6329290294638767,
+      "learning_rate": 2.593331974393966e-06,
+      "loss": 0.882,
+      "step": 1993
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.695589358884439,
+      "learning_rate": 2.588559263632719e-06,
+      "loss": 0.5545,
+      "step": 1994
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3493370661675521,
+      "learning_rate": 2.5837894143189825e-06,
+      "loss": 0.9171,
+      "step": 1995
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5172056248363428,
+      "learning_rate": 2.57902243211272e-06,
+      "loss": 0.8305,
+      "step": 1996
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.286496659663681,
+      "learning_rate": 2.574258322670496e-06,
+      "loss": 0.7721,
+      "step": 1997
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4110130247442307,
+      "learning_rate": 2.5694970916454686e-06,
+      "loss": 0.7911,
+      "step": 1998
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1927708633155314,
+      "learning_rate": 2.5647387446873746e-06,
+      "loss": 0.7924,
+      "step": 1999
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4329834370667136,
+      "learning_rate": 2.5599832874425325e-06,
+      "loss": 0.983,
+      "step": 2000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0353468403684816,
+      "learning_rate": 2.555230725553832e-06,
+      "loss": 0.7482,
+      "step": 2001
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6578208591032251,
+      "learning_rate": 2.550481064660724e-06,
+      "loss": 0.9069,
+      "step": 2002
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9211726309738654,
+      "learning_rate": 2.54573431039922e-06,
+      "loss": 0.9658,
+      "step": 2003
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0693586015756704,
+      "learning_rate": 2.540990468401883e-06,
+      "loss": 0.4383,
+      "step": 2004
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0693803629192602,
+      "learning_rate": 2.536249544297815e-06,
+      "loss": 0.9018,
+      "step": 2005
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9542724765471767,
+      "learning_rate": 2.531511543712662e-06,
+      "loss": 0.8935,
+      "step": 2006
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.188416908292624,
+      "learning_rate": 2.5267764722685995e-06,
+      "loss": 0.9002,
+      "step": 2007
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8348515981783474,
+      "learning_rate": 2.522044335584322e-06,
+      "loss": 0.688,
+      "step": 2008
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3010319538444532,
+      "learning_rate": 2.5173151392750477e-06,
+      "loss": 0.8332,
+      "step": 2009
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0386552657035077,
+      "learning_rate": 2.5125888889525057e-06,
+      "loss": 0.8921,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0087746900865306,
+      "learning_rate": 2.5078655902249233e-06,
+      "loss": 0.7261,
+      "step": 2011
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6652084647212105,
+      "learning_rate": 2.503145248697033e-06,
+      "loss": 0.5843,
+      "step": 2012
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8837902697373446,
+      "learning_rate": 2.4984278699700504e-06,
+      "loss": 0.7292,
+      "step": 2013
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6144750435007447,
+      "learning_rate": 2.4937134596416823e-06,
+      "loss": 0.9895,
+      "step": 2014
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5986888299469089,
+      "learning_rate": 2.489002023306112e-06,
+      "loss": 0.8072,
+      "step": 2015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3645922150347785,
+      "learning_rate": 2.4842935665539873e-06,
+      "loss": 0.8031,
+      "step": 2016
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1200394479907605,
+      "learning_rate": 2.4795880949724296e-06,
+      "loss": 0.9189,
+      "step": 2017
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8258467432033063,
+      "learning_rate": 2.4748856141450132e-06,
+      "loss": 0.7065,
+      "step": 2018
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8314621644475588,
+      "learning_rate": 2.4701861296517623e-06,
+      "loss": 0.9595,
+      "step": 2019
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3628517055418168,
+      "learning_rate": 2.465489647069148e-06,
+      "loss": 0.7752,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.266772452357036,
+      "learning_rate": 2.4607961719700806e-06,
+      "loss": 0.83,
+      "step": 2021
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7717936970575051,
+      "learning_rate": 2.4561057099238973e-06,
+      "loss": 0.5308,
+      "step": 2022
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.742102521605795,
+      "learning_rate": 2.4514182664963633e-06,
+      "loss": 0.702,
+      "step": 2023
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3117924762420108,
+      "learning_rate": 2.4467338472496633e-06,
+      "loss": 0.8277,
+      "step": 2024
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2476710026359297,
+      "learning_rate": 2.4420524577423872e-06,
+      "loss": 0.755,
+      "step": 2025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9714968084652374,
+      "learning_rate": 2.4373741035295357e-06,
+      "loss": 0.888,
+      "step": 2026
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6774722173495927,
+      "learning_rate": 2.432698790162508e-06,
+      "loss": 0.9828,
+      "step": 2027
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0633662154085983,
+      "learning_rate": 2.4280265231890887e-06,
+      "loss": 0.7477,
+      "step": 2028
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.79374951417817,
+      "learning_rate": 2.423357308153454e-06,
+      "loss": 0.6088,
+      "step": 2029
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2050101316579092,
+      "learning_rate": 2.418691150596158e-06,
+      "loss": 0.732,
+      "step": 2030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0514058708301328,
+      "learning_rate": 2.4140280560541214e-06,
+      "loss": 0.88,
+      "step": 2031
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.24783047292958,
+      "learning_rate": 2.409368030060637e-06,
+      "loss": 0.6243,
+      "step": 2032
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9407094417479344,
+      "learning_rate": 2.4047110781453498e-06,
+      "loss": 0.7774,
+      "step": 2033
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5937063132570195,
+      "learning_rate": 2.4000572058342637e-06,
+      "loss": 0.9605,
+      "step": 2034
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9858584830195187,
+      "learning_rate": 2.3954064186497257e-06,
+      "loss": 0.7533,
+      "step": 2035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7213803626019681,
+      "learning_rate": 2.390758722110418e-06,
+      "loss": 0.782,
+      "step": 2036
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8945920085458076,
+      "learning_rate": 2.386114121731361e-06,
+      "loss": 0.985,
+      "step": 2037
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5692698352580092,
+      "learning_rate": 2.3814726230239006e-06,
+      "loss": 0.7589,
+      "step": 2038
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7787189550555421,
+      "learning_rate": 2.3768342314956973e-06,
+      "loss": 0.6031,
+      "step": 2039
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9284137123666689,
+      "learning_rate": 2.3721989526507296e-06,
+      "loss": 0.8264,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7134196816499574,
+      "learning_rate": 2.3675667919892827e-06,
+      "loss": 0.9216,
+      "step": 2041
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8725857158656227,
+      "learning_rate": 2.362937755007935e-06,
+      "loss": 0.6918,
+      "step": 2042
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0944813535376132,
+      "learning_rate": 2.358311847199567e-06,
+      "loss": 0.9257,
+      "step": 2043
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.450815910976798,
+      "learning_rate": 2.353689074053342e-06,
+      "loss": 0.6368,
+      "step": 2044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7886699830112005,
+      "learning_rate": 2.3490694410547015e-06,
+      "loss": 0.6403,
+      "step": 2045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5543093414269864,
+      "learning_rate": 2.3444529536853645e-06,
+      "loss": 0.8799,
+      "step": 2046
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1438299060525436,
+      "learning_rate": 2.339839617423318e-06,
+      "loss": 0.8268,
+      "step": 2047
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7427806193724306,
+      "learning_rate": 2.3352294377428035e-06,
+      "loss": 0.5907,
+      "step": 2048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1155379042169018,
+      "learning_rate": 2.3306224201143242e-06,
+      "loss": 0.7981,
+      "step": 2049
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1713047781315247,
+      "learning_rate": 2.3260185700046295e-06,
+      "loss": 0.9655,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2594681830948358,
+      "learning_rate": 2.321417892876706e-06,
+      "loss": 0.8819,
+      "step": 2051
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6919265316853799,
+      "learning_rate": 2.31682039418978e-06,
+      "loss": 0.9174,
+      "step": 2052
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9062346224720673,
+      "learning_rate": 2.312226079399302e-06,
+      "loss": 0.8665,
+      "step": 2053
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0425334825826984,
+      "learning_rate": 2.307634953956948e-06,
+      "loss": 0.7472,
+      "step": 2054
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3923237150413887,
+      "learning_rate": 2.303047023310611e-06,
+      "loss": 0.8547,
+      "step": 2055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9473232588253852,
+      "learning_rate": 2.298462292904386e-06,
+      "loss": 0.4699,
+      "step": 2056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.276811564528585,
+      "learning_rate": 2.2938807681785764e-06,
+      "loss": 0.7906,
+      "step": 2057
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2792231206736864,
+      "learning_rate": 2.2893024545696822e-06,
+      "loss": 0.887,
+      "step": 2058
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0248207407589744,
+      "learning_rate": 2.2847273575103873e-06,
+      "loss": 0.7975,
+      "step": 2059
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5223184222554462,
+      "learning_rate": 2.2801554824295646e-06,
+      "loss": 1.0814,
+      "step": 2060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.106738703417503,
+      "learning_rate": 2.2755868347522624e-06,
+      "loss": 0.8816,
+      "step": 2061
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1501107034141043,
+      "learning_rate": 2.271021419899696e-06,
+      "loss": 0.9093,
+      "step": 2062
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9349464461409852,
+      "learning_rate": 2.2664592432892485e-06,
+      "loss": 0.7588,
+      "step": 2063
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.60859403965203,
+      "learning_rate": 2.2619003103344607e-06,
+      "loss": 0.8708,
+      "step": 2064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9506573714455907,
+      "learning_rate": 2.2573446264450195e-06,
+      "loss": 0.5248,
+      "step": 2065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.912982989796076,
+      "learning_rate": 2.2527921970267614e-06,
+      "loss": 0.8141,
+      "step": 2066
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7156234352444277,
+      "learning_rate": 2.2482430274816613e-06,
+      "loss": 0.7352,
+      "step": 2067
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0708671955618982,
+      "learning_rate": 2.2436971232078204e-06,
+      "loss": 0.7742,
+      "step": 2068
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.614413373377002,
+      "learning_rate": 2.23915448959947e-06,
+      "loss": 0.9666,
+      "step": 2069
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0122969562806863,
+      "learning_rate": 2.234615132046962e-06,
+      "loss": 0.7711,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9147663909752742,
+      "learning_rate": 2.2300790559367553e-06,
+      "loss": 0.8254,
+      "step": 2071
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1031952139510297,
+      "learning_rate": 2.22554626665142e-06,
+      "loss": 0.8411,
+      "step": 2072
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1511403904583515,
+      "learning_rate": 2.2210167695696205e-06,
+      "loss": 0.8124,
+      "step": 2073
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0064679104005139,
+      "learning_rate": 2.21649057006612e-06,
+      "loss": 0.8938,
+      "step": 2074
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7581993061875385,
+      "learning_rate": 2.211967673511769e-06,
+      "loss": 0.4608,
+      "step": 2075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7111754852353394,
+      "learning_rate": 2.2074480852734926e-06,
+      "loss": 0.9149,
+      "step": 2076
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.632848044675317,
+      "learning_rate": 2.2029318107142954e-06,
+      "loss": 0.773,
+      "step": 2077
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.424190160095501,
+      "learning_rate": 2.1984188551932513e-06,
+      "loss": 0.7485,
+      "step": 2078
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0294070564502176,
+      "learning_rate": 2.1939092240654886e-06,
+      "loss": 0.9205,
+      "step": 2079
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0954220798334724,
+      "learning_rate": 2.1894029226821973e-06,
+      "loss": 0.7076,
+      "step": 2080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1595940405192195,
+      "learning_rate": 2.1848999563906163e-06,
+      "loss": 0.837,
+      "step": 2081
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1013263732161542,
+      "learning_rate": 2.1804003305340217e-06,
+      "loss": 0.9143,
+      "step": 2082
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7334402047121308,
+      "learning_rate": 2.17590405045173e-06,
+      "loss": 0.481,
+      "step": 2083
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1140911037194614,
+      "learning_rate": 2.17141112147909e-06,
+      "loss": 0.9429,
+      "step": 2084
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2390136874660227,
+      "learning_rate": 2.166921548947466e-06,
+      "loss": 0.8204,
+      "step": 2085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9739486204200531,
+      "learning_rate": 2.1624353381842457e-06,
+      "loss": 0.9912,
+      "step": 2086
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9958794871524923,
+      "learning_rate": 2.1579524945128287e-06,
+      "loss": 0.7627,
+      "step": 2087
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3672054626201342,
+      "learning_rate": 2.153473023252613e-06,
+      "loss": 0.7402,
+      "step": 2088
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2163050878674657,
+      "learning_rate": 2.1489969297190017e-06,
+      "loss": 0.9139,
+      "step": 2089
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7674679659451344,
+      "learning_rate": 2.144524219223383e-06,
+      "loss": 0.8727,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8393155311195916,
+      "learning_rate": 2.1400548970731377e-06,
+      "loss": 0.698,
+      "step": 2091
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8601394571753223,
+      "learning_rate": 2.1355889685716225e-06,
+      "loss": 0.8081,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7522916979257017,
+      "learning_rate": 2.1311264390181647e-06,
+      "loss": 0.6498,
+      "step": 2093
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0373767481104474,
+      "learning_rate": 2.126667313708062e-06,
+      "loss": 0.7178,
+      "step": 2094
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4096787154397872,
+      "learning_rate": 2.1222115979325746e-06,
+      "loss": 0.687,
+      "step": 2095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9789253676861411,
+      "learning_rate": 2.1177592969789092e-06,
+      "loss": 1.0175,
+      "step": 2096
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2650366999773184,
+      "learning_rate": 2.113310416130227e-06,
+      "loss": 0.7763,
+      "step": 2097
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.767590580864056,
+      "learning_rate": 2.1088649606656313e-06,
+      "loss": 0.8969,
+      "step": 2098
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1766769976353786,
+      "learning_rate": 2.1044229358601543e-06,
+      "loss": 0.8823,
+      "step": 2099
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.126131187105372,
+      "learning_rate": 2.099984346984763e-06,
+      "loss": 0.8782,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7881643468823694,
+      "learning_rate": 2.095549199306348e-06,
+      "loss": 0.4723,
+      "step": 2101
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2346138566352192,
+      "learning_rate": 2.0911174980877106e-06,
+      "loss": 0.8591,
+      "step": 2102
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1389452840865455,
+      "learning_rate": 2.0866892485875684e-06,
+      "loss": 1.0209,
+      "step": 2103
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2468371115479295,
+      "learning_rate": 2.082264456060542e-06,
+      "loss": 0.6829,
+      "step": 2104
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9748471194316136,
+      "learning_rate": 2.0778431257571453e-06,
+      "loss": 0.938,
+      "step": 2105
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9100141212388755,
+      "learning_rate": 2.0734252629237892e-06,
+      "loss": 0.6873,
+      "step": 2106
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9827851614275706,
+      "learning_rate": 2.06901087280277e-06,
+      "loss": 0.7698,
+      "step": 2107
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4368936051763133,
+      "learning_rate": 2.064599960632259e-06,
+      "loss": 0.9429,
+      "step": 2108
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7658833888530241,
+      "learning_rate": 2.0601925316463047e-06,
+      "loss": 0.6385,
+      "step": 2109
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0133262149425457,
+      "learning_rate": 2.0557885910748177e-06,
+      "loss": 0.8721,
+      "step": 2110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4929422890785522,
+      "learning_rate": 2.0513881441435748e-06,
+      "loss": 0.8371,
+      "step": 2111
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4263099147334162,
+      "learning_rate": 2.0469911960742055e-06,
+      "loss": 0.8884,
+      "step": 2112
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9489625658833941,
+      "learning_rate": 2.0425977520841837e-06,
+      "loss": 0.7435,
+      "step": 2113
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5530449653833549,
+      "learning_rate": 2.03820781738683e-06,
+      "loss": 0.9317,
+      "step": 2114
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9567369997124039,
+      "learning_rate": 2.0338213971913e-06,
+      "loss": 0.8667,
+      "step": 2115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0423651801491989,
+      "learning_rate": 2.0294384967025748e-06,
+      "loss": 0.6488,
+      "step": 2116
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.183966102027982,
+      "learning_rate": 2.0250591211214647e-06,
+      "loss": 0.8271,
+      "step": 2117
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0301817665602286,
+      "learning_rate": 2.0206832756445954e-06,
+      "loss": 0.7817,
+      "step": 2118
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5780094828658037,
+      "learning_rate": 2.0163109654644e-06,
+      "loss": 0.7099,
+      "step": 2119
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2670662266545136,
+      "learning_rate": 2.011942195769122e-06,
+      "loss": 0.7339,
+      "step": 2120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3667485453456756,
+      "learning_rate": 2.007576971742803e-06,
+      "loss": 0.7436,
+      "step": 2121
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4057697879596776,
+      "learning_rate": 2.0032152985652708e-06,
+      "loss": 0.7876,
+      "step": 2122
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4249732054729267,
+      "learning_rate": 1.9988571814121476e-06,
+      "loss": 0.7957,
+      "step": 2123
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.036660670008243,
+      "learning_rate": 1.994502625454834e-06,
+      "loss": 0.8869,
+      "step": 2124
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0305650308537146,
+      "learning_rate": 1.9901516358605012e-06,
+      "loss": 0.7633,
+      "step": 2125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5733508497575637,
+      "learning_rate": 1.9858042177920915e-06,
+      "loss": 0.9362,
+      "step": 2126
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.801805987678441,
+      "learning_rate": 1.9814603764083112e-06,
+      "loss": 0.5285,
+      "step": 2127
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.779676647872439,
+      "learning_rate": 1.977120116863617e-06,
+      "loss": 0.7964,
+      "step": 2128
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8455293138713398,
+      "learning_rate": 1.972783444308221e-06,
+      "loss": 0.8257,
+      "step": 2129
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4501592415243942,
+      "learning_rate": 1.968450363888073e-06,
+      "loss": 0.83,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5272513542144581,
+      "learning_rate": 1.964120880744866e-06,
+      "loss": 0.8709,
+      "step": 2131
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0605635867643433,
+      "learning_rate": 1.959795000016023e-06,
+      "loss": 0.9066,
+      "step": 2132
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.093879664828553,
+      "learning_rate": 1.9554727268346876e-06,
+      "loss": 0.6393,
+      "step": 2133
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9552986121467508,
+      "learning_rate": 1.9511540663297284e-06,
+      "loss": 0.8259,
+      "step": 2134
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0346977729927838,
+      "learning_rate": 1.9468390236257277e-06,
+      "loss": 0.8159,
+      "step": 2135
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.925831891946377,
+      "learning_rate": 1.9425276038429684e-06,
+      "loss": 0.6328,
+      "step": 2136
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0048939165793738,
+      "learning_rate": 1.9382198120974395e-06,
+      "loss": 0.7981,
+      "step": 2137
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9604684170687342,
+      "learning_rate": 1.933915653500826e-06,
+      "loss": 0.7558,
+      "step": 2138
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9029646844796778,
+      "learning_rate": 1.929615133160496e-06,
+      "loss": 0.9386,
+      "step": 2139
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.15694134798398,
+      "learning_rate": 1.925318256179505e-06,
+      "loss": 0.6621,
+      "step": 2140
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4505424962172881,
+      "learning_rate": 1.921025027656587e-06,
+      "loss": 0.9384,
+      "step": 2141
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9445255070929149,
+      "learning_rate": 1.916735452686139e-06,
+      "loss": 0.8677,
+      "step": 2142
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9128067944982278,
+      "learning_rate": 1.9124495363582308e-06,
+      "loss": 0.9324,
+      "step": 2143
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1008775880811192,
+      "learning_rate": 1.9081672837585873e-06,
+      "loss": 0.7436,
+      "step": 2144
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3518319248243669,
+      "learning_rate": 1.903888699968584e-06,
+      "loss": 0.7523,
+      "step": 2145
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6500571677078445,
+      "learning_rate": 1.8996137900652468e-06,
+      "loss": 0.5072,
+      "step": 2146
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.312829258688882,
+      "learning_rate": 1.895342559121241e-06,
+      "loss": 0.9632,
+      "step": 2147
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.659098024412787,
+      "learning_rate": 1.8910750122048638e-06,
+      "loss": 0.8763,
+      "step": 2148
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4143865026546096,
+      "learning_rate": 1.8868111543800456e-06,
+      "loss": 0.8751,
+      "step": 2149
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3987565466994516,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.7917,
+      "step": 2150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.046108071669599,
+      "learning_rate": 1.8782945262388941e-06,
+      "loss": 0.9486,
+      "step": 2151
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.478955157766606,
+      "learning_rate": 1.8740417660285087e-06,
+      "loss": 0.772,
+      "step": 2152
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0619002965811526,
+      "learning_rate": 1.8697927151215545e-06,
+      "loss": 0.8473,
+      "step": 2153
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.627059396006647,
+      "learning_rate": 1.8655473785600125e-06,
+      "loss": 0.6322,
+      "step": 2154
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.288545271294873,
+      "learning_rate": 1.8613057613814584e-06,
+      "loss": 0.8473,
+      "step": 2155
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5044739594642453,
+      "learning_rate": 1.8570678686190463e-06,
+      "loss": 0.8343,
+      "step": 2156
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2889514914706683,
+      "learning_rate": 1.8528337053015188e-06,
+      "loss": 0.6933,
+      "step": 2157
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.957491296736036,
+      "learning_rate": 1.8486032764531918e-06,
+      "loss": 0.9831,
+      "step": 2158
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9413049458029055,
+      "learning_rate": 1.844376587093944e-06,
+      "loss": 0.7114,
+      "step": 2159
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2320262194606764,
+      "learning_rate": 1.8401536422392237e-06,
+      "loss": 0.8634,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2232199588728476,
+      "learning_rate": 1.8359344469000334e-06,
+      "loss": 0.7033,
+      "step": 2161
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4253170054915016,
+      "learning_rate": 1.8317190060829242e-06,
+      "loss": 0.9524,
+      "step": 2162
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9178360050064651,
+      "learning_rate": 1.827507324789995e-06,
+      "loss": 0.4764,
+      "step": 2163
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9582955340905063,
+      "learning_rate": 1.8232994080188848e-06,
+      "loss": 0.862,
+      "step": 2164
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0247054749440632,
+      "learning_rate": 1.8190952607627594e-06,
+      "loss": 0.896,
+      "step": 2165
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.829398107714619,
+      "learning_rate": 1.8148948880103174e-06,
+      "loss": 0.9211,
+      "step": 2166
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2515671761685947,
+      "learning_rate": 1.8106982947457785e-06,
+      "loss": 0.8274,
+      "step": 2167
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9336197995135024,
+      "learning_rate": 1.8065054859488729e-06,
+      "loss": 0.8311,
+      "step": 2168
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0771144401169142,
+      "learning_rate": 1.8023164665948455e-06,
+      "loss": 0.6923,
+      "step": 2169
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0424401639041374,
+      "learning_rate": 1.7981312416544394e-06,
+      "loss": 1.0159,
+      "step": 2170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1162971568617617,
+      "learning_rate": 1.7939498160938995e-06,
+      "loss": 0.7061,
+      "step": 2171
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7242120484497414,
+      "learning_rate": 1.7897721948749625e-06,
+      "loss": 0.552,
+      "step": 2172
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.274599606166249,
+      "learning_rate": 1.7855983829548456e-06,
+      "loss": 0.7365,
+      "step": 2173
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9669919873165157,
+      "learning_rate": 1.7814283852862507e-06,
+      "loss": 0.9749,
+      "step": 2174
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8547236230054864,
+      "learning_rate": 1.7772622068173544e-06,
+      "loss": 0.8866,
+      "step": 2175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.249852126145529,
+      "learning_rate": 1.773099852491796e-06,
+      "loss": 0.7475,
+      "step": 2176
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.110892013981014,
+      "learning_rate": 1.768941327248681e-06,
+      "loss": 0.8346,
+      "step": 2177
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0356841217479411,
+      "learning_rate": 1.764786636022573e-06,
+      "loss": 0.8788,
+      "step": 2178
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7834654857736472,
+      "learning_rate": 1.76063578374348e-06,
+      "loss": 0.8933,
+      "step": 2179
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4595361569007004,
+      "learning_rate": 1.7564887753368604e-06,
+      "loss": 0.5441,
+      "step": 2180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8784933587635942,
+      "learning_rate": 1.7523456157236102e-06,
+      "loss": 0.9367,
+      "step": 2181
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1626659333680525,
+      "learning_rate": 1.7482063098200547e-06,
+      "loss": 0.7581,
+      "step": 2182
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.267393454365419,
+      "learning_rate": 1.7440708625379503e-06,
+      "loss": 0.7661,
+      "step": 2183
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2509273353150299,
+      "learning_rate": 1.739939278784476e-06,
+      "loss": 0.8568,
+      "step": 2184
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.259152409878099,
+      "learning_rate": 1.7358115634622192e-06,
+      "loss": 0.8142,
+      "step": 2185
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2870371614479916,
+      "learning_rate": 1.7316877214691863e-06,
+      "loss": 0.913,
+      "step": 2186
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.57707281755728,
+      "learning_rate": 1.727567757698778e-06,
+      "loss": 0.7742,
+      "step": 2187
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6195621293826834,
+      "learning_rate": 1.7234516770398008e-06,
+      "loss": 0.9497,
+      "step": 2188
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7984097083295578,
+      "learning_rate": 1.7193394843764516e-06,
+      "loss": 0.6133,
+      "step": 2189
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.151909090519514,
+      "learning_rate": 1.7152311845883096e-06,
+      "loss": 0.6849,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0264236758537604,
+      "learning_rate": 1.711126782550339e-06,
+      "loss": 0.9487,
+      "step": 2191
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.249109782406649,
+      "learning_rate": 1.7070262831328799e-06,
+      "loss": 0.8529,
+      "step": 2192
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.627752889265874,
+      "learning_rate": 1.7029296912016346e-06,
+      "loss": 0.8408,
+      "step": 2193
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1236823958662465,
+      "learning_rate": 1.6988370116176766e-06,
+      "loss": 0.7122,
+      "step": 2194
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4541354983258477,
+      "learning_rate": 1.6947482492374346e-06,
+      "loss": 0.7821,
+      "step": 2195
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.23322959709137,
+      "learning_rate": 1.6906634089126844e-06,
+      "loss": 0.9042,
+      "step": 2196
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9565491496470204,
+      "learning_rate": 1.686582495490554e-06,
+      "loss": 0.8069,
+      "step": 2197
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7992804156731971,
+      "learning_rate": 1.6825055138135105e-06,
+      "loss": 0.6086,
+      "step": 2198
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.149800653837,
+      "learning_rate": 1.6784324687193516e-06,
+      "loss": 0.7279,
+      "step": 2199
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5390612585654234,
+      "learning_rate": 1.6743633650412077e-06,
+      "loss": 0.9676,
+      "step": 2200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8843428557443909,
+      "learning_rate": 1.6702982076075326e-06,
+      "loss": 0.7994,
+      "step": 2201
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3905392234664977,
+      "learning_rate": 1.6662370012420931e-06,
+      "loss": 0.7618,
+      "step": 2202
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.153695863195262,
+      "learning_rate": 1.662179750763972e-06,
+      "loss": 0.8654,
+      "step": 2203
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9390096423027076,
+      "learning_rate": 1.658126460987558e-06,
+      "loss": 0.7734,
+      "step": 2204
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5064738736130712,
+      "learning_rate": 1.654077136722535e-06,
+      "loss": 0.9076,
+      "step": 2205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9452064144713062,
+      "learning_rate": 1.6500317827738887e-06,
+      "loss": 0.7313,
+      "step": 2206
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.117030871951949,
+      "learning_rate": 1.6459904039418856e-06,
+      "loss": 0.8576,
+      "step": 2207
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6991066995352259,
+      "learning_rate": 1.641953005022081e-06,
+      "loss": 0.5219,
+      "step": 2208
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5872180103387166,
+      "learning_rate": 1.637919590805308e-06,
+      "loss": 0.6755,
+      "step": 2209
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6710453725031984,
+      "learning_rate": 1.6338901660776662e-06,
+      "loss": 0.9036,
+      "step": 2210
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9639484618535001,
+      "learning_rate": 1.6298647356205255e-06,
+      "loss": 0.7764,
+      "step": 2211
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9860074189407274,
+      "learning_rate": 1.6258433042105164e-06,
+      "loss": 0.8629,
+      "step": 2212
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8307925865451704,
+      "learning_rate": 1.62182587661952e-06,
+      "loss": 0.9912,
+      "step": 2213
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4591614232114465,
+      "learning_rate": 1.6178124576146708e-06,
+      "loss": 0.6857,
+      "step": 2214
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9716876177871105,
+      "learning_rate": 1.6138030519583458e-06,
+      "loss": 0.587,
+      "step": 2215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2096299640116677,
+      "learning_rate": 1.6097976644081564e-06,
+      "loss": 0.7526,
+      "step": 2216
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7435733135721139,
+      "learning_rate": 1.605796299716949e-06,
+      "loss": 0.878,
+      "step": 2217
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4227154787383105,
+      "learning_rate": 1.601798962632799e-06,
+      "loss": 0.6665,
+      "step": 2218
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.377927674290213,
+      "learning_rate": 1.5978056578989953e-06,
+      "loss": 0.7888,
+      "step": 2219
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9732689450514724,
+      "learning_rate": 1.5938163902540483e-06,
+      "loss": 0.9759,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9766511177947368,
+      "learning_rate": 1.5898311644316772e-06,
+      "loss": 0.7961,
+      "step": 2221
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1002711707062518,
+      "learning_rate": 1.5858499851608006e-06,
+      "loss": 0.7656,
+      "step": 2222
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4356888281482287,
+      "learning_rate": 1.5818728571655394e-06,
+      "loss": 0.9824,
+      "step": 2223
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0144808748685545,
+      "learning_rate": 1.5778997851652095e-06,
+      "loss": 0.9013,
+      "step": 2224
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9992371818073146,
+      "learning_rate": 1.573930773874306e-06,
+      "loss": 0.5064,
+      "step": 2225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7836869743301098,
+      "learning_rate": 1.5699658280025143e-06,
+      "loss": 0.765,
+      "step": 2226
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6397634441897322,
+      "learning_rate": 1.5660049522546883e-06,
+      "loss": 0.9508,
+      "step": 2227
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3338330455369494,
+      "learning_rate": 1.5620481513308571e-06,
+      "loss": 0.7368,
+      "step": 2228
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0192334378777737,
+      "learning_rate": 1.5580954299262146e-06,
+      "loss": 0.8618,
+      "step": 2229
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.658571074725776,
+      "learning_rate": 1.5541467927311093e-06,
+      "loss": 0.6748,
+      "step": 2230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0491554666708955,
+      "learning_rate": 1.5502022444310482e-06,
+      "loss": 0.6759,
+      "step": 2231
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.126428018237841,
+      "learning_rate": 1.5462617897066863e-06,
+      "loss": 0.8716,
+      "step": 2232
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8993929634029171,
+      "learning_rate": 1.542325433233815e-06,
+      "loss": 0.8128,
+      "step": 2233
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6515372385432986,
+      "learning_rate": 1.5383931796833702e-06,
+      "loss": 0.5979,
+      "step": 2234
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.074175377329513,
+      "learning_rate": 1.5344650337214163e-06,
+      "loss": 0.7714,
+      "step": 2235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.825423774666699,
+      "learning_rate": 1.5305410000091408e-06,
+      "loss": 0.9602,
+      "step": 2236
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.32288142402475,
+      "learning_rate": 1.5266210832028556e-06,
+      "loss": 0.8927,
+      "step": 2237
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6741962862847649,
+      "learning_rate": 1.5227052879539872e-06,
+      "loss": 0.9027,
+      "step": 2238
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0419586822841556,
+      "learning_rate": 1.5187936189090668e-06,
+      "loss": 0.8439,
+      "step": 2239
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3651575716965387,
+      "learning_rate": 1.5148860807097348e-06,
+      "loss": 0.6911,
+      "step": 2240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.833402482390377,
+      "learning_rate": 1.5109826779927296e-06,
+      "loss": 0.8734,
+      "step": 2241
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3059645269370699,
+      "learning_rate": 1.5070834153898766e-06,
+      "loss": 0.4667,
+      "step": 2242
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8423179986459246,
+      "learning_rate": 1.5031882975280942e-06,
+      "loss": 0.8783,
+      "step": 2243
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3878066884659166,
+      "learning_rate": 1.4992973290293827e-06,
+      "loss": 0.7616,
+      "step": 2244
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.655371432053139,
+      "learning_rate": 1.4954105145108133e-06,
+      "loss": 0.7574,
+      "step": 2245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.404717933671046,
+      "learning_rate": 1.491527858584535e-06,
+      "loss": 1.0669,
+      "step": 2246
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6064424232333796,
+      "learning_rate": 1.4876493658577557e-06,
+      "loss": 0.8894,
+      "step": 2247
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9447083214804298,
+      "learning_rate": 1.4837750409327478e-06,
+      "loss": 0.8809,
+      "step": 2248
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.029048883248124,
+      "learning_rate": 1.479904888406838e-06,
+      "loss": 0.6711,
+      "step": 2249
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7488140980651861,
+      "learning_rate": 1.4760389128723968e-06,
+      "loss": 0.8823,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8154097228227949,
+      "learning_rate": 1.4721771189168444e-06,
+      "loss": 0.5482,
+      "step": 2251
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9315320905713004,
+      "learning_rate": 1.4683195111226378e-06,
+      "loss": 0.7464,
+      "step": 2252
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3506068908199804,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.7908,
+      "step": 2253
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0030401282302235,
+      "learning_rate": 1.460616872323236e-06,
+      "loss": 0.7547,
+      "step": 2254
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5616035193786955,
+      "learning_rate": 1.456771850458097e-06,
+      "loss": 0.9269,
+      "step": 2255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.255422359612142,
+      "learning_rate": 1.4529310330343964e-06,
+      "loss": 0.7681,
+      "step": 2256
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.27370009006186,
+      "learning_rate": 1.4490944246097e-06,
+      "loss": 0.8156,
+      "step": 2257
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.882723210142203,
+      "learning_rate": 1.4452620297365804e-06,
+      "loss": 0.9012,
+      "step": 2258
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7660892140227333,
+      "learning_rate": 1.4414338529626054e-06,
+      "loss": 0.8971,
+      "step": 2259
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8655989416091717,
+      "learning_rate": 1.4376098988303406e-06,
+      "loss": 0.54,
+      "step": 2260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.010010931690868,
+      "learning_rate": 1.4337901718773427e-06,
+      "loss": 0.7594,
+      "step": 2261
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.72090346794506,
+      "learning_rate": 1.4299746766361477e-06,
+      "loss": 0.9122,
+      "step": 2262
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1526351236023136,
+      "learning_rate": 1.4261634176342747e-06,
+      "loss": 0.8371,
+      "step": 2263
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0769635134211175,
+      "learning_rate": 1.4223563993942124e-06,
+      "loss": 0.7857,
+      "step": 2264
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0385662102919753,
+      "learning_rate": 1.41855362643342e-06,
+      "loss": 0.9042,
+      "step": 2265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1277801883924317,
+      "learning_rate": 1.4147551032643192e-06,
+      "loss": 0.6511,
+      "step": 2266
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1922728803817668,
+      "learning_rate": 1.4109608343942855e-06,
+      "loss": 0.9264,
+      "step": 2267
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0573595170419734,
+      "learning_rate": 1.40717082432565e-06,
+      "loss": 0.4873,
+      "step": 2268
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1978209926542616,
+      "learning_rate": 1.40338507755569e-06,
+      "loss": 0.7309,
+      "step": 2269
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9034030048013204,
+      "learning_rate": 1.3996035985766205e-06,
+      "loss": 0.9682,
+      "step": 2270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2581119804244858,
+      "learning_rate": 1.3958263918755955e-06,
+      "loss": 0.8946,
+      "step": 2271
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2581310278148699,
+      "learning_rate": 1.3920534619346998e-06,
+      "loss": 0.8831,
+      "step": 2272
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4890342768857223,
+      "learning_rate": 1.3882848132309395e-06,
+      "loss": 0.7614,
+      "step": 2273
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3939576979685557,
+      "learning_rate": 1.3845204502362442e-06,
+      "loss": 0.902,
+      "step": 2274
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6098051344897655,
+      "learning_rate": 1.3807603774174582e-06,
+      "loss": 0.7775,
+      "step": 2275
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0117500526863032,
+      "learning_rate": 1.3770045992363313e-06,
+      "loss": 0.8619,
+      "step": 2276
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.098387442950193,
+      "learning_rate": 1.3732531201495208e-06,
+      "loss": 0.814,
+      "step": 2277
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9995938630585768,
+      "learning_rate": 1.369505944608583e-06,
+      "loss": 0.8789,
+      "step": 2278
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9963180426757857,
+      "learning_rate": 1.3657630770599633e-06,
+      "loss": 0.6227,
+      "step": 2279
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1204747653959684,
+      "learning_rate": 1.3620245219449984e-06,
+      "loss": 0.7535,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1667101702402443,
+      "learning_rate": 1.35829028369991e-06,
+      "loss": 0.7382,
+      "step": 2281
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.202054859167428,
+      "learning_rate": 1.3545603667557911e-06,
+      "loss": 0.9579,
+      "step": 2282
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9676452436747118,
+      "learning_rate": 1.350834775538613e-06,
+      "loss": 0.791,
+      "step": 2283
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9180606070777863,
+      "learning_rate": 1.34711351446921e-06,
+      "loss": 0.9215,
+      "step": 2284
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1169458923812015,
+      "learning_rate": 1.3433965879632799e-06,
+      "loss": 0.8449,
+      "step": 2285
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3631814484828657,
+      "learning_rate": 1.3396840004313789e-06,
+      "loss": 0.8402,
+      "step": 2286
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.221635772729121,
+      "learning_rate": 1.3359757562789099e-06,
+      "loss": 0.4627,
+      "step": 2287
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7293332796606438,
+      "learning_rate": 1.3322718599061252e-06,
+      "loss": 0.8873,
+      "step": 2288
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9050246798811514,
+      "learning_rate": 1.3285723157081193e-06,
+      "loss": 0.9833,
+      "step": 2289
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.898727149823827,
+      "learning_rate": 1.3248771280748174e-06,
+      "loss": 0.8266,
+      "step": 2290
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0850063107339285,
+      "learning_rate": 1.3211863013909797e-06,
+      "loss": 0.8631,
+      "step": 2291
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1122708297661077,
+      "learning_rate": 1.3174998400361911e-06,
+      "loss": 0.6929,
+      "step": 2292
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.218081160144324,
+      "learning_rate": 1.3138177483848524e-06,
+      "loss": 0.8782,
+      "step": 2293
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0815946132835832,
+      "learning_rate": 1.310140030806184e-06,
+      "loss": 0.8771,
+      "step": 2294
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.258392857526076,
+      "learning_rate": 1.306466691664216e-06,
+      "loss": 0.4957,
+      "step": 2295
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9434417878337424,
+      "learning_rate": 1.302797735317778e-06,
+      "loss": 0.8812,
+      "step": 2296
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2368574883751695,
+      "learning_rate": 1.2991331661205036e-06,
+      "loss": 0.7929,
+      "step": 2297
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3544706699206146,
+      "learning_rate": 1.2954729884208212e-06,
+      "loss": 0.9538,
+      "step": 2298
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3161137387387543,
+      "learning_rate": 1.2918172065619422e-06,
+      "loss": 0.7602,
+      "step": 2299
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6600181140679835,
+      "learning_rate": 1.288165824881868e-06,
+      "loss": 0.9271,
+      "step": 2300
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8911749090613474,
+      "learning_rate": 1.284518847713378e-06,
+      "loss": 0.7883,
+      "step": 2301
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.172981463227542,
+      "learning_rate": 1.28087627938402e-06,
+      "loss": 0.7186,
+      "step": 2302
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0304799535475118,
+      "learning_rate": 1.277238124216117e-06,
+      "loss": 0.8437,
+      "step": 2303
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2723732890398378,
+      "learning_rate": 1.2736043865267495e-06,
+      "loss": 0.7121,
+      "step": 2304
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6384373918721116,
+      "learning_rate": 1.2699750706277604e-06,
+      "loss": 0.6708,
+      "step": 2305
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5101986671519778,
+      "learning_rate": 1.2663501808257444e-06,
+      "loss": 0.7344,
+      "step": 2306
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1413464803810025,
+      "learning_rate": 1.2627297214220413e-06,
+      "loss": 0.7496,
+      "step": 2307
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8159967378907954,
+      "learning_rate": 1.2591136967127383e-06,
+      "loss": 0.8161,
+      "step": 2308
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.398880932392499,
+      "learning_rate": 1.2555021109886589e-06,
+      "loss": 0.7125,
+      "step": 2309
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.886542753279974,
+      "learning_rate": 1.251894968535356e-06,
+      "loss": 0.9685,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.989355458438701,
+      "learning_rate": 1.2482922736331144e-06,
+      "loss": 0.8486,
+      "step": 2311
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6404392372815044,
+      "learning_rate": 1.24469403055694e-06,
+      "loss": 0.9091,
+      "step": 2312
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1958543781850979,
+      "learning_rate": 1.2411002435765535e-06,
+      "loss": 0.546,
+      "step": 2313
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.612665416902427,
+      "learning_rate": 1.2375109169563915e-06,
+      "loss": 0.8027,
+      "step": 2314
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2757097561492727,
+      "learning_rate": 1.2339260549555976e-06,
+      "loss": 0.8497,
+      "step": 2315
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.578673003767657,
+      "learning_rate": 1.2303456618280141e-06,
+      "loss": 0.7745,
+      "step": 2316
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.022899447259873,
+      "learning_rate": 1.226769741822184e-06,
+      "loss": 0.9657,
+      "step": 2317
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.073099317957969,
+      "learning_rate": 1.2231982991813428e-06,
+      "loss": 0.8364,
+      "step": 2318
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8223947533740528,
+      "learning_rate": 1.2196313381434094e-06,
+      "loss": 0.6459,
+      "step": 2319
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9860179164702624,
+      "learning_rate": 1.2160688629409873e-06,
+      "loss": 0.7918,
+      "step": 2320
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8440155057622306,
+      "learning_rate": 1.2125108778013588e-06,
+      "loss": 0.5722,
+      "step": 2321
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5147346490493803,
+      "learning_rate": 1.2089573869464738e-06,
+      "loss": 0.9833,
+      "step": 2322
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0037955950505677,
+      "learning_rate": 1.2054083945929534e-06,
+      "loss": 0.7457,
+      "step": 2323
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3312010365755162,
+      "learning_rate": 1.201863904952077e-06,
+      "loss": 0.7388,
+      "step": 2324
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9755277417526166,
+      "learning_rate": 1.1983239222297837e-06,
+      "loss": 0.9509,
+      "step": 2325
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9234044660721141,
+      "learning_rate": 1.1947884506266655e-06,
+      "loss": 0.6551,
+      "step": 2326
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3832244345217564,
+      "learning_rate": 1.1912574943379574e-06,
+      "loss": 0.944,
+      "step": 2327
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8524004096071586,
+      "learning_rate": 1.1877310575535394e-06,
+      "loss": 0.7971,
+      "step": 2328
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2754167444658913,
+      "learning_rate": 1.1842091444579307e-06,
+      "loss": 0.9713,
+      "step": 2329
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8415272387405162,
+      "learning_rate": 1.1806917592302763e-06,
+      "loss": 0.7228,
+      "step": 2330
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0624036278779272,
+      "learning_rate": 1.1771789060443535e-06,
+      "loss": 0.7166,
+      "step": 2331
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0540249548859664,
+      "learning_rate": 1.1736705890685623e-06,
+      "loss": 0.5413,
+      "step": 2332
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5949473097797748,
+      "learning_rate": 1.1701668124659148e-06,
+      "loss": 0.8674,
+      "step": 2333
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3395159117906856,
+      "learning_rate": 1.166667580394041e-06,
+      "loss": 0.9908,
+      "step": 2334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0548198156592377,
+      "learning_rate": 1.163172897005176e-06,
+      "loss": 0.8109,
+      "step": 2335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1384107201747855,
+      "learning_rate": 1.1596827664461558e-06,
+      "loss": 0.7649,
+      "step": 2336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7461369680146015,
+      "learning_rate": 1.1561971928584158e-06,
+      "loss": 0.9269,
+      "step": 2337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.095614103103187,
+      "learning_rate": 1.1527161803779868e-06,
+      "loss": 0.7297,
+      "step": 2338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1826511639334585,
+      "learning_rate": 1.14923973313548e-06,
+      "loss": 0.9317,
+      "step": 2339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9562268941702745,
+      "learning_rate": 1.1457678552560969e-06,
+      "loss": 0.6115,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.432375759210558,
+      "learning_rate": 1.142300550859612e-06,
+      "loss": 0.8539,
+      "step": 2341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.353177351743942,
+      "learning_rate": 1.1388378240603742e-06,
+      "loss": 0.8407,
+      "step": 2342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7325713219337957,
+      "learning_rate": 1.1353796789673037e-06,
+      "loss": 0.785,
+      "step": 2343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.375789468099245,
+      "learning_rate": 1.1319261196838782e-06,
+      "loss": 0.949,
+      "step": 2344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0094720906979169,
+      "learning_rate": 1.1284771503081377e-06,
+      "loss": 0.7592,
+      "step": 2345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2598476904222036,
+      "learning_rate": 1.1250327749326772e-06,
+      "loss": 0.938,
+      "step": 2346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3052674450924475,
+      "learning_rate": 1.121592997644635e-06,
+      "loss": 0.7049,
+      "step": 2347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.258276964322486,
+      "learning_rate": 1.1181578225256973e-06,
+      "loss": 0.929,
+      "step": 2348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7455842855608721,
+      "learning_rate": 1.1147272536520904e-06,
+      "loss": 0.5562,
+      "step": 2349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7457790312015599,
+      "learning_rate": 1.1113012950945695e-06,
+      "loss": 0.8702,
+      "step": 2350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6245341625021341,
+      "learning_rate": 1.1078799509184246e-06,
+      "loss": 0.905,
+      "step": 2351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9401515244220993,
+      "learning_rate": 1.1044632251834687e-06,
+      "loss": 0.9055,
+      "step": 2352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9769495132231634,
+      "learning_rate": 1.1010511219440307e-06,
+      "loss": 0.7669,
+      "step": 2353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1402928990680292,
+      "learning_rate": 1.0976436452489592e-06,
+      "loss": 0.8165,
+      "step": 2354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9791375139903892,
+      "learning_rate": 1.0942407991416132e-06,
+      "loss": 0.6946,
+      "step": 2355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3042525156872935,
+      "learning_rate": 1.0908425876598512e-06,
+      "loss": 1.0021,
+      "step": 2356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.454396993917433,
+      "learning_rate": 1.0874490148360383e-06,
+      "loss": 0.5547,
+      "step": 2357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9857576888114316,
+      "learning_rate": 1.0840600846970333e-06,
+      "loss": 0.603,
+      "step": 2358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.95662783267656,
+      "learning_rate": 1.0806758012641833e-06,
+      "loss": 0.7741,
+      "step": 2359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4470815689954033,
+      "learning_rate": 1.0772961685533267e-06,
+      "loss": 0.9107,
+      "step": 2360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6742428673471295,
+      "learning_rate": 1.0739211905747777e-06,
+      "loss": 0.9141,
+      "step": 2361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1727744757028575,
+      "learning_rate": 1.0705508713333313e-06,
+      "loss": 0.6866,
+      "step": 2362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0147858184700467,
+      "learning_rate": 1.0671852148282535e-06,
+      "loss": 0.9729,
+      "step": 2363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7597236228137496,
+      "learning_rate": 1.0638242250532754e-06,
+      "loss": 0.7304,
+      "step": 2364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0550704998645823,
+      "learning_rate": 1.0604679059965923e-06,
+      "loss": 0.854,
+      "step": 2365
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7722936135755233,
+      "learning_rate": 1.0571162616408586e-06,
+      "loss": 0.5621,
+      "step": 2366
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.069885117749818,
+      "learning_rate": 1.0537692959631773e-06,
+      "loss": 0.9321,
+      "step": 2367
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.464506696428825,
+      "learning_rate": 1.0504270129351035e-06,
+      "loss": 0.6972,
+      "step": 2368
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3707491941604915,
+      "learning_rate": 1.0470894165226359e-06,
+      "loss": 0.7622,
+      "step": 2369
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2947579169197967,
+      "learning_rate": 1.0437565106862075e-06,
+      "loss": 0.8383,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4705401968710528,
+      "learning_rate": 1.0404282993806897e-06,
+      "loss": 0.7437,
+      "step": 2371
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3408134122761592,
+      "learning_rate": 1.0371047865553847e-06,
+      "loss": 0.9487,
+      "step": 2372
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7966098630383479,
+      "learning_rate": 1.0337859761540125e-06,
+      "loss": 0.8935,
+      "step": 2373
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0036488207692273,
+      "learning_rate": 1.03047187211472e-06,
+      "loss": 0.9839,
+      "step": 2374
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9949978093749406,
+      "learning_rate": 1.0271624783700678e-06,
+      "loss": 0.6756,
+      "step": 2375
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0553984792477502,
+      "learning_rate": 1.0238577988470233e-06,
+      "loss": 0.6602,
+      "step": 2376
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4181608983961633,
+      "learning_rate": 1.0205578374669655e-06,
+      "loss": 0.8833,
+      "step": 2377
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1850033070984103,
+      "learning_rate": 1.0172625981456723e-06,
+      "loss": 0.8265,
+      "step": 2378
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0861380474885987,
+      "learning_rate": 1.0139720847933166e-06,
+      "loss": 0.9184,
+      "step": 2379
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5943867267333425,
+      "learning_rate": 1.0106863013144668e-06,
+      "loss": 0.6651,
+      "step": 2380
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1222680962800449,
+      "learning_rate": 1.0074052516080756e-06,
+      "loss": 0.762,
+      "step": 2381
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3357721906549833,
+      "learning_rate": 1.0041289395674802e-06,
+      "loss": 0.8829,
+      "step": 2382
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.051503139505891,
+      "learning_rate": 1.000857369080398e-06,
+      "loss": 0.7839,
+      "step": 2383
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.985893480447133,
+      "learning_rate": 9.975905440289152e-07,
+      "loss": 0.6077,
+      "step": 2384
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2409952500484631,
+      "learning_rate": 9.943284682894905e-07,
+      "loss": 0.7258,
+      "step": 2385
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6793930183350949,
+      "learning_rate": 9.91071145732948e-07,
+      "loss": 0.9676,
+      "step": 2386
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4400023196668834,
+      "learning_rate": 9.878185802244666e-07,
+      "loss": 0.8104,
+      "step": 2387
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0050833233562484,
+      "learning_rate": 9.84570775623585e-07,
+      "loss": 0.7959,
+      "step": 2388
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1393784390936574,
+      "learning_rate": 9.813277357841928e-07,
+      "loss": 0.9064,
+      "step": 2389
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.079407523900785,
+      "learning_rate": 9.780894645545215e-07,
+      "loss": 0.808,
+      "step": 2390
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5770085454880134,
+      "learning_rate": 9.748559657771472e-07,
+      "loss": 0.8763,
+      "step": 2391
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.149565389165838,
+      "learning_rate": 9.716272432889845e-07,
+      "loss": 0.7762,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.795584087829399,
+      "learning_rate": 9.684033009212752e-07,
+      "loss": 0.7917,
+      "step": 2393
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7292722679624005,
+      "learning_rate": 9.651841424995933e-07,
+      "loss": 0.5128,
+      "step": 2394
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3757507752060572,
+      "learning_rate": 9.619697718438358e-07,
+      "loss": 0.6103,
+      "step": 2395
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.768956229825491,
+      "learning_rate": 9.587601927682162e-07,
+      "loss": 0.9853,
+      "step": 2396
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.434290466603993,
+      "learning_rate": 9.55555409081263e-07,
+      "loss": 0.7562,
+      "step": 2397
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8827510177898042,
+      "learning_rate": 9.52355424585818e-07,
+      "loss": 0.9672,
+      "step": 2398
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6763148496192222,
+      "learning_rate": 9.491602430790209e-07,
+      "loss": 0.9465,
+      "step": 2399
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1900209379285325,
+      "learning_rate": 9.459698683523205e-07,
+      "loss": 0.6813,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8481846073794037,
+      "learning_rate": 9.427843041914541e-07,
+      "loss": 0.5585,
+      "step": 2401
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9988368889205863,
+      "learning_rate": 9.396035543764559e-07,
+      "loss": 0.8286,
+      "step": 2402
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.21399862041315,
+      "learning_rate": 9.364276226816476e-07,
+      "loss": 0.8714,
+      "step": 2403
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2510798408556612,
+      "learning_rate": 9.332565128756283e-07,
+      "loss": 0.65,
+      "step": 2404
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3000796728068569,
+      "learning_rate": 9.300902287212804e-07,
+      "loss": 0.9264,
+      "step": 2405
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2802506805210774,
+      "learning_rate": 9.269287739757604e-07,
+      "loss": 0.8394,
+      "step": 2406
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2092031034359807,
+      "learning_rate": 9.237721523904891e-07,
+      "loss": 0.7507,
+      "step": 2407
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1379488359220626,
+      "learning_rate": 9.206203677111574e-07,
+      "loss": 0.7785,
+      "step": 2408
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4367401693079416,
+      "learning_rate": 9.174734236777156e-07,
+      "loss": 1.016,
+      "step": 2409
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1674566619353695,
+      "learning_rate": 9.143313240243668e-07,
+      "loss": 0.9143,
+      "step": 2410
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8418664053844834,
+      "learning_rate": 9.111940724795692e-07,
+      "loss": 0.5255,
+      "step": 2411
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3247296099424044,
+      "learning_rate": 9.080616727660286e-07,
+      "loss": 0.8179,
+      "step": 2412
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6978295615907156,
+      "learning_rate": 9.049341286006891e-07,
+      "loss": 0.9199,
+      "step": 2413
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3204966731278756,
+      "learning_rate": 9.018114436947373e-07,
+      "loss": 0.6793,
+      "step": 2414
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5200204379908404,
+      "learning_rate": 8.986936217535941e-07,
+      "loss": 0.9479,
+      "step": 2415
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1921185265742145,
+      "learning_rate": 8.955806664769051e-07,
+      "loss": 0.6923,
+      "step": 2416
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3063979384644793,
+      "learning_rate": 8.924725815585461e-07,
+      "loss": 0.6276,
+      "step": 2417
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1897782910307466,
+      "learning_rate": 8.893693706866124e-07,
+      "loss": 0.9493,
+      "step": 2418
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2305850134958418,
+      "learning_rate": 8.862710375434136e-07,
+      "loss": 0.8371,
+      "step": 2419
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7683487422779512,
+      "learning_rate": 8.831775858054747e-07,
+      "loss": 0.5769,
+      "step": 2420
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7745514020402156,
+      "learning_rate": 8.80089019143524e-07,
+      "loss": 0.7453,
+      "step": 2421
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9144351134110872,
+      "learning_rate": 8.770053412224972e-07,
+      "loss": 1.0755,
+      "step": 2422
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2169055685852592,
+      "learning_rate": 8.739265557015287e-07,
+      "loss": 0.7382,
+      "step": 2423
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.767917285743808,
+      "learning_rate": 8.708526662339428e-07,
+      "loss": 0.8972,
+      "step": 2424
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9720253238071261,
+      "learning_rate": 8.677836764672592e-07,
+      "loss": 0.8254,
+      "step": 2425
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3097439707466092,
+      "learning_rate": 8.647195900431832e-07,
+      "loss": 0.7223,
+      "step": 2426
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2786700406866016,
+      "learning_rate": 8.616604105975973e-07,
+      "loss": 0.8487,
+      "step": 2427
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0544389576922255,
+      "learning_rate": 8.586061417605668e-07,
+      "loss": 0.447,
+      "step": 2428
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1427357767101853,
+      "learning_rate": 8.555567871563275e-07,
+      "loss": 0.852,
+      "step": 2429
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8575917293109778,
+      "learning_rate": 8.525123504032817e-07,
+      "loss": 0.7283,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2758605357216581,
+      "learning_rate": 8.494728351140003e-07,
+      "loss": 0.8138,
+      "step": 2431
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.963960872574283,
+      "learning_rate": 8.46438244895213e-07,
+      "loss": 1.0474,
+      "step": 2432
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.005126352083494,
+      "learning_rate": 8.434085833478023e-07,
+      "loss": 0.8864,
+      "step": 2433
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1575903617662777,
+      "learning_rate": 8.403838540668058e-07,
+      "loss": 0.8197,
+      "step": 2434
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5719402232346518,
+      "learning_rate": 8.373640606414097e-07,
+      "loss": 0.7149,
+      "step": 2435
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9008380923943919,
+      "learning_rate": 8.343492066549369e-07,
+      "loss": 0.892,
+      "step": 2436
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9972292450028523,
+      "learning_rate": 8.313392956848559e-07,
+      "loss": 0.546,
+      "step": 2437
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1076401995060166,
+      "learning_rate": 8.283343313027654e-07,
+      "loss": 0.7467,
+      "step": 2438
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3706476435558113,
+      "learning_rate": 8.25334317074396e-07,
+      "loss": 0.7947,
+      "step": 2439
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3734114541827636,
+      "learning_rate": 8.223392565596061e-07,
+      "loss": 0.7421,
+      "step": 2440
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.215855825842251,
+      "learning_rate": 8.193491533123715e-07,
+      "loss": 0.9371,
+      "step": 2441
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0644114539341294,
+      "learning_rate": 8.163640108807897e-07,
+      "loss": 0.8787,
+      "step": 2442
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5640732780133393,
+      "learning_rate": 8.133838328070714e-07,
+      "loss": 0.713,
+      "step": 2443
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3675440945979755,
+      "learning_rate": 8.104086226275332e-07,
+      "loss": 0.9599,
+      "step": 2444
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2694804986157469,
+      "learning_rate": 8.074383838725997e-07,
+      "loss": 0.892,
+      "step": 2445
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9415163541727588,
+      "learning_rate": 8.044731200667966e-07,
+      "loss": 0.5622,
+      "step": 2446
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3070241553248019,
+      "learning_rate": 8.015128347287438e-07,
+      "loss": 0.7092,
+      "step": 2447
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7871605001029577,
+      "learning_rate": 7.985575313711547e-07,
+      "loss": 0.9455,
+      "step": 2448
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4466036832090123,
+      "learning_rate": 7.956072135008336e-07,
+      "loss": 0.6525,
+      "step": 2449
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.550916531965268,
+      "learning_rate": 7.926618846186646e-07,
+      "loss": 0.7742,
+      "step": 2450
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.269310672083397,
+      "learning_rate": 7.897215482196146e-07,
+      "loss": 0.9267,
+      "step": 2451
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3742601172818556,
+      "learning_rate": 7.867862077927269e-07,
+      "loss": 0.6435,
+      "step": 2452
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3621824129390632,
+      "learning_rate": 7.838558668211132e-07,
+      "loss": 0.9134,
+      "step": 2453
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.071327303369928,
+      "learning_rate": 7.809305287819557e-07,
+      "loss": 0.4396,
+      "step": 2454
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.128642394800197,
+      "learning_rate": 7.780101971465015e-07,
+      "loss": 0.7254,
+      "step": 2455
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.919637953584474,
+      "learning_rate": 7.750948753800508e-07,
+      "loss": 0.9486,
+      "step": 2456
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3204669133054163,
+      "learning_rate": 7.721845669419664e-07,
+      "loss": 0.8612,
+      "step": 2457
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.630790445281691,
+      "learning_rate": 7.692792752856564e-07,
+      "loss": 0.9241,
+      "step": 2458
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3052077057560727,
+      "learning_rate": 7.663790038585794e-07,
+      "loss": 0.7472,
+      "step": 2459
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1795484483123213,
+      "learning_rate": 7.634837561022373e-07,
+      "loss": 0.8836,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7153860262220557,
+      "learning_rate": 7.605935354521665e-07,
+      "loss": 0.7528,
+      "step": 2461
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9871185824596266,
+      "learning_rate": 7.577083453379425e-07,
+      "loss": 0.9065,
+      "step": 2462
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5202806712168977,
+      "learning_rate": 7.548281891831715e-07,
+      "loss": 0.6943,
+      "step": 2463
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8959949333831585,
+      "learning_rate": 7.519530704054828e-07,
+      "loss": 0.5424,
+      "step": 2464
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9821287337123545,
+      "learning_rate": 7.490829924165311e-07,
+      "loss": 0.9143,
+      "step": 2465
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.205535732679849,
+      "learning_rate": 7.462179586219897e-07,
+      "loss": 0.7744,
+      "step": 2466
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1063631303276116,
+      "learning_rate": 7.433579724215445e-07,
+      "loss": 0.8497,
+      "step": 2467
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0353637966114024,
+      "learning_rate": 7.405030372088934e-07,
+      "loss": 0.918,
+      "step": 2468
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8640977519679447,
+      "learning_rate": 7.376531563717415e-07,
+      "loss": 0.7716,
+      "step": 2469
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3847151098353092,
+      "learning_rate": 7.348083332917927e-07,
+      "loss": 0.9306,
+      "step": 2470
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3804166260223663,
+      "learning_rate": 7.319685713447533e-07,
+      "loss": 0.8427,
+      "step": 2471
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1779325365335096,
+      "learning_rate": 7.291338739003234e-07,
+      "loss": 0.8251,
+      "step": 2472
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6101088331164868,
+      "learning_rate": 7.263042443221907e-07,
+      "loss": 0.7363,
+      "step": 2473
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.193729346105327,
+      "learning_rate": 7.23479685968031e-07,
+      "loss": 0.7908,
+      "step": 2474
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4879716708812818,
+      "learning_rate": 7.206602021895054e-07,
+      "loss": 0.8903,
+      "step": 2475
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8304033946144136,
+      "learning_rate": 7.178457963322471e-07,
+      "loss": 0.8541,
+      "step": 2476
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.016175165370311,
+      "learning_rate": 7.150364717358699e-07,
+      "loss": 0.9032,
+      "step": 2477
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2709855295380657,
+      "learning_rate": 7.122322317339542e-07,
+      "loss": 0.713,
+      "step": 2478
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6377570920373563,
+      "learning_rate": 7.094330796540483e-07,
+      "loss": 0.9233,
+      "step": 2479
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9878590653708235,
+      "learning_rate": 7.066390188176647e-07,
+      "loss": 0.9208,
+      "step": 2480
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5936887943654459,
+      "learning_rate": 7.038500525402709e-07,
+      "loss": 0.5164,
+      "step": 2481
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3585483905071323,
+      "learning_rate": 7.010661841312921e-07,
+      "loss": 0.8631,
+      "step": 2482
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8315633231430772,
+      "learning_rate": 6.98287416894105e-07,
+      "loss": 0.8767,
+      "step": 2483
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3762529587784504,
+      "learning_rate": 6.955137541260287e-07,
+      "loss": 0.8644,
+      "step": 2484
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6931405861421784,
+      "learning_rate": 6.927451991183292e-07,
+      "loss": 0.7468,
+      "step": 2485
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7933292843277707,
+      "learning_rate": 6.899817551562127e-07,
+      "loss": 0.9223,
+      "step": 2486
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.072175864671769,
+      "learning_rate": 6.872234255188154e-07,
+      "loss": 0.8144,
+      "step": 2487
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3923344255659424,
+      "learning_rate": 6.844702134792086e-07,
+      "loss": 0.7266,
+      "step": 2488
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5167894869616714,
+      "learning_rate": 6.817221223043918e-07,
+      "loss": 0.808,
+      "step": 2489
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9338271301984804,
+      "learning_rate": 6.789791552552838e-07,
+      "loss": 0.4791,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3420420462409237,
+      "learning_rate": 6.762413155867276e-07,
+      "loss": 0.8875,
+      "step": 2491
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3247182064522995,
+      "learning_rate": 6.735086065474799e-07,
+      "loss": 0.7643,
+      "step": 2492
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0603305600189925,
+      "learning_rate": 6.707810313802077e-07,
+      "loss": 0.7917,
+      "step": 2493
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0158506463368167,
+      "learning_rate": 6.680585933214895e-07,
+      "loss": 0.752,
+      "step": 2494
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1519827749274936,
+      "learning_rate": 6.653412956018068e-07,
+      "loss": 0.78,
+      "step": 2495
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3398923188770422,
+      "learning_rate": 6.626291414455388e-07,
+      "loss": 0.9415,
+      "step": 2496
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1171663523879842,
+      "learning_rate": 6.599221340709661e-07,
+      "loss": 0.944,
+      "step": 2497
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5730085960393438,
+      "learning_rate": 6.572202766902569e-07,
+      "loss": 0.9169,
+      "step": 2498
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3294578621594544,
+      "learning_rate": 6.545235725094723e-07,
+      "loss": 0.4056,
+      "step": 2499
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3147620527500556,
+      "learning_rate": 6.518320247285581e-07,
+      "loss": 0.8077,
+      "step": 2500
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.132701507545941,
+      "learning_rate": 6.49145636541339e-07,
+      "loss": 0.8238,
+      "step": 2501
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6034325890977623,
+      "learning_rate": 6.46464411135519e-07,
+      "loss": 0.7712,
+      "step": 2502
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8644652542166643,
+      "learning_rate": 6.437883516926774e-07,
+      "loss": 0.968,
+      "step": 2503
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3190291444729505,
+      "learning_rate": 6.411174613882593e-07,
+      "loss": 0.805,
+      "step": 2504
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3239353327004713,
+      "learning_rate": 6.384517433915794e-07,
+      "loss": 0.6499,
+      "step": 2505
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1664771264334837,
+      "learning_rate": 6.357912008658151e-07,
+      "loss": 0.7498,
+      "step": 2506
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9233719629976624,
+      "learning_rate": 6.331358369680002e-07,
+      "loss": 0.5923,
+      "step": 2507
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3989930717401893,
+      "learning_rate": 6.30485654849024e-07,
+      "loss": 0.9766,
+      "step": 2508
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.231688073176058,
+      "learning_rate": 6.278406576536306e-07,
+      "loss": 0.7898,
+      "step": 2509
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8319742501822631,
+      "learning_rate": 6.252008485204053e-07,
+      "loss": 0.8972,
+      "step": 2510
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7716866771297338,
+      "learning_rate": 6.22566230581782e-07,
+      "loss": 0.8978,
+      "step": 2511
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4539909533637325,
+      "learning_rate": 6.199368069640343e-07,
+      "loss": 0.6551,
+      "step": 2512
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.256650013849379,
+      "learning_rate": 6.173125807872687e-07,
+      "loss": 0.992,
+      "step": 2513
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1330986527155416,
+      "learning_rate": 6.146935551654298e-07,
+      "loss": 0.7387,
+      "step": 2514
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.07907090735829,
+      "learning_rate": 6.12079733206285e-07,
+      "loss": 0.9791,
+      "step": 2515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.350599132472098,
+      "learning_rate": 6.09471118011431e-07,
+      "loss": 0.7621,
+      "step": 2516
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1196406436719715,
+      "learning_rate": 6.068677126762867e-07,
+      "loss": 0.853,
+      "step": 2517
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3340358916459234,
+      "learning_rate": 6.042695202900855e-07,
+      "loss": 0.4322,
+      "step": 2518
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2631654332238806,
+      "learning_rate": 6.016765439358774e-07,
+      "loss": 0.8816,
+      "step": 2519
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9575851657041046,
+      "learning_rate": 5.990887866905237e-07,
+      "loss": 0.9843,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.33427376805103,
+      "learning_rate": 5.965062516246895e-07,
+      "loss": 0.7841,
+      "step": 2521
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.494355311259869,
+      "learning_rate": 5.939289418028455e-07,
+      "loss": 0.8985,
+      "step": 2522
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8071967893441404,
+      "learning_rate": 5.913568602832637e-07,
+      "loss": 0.8716,
+      "step": 2523
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.014836434643359,
+      "learning_rate": 5.887900101180067e-07,
+      "loss": 0.7717,
+      "step": 2524
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3789195911060184,
+      "learning_rate": 5.862283943529351e-07,
+      "loss": 0.9051,
+      "step": 2525
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3227108104248329,
+      "learning_rate": 5.836720160276971e-07,
+      "loss": 0.5816,
+      "step": 2526
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4750493069450232,
+      "learning_rate": 5.811208781757221e-07,
+      "loss": 0.9124,
+      "step": 2527
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8536069269041333,
+      "learning_rate": 5.785749838242261e-07,
+      "loss": 0.8605,
+      "step": 2528
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4161161953443637,
+      "learning_rate": 5.760343359942011e-07,
+      "loss": 0.8873,
+      "step": 2529
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4990020447262482,
+      "learning_rate": 5.734989377004119e-07,
+      "loss": 0.8173,
+      "step": 2530
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2186513764915625,
+      "learning_rate": 5.709687919513974e-07,
+      "loss": 0.7653,
+      "step": 2531
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4530072862016132,
+      "learning_rate": 5.684439017494614e-07,
+      "loss": 0.823,
+      "step": 2532
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1678610978614183,
+      "learning_rate": 5.659242700906719e-07,
+      "loss": 0.6481,
+      "step": 2533
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6070921849925568,
+      "learning_rate": 5.63409899964858e-07,
+      "loss": 0.9878,
+      "step": 2534
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6817220941026081,
+      "learning_rate": 5.609007943556028e-07,
+      "loss": 0.7327,
+      "step": 2535
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2070055383720713,
+      "learning_rate": 5.583969562402447e-07,
+      "loss": 0.7993,
+      "step": 2536
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.464390214043122,
+      "learning_rate": 5.558983885898739e-07,
+      "loss": 0.7584,
+      "step": 2537
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5268418031112772,
+      "learning_rate": 5.534050943693197e-07,
+      "loss": 0.8934,
+      "step": 2538
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2920636202174511,
+      "learning_rate": 5.509170765371608e-07,
+      "loss": 0.7586,
+      "step": 2539
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3066151533465031,
+      "learning_rate": 5.484343380457124e-07,
+      "loss": 0.8136,
+      "step": 2540
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8825834723517554,
+      "learning_rate": 5.459568818410221e-07,
+      "loss": 0.8921,
+      "step": 2541
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.424820146310035,
+      "learning_rate": 5.434847108628749e-07,
+      "loss": 0.827,
+      "step": 2542
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1598920315406398,
+      "learning_rate": 5.410178280447814e-07,
+      "loss": 0.4044,
+      "step": 2543
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2492794057464023,
+      "learning_rate": 5.385562363139757e-07,
+      "loss": 0.8842,
+      "step": 2544
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2541698106721997,
+      "learning_rate": 5.360999385914156e-07,
+      "loss": 0.7851,
+      "step": 2545
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.730326306886756,
+      "learning_rate": 5.336489377917786e-07,
+      "loss": 0.8641,
+      "step": 2546
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7566236656982546,
+      "learning_rate": 5.312032368234527e-07,
+      "loss": 0.9082,
+      "step": 2547
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.707747154147361,
+      "learning_rate": 5.287628385885385e-07,
+      "loss": 0.696,
+      "step": 2548
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9643633910069401,
+      "learning_rate": 5.263277459828475e-07,
+      "loss": 1.007,
+      "step": 2549
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8878771634112432,
+      "learning_rate": 5.2389796189589e-07,
+      "loss": 0.776,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0491009982257882,
+      "learning_rate": 5.214734892108813e-07,
+      "loss": 0.8311,
+      "step": 2551
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.221925754378292,
+      "learning_rate": 5.190543308047336e-07,
+      "loss": 0.535,
+      "step": 2552
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2945996244696338,
+      "learning_rate": 5.166404895480509e-07,
+      "loss": 0.9175,
+      "step": 2553
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5809084055469154,
+      "learning_rate": 5.1423196830513e-07,
+      "loss": 0.7298,
+      "step": 2554
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2829183702187854,
+      "learning_rate": 5.11828769933953e-07,
+      "loss": 0.71,
+      "step": 2555
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3669668383718663,
+      "learning_rate": 5.094308972861878e-07,
+      "loss": 0.9086,
+      "step": 2556
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5804824869219238,
+      "learning_rate": 5.070383532071826e-07,
+      "loss": 0.752,
+      "step": 2557
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.352299737397528,
+      "learning_rate": 5.0465114053596e-07,
+      "loss": 0.9004,
+      "step": 2558
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6939503325010052,
+      "learning_rate": 5.022692621052189e-07,
+      "loss": 0.8624,
+      "step": 2559
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3433444228417872,
+      "learning_rate": 4.998927207413295e-07,
+      "loss": 0.9749,
+      "step": 2560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8911173793592974,
+      "learning_rate": 4.975215192643246e-07,
+      "loss": 0.6645,
+      "step": 2561
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8313208861688883,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.66,
+      "step": 2562
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2744031786336527,
+      "learning_rate": 4.927951472194298e-07,
+      "loss": 0.851,
+      "step": 2563
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5620718861556002,
+      "learning_rate": 4.904399822599143e-07,
+      "loss": 0.8601,
+      "step": 2564
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2681520058309828,
+      "learning_rate": 4.88090168404029e-07,
+      "loss": 0.7409,
+      "step": 2565
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5192034888788317,
+      "learning_rate": 4.857457084400957e-07,
+      "loss": 0.7763,
+      "step": 2566
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4527106230482187,
+      "learning_rate": 4.834066051500785e-07,
+      "loss": 0.7894,
+      "step": 2567
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.309316048385762,
+      "learning_rate": 4.81072861309591e-07,
+      "loss": 0.8275,
+      "step": 2568
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.258101393803827,
+      "learning_rate": 4.787444796878843e-07,
+      "loss": 0.692,
+      "step": 2569
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.02150093236804,
+      "learning_rate": 4.7642146304784475e-07,
+      "loss": 0.6101,
+      "step": 2570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9874817553784718,
+      "learning_rate": 4.7410381414599635e-07,
+      "loss": 0.7692,
+      "step": 2571
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9725539748676338,
+      "learning_rate": 4.717915357324926e-07,
+      "loss": 0.969,
+      "step": 2572
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3059456165394752,
+      "learning_rate": 4.6948463055111193e-07,
+      "loss": 0.6611,
+      "step": 2573
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.111928279023787,
+      "learning_rate": 4.6718310133926084e-07,
+      "loss": 0.7993,
+      "step": 2574
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3458259026280677,
+      "learning_rate": 4.648869508279613e-07,
+      "loss": 0.9022,
+      "step": 2575
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3840511848579722,
+      "learning_rate": 4.625961817418578e-07,
+      "loss": 0.7794,
+      "step": 2576
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4864110945031077,
+      "learning_rate": 4.603107967992082e-07,
+      "loss": 0.8821,
+      "step": 2577
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7079723209390272,
+      "learning_rate": 4.5803079871187816e-07,
+      "loss": 0.7753,
+      "step": 2578
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3276449676046578,
+      "learning_rate": 4.557561901853452e-07,
+      "loss": 0.92,
+      "step": 2579
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0470813362444034,
+      "learning_rate": 4.5348697391869034e-07,
+      "loss": 0.3982,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3758581061209418,
+      "learning_rate": 4.5122315260459347e-07,
+      "loss": 0.6637,
+      "step": 2581
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5700523182115533,
+      "learning_rate": 4.4896472892933693e-07,
+      "loss": 0.8942,
+      "step": 2582
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0981440676895904,
+      "learning_rate": 4.4671170557279597e-07,
+      "loss": 0.7967,
+      "step": 2583
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9423782464995006,
+      "learning_rate": 4.4446408520843696e-07,
+      "loss": 0.9528,
+      "step": 2584
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7052630794207033,
+      "learning_rate": 4.4222187050331657e-07,
+      "loss": 0.9453,
+      "step": 2585
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4825024489531355,
+      "learning_rate": 4.39985064118077e-07,
+      "loss": 0.8064,
+      "step": 2586
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9464497457984802,
+      "learning_rate": 4.3775366870694114e-07,
+      "loss": 0.4403,
+      "step": 2587
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.540306395850577,
+      "learning_rate": 4.355276869177127e-07,
+      "loss": 0.8197,
+      "step": 2588
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8364558067549375,
+      "learning_rate": 4.333071213917722e-07,
+      "loss": 0.7955,
+      "step": 2589
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0876391473465792,
+      "learning_rate": 4.310919747640707e-07,
+      "loss": 0.6868,
+      "step": 2590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5754038178115388,
+      "learning_rate": 4.288822496631323e-07,
+      "loss": 1.0056,
+      "step": 2591
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7764085310633109,
+      "learning_rate": 4.266779487110434e-07,
+      "loss": 0.7429,
+      "step": 2592
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2633171674960981,
+      "learning_rate": 4.244790745234584e-07,
+      "loss": 0.7722,
+      "step": 2593
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.151552050545388,
+      "learning_rate": 4.222856297095912e-07,
+      "loss": 0.8117,
+      "step": 2594
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8574415419736506,
+      "learning_rate": 4.200976168722104e-07,
+      "loss": 0.9719,
+      "step": 2595
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9447364295695428,
+      "learning_rate": 4.179150386076425e-07,
+      "loss": 0.5651,
+      "step": 2596
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5086483354580267,
+      "learning_rate": 4.157378975057647e-07,
+      "loss": 0.7522,
+      "step": 2597
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4692259235827902,
+      "learning_rate": 4.135661961499987e-07,
+      "loss": 0.9872,
+      "step": 2598
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0469630491870547,
+      "learning_rate": 4.113999371173155e-07,
+      "loss": 0.8815,
+      "step": 2599
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2141514043743913,
+      "learning_rate": 4.092391229782278e-07,
+      "loss": 0.6856,
+      "step": 2600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.588220205824226,
+      "learning_rate": 4.07083756296785e-07,
+      "loss": 0.9539,
+      "step": 2601
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5319301421528126,
+      "learning_rate": 4.0493383963057354e-07,
+      "loss": 0.6563,
+      "step": 2602
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.269545594617609,
+      "learning_rate": 4.027893755307144e-07,
+      "loss": 0.8036,
+      "step": 2603
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5183025731120727,
+      "learning_rate": 4.0065036654185543e-07,
+      "loss": 0.8163,
+      "step": 2604
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2989881523632163,
+      "learning_rate": 3.9851681520217346e-07,
+      "loss": 0.8449,
+      "step": 2605
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8052545180303773,
+      "learning_rate": 3.9638872404337057e-07,
+      "loss": 0.5588,
+      "step": 2606
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.264747635633115,
+      "learning_rate": 3.942660955906652e-07,
+      "loss": 0.8793,
+      "step": 2607
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.836495769130227,
+      "learning_rate": 3.921489323627975e-07,
+      "loss": 1.0434,
+      "step": 2608
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.316490612196402,
+      "learning_rate": 3.900372368720223e-07,
+      "loss": 0.7232,
+      "step": 2609
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8774964034626762,
+      "learning_rate": 3.8793101162410417e-07,
+      "loss": 0.8986,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.315938069844575,
+      "learning_rate": 3.8583025911831885e-07,
+      "loss": 0.8047,
+      "step": 2611
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9206945451106001,
+      "learning_rate": 3.8373498184744564e-07,
+      "loss": 0.6942,
+      "step": 2612
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.428338702142425,
+      "learning_rate": 3.816451822977696e-07,
+      "loss": 0.7985,
+      "step": 2613
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7158094454678,
+      "learning_rate": 3.795608629490738e-07,
+      "loss": 0.4738,
+      "step": 2614
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9669407880254337,
+      "learning_rate": 3.774820262746398e-07,
+      "loss": 0.8941,
+      "step": 2615
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9404514700846816,
+      "learning_rate": 3.754086747412422e-07,
+      "loss": 0.7422,
+      "step": 2616
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3491579028610894,
+      "learning_rate": 3.733408108091485e-07,
+      "loss": 0.9446,
+      "step": 2617
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2016595161851682,
+      "learning_rate": 3.712784369321121e-07,
+      "loss": 0.9242,
+      "step": 2618
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0418209050573344,
+      "learning_rate": 3.6922155555737415e-07,
+      "loss": 0.8797,
+      "step": 2619
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9548362943548335,
+      "learning_rate": 3.6717016912565575e-07,
+      "loss": 0.8788,
+      "step": 2620
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7988856000096167,
+      "learning_rate": 3.6512428007116076e-07,
+      "loss": 0.7274,
+      "step": 2621
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.113828925707038,
+      "learning_rate": 3.6308389082156835e-07,
+      "loss": 0.9286,
+      "step": 2622
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0412444144054664,
+      "learning_rate": 3.6104900379802966e-07,
+      "loss": 0.4107,
+      "step": 2623
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2911839525173183,
+      "learning_rate": 3.5901962141516975e-07,
+      "loss": 0.7393,
+      "step": 2624
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0518374180781103,
+      "learning_rate": 3.569957460810819e-07,
+      "loss": 0.8024,
+      "step": 2625
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.32965130865643,
+      "learning_rate": 3.549773801973211e-07,
+      "loss": 0.6666,
+      "step": 2626
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.212231599328838,
+      "learning_rate": 3.5296452615890763e-07,
+      "loss": 0.9432,
+      "step": 2627
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3608260329875685,
+      "learning_rate": 3.5095718635432206e-07,
+      "loss": 0.8546,
+      "step": 2628
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5581576962879526,
+      "learning_rate": 3.4895536316549847e-07,
+      "loss": 0.8538,
+      "step": 2629
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.6614150120838,
+      "learning_rate": 3.469590589678284e-07,
+      "loss": 0.7863,
+      "step": 2630
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4926557661612472,
+      "learning_rate": 3.4496827613015206e-07,
+      "loss": 0.8907,
+      "step": 2631
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8039499945880891,
+      "learning_rate": 3.4298301701475855e-07,
+      "loss": 0.4808,
+      "step": 2632
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1905722669358323,
+      "learning_rate": 3.41003283977383e-07,
+      "loss": 0.7487,
+      "step": 2633
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6986371991107881,
+      "learning_rate": 3.3902907936720353e-07,
+      "loss": 0.9495,
+      "step": 2634
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4869474406248189,
+      "learning_rate": 3.370604055268356e-07,
+      "loss": 0.7212,
+      "step": 2635
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8779564155646393,
+      "learning_rate": 3.3509726479233506e-07,
+      "loss": 0.8381,
+      "step": 2636
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2264711868082663,
+      "learning_rate": 3.3313965949319124e-07,
+      "loss": 0.8715,
+      "step": 2637
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1683765059818958,
+      "learning_rate": 3.3118759195232273e-07,
+      "loss": 0.7021,
+      "step": 2638
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5641823064696316,
+      "learning_rate": 3.2924106448607997e-07,
+      "loss": 0.892,
+      "step": 2639
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.532593479623006,
+      "learning_rate": 3.27300079404238e-07,
+      "loss": 0.4987,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5364680052249882,
+      "learning_rate": 3.253646390099946e-07,
+      "loss": 0.9574,
+      "step": 2641
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7494225104896928,
+      "learning_rate": 3.234347455999709e-07,
+      "loss": 0.8853,
+      "step": 2642
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7670760889127708,
+      "learning_rate": 3.2151040146420174e-07,
+      "loss": 0.8248,
+      "step": 2643
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3612872727667016,
+      "learning_rate": 3.195916088861406e-07,
+      "loss": 0.8657,
+      "step": 2644
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.61232949463098,
+      "learning_rate": 3.176783701426528e-07,
+      "loss": 0.7624,
+      "step": 2645
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7512751376725006,
+      "learning_rate": 3.157706875040112e-07,
+      "loss": 0.8892,
+      "step": 2646
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1279304547016331,
+      "learning_rate": 3.138685632338978e-07,
+      "loss": 0.8946,
+      "step": 2647
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0156001862065982,
+      "learning_rate": 3.119719995893994e-07,
+      "loss": 0.8626,
+      "step": 2648
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1047225743161706,
+      "learning_rate": 3.1008099882100183e-07,
+      "loss": 0.7032,
+      "step": 2649
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8055209712012161,
+      "learning_rate": 3.0819556317259304e-07,
+      "loss": 0.5099,
+      "step": 2650
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8978654218495539,
+      "learning_rate": 3.0631569488145605e-07,
+      "loss": 0.8879,
+      "step": 2651
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.56163387406707,
+      "learning_rate": 3.0444139617826605e-07,
+      "loss": 0.7315,
+      "step": 2652
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2639097250146507,
+      "learning_rate": 3.0257266928709174e-07,
+      "loss": 0.9328,
+      "step": 2653
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.096583484777574,
+      "learning_rate": 3.0070951642538925e-07,
+      "loss": 0.8307,
+      "step": 2654
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.962268040329778,
+      "learning_rate": 2.988519398040002e-07,
+      "loss": 0.8051,
+      "step": 2655
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.318280239576707,
+      "learning_rate": 2.969999416271496e-07,
+      "loss": 0.9379,
+      "step": 2656
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7944453955910211,
+      "learning_rate": 2.95153524092443e-07,
+      "loss": 0.8111,
+      "step": 2657
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4912483684888345,
+      "learning_rate": 2.9331268939086334e-07,
+      "loss": 0.8889,
+      "step": 2658
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0334675316335602,
+      "learning_rate": 2.91477439706771e-07,
+      "loss": 0.7182,
+      "step": 2659
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5064936448328934,
+      "learning_rate": 2.896477772178957e-07,
+      "loss": 0.9439,
+      "step": 2660
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3714749272209092,
+      "learning_rate": 2.878237040953402e-07,
+      "loss": 0.7028,
+      "step": 2661
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1420997964952955,
+      "learning_rate": 2.860052225035742e-07,
+      "loss": 0.8369,
+      "step": 2662
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.98634064920856,
+      "learning_rate": 2.8419233460043106e-07,
+      "loss": 0.8062,
+      "step": 2663
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.757723335554425,
+      "learning_rate": 2.823850425371083e-07,
+      "loss": 0.8319,
+      "step": 2664
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0577128772334607,
+      "learning_rate": 2.8058334845816214e-07,
+      "loss": 0.8647,
+      "step": 2665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.348419199656031,
+      "learning_rate": 2.787872545015069e-07,
+      "loss": 0.8924,
+      "step": 2666
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0927291339598486,
+      "learning_rate": 2.7699676279841104e-07,
+      "loss": 0.586,
+      "step": 2667
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4509141583941134,
+      "learning_rate": 2.752118754734967e-07,
+      "loss": 0.7888,
+      "step": 2668
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6815397535294638,
+      "learning_rate": 2.734325946447336e-07,
+      "loss": 0.8181,
+      "step": 2669
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.244936753957303,
+      "learning_rate": 2.716589224234406e-07,
+      "loss": 0.8886,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.406583885789427,
+      "learning_rate": 2.698908609142814e-07,
+      "loss": 0.7041,
+      "step": 2671
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8948786614356602,
+      "learning_rate": 2.681284122152594e-07,
+      "loss": 0.9313,
+      "step": 2672
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3252904412145894,
+      "learning_rate": 2.663715784177201e-07,
+      "loss": 0.7806,
+      "step": 2673
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8280351031431339,
+      "learning_rate": 2.646203616063475e-07,
+      "loss": 0.6111,
+      "step": 2674
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1907461322283228,
+      "learning_rate": 2.628747638591561e-07,
+      "loss": 0.9705,
+      "step": 2675
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8008811537584114,
+      "learning_rate": 2.6113478724749785e-07,
+      "loss": 0.3866,
+      "step": 2676
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3463737178099071,
+      "learning_rate": 2.594004338360495e-07,
+      "loss": 0.934,
+      "step": 2677
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1515492778143783,
+      "learning_rate": 2.576717056828193e-07,
+      "loss": 0.7157,
+      "step": 2678
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6211610454378487,
+      "learning_rate": 2.559486048391391e-07,
+      "loss": 0.7337,
+      "step": 2679
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.237943686050515,
+      "learning_rate": 2.542311333496622e-07,
+      "loss": 0.7678,
+      "step": 2680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.128844800171527,
+      "learning_rate": 2.525192932523635e-07,
+      "loss": 0.7852,
+      "step": 2681
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1653773964017915,
+      "learning_rate": 2.5081308657853576e-07,
+      "loss": 0.8894,
+      "step": 2682
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7320890249144278,
+      "learning_rate": 2.4911251535278515e-07,
+      "loss": 0.9252,
+      "step": 2683
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4483951444661212,
+      "learning_rate": 2.474175815930324e-07,
+      "loss": 0.8586,
+      "step": 2684
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0816246090252573,
+      "learning_rate": 2.457282873105099e-07,
+      "loss": 0.4372,
+      "step": 2685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9076215417844995,
+      "learning_rate": 2.4404463450975415e-07,
+      "loss": 0.7595,
+      "step": 2686
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.109046375961518,
+      "learning_rate": 2.423666251886114e-07,
+      "loss": 0.791,
+      "step": 2687
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8937378798553781,
+      "learning_rate": 2.406942613382296e-07,
+      "loss": 0.9573,
+      "step": 2688
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3478164851725647,
+      "learning_rate": 2.3902754494305734e-07,
+      "loss": 0.9077,
+      "step": 2689
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8611408984208746,
+      "learning_rate": 2.3736647798084268e-07,
+      "loss": 0.8201,
+      "step": 2690
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7508442940077549,
+      "learning_rate": 2.3571106242262997e-07,
+      "loss": 0.814,
+      "step": 2691
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2873843580683795,
+      "learning_rate": 2.340613002327563e-07,
+      "loss": 0.6386,
+      "step": 2692
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4877778479832204,
+      "learning_rate": 2.3241719336885182e-07,
+      "loss": 0.5405,
+      "step": 2693
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.768355274669733,
+      "learning_rate": 2.307787437818365e-07,
+      "loss": 0.96,
+      "step": 2694
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1451211377964616,
+      "learning_rate": 2.2914595341591517e-07,
+      "loss": 0.8071,
+      "step": 2695
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4088988975050831,
+      "learning_rate": 2.275188242085785e-07,
+      "loss": 0.8704,
+      "step": 2696
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8536693088323359,
+      "learning_rate": 2.2589735809059986e-07,
+      "loss": 0.86,
+      "step": 2697
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.029995695039867,
+      "learning_rate": 2.2428155698603182e-07,
+      "loss": 0.6683,
+      "step": 2698
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0637082649245424,
+      "learning_rate": 2.2267142281220678e-07,
+      "loss": 1.0052,
+      "step": 2699
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.500196284620567,
+      "learning_rate": 2.2106695747973027e-07,
+      "loss": 0.7536,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7181836595335211,
+      "learning_rate": 2.1946816289248163e-07,
+      "loss": 0.969,
+      "step": 2701
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1300756768563276,
+      "learning_rate": 2.1787504094761268e-07,
+      "loss": 0.732,
+      "step": 2702
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9440183717407935,
+      "learning_rate": 2.162875935355424e-07,
+      "loss": 0.5243,
+      "step": 2703
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3445487212113068,
+      "learning_rate": 2.147058225399573e-07,
+      "loss": 0.6981,
+      "step": 2704
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0373516663363105,
+      "learning_rate": 2.1312972983780823e-07,
+      "loss": 0.8654,
+      "step": 2705
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5322524314847275,
+      "learning_rate": 2.115593172993058e-07,
+      "loss": 0.9121,
+      "step": 2706
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9699543892485092,
+      "learning_rate": 2.0999458678792384e-07,
+      "loss": 0.8491,
+      "step": 2707
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3891819548225708,
+      "learning_rate": 2.0843554016039326e-07,
+      "loss": 0.9043,
+      "step": 2708
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8185285451078111,
+      "learning_rate": 2.0688217926669695e-07,
+      "loss": 0.8791,
+      "step": 2709
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3692391542828637,
+      "learning_rate": 2.05334505950075e-07,
+      "loss": 0.8617,
+      "step": 2710
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1268558914274482,
+      "learning_rate": 2.0379252204701715e-07,
+      "loss": 0.7171,
+      "step": 2711
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7812773908286235,
+      "learning_rate": 2.0225622938726097e-07,
+      "loss": 0.6252,
+      "step": 2712
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.224911191183341,
+      "learning_rate": 2.007256297937926e-07,
+      "loss": 0.8727,
+      "step": 2713
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1247883465443187,
+      "learning_rate": 1.9920072508284204e-07,
+      "loss": 0.7788,
+      "step": 2714
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0348190940535864,
+      "learning_rate": 1.9768151706388016e-07,
+      "loss": 0.9084,
+      "step": 2715
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1055918897070227,
+      "learning_rate": 1.961680075396205e-07,
+      "loss": 0.7573,
+      "step": 2716
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.133863957282143,
+      "learning_rate": 1.9466019830601257e-07,
+      "loss": 0.7605,
+      "step": 2717
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4499029231231835,
+      "learning_rate": 1.9315809115224348e-07,
+      "loss": 0.8533,
+      "step": 2718
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0703910942557104,
+      "learning_rate": 1.9166168786073303e-07,
+      "loss": 0.7337,
+      "step": 2719
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6000480850439953,
+      "learning_rate": 1.9017099020713358e-07,
+      "loss": 0.6344,
+      "step": 2720
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6821691857990759,
+      "learning_rate": 1.8868599996032577e-07,
+      "loss": 0.8582,
+      "step": 2721
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0853464378564273,
+      "learning_rate": 1.8720671888242058e-07,
+      "loss": 0.9131,
+      "step": 2722
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.005313108250883,
+      "learning_rate": 1.857331487287506e-07,
+      "loss": 0.7679,
+      "step": 2723
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0551623693748535,
+      "learning_rate": 1.8426529124787372e-07,
+      "loss": 0.9143,
+      "step": 2724
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.164534107841947,
+      "learning_rate": 1.8280314818157118e-07,
+      "loss": 0.7628,
+      "step": 2725
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1555443535835623,
+      "learning_rate": 1.81346721264839e-07,
+      "loss": 0.7927,
+      "step": 2726
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.678958259665459,
+      "learning_rate": 1.7989601222589358e-07,
+      "loss": 0.8605,
+      "step": 2727
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8194971928180865,
+      "learning_rate": 1.7845102278616578e-07,
+      "loss": 0.9618,
+      "step": 2728
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1070980135977966,
+      "learning_rate": 1.7701175466029895e-07,
+      "loss": 0.5115,
+      "step": 2729
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.082398121797478,
+      "learning_rate": 1.75578209556147e-07,
+      "loss": 0.7602,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3233433072764758,
+      "learning_rate": 1.7415038917477478e-07,
+      "loss": 0.7592,
+      "step": 2731
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7603468281879102,
+      "learning_rate": 1.7272829521045142e-07,
+      "loss": 0.9079,
+      "step": 2732
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3310756442803844,
+      "learning_rate": 1.7131192935065433e-07,
+      "loss": 0.9207,
+      "step": 2733
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.432863365543643,
+      "learning_rate": 1.699012932760602e-07,
+      "loss": 0.8256,
+      "step": 2734
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2019899740262794,
+      "learning_rate": 1.6849638866054896e-07,
+      "loss": 0.8681,
+      "step": 2735
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0362999890463431,
+      "learning_rate": 1.6709721717120042e-07,
+      "loss": 0.73,
+      "step": 2736
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4008844037260764,
+      "learning_rate": 1.657037804682876e-07,
+      "loss": 0.8586,
+      "step": 2737
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8131916555986421,
+      "learning_rate": 1.6431608020528233e-07,
+      "loss": 0.5088,
+      "step": 2738
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1737569598227815,
+      "learning_rate": 1.629341180288485e-07,
+      "loss": 0.9133,
+      "step": 2739
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3592387918315283,
+      "learning_rate": 1.6155789557883994e-07,
+      "loss": 0.7353,
+      "step": 2740
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3910768653046546,
+      "learning_rate": 1.6018741448830033e-07,
+      "loss": 0.8418,
+      "step": 2741
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1381087630479967,
+      "learning_rate": 1.5882267638346217e-07,
+      "loss": 0.8105,
+      "step": 2742
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3239099809529415,
+      "learning_rate": 1.574636828837395e-07,
+      "loss": 0.7096,
+      "step": 2743
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4997744399914963,
+      "learning_rate": 1.5611043560173345e-07,
+      "loss": 0.9191,
+      "step": 2744
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8359550538385863,
+      "learning_rate": 1.5476293614322512e-07,
+      "loss": 0.878,
+      "step": 2745
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.359484406423948,
+      "learning_rate": 1.5342118610717438e-07,
+      "loss": 1.0256,
+      "step": 2746
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6486483587835274,
+      "learning_rate": 1.520851870857204e-07,
+      "loss": 0.6357,
+      "step": 2747
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5645601518660328,
+      "learning_rate": 1.507549406641773e-07,
+      "loss": 0.7689,
+      "step": 2748
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9902596907420013,
+      "learning_rate": 1.4943044842103238e-07,
+      "loss": 0.853,
+      "step": 2749
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.400496082419666,
+      "learning_rate": 1.4811171192794628e-07,
+      "loss": 0.7175,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4144483947692645,
+      "learning_rate": 1.4679873274975e-07,
+      "loss": 0.7187,
+      "step": 2751
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.34712934318542,
+      "learning_rate": 1.4549151244444003e-07,
+      "loss": 0.8035,
+      "step": 2752
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0405940886163436,
+      "learning_rate": 1.4419005256318286e-07,
+      "loss": 0.88,
+      "step": 2753
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1716276751353585,
+      "learning_rate": 1.42894354650307e-07,
+      "loss": 0.7532,
+      "step": 2754
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7373243711532824,
+      "learning_rate": 1.4160442024330533e-07,
+      "loss": 0.7798,
+      "step": 2755
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8435155294559333,
+      "learning_rate": 1.403202508728313e-07,
+      "loss": 0.5962,
+      "step": 2756
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1825868007187057,
+      "learning_rate": 1.3904184806269705e-07,
+      "loss": 0.7366,
+      "step": 2757
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5994068452164278,
+      "learning_rate": 1.3776921332987193e-07,
+      "loss": 0.9596,
+      "step": 2758
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.01854784544051,
+      "learning_rate": 1.3650234818448182e-07,
+      "loss": 0.683,
+      "step": 2759
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4836020926047175,
+      "learning_rate": 1.3524125412980482e-07,
+      "loss": 0.9585,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.291686827417036,
+      "learning_rate": 1.3398593266227223e-07,
+      "loss": 0.7816,
+      "step": 2761
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7847349422196483,
+      "learning_rate": 1.3273638527146638e-07,
+      "loss": 0.7896,
+      "step": 2762
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.916900923840311,
+      "learning_rate": 1.3149261344011456e-07,
+      "loss": 0.8997,
+      "step": 2763
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1383519338225176,
+      "learning_rate": 1.3025461864409395e-07,
+      "loss": 0.6844,
+      "step": 2764
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8048809844510697,
+      "learning_rate": 1.2902240235242614e-07,
+      "loss": 0.6021,
+      "step": 2765
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0140358528007924,
+      "learning_rate": 1.277959660272743e-07,
+      "loss": 0.6568,
+      "step": 2766
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5501260390590794,
+      "learning_rate": 1.265753111239443e-07,
+      "loss": 0.6727,
+      "step": 2767
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3055283965153037,
+      "learning_rate": 1.253604390908819e-07,
+      "loss": 0.8678,
+      "step": 2768
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1577868973660936,
+      "learning_rate": 1.2415135136967015e-07,
+      "loss": 0.7871,
+      "step": 2769
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9965066907383104,
+      "learning_rate": 1.2294804939502746e-07,
+      "loss": 0.9919,
+      "step": 2770
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7453024542162656,
+      "learning_rate": 1.2175053459481e-07,
+      "loss": 0.9242,
+      "step": 2771
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.792593540406242,
+      "learning_rate": 1.2055880839000277e-07,
+      "loss": 0.8683,
+      "step": 2772
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8780490052553268,
+      "learning_rate": 1.1937287219472515e-07,
+      "loss": 0.4764,
+      "step": 2773
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.572439546137704,
+      "learning_rate": 1.1819272741622367e-07,
+      "loss": 0.77,
+      "step": 2774
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6028433284299646,
+      "learning_rate": 1.1701837545487426e-07,
+      "loss": 0.7625,
+      "step": 2775
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4051504218425173,
+      "learning_rate": 1.1584981770417891e-07,
+      "loss": 0.7717,
+      "step": 2776
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5514585939316918,
+      "learning_rate": 1.1468705555076287e-07,
+      "loss": 0.9766,
+      "step": 2777
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5954583559938271,
+      "learning_rate": 1.1353009037437523e-07,
+      "loss": 0.7514,
+      "step": 2778
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5557266932818827,
+      "learning_rate": 1.1237892354788726e-07,
+      "loss": 0.7565,
+      "step": 2779
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0523161757593718,
+      "learning_rate": 1.1123355643728795e-07,
+      "loss": 0.8161,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.877720929631673,
+      "learning_rate": 1.1009399040168455e-07,
+      "loss": 0.98,
+      "step": 2781
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8200839716028724,
+      "learning_rate": 1.0896022679330265e-07,
+      "loss": 0.5908,
+      "step": 2782
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.492926507357318,
+      "learning_rate": 1.078322669574794e-07,
+      "loss": 0.756,
+      "step": 2783
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.672312280011121,
+      "learning_rate": 1.0671011223266803e-07,
+      "loss": 0.9149,
+      "step": 2784
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1191277377157092,
+      "learning_rate": 1.0559376395043285e-07,
+      "loss": 0.8994,
+      "step": 2785
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1915379652867963,
+      "learning_rate": 1.0448322343544537e-07,
+      "loss": 0.6998,
+      "step": 2786
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7932227448501694,
+      "learning_rate": 1.0337849200548978e-07,
+      "loss": 0.9484,
+      "step": 2787
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5141511480163188,
+      "learning_rate": 1.0227957097145413e-07,
+      "loss": 0.608,
+      "step": 2788
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.152302688493958,
+      "learning_rate": 1.0118646163733203e-07,
+      "loss": 0.8456,
+      "step": 2789
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7823115205185618,
+      "learning_rate": 1.0009916530022256e-07,
+      "loss": 0.7747,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.038014928455321,
+      "learning_rate": 9.901768325032535e-08,
+      "loss": 0.7265,
+      "step": 2791
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8818206087893566,
+      "learning_rate": 9.794201677094162e-08,
+      "loss": 0.585,
+      "step": 2792
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2547416577637318,
+      "learning_rate": 9.687216713847203e-08,
+      "loss": 0.8381,
+      "step": 2793
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.390347746136916,
+      "learning_rate": 9.580813562241276e-08,
+      "loss": 1.0134,
+      "step": 2794
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7077790688127303,
+      "learning_rate": 9.474992348535884e-08,
+      "loss": 0.7519,
+      "step": 2795
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8618027450605499,
+      "learning_rate": 9.369753198299859e-08,
+      "loss": 0.9083,
+      "step": 2796
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5474406994215173,
+      "learning_rate": 9.265096236411364e-08,
+      "loss": 0.812,
+      "step": 2797
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1086692199016546,
+      "learning_rate": 9.161021587057728e-08,
+      "loss": 0.6945,
+      "step": 2798
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2845364699103805,
+      "learning_rate": 9.05752937373533e-08,
+      "loss": 0.8298,
+      "step": 2799
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7274748732133787,
+      "learning_rate": 8.95461971924938e-08,
+      "loss": 0.503,
+      "step": 2800
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.579341626939155,
+      "learning_rate": 8.852292745713864e-08,
+      "loss": 0.9023,
+      "step": 2801
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8792153572738006,
+      "learning_rate": 8.75054857455132e-08,
+      "loss": 0.7504,
+      "step": 2802
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.75396479889267,
+      "learning_rate": 8.649387326492676e-08,
+      "loss": 0.983,
+      "step": 2803
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6286949413983647,
+      "learning_rate": 8.54880912157724e-08,
+      "loss": 0.9089,
+      "step": 2804
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.016504151710561,
+      "learning_rate": 8.448814079152489e-08,
+      "loss": 0.9024,
+      "step": 2805
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1466154643421527,
+      "learning_rate": 8.34940231787379e-08,
+      "loss": 0.891,
+      "step": 2806
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4318311378856534,
+      "learning_rate": 8.2505739557045e-08,
+      "loss": 0.6958,
+      "step": 2807
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.405923090227432,
+      "learning_rate": 8.152329109915646e-08,
+      "loss": 0.9092,
+      "step": 2808
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0913619414597642,
+      "learning_rate": 8.054667897085867e-08,
+      "loss": 0.5298,
+      "step": 2809
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.06654524764049,
+      "learning_rate": 7.957590433101293e-08,
+      "loss": 0.6632,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.683822742638605,
+      "learning_rate": 7.861096833155225e-08,
+      "loss": 0.7473,
+      "step": 2811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 27.85795034761273,
+      "learning_rate": 7.765187211748348e-08,
+      "loss": 0.7204,
+      "step": 2812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3082184099523753,
+      "learning_rate": 7.66986168268824e-08,
+      "loss": 0.9326,
+      "step": 2813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.285408237136161,
+      "learning_rate": 7.575120359089416e-08,
+      "loss": 0.8532,
+      "step": 2814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7276784592250434,
+      "learning_rate": 7.480963353373227e-08,
+      "loss": 0.8137,
+      "step": 2815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2323718055539288,
+      "learning_rate": 7.387390777267689e-08,
+      "loss": 0.8,
+      "step": 2816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4885273664219254,
+      "learning_rate": 7.294402741807094e-08,
+      "loss": 0.8604,
+      "step": 2817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8279071472818741,
+      "learning_rate": 7.201999357332346e-08,
+      "loss": 0.5293,
+      "step": 2818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6032945109568746,
+      "learning_rate": 7.110180733490624e-08,
+      "loss": 0.7402,
+      "step": 2819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7492380973876093,
+      "learning_rate": 7.018946979234997e-08,
+      "loss": 0.9389,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3966298405463409,
+      "learning_rate": 6.928298202824646e-08,
+      "loss": 0.6351,
+      "step": 2821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3345980627681735,
+      "learning_rate": 6.838234511824748e-08,
+      "loss": 0.917,
+      "step": 2822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6295679723547318,
+      "learning_rate": 6.748756013105928e-08,
+      "loss": 0.7686,
+      "step": 2823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1391050300703223,
+      "learning_rate": 6.659862812844586e-08,
+      "loss": 0.6165,
+      "step": 2824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5156717776869415,
+      "learning_rate": 6.571555016522624e-08,
+      "loss": 0.8976,
+      "step": 2825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.082625849926957,
+      "learning_rate": 6.483832728927219e-08,
+      "loss": 0.5173,
+      "step": 2826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.492391202028404,
+      "learning_rate": 6.396696054150719e-08,
+      "loss": 0.9167,
+      "step": 2827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4373134743707778,
+      "learning_rate": 6.310145095590803e-08,
+      "loss": 0.8713,
+      "step": 2828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6796085693070086,
+      "learning_rate": 6.224179955949871e-08,
+      "loss": 0.8767,
+      "step": 2829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0452896037492665,
+      "learning_rate": 6.138800737235384e-08,
+      "loss": 0.9183,
+      "step": 2830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.070237448512835,
+      "learning_rate": 6.054007540759299e-08,
+      "loss": 0.7427,
+      "step": 2831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5334890357637427,
+      "learning_rate": 5.96980046713852e-08,
+      "loss": 0.9314,
+      "step": 2832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7766748056186943,
+      "learning_rate": 5.8861796162942294e-08,
+      "loss": 0.8632,
+      "step": 2833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2322316939159468,
+      "learning_rate": 5.803145087451945e-08,
+      "loss": 0.8742,
+      "step": 2834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0169452088112538,
+      "learning_rate": 5.7206969791416265e-08,
+      "loss": 0.6839,
+      "step": 2835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.273736606266712,
+      "learning_rate": 5.638835389197295e-08,
+      "loss": 0.4673,
+      "step": 2836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5591947561359523,
+      "learning_rate": 5.557560414756913e-08,
+      "loss": 0.9086,
+      "step": 2837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3436444420084304,
+      "learning_rate": 5.476872152262558e-08,
+      "loss": 0.7817,
+      "step": 2838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0192235068101376,
+      "learning_rate": 5.396770697459919e-08,
+      "loss": 0.8877,
+      "step": 2839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4278205395842931,
+      "learning_rate": 5.317256145398408e-08,
+      "loss": 0.8764,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1030238393235807,
+      "learning_rate": 5.238328590431163e-08,
+      "loss": 0.8224,
+      "step": 2841
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3022857617886654,
+      "learning_rate": 5.159988126214543e-08,
+      "loss": 0.9315,
+      "step": 2842
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0357533494944793,
+      "learning_rate": 5.082234845708467e-08,
+      "loss": 0.7813,
+      "step": 2843
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.109572628711127,
+      "learning_rate": 5.005068841175964e-08,
+      "loss": 0.8474,
+      "step": 2844
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7099904152391837,
+      "learning_rate": 4.928490204183234e-08,
+      "loss": 0.7453,
+      "step": 2845
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2704347657423485,
+      "learning_rate": 4.852499025599533e-08,
+      "loss": 0.8835,
+      "step": 2846
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.949666744239215,
+      "learning_rate": 4.7770953955969537e-08,
+      "loss": 0.6689,
+      "step": 2847
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6891023201065782,
+      "learning_rate": 4.702279403650534e-08,
+      "loss": 0.8529,
+      "step": 2848
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9072247040319394,
+      "learning_rate": 4.628051138537759e-08,
+      "loss": 0.7541,
+      "step": 2849
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4752473212325041,
+      "learning_rate": 4.5544106883390614e-08,
+      "loss": 0.8493,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1882811812702956,
+      "learning_rate": 4.481358140437042e-08,
+      "loss": 0.8976,
+      "step": 2851
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2822194977728942,
+      "learning_rate": 4.4088935815168045e-08,
+      "loss": 0.8947,
+      "step": 2852
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0386455894656708,
+      "learning_rate": 4.337017097565899e-08,
+      "loss": 0.5956,
+      "step": 2853
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6133590573331489,
+      "learning_rate": 4.265728773873767e-08,
+      "loss": 0.7975,
+      "step": 2854
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.798738191382791,
+      "learning_rate": 4.195028695032133e-08,
+      "loss": 0.7698,
+      "step": 2855
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0311895499002737,
+      "learning_rate": 4.124916944934665e-08,
+      "loss": 0.8482,
+      "step": 2856
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7798772971991057,
+      "learning_rate": 4.05539360677687e-08,
+      "loss": 0.6986,
+      "step": 2857
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7563047826170088,
+      "learning_rate": 3.986458763056089e-08,
+      "loss": 0.9435,
+      "step": 2858
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7522744583234662,
+      "learning_rate": 3.91811249557128e-08,
+      "loss": 0.7607,
+      "step": 2859
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.133195514167576,
+      "learning_rate": 3.850354885423069e-08,
+      "loss": 0.7754,
+      "step": 2860
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3682552109118689,
+      "learning_rate": 3.783186013013473e-08,
+      "loss": 0.831,
+      "step": 2861
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8665705918432813,
+      "learning_rate": 3.716605958046071e-08,
+      "loss": 0.3939,
+      "step": 2862
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3623168549188127,
+      "learning_rate": 3.6506147995254985e-08,
+      "loss": 0.9267,
+      "step": 2863
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.274291052384202,
+      "learning_rate": 3.585212615757838e-08,
+      "loss": 0.7229,
+      "step": 2864
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2012770872292033,
+      "learning_rate": 3.5203994843501786e-08,
+      "loss": 0.7463,
+      "step": 2865
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9014645598010784,
+      "learning_rate": 3.456175482210611e-08,
+      "loss": 0.753,
+      "step": 2866
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5222153758917845,
+      "learning_rate": 3.392540685548229e-08,
+      "loss": 0.8704,
+      "step": 2867
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.324973446678453,
+      "learning_rate": 3.329495169872854e-08,
+      "loss": 0.9097,
+      "step": 2868
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1807499635893288,
+      "learning_rate": 3.2670390099951985e-08,
+      "loss": 0.936,
+      "step": 2869
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7997723572284332,
+      "learning_rate": 3.205172280026536e-08,
+      "loss": 0.8134,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5944884025579937,
+      "learning_rate": 3.143895053378698e-08,
+      "loss": 0.433,
+      "step": 2871
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.095347008143277,
+      "learning_rate": 3.0832074027641324e-08,
+      "loss": 0.7995,
+      "step": 2872
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0678854888983875,
+      "learning_rate": 3.023109400195623e-08,
+      "loss": 0.8272,
+      "step": 2873
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0799860356581101,
+      "learning_rate": 2.9636011169861812e-08,
+      "loss": 0.8478,
+      "step": 2874
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2217083283697843,
+      "learning_rate": 2.9046826237491554e-08,
+      "loss": 0.9538,
+      "step": 2875
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2891896510018948,
+      "learning_rate": 2.846353990398065e-08,
+      "loss": 0.6506,
+      "step": 2876
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0517548748439856,
+      "learning_rate": 2.7886152861463233e-08,
+      "loss": 0.8227,
+      "step": 2877
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2062685686207395,
+      "learning_rate": 2.7314665795075135e-08,
+      "loss": 0.6312,
+      "step": 2878
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4127536597105737,
+      "learning_rate": 2.6749079382950572e-08,
+      "loss": 0.5643,
+      "step": 2879
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.005444399484631,
+      "learning_rate": 2.6189394296221583e-08,
+      "loss": 0.9918,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1977301856256157,
+      "learning_rate": 2.5635611199018028e-08,
+      "loss": 0.7519,
+      "step": 2881
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3012197035973743,
+      "learning_rate": 2.508773074846649e-08,
+      "loss": 0.8648,
+      "step": 2882
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7303618861183823,
+      "learning_rate": 2.4545753594688582e-08,
+      "loss": 0.8718,
+      "step": 2883
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.065032589993988,
+      "learning_rate": 2.4009680380802093e-08,
+      "loss": 0.7984,
+      "step": 2884
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0413106854330325,
+      "learning_rate": 2.347951174291818e-08,
+      "loss": 0.9915,
+      "step": 2885
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4807755200838408,
+      "learning_rate": 2.295524831014251e-08,
+      "loss": 0.7589,
+      "step": 2886
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5431177848399211,
+      "learning_rate": 2.2436890704572444e-08,
+      "loss": 1.0315,
+      "step": 2887
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0700724112399451,
+      "learning_rate": 2.192443954129819e-08,
+      "loss": 0.7125,
+      "step": 2888
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8136741525945359,
+      "learning_rate": 2.1417895428401114e-08,
+      "loss": 0.4935,
+      "step": 2889
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5636297894332185,
+      "learning_rate": 2.0917258966953735e-08,
+      "loss": 0.8381,
+      "step": 2890
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1989555632725524,
+      "learning_rate": 2.0422530751016968e-08,
+      "loss": 0.8872,
+      "step": 2891
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2335344838745907,
+      "learning_rate": 1.9933711367641772e-08,
+      "loss": 0.9419,
+      "step": 2892
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.230602926271679,
+      "learning_rate": 1.9450801396869167e-08,
+      "loss": 0.803,
+      "step": 2893
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5477305280215763,
+      "learning_rate": 1.8973801411724668e-08,
+      "loss": 0.9493,
+      "step": 2894
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8008213077354897,
+      "learning_rate": 1.8502711978223843e-08,
+      "loss": 0.8498,
+      "step": 2895
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3545971628784876,
+      "learning_rate": 1.8037533655367866e-08,
+      "loss": 0.8867,
+      "step": 2896
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0517615949543306,
+      "learning_rate": 1.757826699514298e-08,
+      "loss": 0.7488,
+      "step": 2897
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0311265549693145,
+      "learning_rate": 1.7124912542520468e-08,
+      "loss": 0.6,
+      "step": 2898
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7500536823757533,
+      "learning_rate": 1.6677470835458343e-08,
+      "loss": 0.963,
+      "step": 2899
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5259791890684848,
+      "learning_rate": 1.623594240489579e-08,
+      "loss": 0.794,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1018966860245385,
+      "learning_rate": 1.580032777475593e-08,
+      "loss": 0.8865,
+      "step": 2901
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.146786847307999,
+      "learning_rate": 1.537062746194584e-08,
+      "loss": 0.7938,
+      "step": 2902
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7870659166667973,
+      "learning_rate": 1.4946841976352654e-08,
+      "loss": 0.6718,
+      "step": 2903
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.331851712607583,
+      "learning_rate": 1.4528971820846894e-08,
+      "loss": 0.9135,
+      "step": 2904
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.23018629511565,
+      "learning_rate": 1.411701749127803e-08,
+      "loss": 0.7249,
+      "step": 2905
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7801736574790018,
+      "learning_rate": 1.3710979476476705e-08,
+      "loss": 0.6199,
+      "step": 2906
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.827381685492298,
+      "learning_rate": 1.3310858258253068e-08,
+      "loss": 0.875,
+      "step": 2907
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.893561026034762,
+      "learning_rate": 1.2916654311396214e-08,
+      "loss": 0.9098,
+      "step": 2908
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8709030249230973,
+      "learning_rate": 1.2528368103674194e-08,
+      "loss": 0.7535,
+      "step": 2909
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6431291058180384,
+      "learning_rate": 1.2146000095831777e-08,
+      "loss": 0.8989,
+      "step": 2910
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0566883476345823,
+      "learning_rate": 1.176955074159214e-08,
+      "loss": 0.7943,
+      "step": 2911
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.109934382450758,
+      "learning_rate": 1.1399020487655176e-08,
+      "loss": 0.7859,
+      "step": 2912
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.733394302424562,
+      "learning_rate": 1.1034409773697519e-08,
+      "loss": 0.9735,
+      "step": 2913
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2956573645769227,
+      "learning_rate": 1.0675719032370303e-08,
+      "loss": 0.75,
+      "step": 2914
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6918120859148756,
+      "learning_rate": 1.0322948689300838e-08,
+      "loss": 0.4927,
+      "step": 2915
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2883447718342205,
+      "learning_rate": 9.976099163091502e-09,
+      "loss": 0.7585,
+      "step": 2916
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5600880374809334,
+      "learning_rate": 9.635170865318067e-09,
+      "loss": 0.8194,
+      "step": 2917
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1889416102528128,
+      "learning_rate": 9.300164200530815e-09,
+      "loss": 0.8717,
+      "step": 2918
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7682347412498541,
+      "learning_rate": 8.971079566252871e-09,
+      "loss": 0.9329,
+      "step": 2919
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.345576386963569,
+      "learning_rate": 8.647917352980762e-09,
+      "loss": 0.775,
+      "step": 2920
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.066589291927104,
+      "learning_rate": 8.330677944183297e-09,
+      "loss": 0.8994,
+      "step": 2921
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.683050792527598,
+      "learning_rate": 8.019361716299912e-09,
+      "loss": 0.8441,
+      "step": 2922
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0362625832344174,
+      "learning_rate": 7.713969038743441e-09,
+      "loss": 0.8499,
+      "step": 2923
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9569985150101002,
+      "learning_rate": 7.4145002738967856e-09,
+      "loss": 0.477,
+      "step": 2924
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4028261734870782,
+      "learning_rate": 7.120955777112915e-09,
+      "loss": 0.9098,
+      "step": 2925
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8647500053818427,
+      "learning_rate": 6.833335896716531e-09,
+      "loss": 0.7264,
+      "step": 2926
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.321284144827646,
+      "learning_rate": 6.551640974000739e-09,
+      "loss": 0.8208,
+      "step": 2927
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3705063716347379,
+      "learning_rate": 6.275871343228157e-09,
+      "loss": 0.7627,
+      "step": 2928
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7297025012851857,
+      "learning_rate": 6.00602733163147e-09,
+      "loss": 0.7381,
+      "step": 2929
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2609896829856952,
+      "learning_rate": 5.7421092594101004e-09,
+      "loss": 0.9043,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.610539373400674,
+      "learning_rate": 5.484117439734094e-09,
+      "loss": 0.9149,
+      "step": 2931
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0672127847733477,
+      "learning_rate": 5.232052178738567e-09,
+      "loss": 0.8878,
+      "step": 2932
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7168059806707988,
+      "learning_rate": 4.985913775528151e-09,
+      "loss": 0.6174,
+      "step": 2933
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.446604754777154,
+      "learning_rate": 4.745702522174211e-09,
+      "loss": 0.8191,
+      "step": 2934
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.165136682404569,
+      "learning_rate": 4.511418703713743e-09,
+      "loss": 0.8733,
+      "step": 2935
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5725329575693863,
+      "learning_rate": 4.283062598152143e-09,
+      "loss": 0.6797,
+      "step": 2936
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6882600516260495,
+      "learning_rate": 4.060634476458769e-09,
+      "loss": 0.7946,
+      "step": 2937
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4242659467270142,
+      "learning_rate": 3.844134602570826e-09,
+      "loss": 0.8425,
+      "step": 2938
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6034925279341545,
+      "learning_rate": 3.633563233388926e-09,
+      "loss": 0.8837,
+      "step": 2939
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2070505481980105,
+      "learning_rate": 3.428920618780418e-09,
+      "loss": 0.7761,
+      "step": 2940
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4832983700162992,
+      "learning_rate": 3.230207001576613e-09,
+      "loss": 0.7375,
+      "step": 2941
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.996797601674895,
+      "learning_rate": 3.037422617573893e-09,
+      "loss": 0.5617,
+      "step": 2942
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2635392501325184,
+      "learning_rate": 2.8505676955326023e-09,
+      "loss": 0.7151,
+      "step": 2943
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6930781222192645,
+      "learning_rate": 2.6696424571776015e-09,
+      "loss": 0.9819,
+      "step": 2944
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2146570606139524,
+      "learning_rate": 2.4946471171966025e-09,
+      "loss": 0.6531,
+      "step": 2945
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1967116918425778,
+      "learning_rate": 2.3255818832423894e-09,
+      "loss": 0.9456,
+      "step": 2946
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1064468117369322,
+      "learning_rate": 2.162446955928932e-09,
+      "loss": 0.7448,
+      "step": 2947
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3392700854942934,
+      "learning_rate": 2.0052425288347166e-09,
+      "loss": 0.8122,
+      "step": 2948
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5276281552099131,
+      "learning_rate": 1.8539687885010815e-09,
+      "loss": 0.8553,
+      "step": 2949
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2089048876969786,
+      "learning_rate": 1.7086259144305507e-09,
+      "loss": 0.6763,
+      "step": 2950
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8368476029048906,
+      "learning_rate": 1.5692140790896093e-09,
+      "loss": 0.6445,
+      "step": 2951
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6081498999207793,
+      "learning_rate": 1.4357334479059292e-09,
+      "loss": 0.6142,
+      "step": 2952
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5716635298682105,
+      "learning_rate": 1.3081841792694783e-09,
+      "loss": 0.7138,
+      "step": 2953
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1074473932686064,
+      "learning_rate": 1.1865664245314101e-09,
+      "loss": 0.8846,
+      "step": 2954
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.198985351180575,
+      "learning_rate": 1.0708803280057301e-09,
+      "loss": 0.8783,
+      "step": 2955
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0815136427674346,
+      "learning_rate": 9.61126026966519e-10,
+      "loss": 0.9966,
+      "step": 2956
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.752917184657465,
+      "learning_rate": 8.573036516501542e-10,
+      "loss": 0.8987,
+      "step": 2957
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.22868197824113,
+      "learning_rate": 7.594133252530888e-10,
+      "loss": 0.8518,
+      "step": 2958
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0818977706460005,
+      "learning_rate": 6.674551639340721e-10,
+      "loss": 0.4442,
+      "step": 2959
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.017264109910308,
+      "learning_rate": 5.814292768108187e-10,
+      "loss": 0.7774,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6101903519683818,
+      "learning_rate": 5.013357659638951e-10,
+      "loss": 0.7748,
+      "step": 2961
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1837883936490887,
+      "learning_rate": 4.2717472643227785e-10,
+      "loss": 0.7101,
+      "step": 2962
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.37741727601566,
+      "learning_rate": 3.5894624621723997e-10,
+      "loss": 1.0004,
+      "step": 2963
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.845416875256494,
+      "learning_rate": 2.9665040627901985e-10,
+      "loss": 0.7681,
+      "step": 2964
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6916461976655706,
+      "learning_rate": 2.4028728053959725e-10,
+      "loss": 0.7667,
+      "step": 2965
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9592156905055158,
+      "learning_rate": 1.8985693587880715e-10,
+      "loss": 0.7785,
+      "step": 2966
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8003383457772132,
+      "learning_rate": 1.453594321393359e-10,
+      "loss": 0.9904,
+      "step": 2967
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0036610559203596,
+      "learning_rate": 1.067948221217252e-10,
+      "loss": 0.6311,
+      "step": 2968
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6793011824840998,
+      "learning_rate": 7.416315158714771e-11,
+      "loss": 0.686,
+      "step": 2969
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8830312738874864,
+      "learning_rate": 4.746445925740695e-11,
+      "loss": 0.9331,
+      "step": 2970
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9593521461127539,
+      "learning_rate": 2.669877681271693e-11,
+      "loss": 0.8856,
+      "step": 2971
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6446079350898846,
+      "learning_rate": 1.1866128895032803e-11,
+      "loss": 0.6982,
+      "step": 2972
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.529047158703334,
+      "learning_rate": 2.9665331036099475e-12,
+      "loss": 0.9038,
+      "step": 2973
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.376137049809044,
+      "learning_rate": 0.0,
+      "loss": 0.5241,
+      "step": 2974
+    },
+    {
+      "epoch": 1.0,
+      "step": 2974,
+      "total_flos": 0.0,
+      "train_loss": 0.015144473065765576,
+      "train_runtime": 3680.2552,
+      "train_samples_per_second": 1657.147,
+      "train_steps_per_second": 0.808
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2974,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50,
+  "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": 2,
+  "trial_name": null,
+  "trial_params": null
+}