diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5419 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "global_step": 899,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9977753058954393e-05,
+      "loss": 0.3597,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.995550611790879e-05,
+      "loss": 0.3028,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9933259176863182e-05,
+      "loss": 0.2636,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9911012235817577e-05,
+      "loss": 0.2089,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.988876529477197e-05,
+      "loss": 0.2438,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9866518353726366e-05,
+      "loss": 0.19,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9844271412680757e-05,
+      "loss": 0.1884,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.982202447163515e-05,
+      "loss": 0.2393,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9799777530589546e-05,
+      "loss": 0.1862,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9777530589543937e-05,
+      "loss": 0.126,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9755283648498332e-05,
+      "loss": 0.1401,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9733036707452726e-05,
+      "loss": 0.1846,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.971078976640712e-05,
+      "loss": 0.1542,
+      "step": 13
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9688542825361516e-05,
+      "loss": 0.1631,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.966629588431591e-05,
+      "loss": 0.1525,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.96440489432703e-05,
+      "loss": 0.1558,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9621802002224696e-05,
+      "loss": 0.201,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.959955506117909e-05,
+      "loss": 0.2093,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9577308120133482e-05,
+      "loss": 0.1792,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9555061179087876e-05,
+      "loss": 0.1571,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.953281423804227e-05,
+      "loss": 0.1716,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9510567296996666e-05,
+      "loss": 0.1548,
+      "step": 22
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9488320355951057e-05,
+      "loss": 0.0918,
+      "step": 23
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9466073414905455e-05,
+      "loss": 0.1617,
+      "step": 24
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9443826473859846e-05,
+      "loss": 0.1641,
+      "step": 25
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9421579532814237e-05,
+      "loss": 0.1177,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9399332591768635e-05,
+      "loss": 0.1971,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9377085650723026e-05,
+      "loss": 0.1493,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.174,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9332591768631815e-05,
+      "loss": 0.1866,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.931034482758621e-05,
+      "loss": 0.1819,
+      "step": 31
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.92880978865406e-05,
+      "loss": 0.1585,
+      "step": 32
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9265850945494996e-05,
+      "loss": 0.1615,
+      "step": 33
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.924360400444939e-05,
+      "loss": 0.1918,
+      "step": 34
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.922135706340378e-05,
+      "loss": 0.233,
+      "step": 35
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9199110122358176e-05,
+      "loss": 0.1953,
+      "step": 36
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.917686318131257e-05,
+      "loss": 0.1619,
+      "step": 37
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9154616240266965e-05,
+      "loss": 0.1466,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.913236929922136e-05,
+      "loss": 0.2283,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9110122358175754e-05,
+      "loss": 0.141,
+      "step": 40
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9087875417130146e-05,
+      "loss": 0.1253,
+      "step": 41
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.906562847608454e-05,
+      "loss": 0.1218,
+      "step": 42
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9043381535038935e-05,
+      "loss": 0.228,
+      "step": 43
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9021134593993326e-05,
+      "loss": 0.1762,
+      "step": 44
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.899888765294772e-05,
+      "loss": 0.146,
+      "step": 45
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.8976640711902115e-05,
+      "loss": 0.0873,
+      "step": 46
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.895439377085651e-05,
+      "loss": 0.2073,
+      "step": 47
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.89321468298109e-05,
+      "loss": 0.1531,
+      "step": 48
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.89098998887653e-05,
+      "loss": 0.1897,
+      "step": 49
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.888765294771969e-05,
+      "loss": 0.1044,
+      "step": 50
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.886540600667408e-05,
+      "loss": 0.1517,
+      "step": 51
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.884315906562848e-05,
+      "loss": 0.1919,
+      "step": 52
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.882091212458287e-05,
+      "loss": 0.1601,
+      "step": 53
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.8798665183537265e-05,
+      "loss": 0.1452,
+      "step": 54
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.877641824249166e-05,
+      "loss": 0.0979,
+      "step": 55
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.8754171301446054e-05,
+      "loss": 0.1448,
+      "step": 56
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.8731924360400445e-05,
+      "loss": 0.1391,
+      "step": 57
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 0.1886,
+      "step": 58
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.8687430478309235e-05,
+      "loss": 0.1399,
+      "step": 59
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.8665183537263626e-05,
+      "loss": 0.1223,
+      "step": 60
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.864293659621802e-05,
+      "loss": 0.1544,
+      "step": 61
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.8620689655172415e-05,
+      "loss": 0.1103,
+      "step": 62
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.859844271412681e-05,
+      "loss": 0.1665,
+      "step": 63
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.8576195773081204e-05,
+      "loss": 0.0724,
+      "step": 64
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.85539488320356e-05,
+      "loss": 0.0927,
+      "step": 65
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.853170189098999e-05,
+      "loss": 0.1999,
+      "step": 66
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.8509454949944384e-05,
+      "loss": 0.318,
+      "step": 67
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.848720800889878e-05,
+      "loss": 0.1065,
+      "step": 68
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.846496106785317e-05,
+      "loss": 0.119,
+      "step": 69
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8442714126807565e-05,
+      "loss": 0.2425,
+      "step": 70
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.842046718576196e-05,
+      "loss": 0.0852,
+      "step": 71
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8398220244716354e-05,
+      "loss": 0.1443,
+      "step": 72
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8375973303670745e-05,
+      "loss": 0.1345,
+      "step": 73
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8353726362625143e-05,
+      "loss": 0.1492,
+      "step": 74
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8331479421579534e-05,
+      "loss": 0.1651,
+      "step": 75
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8309232480533926e-05,
+      "loss": 0.2034,
+      "step": 76
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8286985539488324e-05,
+      "loss": 0.1517,
+      "step": 77
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8264738598442715e-05,
+      "loss": 0.1428,
+      "step": 78
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.824249165739711e-05,
+      "loss": 0.1755,
+      "step": 79
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8220244716351504e-05,
+      "loss": 0.1029,
+      "step": 80
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.81979977753059e-05,
+      "loss": 0.1242,
+      "step": 81
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.817575083426029e-05,
+      "loss": 0.2067,
+      "step": 82
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8153503893214684e-05,
+      "loss": 0.1591,
+      "step": 83
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.813125695216908e-05,
+      "loss": 0.1465,
+      "step": 84
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.810901001112347e-05,
+      "loss": 0.1462,
+      "step": 85
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.8086763070077865e-05,
+      "loss": 0.1751,
+      "step": 86
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 0.1788,
+      "step": 87
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.8042269187986654e-05,
+      "loss": 0.0804,
+      "step": 88
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.802002224694105e-05,
+      "loss": 0.1187,
+      "step": 89
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.7997775305895443e-05,
+      "loss": 0.1467,
+      "step": 90
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.7975528364849834e-05,
+      "loss": 0.0992,
+      "step": 91
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.795328142380423e-05,
+      "loss": 0.1486,
+      "step": 92
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.7931034482758623e-05,
+      "loss": 0.1994,
+      "step": 93
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.7908787541713015e-05,
+      "loss": 0.1112,
+      "step": 94
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.788654060066741e-05,
+      "loss": 0.161,
+      "step": 95
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.7864293659621804e-05,
+      "loss": 0.1667,
+      "step": 96
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.7842046718576198e-05,
+      "loss": 0.1416,
+      "step": 97
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.781979977753059e-05,
+      "loss": 0.1986,
+      "step": 98
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.7797552836484984e-05,
+      "loss": 0.0908,
+      "step": 99
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.777530589543938e-05,
+      "loss": 0.2254,
+      "step": 100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.775305895439377e-05,
+      "loss": 0.1893,
+      "step": 101
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.7730812013348168e-05,
+      "loss": 0.1303,
+      "step": 102
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.770856507230256e-05,
+      "loss": 0.1742,
+      "step": 103
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.7686318131256954e-05,
+      "loss": 0.1282,
+      "step": 104
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.7664071190211348e-05,
+      "loss": 0.1307,
+      "step": 105
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.7641824249165743e-05,
+      "loss": 0.0995,
+      "step": 106
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.7619577308120134e-05,
+      "loss": 0.1242,
+      "step": 107
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.759733036707453e-05,
+      "loss": 0.1959,
+      "step": 108
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.7575083426028923e-05,
+      "loss": 0.1665,
+      "step": 109
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.7552836484983314e-05,
+      "loss": 0.2246,
+      "step": 110
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.753058954393771e-05,
+      "loss": 0.139,
+      "step": 111
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.7508342602892103e-05,
+      "loss": 0.1272,
+      "step": 112
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.7486095661846498e-05,
+      "loss": 0.0899,
+      "step": 113
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.7463848720800893e-05,
+      "loss": 0.1121,
+      "step": 114
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.7441601779755287e-05,
+      "loss": 0.2018,
+      "step": 115
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 0.1133,
+      "step": 116
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.7397107897664073e-05,
+      "loss": 0.1117,
+      "step": 117
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.7374860956618468e-05,
+      "loss": 0.1286,
+      "step": 118
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.735261401557286e-05,
+      "loss": 0.1093,
+      "step": 119
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.7330367074527253e-05,
+      "loss": 0.1698,
+      "step": 120
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.7308120133481648e-05,
+      "loss": 0.1158,
+      "step": 121
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.7285873192436043e-05,
+      "loss": 0.2543,
+      "step": 122
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.7263626251390434e-05,
+      "loss": 0.1264,
+      "step": 123
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.7241379310344828e-05,
+      "loss": 0.1085,
+      "step": 124
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.7219132369299223e-05,
+      "loss": 0.1683,
+      "step": 125
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.7196885428253614e-05,
+      "loss": 0.1148,
+      "step": 126
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.7174638487208012e-05,
+      "loss": 0.1611,
+      "step": 127
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.7152391546162403e-05,
+      "loss": 0.219,
+      "step": 128
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.7130144605116798e-05,
+      "loss": 0.246,
+      "step": 129
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.7107897664071192e-05,
+      "loss": 0.1036,
+      "step": 130
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.7085650723025587e-05,
+      "loss": 0.1537,
+      "step": 131
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.7063403781979978e-05,
+      "loss": 0.147,
+      "step": 132
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.7041156840934373e-05,
+      "loss": 0.1204,
+      "step": 133
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.7018909899888767e-05,
+      "loss": 0.1681,
+      "step": 134
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.699666295884316e-05,
+      "loss": 0.1554,
+      "step": 135
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.6974416017797553e-05,
+      "loss": 0.1023,
+      "step": 136
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.6952169076751948e-05,
+      "loss": 0.1113,
+      "step": 137
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.6929922135706342e-05,
+      "loss": 0.172,
+      "step": 138
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.6907675194660737e-05,
+      "loss": 0.0839,
+      "step": 139
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.688542825361513e-05,
+      "loss": 0.1646,
+      "step": 140
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.6863181312569523e-05,
+      "loss": 0.1958,
+      "step": 141
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.6840934371523917e-05,
+      "loss": 0.0836,
+      "step": 142
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.6818687430478312e-05,
+      "loss": 0.1434,
+      "step": 143
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.6796440489432703e-05,
+      "loss": 0.1552,
+      "step": 144
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 0.1122,
+      "step": 145
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.6751946607341492e-05,
+      "loss": 0.1263,
+      "step": 146
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.6729699666295887e-05,
+      "loss": 0.0977,
+      "step": 147
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.6707452725250278e-05,
+      "loss": 0.1238,
+      "step": 148
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.6685205784204673e-05,
+      "loss": 0.1962,
+      "step": 149
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.6662958843159067e-05,
+      "loss": 0.1497,
+      "step": 150
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.664071190211346e-05,
+      "loss": 0.1162,
+      "step": 151
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.6618464961067856e-05,
+      "loss": 0.1397,
+      "step": 152
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.6596218020022247e-05,
+      "loss": 0.1656,
+      "step": 153
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.6573971078976642e-05,
+      "loss": 0.2044,
+      "step": 154
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.6551724137931037e-05,
+      "loss": 0.142,
+      "step": 155
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.652947719688543e-05,
+      "loss": 0.1261,
+      "step": 156
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.6507230255839822e-05,
+      "loss": 0.1906,
+      "step": 157
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.6484983314794217e-05,
+      "loss": 0.1451,
+      "step": 158
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.646273637374861e-05,
+      "loss": 0.1163,
+      "step": 159
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.6440489432703003e-05,
+      "loss": 0.0779,
+      "step": 160
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.6418242491657397e-05,
+      "loss": 0.1511,
+      "step": 161
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.6395995550611792e-05,
+      "loss": 0.1401,
+      "step": 162
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.6373748609566187e-05,
+      "loss": 0.1894,
+      "step": 163
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.635150166852058e-05,
+      "loss": 0.1263,
+      "step": 164
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.6329254727474972e-05,
+      "loss": 0.1233,
+      "step": 165
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.6307007786429367e-05,
+      "loss": 0.1428,
+      "step": 166
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.628476084538376e-05,
+      "loss": 0.1615,
+      "step": 167
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.6262513904338156e-05,
+      "loss": 0.195,
+      "step": 168
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.6240266963292547e-05,
+      "loss": 0.1727,
+      "step": 169
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.6218020022246942e-05,
+      "loss": 0.1086,
+      "step": 170
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.6195773081201336e-05,
+      "loss": 0.1692,
+      "step": 171
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.617352614015573e-05,
+      "loss": 0.1243,
+      "step": 172
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.6151279199110122e-05,
+      "loss": 0.1263,
+      "step": 173
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.1632,
+      "step": 174
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.610678531701891e-05,
+      "loss": 0.1465,
+      "step": 175
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.6084538375973303e-05,
+      "loss": 0.1518,
+      "step": 176
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.60622914349277e-05,
+      "loss": 0.1119,
+      "step": 177
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.6040044493882092e-05,
+      "loss": 0.164,
+      "step": 178
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.6017797552836486e-05,
+      "loss": 0.1423,
+      "step": 179
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.599555061179088e-05,
+      "loss": 0.1921,
+      "step": 180
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.5973303670745275e-05,
+      "loss": 0.0758,
+      "step": 181
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.5951056729699667e-05,
+      "loss": 0.158,
+      "step": 182
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.592880978865406e-05,
+      "loss": 0.0748,
+      "step": 183
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.5906562847608456e-05,
+      "loss": 0.0855,
+      "step": 184
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.5884315906562847e-05,
+      "loss": 0.2064,
+      "step": 185
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.586206896551724e-05,
+      "loss": 0.1268,
+      "step": 186
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.5839822024471636e-05,
+      "loss": 0.1795,
+      "step": 187
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.581757508342603e-05,
+      "loss": 0.1271,
+      "step": 188
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.5795328142380425e-05,
+      "loss": 0.1689,
+      "step": 189
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.5773081201334817e-05,
+      "loss": 0.1817,
+      "step": 190
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.575083426028921e-05,
+      "loss": 0.1271,
+      "step": 191
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.5728587319243606e-05,
+      "loss": 0.1685,
+      "step": 192
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.5706340378198e-05,
+      "loss": 0.1368,
+      "step": 193
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.568409343715239e-05,
+      "loss": 0.088,
+      "step": 194
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.5661846496106786e-05,
+      "loss": 0.0971,
+      "step": 195
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.563959955506118e-05,
+      "loss": 0.1124,
+      "step": 196
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.5617352614015575e-05,
+      "loss": 0.1521,
+      "step": 197
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.5595105672969966e-05,
+      "loss": 0.1711,
+      "step": 198
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.557285873192436e-05,
+      "loss": 0.1651,
+      "step": 199
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.5550611790878756e-05,
+      "loss": 0.167,
+      "step": 200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.552836484983315e-05,
+      "loss": 0.0826,
+      "step": 201
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.5506117908787545e-05,
+      "loss": 0.1001,
+      "step": 202
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 0.1263,
+      "step": 203
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.546162402669633e-05,
+      "loss": 0.0986,
+      "step": 204
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.5439377085650725e-05,
+      "loss": 0.1578,
+      "step": 205
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.541713014460512e-05,
+      "loss": 0.1387,
+      "step": 206
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.539488320355951e-05,
+      "loss": 0.1384,
+      "step": 207
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.5372636262513906e-05,
+      "loss": 0.1679,
+      "step": 208
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.53503893214683e-05,
+      "loss": 0.2111,
+      "step": 209
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.532814238042269e-05,
+      "loss": 0.1777,
+      "step": 210
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.5305895439377086e-05,
+      "loss": 0.1638,
+      "step": 211
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.528364849833148e-05,
+      "loss": 0.1473,
+      "step": 212
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.5261401557285875e-05,
+      "loss": 0.1365,
+      "step": 213
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.5239154616240268e-05,
+      "loss": 0.1964,
+      "step": 214
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.521690767519466e-05,
+      "loss": 0.1425,
+      "step": 215
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.5194660734149055e-05,
+      "loss": 0.1283,
+      "step": 216
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.5172413793103448e-05,
+      "loss": 0.1776,
+      "step": 217
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.5150166852057845e-05,
+      "loss": 0.1678,
+      "step": 218
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.5127919911012236e-05,
+      "loss": 0.104,
+      "step": 219
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.5105672969966632e-05,
+      "loss": 0.1697,
+      "step": 220
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.5083426028921025e-05,
+      "loss": 0.1292,
+      "step": 221
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.506117908787542e-05,
+      "loss": 0.1151,
+      "step": 222
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.5038932146829812e-05,
+      "loss": 0.1662,
+      "step": 223
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.5016685205784205e-05,
+      "loss": 0.1118,
+      "step": 224
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.49944382647386e-05,
+      "loss": 0.1394,
+      "step": 225
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.4972191323692993e-05,
+      "loss": 0.1814,
+      "step": 226
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.4949944382647387e-05,
+      "loss": 0.1291,
+      "step": 227
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.492769744160178e-05,
+      "loss": 0.1509,
+      "step": 228
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.4905450500556175e-05,
+      "loss": 0.124,
+      "step": 229
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.4883203559510568e-05,
+      "loss": 0.1561,
+      "step": 230
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.486095661846496e-05,
+      "loss": 0.1941,
+      "step": 231
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 0.1522,
+      "step": 232
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.481646273637375e-05,
+      "loss": 0.1267,
+      "step": 233
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.4794215795328144e-05,
+      "loss": 0.1505,
+      "step": 234
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.4771968854282537e-05,
+      "loss": 0.127,
+      "step": 235
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.4749721913236932e-05,
+      "loss": 0.1281,
+      "step": 236
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.4727474972191325e-05,
+      "loss": 0.1379,
+      "step": 237
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.470522803114572e-05,
+      "loss": 0.1518,
+      "step": 238
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.4682981090100112e-05,
+      "loss": 0.1511,
+      "step": 239
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.4660734149054505e-05,
+      "loss": 0.2012,
+      "step": 240
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.46384872080089e-05,
+      "loss": 0.1492,
+      "step": 241
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.4616240266963293e-05,
+      "loss": 0.1444,
+      "step": 242
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.4593993325917689e-05,
+      "loss": 0.0894,
+      "step": 243
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.457174638487208e-05,
+      "loss": 0.115,
+      "step": 244
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.4549499443826476e-05,
+      "loss": 0.155,
+      "step": 245
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.452725250278087e-05,
+      "loss": 0.1707,
+      "step": 246
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.4505005561735264e-05,
+      "loss": 0.1388,
+      "step": 247
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.4482758620689657e-05,
+      "loss": 0.1635,
+      "step": 248
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.446051167964405e-05,
+      "loss": 0.126,
+      "step": 249
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.4438264738598444e-05,
+      "loss": 0.0949,
+      "step": 250
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.4416017797552837e-05,
+      "loss": 0.0982,
+      "step": 251
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.4393770856507232e-05,
+      "loss": 0.0994,
+      "step": 252
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.4371523915461624e-05,
+      "loss": 0.1265,
+      "step": 253
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.4349276974416019e-05,
+      "loss": 0.1341,
+      "step": 254
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.4327030033370412e-05,
+      "loss": 0.136,
+      "step": 255
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.4304783092324805e-05,
+      "loss": 0.1367,
+      "step": 256
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.4282536151279201e-05,
+      "loss": 0.1139,
+      "step": 257
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.4260289210233594e-05,
+      "loss": 0.125,
+      "step": 258
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.4238042269187989e-05,
+      "loss": 0.1224,
+      "step": 259
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.4215795328142381e-05,
+      "loss": 0.1866,
+      "step": 260
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 0.0824,
+      "step": 261
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.4171301446051169e-05,
+      "loss": 0.1221,
+      "step": 262
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.4149054505005564e-05,
+      "loss": 0.1397,
+      "step": 263
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.4126807563959956e-05,
+      "loss": 0.1498,
+      "step": 264
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.410456062291435e-05,
+      "loss": 0.194,
+      "step": 265
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.4082313681868744e-05,
+      "loss": 0.2462,
+      "step": 266
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.4060066740823137e-05,
+      "loss": 0.1401,
+      "step": 267
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.4037819799777533e-05,
+      "loss": 0.1145,
+      "step": 268
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.4015572858731924e-05,
+      "loss": 0.1187,
+      "step": 269
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.399332591768632e-05,
+      "loss": 0.0934,
+      "step": 270
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.3971078976640713e-05,
+      "loss": 0.1304,
+      "step": 271
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.3948832035595108e-05,
+      "loss": 0.1761,
+      "step": 272
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.3926585094549501e-05,
+      "loss": 0.1373,
+      "step": 273
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.3904338153503894e-05,
+      "loss": 0.1008,
+      "step": 274
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.3882091212458288e-05,
+      "loss": 0.1393,
+      "step": 275
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.3859844271412681e-05,
+      "loss": 0.1058,
+      "step": 276
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.3837597330367076e-05,
+      "loss": 0.1129,
+      "step": 277
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.3815350389321469e-05,
+      "loss": 0.1152,
+      "step": 278
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.3793103448275863e-05,
+      "loss": 0.1385,
+      "step": 279
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.3770856507230256e-05,
+      "loss": 0.1865,
+      "step": 280
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.3748609566184649e-05,
+      "loss": 0.1783,
+      "step": 281
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.3726362625139045e-05,
+      "loss": 0.1278,
+      "step": 282
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.3704115684093438e-05,
+      "loss": 0.1797,
+      "step": 283
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.3681868743047833e-05,
+      "loss": 0.1101,
+      "step": 284
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.3659621802002226e-05,
+      "loss": 0.1394,
+      "step": 285
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.363737486095662e-05,
+      "loss": 0.1645,
+      "step": 286
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.3615127919911013e-05,
+      "loss": 0.0854,
+      "step": 287
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.3592880978865408e-05,
+      "loss": 0.0963,
+      "step": 288
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.35706340378198e-05,
+      "loss": 0.1229,
+      "step": 289
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 0.1717,
+      "step": 290
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.3526140155728588e-05,
+      "loss": 0.1502,
+      "step": 291
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.3503893214682981e-05,
+      "loss": 0.1245,
+      "step": 292
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.3481646273637377e-05,
+      "loss": 0.0957,
+      "step": 293
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.3459399332591769e-05,
+      "loss": 0.0829,
+      "step": 294
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.3437152391546165e-05,
+      "loss": 0.2046,
+      "step": 295
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.3414905450500558e-05,
+      "loss": 0.1544,
+      "step": 296
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.339265850945495e-05,
+      "loss": 0.1204,
+      "step": 297
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.3370411568409345e-05,
+      "loss": 0.1099,
+      "step": 298
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.3348164627363738e-05,
+      "loss": 0.1369,
+      "step": 299
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.3325917686318133e-05,
+      "loss": 0.0971,
+      "step": 300
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.3303670745272526e-05,
+      "loss": 0.2003,
+      "step": 301
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.328142380422692e-05,
+      "loss": 0.1246,
+      "step": 302
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.3259176863181313e-05,
+      "loss": 0.1783,
+      "step": 303
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.3236929922135708e-05,
+      "loss": 0.1114,
+      "step": 304
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.32146829810901e-05,
+      "loss": 0.1093,
+      "step": 305
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.3192436040044493e-05,
+      "loss": 0.1327,
+      "step": 306
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.317018909899889e-05,
+      "loss": 0.1762,
+      "step": 307
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.3147942157953283e-05,
+      "loss": 0.1508,
+      "step": 308
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.3125695216907677e-05,
+      "loss": 0.1401,
+      "step": 309
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.310344827586207e-05,
+      "loss": 0.1276,
+      "step": 310
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.3081201334816465e-05,
+      "loss": 0.1106,
+      "step": 311
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.3058954393770857e-05,
+      "loss": 0.1894,
+      "step": 312
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.3036707452725252e-05,
+      "loss": 0.1045,
+      "step": 313
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.3014460511679645e-05,
+      "loss": 0.162,
+      "step": 314
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.2992213570634038e-05,
+      "loss": 0.1029,
+      "step": 315
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.2969966629588432e-05,
+      "loss": 0.1472,
+      "step": 316
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.2947719688542825e-05,
+      "loss": 0.1846,
+      "step": 317
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.2925472747497222e-05,
+      "loss": 0.1557,
+      "step": 318
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.1099,
+      "step": 319
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.2880978865406009e-05,
+      "loss": 0.1306,
+      "step": 320
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.2858731924360402e-05,
+      "loss": 0.089,
+      "step": 321
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.2836484983314795e-05,
+      "loss": 0.1697,
+      "step": 322
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.281423804226919e-05,
+      "loss": 0.1208,
+      "step": 323
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.2791991101223582e-05,
+      "loss": 0.0941,
+      "step": 324
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.2769744160177977e-05,
+      "loss": 0.1103,
+      "step": 325
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.274749721913237e-05,
+      "loss": 0.1815,
+      "step": 326
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.2725250278086764e-05,
+      "loss": 0.174,
+      "step": 327
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.2703003337041157e-05,
+      "loss": 0.1094,
+      "step": 328
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.2680756395995552e-05,
+      "loss": 0.1329,
+      "step": 329
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.2658509454949945e-05,
+      "loss": 0.102,
+      "step": 330
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.2636262513904338e-05,
+      "loss": 0.1236,
+      "step": 331
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.2614015572858734e-05,
+      "loss": 0.123,
+      "step": 332
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.2591768631813127e-05,
+      "loss": 0.0853,
+      "step": 333
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.2569521690767521e-05,
+      "loss": 0.1865,
+      "step": 334
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.2547274749721914e-05,
+      "loss": 0.1145,
+      "step": 335
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.2525027808676309e-05,
+      "loss": 0.1631,
+      "step": 336
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.2502780867630702e-05,
+      "loss": 0.1286,
+      "step": 337
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.2480533926585096e-05,
+      "loss": 0.1702,
+      "step": 338
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.245828698553949e-05,
+      "loss": 0.1594,
+      "step": 339
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.2436040044493882e-05,
+      "loss": 0.1371,
+      "step": 340
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.2413793103448277e-05,
+      "loss": 0.1514,
+      "step": 341
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.239154616240267e-05,
+      "loss": 0.1147,
+      "step": 342
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.2369299221357066e-05,
+      "loss": 0.1334,
+      "step": 343
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.2347052280311459e-05,
+      "loss": 0.0993,
+      "step": 344
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.2324805339265853e-05,
+      "loss": 0.1651,
+      "step": 345
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.2302558398220246e-05,
+      "loss": 0.1527,
+      "step": 346
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.2280311457174639e-05,
+      "loss": 0.1341,
+      "step": 347
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 0.0667,
+      "step": 348
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.2235817575083427e-05,
+      "loss": 0.1488,
+      "step": 349
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.2213570634037821e-05,
+      "loss": 0.1568,
+      "step": 350
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.2191323692992214e-05,
+      "loss": 0.1596,
+      "step": 351
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.2169076751946609e-05,
+      "loss": 0.1998,
+      "step": 352
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.2146829810901001e-05,
+      "loss": 0.167,
+      "step": 353
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.2124582869855396e-05,
+      "loss": 0.1583,
+      "step": 354
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.2102335928809789e-05,
+      "loss": 0.1266,
+      "step": 355
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.2080088987764182e-05,
+      "loss": 0.1191,
+      "step": 356
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.2057842046718578e-05,
+      "loss": 0.1002,
+      "step": 357
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.2035595105672971e-05,
+      "loss": 0.1033,
+      "step": 358
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.2013348164627366e-05,
+      "loss": 0.0964,
+      "step": 359
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.1991101223581758e-05,
+      "loss": 0.1612,
+      "step": 360
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.1968854282536153e-05,
+      "loss": 0.2018,
+      "step": 361
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.1946607341490546e-05,
+      "loss": 0.1069,
+      "step": 362
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.1924360400444939e-05,
+      "loss": 0.1268,
+      "step": 363
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.1902113459399333e-05,
+      "loss": 0.1485,
+      "step": 364
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.1879866518353726e-05,
+      "loss": 0.1429,
+      "step": 365
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.1857619577308121e-05,
+      "loss": 0.1374,
+      "step": 366
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.1835372636262514e-05,
+      "loss": 0.1323,
+      "step": 367
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.181312569521691e-05,
+      "loss": 0.1134,
+      "step": 368
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.1790878754171303e-05,
+      "loss": 0.149,
+      "step": 369
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.1768631813125698e-05,
+      "loss": 0.0928,
+      "step": 370
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.174638487208009e-05,
+      "loss": 0.0988,
+      "step": 371
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.1724137931034483e-05,
+      "loss": 0.1223,
+      "step": 372
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.1701890989988878e-05,
+      "loss": 0.1102,
+      "step": 373
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.167964404894327e-05,
+      "loss": 0.2099,
+      "step": 374
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.1657397107897665e-05,
+      "loss": 0.1,
+      "step": 375
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.1635150166852058e-05,
+      "loss": 0.1355,
+      "step": 376
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 0.1541,
+      "step": 377
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.1590656284760846e-05,
+      "loss": 0.1624,
+      "step": 378
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.1568409343715242e-05,
+      "loss": 0.1106,
+      "step": 379
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.1546162402669633e-05,
+      "loss": 0.1526,
+      "step": 380
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.1523915461624026e-05,
+      "loss": 0.1516,
+      "step": 381
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.1501668520578422e-05,
+      "loss": 0.1906,
+      "step": 382
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.1479421579532815e-05,
+      "loss": 0.1259,
+      "step": 383
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.145717463848721e-05,
+      "loss": 0.1446,
+      "step": 384
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.1434927697441603e-05,
+      "loss": 0.0954,
+      "step": 385
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.1412680756395997e-05,
+      "loss": 0.1493,
+      "step": 386
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.139043381535039e-05,
+      "loss": 0.0875,
+      "step": 387
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.1368186874304783e-05,
+      "loss": 0.0963,
+      "step": 388
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.1345939933259178e-05,
+      "loss": 0.0652,
+      "step": 389
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.132369299221357e-05,
+      "loss": 0.1669,
+      "step": 390
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.1301446051167965e-05,
+      "loss": 0.1409,
+      "step": 391
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.1279199110122358e-05,
+      "loss": 0.1069,
+      "step": 392
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.1256952169076754e-05,
+      "loss": 0.1092,
+      "step": 393
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.1234705228031147e-05,
+      "loss": 0.1222,
+      "step": 394
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.1212458286985542e-05,
+      "loss": 0.1234,
+      "step": 395
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.1190211345939935e-05,
+      "loss": 0.074,
+      "step": 396
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.1167964404894328e-05,
+      "loss": 0.0612,
+      "step": 397
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.1145717463848722e-05,
+      "loss": 0.1779,
+      "step": 398
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.1123470522803115e-05,
+      "loss": 0.1303,
+      "step": 399
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.110122358175751e-05,
+      "loss": 0.1453,
+      "step": 400
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.1078976640711903e-05,
+      "loss": 0.0602,
+      "step": 401
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.1056729699666297e-05,
+      "loss": 0.1564,
+      "step": 402
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.103448275862069e-05,
+      "loss": 0.1226,
+      "step": 403
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.1012235817575086e-05,
+      "loss": 0.1334,
+      "step": 404
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.0989988876529477e-05,
+      "loss": 0.1854,
+      "step": 405
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 0.1344,
+      "step": 406
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.0945494994438267e-05,
+      "loss": 0.157,
+      "step": 407
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.092324805339266e-05,
+      "loss": 0.1419,
+      "step": 408
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.0901001112347054e-05,
+      "loss": 0.1884,
+      "step": 409
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.0878754171301447e-05,
+      "loss": 0.1676,
+      "step": 410
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.0856507230255842e-05,
+      "loss": 0.1173,
+      "step": 411
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.0834260289210234e-05,
+      "loss": 0.0992,
+      "step": 412
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.0812013348164627e-05,
+      "loss": 0.1187,
+      "step": 413
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.0789766407119022e-05,
+      "loss": 0.1282,
+      "step": 414
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.0767519466073415e-05,
+      "loss": 0.1508,
+      "step": 415
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.074527252502781e-05,
+      "loss": 0.1056,
+      "step": 416
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.0723025583982202e-05,
+      "loss": 0.1837,
+      "step": 417
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.0700778642936599e-05,
+      "loss": 0.1114,
+      "step": 418
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.0678531701890991e-05,
+      "loss": 0.1043,
+      "step": 419
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.0656284760845386e-05,
+      "loss": 0.0863,
+      "step": 420
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.0634037819799779e-05,
+      "loss": 0.1171,
+      "step": 421
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.0611790878754172e-05,
+      "loss": 0.1687,
+      "step": 422
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.0589543937708566e-05,
+      "loss": 0.1668,
+      "step": 423
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.056729699666296e-05,
+      "loss": 0.1114,
+      "step": 424
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.0545050055617354e-05,
+      "loss": 0.1236,
+      "step": 425
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.0522803114571747e-05,
+      "loss": 0.0799,
+      "step": 426
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.0500556173526141e-05,
+      "loss": 0.1358,
+      "step": 427
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.0478309232480534e-05,
+      "loss": 0.0837,
+      "step": 428
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.0456062291434927e-05,
+      "loss": 0.1112,
+      "step": 429
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.0433815350389322e-05,
+      "loss": 0.0838,
+      "step": 430
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.0411568409343715e-05,
+      "loss": 0.1558,
+      "step": 431
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.0389321468298111e-05,
+      "loss": 0.1595,
+      "step": 432
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.0367074527252504e-05,
+      "loss": 0.1729,
+      "step": 433
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.0344827586206898e-05,
+      "loss": 0.0923,
+      "step": 434
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 0.1395,
+      "step": 435
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.0300333704115686e-05,
+      "loss": 0.1475,
+      "step": 436
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0278086763070079e-05,
+      "loss": 0.104,
+      "step": 437
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0255839822024472e-05,
+      "loss": 0.0751,
+      "step": 438
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0233592880978866e-05,
+      "loss": 0.1793,
+      "step": 439
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0211345939933259e-05,
+      "loss": 0.1233,
+      "step": 440
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0189098998887654e-05,
+      "loss": 0.1875,
+      "step": 441
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0166852057842047e-05,
+      "loss": 0.0859,
+      "step": 442
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0144605116796443e-05,
+      "loss": 0.136,
+      "step": 443
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0122358175750836e-05,
+      "loss": 0.0792,
+      "step": 444
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.010011123470523e-05,
+      "loss": 0.1235,
+      "step": 445
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0077864293659623e-05,
+      "loss": 0.1159,
+      "step": 446
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0055617352614016e-05,
+      "loss": 0.2034,
+      "step": 447
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.003337041156841e-05,
+      "loss": 0.1197,
+      "step": 448
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0011123470522804e-05,
+      "loss": 0.1617,
+      "step": 449
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.988876529477196e-06,
+      "loss": 0.1081,
+      "step": 450
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.966629588431591e-06,
+      "loss": 0.1349,
+      "step": 451
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.944382647385986e-06,
+      "loss": 0.1195,
+      "step": 452
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.922135706340378e-06,
+      "loss": 0.1047,
+      "step": 453
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.899888765294773e-06,
+      "loss": 0.1124,
+      "step": 454
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.877641824249166e-06,
+      "loss": 0.2128,
+      "step": 455
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.85539488320356e-06,
+      "loss": 0.0932,
+      "step": 456
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.833147942157955e-06,
+      "loss": 0.174,
+      "step": 457
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.810901001112348e-06,
+      "loss": 0.0757,
+      "step": 458
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.788654060066741e-06,
+      "loss": 0.1388,
+      "step": 459
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.766407119021135e-06,
+      "loss": 0.0941,
+      "step": 460
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.744160177975528e-06,
+      "loss": 0.0583,
+      "step": 461
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.721913236929923e-06,
+      "loss": 0.1737,
+      "step": 462
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.699666295884318e-06,
+      "loss": 0.1694,
+      "step": 463
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.1221,
+      "step": 464
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.655172413793105e-06,
+      "loss": 0.1675,
+      "step": 465
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.632925472747498e-06,
+      "loss": 0.168,
+      "step": 466
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.61067853170189e-06,
+      "loss": 0.1342,
+      "step": 467
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.588431590656285e-06,
+      "loss": 0.1179,
+      "step": 468
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.56618464961068e-06,
+      "loss": 0.1226,
+      "step": 469
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.543937708565073e-06,
+      "loss": 0.1451,
+      "step": 470
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.521690767519467e-06,
+      "loss": 0.0914,
+      "step": 471
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.49944382647386e-06,
+      "loss": 0.1138,
+      "step": 472
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.477196885428255e-06,
+      "loss": 0.2265,
+      "step": 473
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.45494994438265e-06,
+      "loss": 0.1218,
+      "step": 474
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.43270300333704e-06,
+      "loss": 0.1603,
+      "step": 475
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.410456062291435e-06,
+      "loss": 0.1883,
+      "step": 476
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.38820912124583e-06,
+      "loss": 0.1457,
+      "step": 477
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.365962180200223e-06,
+      "loss": 0.1743,
+      "step": 478
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.343715239154617e-06,
+      "loss": 0.181,
+      "step": 479
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.32146829810901e-06,
+      "loss": 0.1338,
+      "step": 480
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.299221357063405e-06,
+      "loss": 0.1293,
+      "step": 481
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.2769744160178e-06,
+      "loss": 0.0753,
+      "step": 482
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.254727474972192e-06,
+      "loss": 0.0954,
+      "step": 483
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.232480533926585e-06,
+      "loss": 0.1116,
+      "step": 484
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.21023359288098e-06,
+      "loss": 0.163,
+      "step": 485
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.187986651835373e-06,
+      "loss": 0.1193,
+      "step": 486
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.165739710789767e-06,
+      "loss": 0.1505,
+      "step": 487
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.143492769744162e-06,
+      "loss": 0.2265,
+      "step": 488
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.121245828698555e-06,
+      "loss": 0.0593,
+      "step": 489
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.09899888765295e-06,
+      "loss": 0.1985,
+      "step": 490
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.076751946607342e-06,
+      "loss": 0.1006,
+      "step": 491
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.054505005561735e-06,
+      "loss": 0.1235,
+      "step": 492
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.0804,
+      "step": 493
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.010011123470524e-06,
+      "loss": 0.146,
+      "step": 494
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.987764182424917e-06,
+      "loss": 0.0699,
+      "step": 495
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.965517241379312e-06,
+      "loss": 0.1299,
+      "step": 496
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.943270300333705e-06,
+      "loss": 0.1434,
+      "step": 497
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.921023359288099e-06,
+      "loss": 0.1083,
+      "step": 498
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.898776418242492e-06,
+      "loss": 0.1202,
+      "step": 499
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.876529477196885e-06,
+      "loss": 0.16,
+      "step": 500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.85428253615128e-06,
+      "loss": 0.1725,
+      "step": 501
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.832035595105674e-06,
+      "loss": 0.084,
+      "step": 502
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.809788654060067e-06,
+      "loss": 0.182,
+      "step": 503
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.787541713014462e-06,
+      "loss": 0.0851,
+      "step": 504
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.765294771968854e-06,
+      "loss": 0.1367,
+      "step": 505
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.743047830923249e-06,
+      "loss": 0.1441,
+      "step": 506
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.720800889877644e-06,
+      "loss": 0.1773,
+      "step": 507
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.698553948832036e-06,
+      "loss": 0.1206,
+      "step": 508
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.67630700778643e-06,
+      "loss": 0.1788,
+      "step": 509
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.654060066740824e-06,
+      "loss": 0.1122,
+      "step": 510
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.631813125695217e-06,
+      "loss": 0.1137,
+      "step": 511
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.609566184649611e-06,
+      "loss": 0.109,
+      "step": 512
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.587319243604006e-06,
+      "loss": 0.1863,
+      "step": 513
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.565072302558399e-06,
+      "loss": 0.1014,
+      "step": 514
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.542825361512793e-06,
+      "loss": 0.1367,
+      "step": 515
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.520578420467186e-06,
+      "loss": 0.1078,
+      "step": 516
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.49833147942158e-06,
+      "loss": 0.1548,
+      "step": 517
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.476084538375974e-06,
+      "loss": 0.1535,
+      "step": 518
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.453837597330368e-06,
+      "loss": 0.0947,
+      "step": 519
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.431590656284761e-06,
+      "loss": 0.1175,
+      "step": 520
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.409343715239156e-06,
+      "loss": 0.1945,
+      "step": 521
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.095,
+      "step": 522
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.364849833147943e-06,
+      "loss": 0.0834,
+      "step": 523
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.342602892102336e-06,
+      "loss": 0.1149,
+      "step": 524
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.32035595105673e-06,
+      "loss": 0.0997,
+      "step": 525
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.298109010011124e-06,
+      "loss": 0.1504,
+      "step": 526
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.275862068965518e-06,
+      "loss": 0.1913,
+      "step": 527
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.253615127919911e-06,
+      "loss": 0.0994,
+      "step": 528
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.231368186874306e-06,
+      "loss": 0.0657,
+      "step": 529
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.209121245828699e-06,
+      "loss": 0.1146,
+      "step": 530
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.186874304783093e-06,
+      "loss": 0.1536,
+      "step": 531
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.164627363737486e-06,
+      "loss": 0.1366,
+      "step": 532
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.14238042269188e-06,
+      "loss": 0.1267,
+      "step": 533
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 8.120133481646274e-06,
+      "loss": 0.1135,
+      "step": 534
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.097886540600668e-06,
+      "loss": 0.0926,
+      "step": 535
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.075639599555061e-06,
+      "loss": 0.1566,
+      "step": 536
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.053392658509456e-06,
+      "loss": 0.124,
+      "step": 537
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.03114571746385e-06,
+      "loss": 0.1865,
+      "step": 538
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 8.008898776418243e-06,
+      "loss": 0.1176,
+      "step": 539
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.986651835372638e-06,
+      "loss": 0.1236,
+      "step": 540
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.96440489432703e-06,
+      "loss": 0.142,
+      "step": 541
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.942157953281424e-06,
+      "loss": 0.1503,
+      "step": 542
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.919911012235818e-06,
+      "loss": 0.1745,
+      "step": 543
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.897664071190213e-06,
+      "loss": 0.1531,
+      "step": 544
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.875417130144606e-06,
+      "loss": 0.1299,
+      "step": 545
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.853170189099e-06,
+      "loss": 0.1245,
+      "step": 546
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.830923248053393e-06,
+      "loss": 0.1316,
+      "step": 547
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.808676307007788e-06,
+      "loss": 0.123,
+      "step": 548
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.78642936596218e-06,
+      "loss": 0.0601,
+      "step": 549
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.764182424916575e-06,
+      "loss": 0.2135,
+      "step": 550
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.0962,
+      "step": 551
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.719688542825363e-06,
+      "loss": 0.0931,
+      "step": 552
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.697441601779755e-06,
+      "loss": 0.1949,
+      "step": 553
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.67519466073415e-06,
+      "loss": 0.1218,
+      "step": 554
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.652947719688543e-06,
+      "loss": 0.1151,
+      "step": 555
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.630700778642938e-06,
+      "loss": 0.1216,
+      "step": 556
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.60845383759733e-06,
+      "loss": 0.0899,
+      "step": 557
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.586206896551724e-06,
+      "loss": 0.1765,
+      "step": 558
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.563959955506118e-06,
+      "loss": 0.117,
+      "step": 559
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.5417130144605125e-06,
+      "loss": 0.1227,
+      "step": 560
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 7.519466073414906e-06,
+      "loss": 0.1312,
+      "step": 561
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.4972191323693e-06,
+      "loss": 0.1302,
+      "step": 562
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.474972191323694e-06,
+      "loss": 0.0877,
+      "step": 563
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.452725250278087e-06,
+      "loss": 0.1555,
+      "step": 564
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.43047830923248e-06,
+      "loss": 0.1694,
+      "step": 565
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.408231368186875e-06,
+      "loss": 0.1208,
+      "step": 566
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.385984427141269e-06,
+      "loss": 0.158,
+      "step": 567
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.363737486095662e-06,
+      "loss": 0.1357,
+      "step": 568
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.341490545050056e-06,
+      "loss": 0.1652,
+      "step": 569
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.31924360400445e-06,
+      "loss": 0.0939,
+      "step": 570
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.296996662958844e-06,
+      "loss": 0.1393,
+      "step": 571
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.274749721913238e-06,
+      "loss": 0.1975,
+      "step": 572
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.252502780867632e-06,
+      "loss": 0.113,
+      "step": 573
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.230255839822025e-06,
+      "loss": 0.1865,
+      "step": 574
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.2080088987764185e-06,
+      "loss": 0.1172,
+      "step": 575
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.185761957730812e-06,
+      "loss": 0.1493,
+      "step": 576
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.163515016685206e-06,
+      "loss": 0.0838,
+      "step": 577
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.1412680756396006e-06,
+      "loss": 0.1406,
+      "step": 578
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 7.119021134593994e-06,
+      "loss": 0.0625,
+      "step": 579
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.0948,
+      "step": 580
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.074527252502782e-06,
+      "loss": 0.1107,
+      "step": 581
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.052280311457175e-06,
+      "loss": 0.1588,
+      "step": 582
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.030033370411568e-06,
+      "loss": 0.0948,
+      "step": 583
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 7.007786429365962e-06,
+      "loss": 0.1782,
+      "step": 584
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 6.985539488320357e-06,
+      "loss": 0.1876,
+      "step": 585
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 6.9632925472747504e-06,
+      "loss": 0.1564,
+      "step": 586
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 6.941045606229144e-06,
+      "loss": 0.1393,
+      "step": 587
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 6.918798665183538e-06,
+      "loss": 0.0944,
+      "step": 588
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 0.1289,
+      "step": 589
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.8743047830923245e-06,
+      "loss": 0.0833,
+      "step": 590
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.852057842046719e-06,
+      "loss": 0.1976,
+      "step": 591
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.829810901001113e-06,
+      "loss": 0.0764,
+      "step": 592
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.807563959955507e-06,
+      "loss": 0.1371,
+      "step": 593
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.7853170189099e-06,
+      "loss": 0.1103,
+      "step": 594
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.763070077864294e-06,
+      "loss": 0.1295,
+      "step": 595
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.740823136818689e-06,
+      "loss": 0.1438,
+      "step": 596
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 6.718576195773082e-06,
+      "loss": 0.1196,
+      "step": 597
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.696329254727475e-06,
+      "loss": 0.1536,
+      "step": 598
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.674082313681869e-06,
+      "loss": 0.1171,
+      "step": 599
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.651835372636263e-06,
+      "loss": 0.0995,
+      "step": 600
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.6295884315906565e-06,
+      "loss": 0.128,
+      "step": 601
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.60734149054505e-06,
+      "loss": 0.1729,
+      "step": 602
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.585094549499445e-06,
+      "loss": 0.1814,
+      "step": 603
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.5628476084538385e-06,
+      "loss": 0.1303,
+      "step": 604
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.540600667408232e-06,
+      "loss": 0.1543,
+      "step": 605
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 6.518353726362626e-06,
+      "loss": 0.139,
+      "step": 606
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.496106785317019e-06,
+      "loss": 0.0994,
+      "step": 607
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.473859844271413e-06,
+      "loss": 0.1305,
+      "step": 608
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.1032,
+      "step": 609
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.429365962180201e-06,
+      "loss": 0.1587,
+      "step": 610
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.407119021134595e-06,
+      "loss": 0.0986,
+      "step": 611
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.3848720800889884e-06,
+      "loss": 0.1344,
+      "step": 612
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.362625139043382e-06,
+      "loss": 0.1832,
+      "step": 613
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.340378197997776e-06,
+      "loss": 0.081,
+      "step": 614
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.318131256952169e-06,
+      "loss": 0.1609,
+      "step": 615
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.295884315906563e-06,
+      "loss": 0.1742,
+      "step": 616
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.273637374860957e-06,
+      "loss": 0.1004,
+      "step": 617
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.251390433815351e-06,
+      "loss": 0.1308,
+      "step": 618
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.229143492769745e-06,
+      "loss": 0.1363,
+      "step": 619
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.206896551724138e-06,
+      "loss": 0.1272,
+      "step": 620
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.184649610678533e-06,
+      "loss": 0.1451,
+      "step": 621
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.162402669632927e-06,
+      "loss": 0.1717,
+      "step": 622
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.1401557285873195e-06,
+      "loss": 0.0978,
+      "step": 623
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.117908787541713e-06,
+      "loss": 0.0851,
+      "step": 624
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.095661846496107e-06,
+      "loss": 0.1123,
+      "step": 625
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.073414905450501e-06,
+      "loss": 0.1026,
+      "step": 626
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.0511679644048945e-06,
+      "loss": 0.1699,
+      "step": 627
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.028921023359289e-06,
+      "loss": 0.1389,
+      "step": 628
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.006674082313683e-06,
+      "loss": 0.1814,
+      "step": 629
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.9844271412680765e-06,
+      "loss": 0.0873,
+      "step": 630
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.962180200222469e-06,
+      "loss": 0.1457,
+      "step": 631
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.939933259176863e-06,
+      "loss": 0.1748,
+      "step": 632
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.917686318131257e-06,
+      "loss": 0.0879,
+      "step": 633
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.8954393770856515e-06,
+      "loss": 0.1342,
+      "step": 634
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.873192436040045e-06,
+      "loss": 0.1325,
+      "step": 635
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.850945494994439e-06,
+      "loss": 0.1546,
+      "step": 636
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.828698553948833e-06,
+      "loss": 0.0895,
+      "step": 637
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.1967,
+      "step": 638
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.784204671857621e-06,
+      "loss": 0.1346,
+      "step": 639
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.761957730812013e-06,
+      "loss": 0.1509,
+      "step": 640
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.739710789766408e-06,
+      "loss": 0.0653,
+      "step": 641
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.717463848720801e-06,
+      "loss": 0.1007,
+      "step": 642
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.695216907675195e-06,
+      "loss": 0.1161,
+      "step": 643
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.672969966629589e-06,
+      "loss": 0.1221,
+      "step": 644
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.6507230255839826e-06,
+      "loss": 0.1192,
+      "step": 645
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.628476084538377e-06,
+      "loss": 0.1577,
+      "step": 646
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.606229143492771e-06,
+      "loss": 0.121,
+      "step": 647
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.583982202447164e-06,
+      "loss": 0.1169,
+      "step": 648
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5617352614015575e-06,
+      "loss": 0.1256,
+      "step": 649
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.539488320355951e-06,
+      "loss": 0.1681,
+      "step": 650
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.517241379310345e-06,
+      "loss": 0.051,
+      "step": 651
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.494994438264739e-06,
+      "loss": 0.0876,
+      "step": 652
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.472747497219133e-06,
+      "loss": 0.0998,
+      "step": 653
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.450500556173527e-06,
+      "loss": 0.1544,
+      "step": 654
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.428253615127921e-06,
+      "loss": 0.1637,
+      "step": 655
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.406006674082314e-06,
+      "loss": 0.1538,
+      "step": 656
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.383759733036707e-06,
+      "loss": 0.157,
+      "step": 657
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.361512791991101e-06,
+      "loss": 0.081,
+      "step": 658
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.339265850945496e-06,
+      "loss": 0.1512,
+      "step": 659
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.3170189098998895e-06,
+      "loss": 0.1404,
+      "step": 660
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.294771968854283e-06,
+      "loss": 0.1109,
+      "step": 661
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.272525027808677e-06,
+      "loss": 0.1473,
+      "step": 662
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.250278086763071e-06,
+      "loss": 0.0665,
+      "step": 663
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.2280311457174636e-06,
+      "loss": 0.0949,
+      "step": 664
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.205784204671857e-06,
+      "loss": 0.1562,
+      "step": 665
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.183537263626252e-06,
+      "loss": 0.1436,
+      "step": 666
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.1641,
+      "step": 667
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.139043381535039e-06,
+      "loss": 0.1152,
+      "step": 668
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.116796440489433e-06,
+      "loss": 0.1611,
+      "step": 669
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.094549499443827e-06,
+      "loss": 0.1225,
+      "step": 670
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.072302558398221e-06,
+      "loss": 0.2414,
+      "step": 671
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.050055617352615e-06,
+      "loss": 0.1613,
+      "step": 672
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.027808676307008e-06,
+      "loss": 0.134,
+      "step": 673
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.005561735261402e-06,
+      "loss": 0.0747,
+      "step": 674
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.9833147942157955e-06,
+      "loss": 0.1797,
+      "step": 675
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.961067853170189e-06,
+      "loss": 0.1408,
+      "step": 676
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.938820912124583e-06,
+      "loss": 0.0535,
+      "step": 677
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.9165739710789776e-06,
+      "loss": 0.0862,
+      "step": 678
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.8943270300333704e-06,
+      "loss": 0.0843,
+      "step": 679
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.872080088987764e-06,
+      "loss": 0.0879,
+      "step": 680
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.849833147942159e-06,
+      "loss": 0.1176,
+      "step": 681
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.8275862068965525e-06,
+      "loss": 0.1298,
+      "step": 682
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.805339265850945e-06,
+      "loss": 0.0729,
+      "step": 683
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.78309232480534e-06,
+      "loss": 0.1257,
+      "step": 684
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.760845383759734e-06,
+      "loss": 0.1531,
+      "step": 685
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.7385984427141274e-06,
+      "loss": 0.1339,
+      "step": 686
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.71635150166852e-06,
+      "loss": 0.0858,
+      "step": 687
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.694104560622915e-06,
+      "loss": 0.1735,
+      "step": 688
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.671857619577309e-06,
+      "loss": 0.1912,
+      "step": 689
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.649610678531702e-06,
+      "loss": 0.1782,
+      "step": 690
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.627363737486096e-06,
+      "loss": 0.0905,
+      "step": 691
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.60511679644049e-06,
+      "loss": 0.1193,
+      "step": 692
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.582869855394884e-06,
+      "loss": 0.1422,
+      "step": 693
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.560622914349277e-06,
+      "loss": 0.1194,
+      "step": 694
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.538375973303671e-06,
+      "loss": 0.1289,
+      "step": 695
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 0.1367,
+      "step": 696
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.4938820912124585e-06,
+      "loss": 0.1569,
+      "step": 697
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.471635150166852e-06,
+      "loss": 0.0773,
+      "step": 698
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.449388209121246e-06,
+      "loss": 0.1164,
+      "step": 699
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.42714126807564e-06,
+      "loss": 0.1482,
+      "step": 700
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.4048943270300335e-06,
+      "loss": 0.0938,
+      "step": 701
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.382647385984427e-06,
+      "loss": 0.1266,
+      "step": 702
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.360400444938822e-06,
+      "loss": 0.1223,
+      "step": 703
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.338153503893215e-06,
+      "loss": 0.108,
+      "step": 704
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.3159065628476084e-06,
+      "loss": 0.0668,
+      "step": 705
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.293659621802003e-06,
+      "loss": 0.1183,
+      "step": 706
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.271412680756397e-06,
+      "loss": 0.1405,
+      "step": 707
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.24916573971079e-06,
+      "loss": 0.1986,
+      "step": 708
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.226918798665184e-06,
+      "loss": 0.1068,
+      "step": 709
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.204671857619578e-06,
+      "loss": 0.0765,
+      "step": 710
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.182424916573972e-06,
+      "loss": 0.1545,
+      "step": 711
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.160177975528365e-06,
+      "loss": 0.1233,
+      "step": 712
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.137931034482759e-06,
+      "loss": 0.0831,
+      "step": 713
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.115684093437153e-06,
+      "loss": 0.1246,
+      "step": 714
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.093437152391547e-06,
+      "loss": 0.1768,
+      "step": 715
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.07119021134594e-06,
+      "loss": 0.1872,
+      "step": 716
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.048943270300334e-06,
+      "loss": 0.0938,
+      "step": 717
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.026696329254728e-06,
+      "loss": 0.1394,
+      "step": 718
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.004449388209122e-06,
+      "loss": 0.1576,
+      "step": 719
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.982202447163515e-06,
+      "loss": 0.182,
+      "step": 720
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.959955506117909e-06,
+      "loss": 0.1301,
+      "step": 721
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.937708565072303e-06,
+      "loss": 0.0663,
+      "step": 722
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.9154616240266965e-06,
+      "loss": 0.0798,
+      "step": 723
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.89321468298109e-06,
+      "loss": 0.1204,
+      "step": 724
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 0.1104,
+      "step": 725
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.848720800889878e-06,
+      "loss": 0.1855,
+      "step": 726
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.8264738598442715e-06,
+      "loss": 0.0931,
+      "step": 727
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.804226918798665e-06,
+      "loss": 0.1468,
+      "step": 728
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.781979977753059e-06,
+      "loss": 0.1151,
+      "step": 729
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.759733036707453e-06,
+      "loss": 0.1115,
+      "step": 730
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.737486095661847e-06,
+      "loss": 0.0745,
+      "step": 731
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.71523915461624e-06,
+      "loss": 0.0635,
+      "step": 732
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6929922135706343e-06,
+      "loss": 0.0787,
+      "step": 733
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.670745272525028e-06,
+      "loss": 0.1109,
+      "step": 734
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.648498331479422e-06,
+      "loss": 0.1109,
+      "step": 735
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.626251390433816e-06,
+      "loss": 0.088,
+      "step": 736
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6040044493882093e-06,
+      "loss": 0.068,
+      "step": 737
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.581757508342603e-06,
+      "loss": 0.1372,
+      "step": 738
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.559510567296997e-06,
+      "loss": 0.1909,
+      "step": 739
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.537263626251391e-06,
+      "loss": 0.1157,
+      "step": 740
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.515016685205784e-06,
+      "loss": 0.1359,
+      "step": 741
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.4927697441601784e-06,
+      "loss": 0.1684,
+      "step": 742
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.470522803114572e-06,
+      "loss": 0.1227,
+      "step": 743
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 0.1014,
+      "step": 744
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.4260289210233596e-06,
+      "loss": 0.0783,
+      "step": 745
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.4037819799777533e-06,
+      "loss": 0.1381,
+      "step": 746
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.381535038932147e-06,
+      "loss": 0.1601,
+      "step": 747
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.359288097886541e-06,
+      "loss": 0.1679,
+      "step": 748
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.3370411568409345e-06,
+      "loss": 0.1359,
+      "step": 749
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.3147942157953282e-06,
+      "loss": 0.1307,
+      "step": 750
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.2925472747497224e-06,
+      "loss": 0.1425,
+      "step": 751
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.270300333704116e-06,
+      "loss": 0.1993,
+      "step": 752
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.2480533926585095e-06,
+      "loss": 0.2172,
+      "step": 753
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 0.0741,
+      "step": 754
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.2035595105672973e-06,
+      "loss": 0.1076,
+      "step": 755
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.181312569521691e-06,
+      "loss": 0.1392,
+      "step": 756
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.1590656284760844e-06,
+      "loss": 0.1159,
+      "step": 757
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.1368186874304786e-06,
+      "loss": 0.1352,
+      "step": 758
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.1145717463848723e-06,
+      "loss": 0.1347,
+      "step": 759
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0923248053392665e-06,
+      "loss": 0.1168,
+      "step": 760
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0700778642936598e-06,
+      "loss": 0.1058,
+      "step": 761
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0478309232480535e-06,
+      "loss": 0.125,
+      "step": 762
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0255839822024472e-06,
+      "loss": 0.1155,
+      "step": 763
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.0033370411568414e-06,
+      "loss": 0.1313,
+      "step": 764
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9810901001112347e-06,
+      "loss": 0.1777,
+      "step": 765
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9588431590656284e-06,
+      "loss": 0.1185,
+      "step": 766
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9365962180200226e-06,
+      "loss": 0.0861,
+      "step": 767
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9143492769744163e-06,
+      "loss": 0.1932,
+      "step": 768
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8921023359288105e-06,
+      "loss": 0.1722,
+      "step": 769
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.869855394883204e-06,
+      "loss": 0.1382,
+      "step": 770
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8476084538375975e-06,
+      "loss": 0.1256,
+      "step": 771
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8253615127919913e-06,
+      "loss": 0.0563,
+      "step": 772
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.8031145717463854e-06,
+      "loss": 0.1465,
+      "step": 773
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.7808676307007788e-06,
+      "loss": 0.1009,
+      "step": 774
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.7586206896551725e-06,
+      "loss": 0.1098,
+      "step": 775
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.7363737486095667e-06,
+      "loss": 0.0802,
+      "step": 776
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.7141268075639604e-06,
+      "loss": 0.1427,
+      "step": 777
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6918798665183537e-06,
+      "loss": 0.2159,
+      "step": 778
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.669632925472748e-06,
+      "loss": 0.1411,
+      "step": 779
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6473859844271416e-06,
+      "loss": 0.1076,
+      "step": 780
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6251390433815353e-06,
+      "loss": 0.0909,
+      "step": 781
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.6028921023359286e-06,
+      "loss": 0.1406,
+      "step": 782
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 0.1497,
+      "step": 783
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.5583982202447165e-06,
+      "loss": 0.2158,
+      "step": 784
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.5361512791991107e-06,
+      "loss": 0.1208,
+      "step": 785
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.513904338153504e-06,
+      "loss": 0.0859,
+      "step": 786
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4916573971078977e-06,
+      "loss": 0.1194,
+      "step": 787
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4694104560622915e-06,
+      "loss": 0.1062,
+      "step": 788
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4471635150166852e-06,
+      "loss": 0.1352,
+      "step": 789
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4249165739710794e-06,
+      "loss": 0.1686,
+      "step": 790
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4026696329254727e-06,
+      "loss": 0.1032,
+      "step": 791
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.380422691879867e-06,
+      "loss": 0.1714,
+      "step": 792
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.35817575083426e-06,
+      "loss": 0.0859,
+      "step": 793
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.3359288097886543e-06,
+      "loss": 0.1005,
+      "step": 794
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.313681868743048e-06,
+      "loss": 0.1104,
+      "step": 795
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.291434927697442e-06,
+      "loss": 0.1599,
+      "step": 796
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.2691879866518355e-06,
+      "loss": 0.1361,
+      "step": 797
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.2469410456062293e-06,
+      "loss": 0.146,
+      "step": 798
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.224694104560623e-06,
+      "loss": 0.1488,
+      "step": 799
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.2024471635150167e-06,
+      "loss": 0.1307,
+      "step": 800
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.180200222469411e-06,
+      "loss": 0.0982,
+      "step": 801
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.1579532814238042e-06,
+      "loss": 0.1512,
+      "step": 802
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.1357063403781984e-06,
+      "loss": 0.1409,
+      "step": 803
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.113459399332592e-06,
+      "loss": 0.1334,
+      "step": 804
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.091212458286986e-06,
+      "loss": 0.126,
+      "step": 805
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.0689655172413796e-06,
+      "loss": 0.1499,
+      "step": 806
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.0467185761957733e-06,
+      "loss": 0.1211,
+      "step": 807
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.024471635150167e-06,
+      "loss": 0.1054,
+      "step": 808
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.002224694104561e-06,
+      "loss": 0.2367,
+      "step": 809
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9799777530589545e-06,
+      "loss": 0.1302,
+      "step": 810
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9577308120133483e-06,
+      "loss": 0.1165,
+      "step": 811
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 0.1599,
+      "step": 812
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9132369299221357e-06,
+      "loss": 0.1346,
+      "step": 813
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8909899888765295e-06,
+      "loss": 0.1285,
+      "step": 814
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8687430478309234e-06,
+      "loss": 0.1147,
+      "step": 815
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8464961067853172e-06,
+      "loss": 0.0707,
+      "step": 816
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.824249165739711e-06,
+      "loss": 0.1767,
+      "step": 817
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8020022246941046e-06,
+      "loss": 0.149,
+      "step": 818
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.7797552836484986e-06,
+      "loss": 0.1751,
+      "step": 819
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.757508342602892e-06,
+      "loss": 0.0821,
+      "step": 820
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.735261401557286e-06,
+      "loss": 0.1253,
+      "step": 821
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.7130144605116798e-06,
+      "loss": 0.1504,
+      "step": 822
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6907675194660735e-06,
+      "loss": 0.1373,
+      "step": 823
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6685205784204673e-06,
+      "loss": 0.1826,
+      "step": 824
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6462736373748612e-06,
+      "loss": 0.1581,
+      "step": 825
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6240266963292547e-06,
+      "loss": 0.1321,
+      "step": 826
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6017797552836487e-06,
+      "loss": 0.1322,
+      "step": 827
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.5795328142380422e-06,
+      "loss": 0.1273,
+      "step": 828
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.5572858731924361e-06,
+      "loss": 0.111,
+      "step": 829
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.5350389321468299e-06,
+      "loss": 0.0755,
+      "step": 830
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.5127919911012236e-06,
+      "loss": 0.0928,
+      "step": 831
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4905450500556174e-06,
+      "loss": 0.0978,
+      "step": 832
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4682981090100113e-06,
+      "loss": 0.1427,
+      "step": 833
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4460511679644053e-06,
+      "loss": 0.0859,
+      "step": 834
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4238042269187988e-06,
+      "loss": 0.1575,
+      "step": 835
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4015572858731927e-06,
+      "loss": 0.1221,
+      "step": 836
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3793103448275862e-06,
+      "loss": 0.1623,
+      "step": 837
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3570634037819802e-06,
+      "loss": 0.1594,
+      "step": 838
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.334816462736374e-06,
+      "loss": 0.1627,
+      "step": 839
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3125695216907677e-06,
+      "loss": 0.1423,
+      "step": 840
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 0.1432,
+      "step": 841
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2680756395995554e-06,
+      "loss": 0.133,
+      "step": 842
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2458286985539489e-06,
+      "loss": 0.0962,
+      "step": 843
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2235817575083426e-06,
+      "loss": 0.1366,
+      "step": 844
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.2013348164627363e-06,
+      "loss": 0.1297,
+      "step": 845
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.17908787541713e-06,
+      "loss": 0.0794,
+      "step": 846
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.156840934371524e-06,
+      "loss": 0.0892,
+      "step": 847
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.1345939933259178e-06,
+      "loss": 0.141,
+      "step": 848
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.1123470522803115e-06,
+      "loss": 0.0575,
+      "step": 849
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0901001112347055e-06,
+      "loss": 0.1203,
+      "step": 850
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0678531701890992e-06,
+      "loss": 0.0925,
+      "step": 851
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.045606229143493e-06,
+      "loss": 0.1518,
+      "step": 852
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0233592880978867e-06,
+      "loss": 0.168,
+      "step": 853
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0011123470522804e-06,
+      "loss": 0.107,
+      "step": 854
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.788654060066741e-07,
+      "loss": 0.1435,
+      "step": 855
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.566184649610679e-07,
+      "loss": 0.1292,
+      "step": 856
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.343715239154617e-07,
+      "loss": 0.0748,
+      "step": 857
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.121245828698556e-07,
+      "loss": 0.1233,
+      "step": 858
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.898776418242493e-07,
+      "loss": 0.1081,
+      "step": 859
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.67630700778643e-07,
+      "loss": 0.1432,
+      "step": 860
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.453837597330368e-07,
+      "loss": 0.2187,
+      "step": 861
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.231368186874306e-07,
+      "loss": 0.1533,
+      "step": 862
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.008898776418243e-07,
+      "loss": 0.1128,
+      "step": 863
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.786429365962181e-07,
+      "loss": 0.1407,
+      "step": 864
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.563959955506118e-07,
+      "loss": 0.118,
+      "step": 865
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.341490545050057e-07,
+      "loss": 0.1084,
+      "step": 866
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.119021134593994e-07,
+      "loss": 0.0935,
+      "step": 867
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.896551724137931e-07,
+      "loss": 0.1299,
+      "step": 868
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.67408231368187e-07,
+      "loss": 0.1639,
+      "step": 869
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 0.1489,
+      "step": 870
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.229143492769744e-07,
+      "loss": 0.1268,
+      "step": 871
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.006674082313682e-07,
+      "loss": 0.1215,
+      "step": 872
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.78420467185762e-07,
+      "loss": 0.1177,
+      "step": 873
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.561735261401558e-07,
+      "loss": 0.1334,
+      "step": 874
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.339265850945496e-07,
+      "loss": 0.105,
+      "step": 875
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.116796440489433e-07,
+      "loss": 0.199,
+      "step": 876
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.894327030033371e-07,
+      "loss": 0.1091,
+      "step": 877
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.6718576195773085e-07,
+      "loss": 0.091,
+      "step": 878
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.4493882091212464e-07,
+      "loss": 0.1135,
+      "step": 879
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.226918798665184e-07,
+      "loss": 0.0831,
+      "step": 880
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.0044493882091217e-07,
+      "loss": 0.1089,
+      "step": 881
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.781979977753059e-07,
+      "loss": 0.1288,
+      "step": 882
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.559510567296997e-07,
+      "loss": 0.0766,
+      "step": 883
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.337041156840935e-07,
+      "loss": 0.1562,
+      "step": 884
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.114571746384872e-07,
+      "loss": 0.1062,
+      "step": 885
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.89210233592881e-07,
+      "loss": 0.076,
+      "step": 886
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.669632925472748e-07,
+      "loss": 0.0614,
+      "step": 887
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.4471635150166853e-07,
+      "loss": 0.1095,
+      "step": 888
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2246941045606232e-07,
+      "loss": 0.1276,
+      "step": 889
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.0022246941045608e-07,
+      "loss": 0.1209,
+      "step": 890
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.7797552836484985e-07,
+      "loss": 0.1023,
+      "step": 891
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.557285873192436e-07,
+      "loss": 0.1354,
+      "step": 892
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.334816462736374e-07,
+      "loss": 0.0939,
+      "step": 893
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.1123470522803116e-07,
+      "loss": 0.1586,
+      "step": 894
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.898776418242492e-08,
+      "loss": 0.1408,
+      "step": 895
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.67408231368187e-08,
+      "loss": 0.1975,
+      "step": 896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.449388209121246e-08,
+      "loss": 0.1062,
+      "step": 897
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.224694104560623e-08,
+      "loss": 0.1045,
+      "step": 898
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0,
+      "loss": 0.468,
+      "step": 899
+    },
+    {
+      "epoch": 1.0,
+      "step": 899,
+      "total_flos": 1.512441279525888e+16,
+      "train_loss": 0.13692307832326586,
+      "train_runtime": 237.2102,
+      "train_samples_per_second": 969.317,
+      "train_steps_per_second": 3.79
+    }
+  ],
+  "max_steps": 899,
+  "num_train_epochs": 1,
+  "total_flos": 1.512441279525888e+16,
+  "trial_name": null,
+  "trial_params": null
+}