diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7417 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9995943204868154,
+  "global_step": 1232,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 0.8354,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0810810810810812e-05,
+      "loss": 0.8354,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6216216216216218e-05,
+      "loss": 0.7924,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1621621621621624e-05,
+      "loss": 0.8145,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.702702702702703e-05,
+      "loss": 0.821,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2432432432432436e-05,
+      "loss": 0.8119,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.783783783783784e-05,
+      "loss": 0.7836,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324324324324325e-05,
+      "loss": 0.7965,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8648648648648654e-05,
+      "loss": 0.8035,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.405405405405406e-05,
+      "loss": 0.8183,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9459459459459466e-05,
+      "loss": 0.7871,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.486486486486487e-05,
+      "loss": 0.8303,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.027027027027028e-05,
+      "loss": 0.7942,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.567567567567568e-05,
+      "loss": 0.8151,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.108108108108109e-05,
+      "loss": 0.7705,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.64864864864865e-05,
+      "loss": 0.8099,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.18918918918919e-05,
+      "loss": 0.8121,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.729729729729731e-05,
+      "loss": 0.7903,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001027027027027027,
+      "loss": 0.8207,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00010810810810810812,
+      "loss": 0.8,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00011351351351351351,
+      "loss": 0.8475,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00011891891891891893,
+      "loss": 0.8314,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00012432432432432433,
+      "loss": 0.8182,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00012972972972972974,
+      "loss": 0.7838,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00013513513513513514,
+      "loss": 0.8174,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00014054054054054056,
+      "loss": 0.8183,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00014594594594594595,
+      "loss": 0.8249,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00015135135135135137,
+      "loss": 0.8258,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00015675675675675676,
+      "loss": 0.8111,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00016216216216216218,
+      "loss": 0.8134,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00016756756756756757,
+      "loss": 0.7814,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000172972972972973,
+      "loss": 0.8227,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00017837837837837839,
+      "loss": 0.8023,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001837837837837838,
+      "loss": 0.796,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001891891891891892,
+      "loss": 0.8162,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019459459459459462,
+      "loss": 0.8131,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002,
+      "loss": 0.8094,
+      "step": 37
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001999996544318651,
+      "loss": 0.759,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019999861772984878,
+      "loss": 0.8087,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019999688990111601,
+      "loss": 0.7988,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019999447095760847,
+      "loss": 0.8245,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019999136091604434,
+      "loss": 0.8331,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019998755979791827,
+      "loss": 0.8376,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001999830676295012,
+      "loss": 0.7836,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019997788444184002,
+      "loss": 0.7823,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019997201027075772,
+      "loss": 0.8507,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019996544515685281,
+      "loss": 0.8166,
+      "step": 47
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019995818914549912,
+      "loss": 0.813,
+      "step": 48
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019995024228684565,
+      "loss": 0.7917,
+      "step": 49
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.000199941604635816,
+      "loss": 0.8279,
+      "step": 50
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001999322762521081,
+      "loss": 0.804,
+      "step": 51
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019992225720019376,
+      "loss": 0.7767,
+      "step": 52
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019991154754931832,
+      "loss": 0.8241,
+      "step": 53
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019990014737350005,
+      "loss": 0.8239,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001998880567515297,
+      "loss": 0.8367,
+      "step": 55
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019987527576696996,
+      "loss": 0.8414,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019986180450815485,
+      "loss": 0.8095,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001998476430681891,
+      "loss": 0.8109,
+      "step": 58
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019983279154494758,
+      "loss": 0.8337,
+      "step": 59
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019981725004107454,
+      "loss": 0.8151,
+      "step": 60
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019980101866398298,
+      "loss": 0.8146,
+      "step": 61
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019978409752585376,
+      "loss": 0.8089,
+      "step": 62
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001997664867436351,
+      "loss": 0.7481,
+      "step": 63
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019974818643904142,
+      "loss": 0.8318,
+      "step": 64
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019972919673855283,
+      "loss": 0.8048,
+      "step": 65
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019970951777341397,
+      "loss": 0.8072,
+      "step": 66
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019968914967963337,
+      "loss": 0.8484,
+      "step": 67
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019966809259798227,
+      "loss": 0.8,
+      "step": 68
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001996463466739938,
+      "loss": 0.8482,
+      "step": 69
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000199623912057962,
+      "loss": 0.7783,
+      "step": 70
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019960078890494053,
+      "loss": 0.7821,
+      "step": 71
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019957697737474196,
+      "loss": 0.8258,
+      "step": 72
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019955247763193637,
+      "loss": 0.8314,
+      "step": 73
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001995272898458504,
+      "loss": 0.8001,
+      "step": 74
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019950141419056597,
+      "loss": 0.8312,
+      "step": 75
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001994748508449191,
+      "loss": 0.8579,
+      "step": 76
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019944759999249872,
+      "loss": 0.8145,
+      "step": 77
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019941966182164533,
+      "loss": 0.7798,
+      "step": 78
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019939103652544984,
+      "loss": 0.8302,
+      "step": 79
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019936172430175196,
+      "loss": 0.8314,
+      "step": 80
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019933172535313914,
+      "loss": 0.7843,
+      "step": 81
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.000199301039886945,
+      "loss": 0.8451,
+      "step": 82
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001992696681152479,
+      "loss": 0.8178,
+      "step": 83
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019923761025486956,
+      "loss": 0.8076,
+      "step": 84
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019920486652737348,
+      "loss": 0.8166,
+      "step": 85
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019917143715906345,
+      "loss": 0.8414,
+      "step": 86
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001991373223809819,
+      "loss": 0.8037,
+      "step": 87
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001991025224289085,
+      "loss": 0.8336,
+      "step": 88
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019906703754335832,
+      "loss": 0.8059,
+      "step": 89
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001990308679695802,
+      "loss": 0.7988,
+      "step": 90
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019899401395755533,
+      "loss": 0.8039,
+      "step": 91
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019895647576199506,
+      "loss": 0.8383,
+      "step": 92
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019891825364233947,
+      "loss": 0.8235,
+      "step": 93
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019887934786275553,
+      "loss": 0.8039,
+      "step": 94
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019883975869213516,
+      "loss": 0.7928,
+      "step": 95
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019879948640409348,
+      "loss": 0.8195,
+      "step": 96
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019875853127696692,
+      "loss": 0.8388,
+      "step": 97
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019871689359381116,
+      "loss": 0.7843,
+      "step": 98
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001986745736423994,
+      "loss": 0.8215,
+      "step": 99
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001986315717152201,
+      "loss": 0.7783,
+      "step": 100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001985878881094752,
+      "loss": 0.8044,
+      "step": 101
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019854352312707798,
+      "loss": 0.7881,
+      "step": 102
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019849847707465088,
+      "loss": 0.8362,
+      "step": 103
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0001984527502635235,
+      "loss": 0.8028,
+      "step": 104
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001984063430097305,
+      "loss": 0.7939,
+      "step": 105
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019835925563400914,
+      "loss": 0.8208,
+      "step": 106
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001983114884617974,
+      "loss": 0.802,
+      "step": 107
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019826304182323155,
+      "loss": 0.8117,
+      "step": 108
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019821391605314383,
+      "loss": 0.8286,
+      "step": 109
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019816411149106028,
+      "loss": 0.7909,
+      "step": 110
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019811362848119832,
+      "loss": 0.8253,
+      "step": 111
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001980624673724643,
+      "loss": 0.813,
+      "step": 112
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019801062851845122,
+      "loss": 0.8087,
+      "step": 113
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019795811227743619,
+      "loss": 0.797,
+      "step": 114
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.000197904919012378,
+      "loss": 0.8031,
+      "step": 115
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001978510490909146,
+      "loss": 0.774,
+      "step": 116
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019779650288536058,
+      "loss": 0.7912,
+      "step": 117
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019774128077270451,
+      "loss": 0.8259,
+      "step": 118
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019768538313460647,
+      "loss": 0.8164,
+      "step": 119
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019762881035739532,
+      "loss": 0.8253,
+      "step": 120
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019757156283206598,
+      "loss": 0.8177,
+      "step": 121
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019751364095427692,
+      "loss": 0.8345,
+      "step": 122
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019745504512434724,
+      "loss": 0.8218,
+      "step": 123
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019739577574725395,
+      "loss": 0.827,
+      "step": 124
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019733583323262925,
+      "loss": 0.791,
+      "step": 125
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019727521799475754,
+      "loss": 0.8005,
+      "step": 126
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019721393045257277,
+      "loss": 0.8125,
+      "step": 127
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019715197102965535,
+      "loss": 0.7679,
+      "step": 128
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001970893401542293,
+      "loss": 0.8662,
+      "step": 129
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019702603825915934,
+      "loss": 0.7922,
+      "step": 130
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019696206578194786,
+      "loss": 0.8084,
+      "step": 131
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019689742316473182,
+      "loss": 0.8376,
+      "step": 132
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019683211085427975,
+      "loss": 0.8121,
+      "step": 133
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001967661293019888,
+      "loss": 0.7874,
+      "step": 134
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019669947896388136,
+      "loss": 0.7887,
+      "step": 135
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019663216030060208,
+      "loss": 0.8491,
+      "step": 136
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001965641737774147,
+      "loss": 0.8053,
+      "step": 137
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001964955198641987,
+      "loss": 0.7842,
+      "step": 138
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019642619903544623,
+      "loss": 0.7935,
+      "step": 139
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019635621177025857,
+      "loss": 0.7847,
+      "step": 140
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019628555855234322,
+      "loss": 0.8045,
+      "step": 141
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019621423987001014,
+      "loss": 0.8402,
+      "step": 142
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001961422562161686,
+      "loss": 0.8336,
+      "step": 143
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019606960808832375,
+      "loss": 0.7739,
+      "step": 144
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019599629598857315,
+      "loss": 0.776,
+      "step": 145
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019592232042360333,
+      "loss": 0.8221,
+      "step": 146
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019584768190468625,
+      "loss": 0.817,
+      "step": 147
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019577238094767573,
+      "loss": 0.8437,
+      "step": 148
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019569641807300406,
+      "loss": 0.8084,
+      "step": 149
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001956197938056782,
+      "loss": 0.8358,
+      "step": 150
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019554250867527624,
+      "loss": 0.7966,
+      "step": 151
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019546456321594376,
+      "loss": 0.8127,
+      "step": 152
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019538595796639007,
+      "loss": 0.8311,
+      "step": 153
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001953066934698846,
+      "loss": 0.8334,
+      "step": 154
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.000195226770274253,
+      "loss": 0.8005,
+      "step": 155
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019514618893187353,
+      "loss": 0.7808,
+      "step": 156
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019506494999967298,
+      "loss": 0.8005,
+      "step": 157
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019498305403912314,
+      "loss": 0.8355,
+      "step": 158
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001949005016162367,
+      "loss": 0.8444,
+      "step": 159
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019481729330156334,
+      "loss": 0.8192,
+      "step": 160
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019473342967018593,
+      "loss": 0.8237,
+      "step": 161
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019464891130171647,
+      "loss": 0.8369,
+      "step": 162
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019456373878029202,
+      "loss": 0.8277,
+      "step": 163
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001944779126945708,
+      "loss": 0.8216,
+      "step": 164
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019439143363772802,
+      "loss": 0.7889,
+      "step": 165
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019430430220745176,
+      "loss": 0.7941,
+      "step": 166
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.000194216519005939,
+      "loss": 0.8167,
+      "step": 167
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019412808463989121,
+      "loss": 0.8245,
+      "step": 168
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019403899972051045,
+      "loss": 0.8126,
+      "step": 169
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019394926486349486,
+      "loss": 0.7887,
+      "step": 170
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001938588806890346,
+      "loss": 0.7763,
+      "step": 171
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019376784782180746,
+      "loss": 0.7994,
+      "step": 172
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019367616689097467,
+      "loss": 0.8273,
+      "step": 173
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019358383853017628,
+      "loss": 0.7773,
+      "step": 174
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019349086337752715,
+      "loss": 0.8255,
+      "step": 175
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019339724207561231,
+      "loss": 0.7879,
+      "step": 176
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019330297527148246,
+      "loss": 0.8149,
+      "step": 177
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019320806361664973,
+      "loss": 0.8388,
+      "step": 178
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.000193112507767083,
+      "loss": 0.7944,
+      "step": 179
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019301630838320332,
+      "loss": 0.7852,
+      "step": 180
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019291946612987962,
+      "loss": 0.8339,
+      "step": 181
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001928219816764238,
+      "loss": 0.8043,
+      "step": 182
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019272385569658627,
+      "loss": 0.8234,
+      "step": 183
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019262508886855125,
+      "loss": 0.7766,
+      "step": 184
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019252568187493215,
+      "loss": 0.8006,
+      "step": 185
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019242563540276675,
+      "loss": 0.8522,
+      "step": 186
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019232495014351246,
+      "loss": 0.8177,
+      "step": 187
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001922236267930417,
+      "loss": 0.8445,
+      "step": 188
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019212166605163682,
+      "loss": 0.812,
+      "step": 189
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019201906862398551,
+      "loss": 0.8058,
+      "step": 190
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019191583521917585,
+      "loss": 0.7863,
+      "step": 191
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019181196655069127,
+      "loss": 0.7901,
+      "step": 192
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001917074633364059,
+      "loss": 0.8032,
+      "step": 193
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019160232629857925,
+      "loss": 0.8499,
+      "step": 194
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001914965561638516,
+      "loss": 0.77,
+      "step": 195
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019139015366323868,
+      "loss": 0.8305,
+      "step": 196
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019128311953212678,
+      "loss": 0.8315,
+      "step": 197
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001911754545102676,
+      "loss": 0.8202,
+      "step": 198
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019106715934177315,
+      "loss": 0.8381,
+      "step": 199
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019095823477511062,
+      "loss": 0.8058,
+      "step": 200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019084868156309717,
+      "loss": 0.7723,
+      "step": 201
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019073850046289484,
+      "loss": 0.8441,
+      "step": 202
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001906276922360051,
+      "loss": 0.8049,
+      "step": 203
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001905162576482639,
+      "loss": 0.8108,
+      "step": 204
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.000190404197469836,
+      "loss": 0.8261,
+      "step": 205
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00019029151247520998,
+      "loss": 0.8247,
+      "step": 206
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001901782034431927,
+      "loss": 0.8483,
+      "step": 207
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00019006427115690397,
+      "loss": 0.8422,
+      "step": 208
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00018994971640377112,
+      "loss": 0.812,
+      "step": 209
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00018983453997552368,
+      "loss": 0.8223,
+      "step": 210
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00018971874266818762,
+      "loss": 0.8143,
+      "step": 211
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00018960232528208022,
+      "loss": 0.8238,
+      "step": 212
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001894852886218042,
+      "loss": 0.8415,
+      "step": 213
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00018936763349624238,
+      "loss": 0.8349,
+      "step": 214
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.000189249360718552,
+      "loss": 0.8235,
+      "step": 215
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018913047110615916,
+      "loss": 0.822,
+      "step": 216
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018901096548075305,
+      "loss": 0.7927,
+      "step": 217
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018889084466828043,
+      "loss": 0.809,
+      "step": 218
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018877010949893975,
+      "loss": 0.8253,
+      "step": 219
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018864876080717561,
+      "loss": 0.8112,
+      "step": 220
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001885267994316728,
+      "loss": 0.8108,
+      "step": 221
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018840422621535066,
+      "loss": 0.8238,
+      "step": 222
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001882810420053571,
+      "loss": 0.8151,
+      "step": 223
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018815724765306287,
+      "loss": 0.8397,
+      "step": 224
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018803284401405564,
+      "loss": 0.8134,
+      "step": 225
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018790783194813412,
+      "loss": 0.8182,
+      "step": 226
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018778221231930203,
+      "loss": 0.8279,
+      "step": 227
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00018765598599576215,
+      "loss": 0.8235,
+      "step": 228
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018752915384991043,
+      "loss": 0.7918,
+      "step": 229
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018740171675832982,
+      "loss": 0.8324,
+      "step": 230
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018727367560178424,
+      "loss": 0.8109,
+      "step": 231
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001871450312652126,
+      "loss": 0.8132,
+      "step": 232
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001870157846377226,
+      "loss": 0.8024,
+      "step": 233
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001868859366125845,
+      "loss": 0.7999,
+      "step": 234
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001867554880872251,
+      "loss": 0.8529,
+      "step": 235
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001866244399632215,
+      "loss": 0.8723,
+      "step": 236
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018649279314629483,
+      "loss": 0.788,
+      "step": 237
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018636054854630395,
+      "loss": 0.8453,
+      "step": 238
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018622770707723928,
+      "loss": 0.8267,
+      "step": 239
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018609426965721638,
+      "loss": 0.7986,
+      "step": 240
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018596023720846968,
+      "loss": 0.7698,
+      "step": 241
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018582561065734604,
+      "loss": 0.826,
+      "step": 242
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018569039093429835,
+      "loss": 0.8124,
+      "step": 243
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001855545789738792,
+      "loss": 0.7951,
+      "step": 244
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018541817571473429,
+      "loss": 0.8107,
+      "step": 245
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018528118209959599,
+      "loss": 0.8023,
+      "step": 246
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001851435990752769,
+      "loss": 0.804,
+      "step": 247
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018500542759266323,
+      "loss": 0.8005,
+      "step": 248
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018486666860670817,
+      "loss": 0.8283,
+      "step": 249
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001847273230764254,
+      "loss": 0.8086,
+      "step": 250
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001845873919648824,
+      "loss": 0.7998,
+      "step": 251
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00018444687623919386,
+      "loss": 0.8127,
+      "step": 252
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018430577687051494,
+      "loss": 0.8219,
+      "step": 253
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018416409483403454,
+      "loss": 0.795,
+      "step": 254
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018402183110896857,
+      "loss": 0.8293,
+      "step": 255
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001838789866785533,
+      "loss": 0.8239,
+      "step": 256
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001837355625300383,
+      "loss": 0.8292,
+      "step": 257
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018359155965468,
+      "loss": 0.7951,
+      "step": 258
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018344697904773444,
+      "loss": 0.8206,
+      "step": 259
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001833018217084506,
+      "loss": 0.8296,
+      "step": 260
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018315608864006354,
+      "loss": 0.8055,
+      "step": 261
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018300978084978735,
+      "loss": 0.8238,
+      "step": 262
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018286289934880826,
+      "loss": 0.825,
+      "step": 263
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00018271544515227755,
+      "loss": 0.7884,
+      "step": 264
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018256741927930462,
+      "loss": 0.7869,
+      "step": 265
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001824188227529501,
+      "loss": 0.8214,
+      "step": 266
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018226965660021836,
+      "loss": 0.802,
+      "step": 267
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001821199218520508,
+      "loss": 0.8181,
+      "step": 268
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018196961954331862,
+      "loss": 0.8012,
+      "step": 269
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018181875071281555,
+      "loss": 0.7895,
+      "step": 270
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018166731640325084,
+      "loss": 0.8053,
+      "step": 271
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018151531766124186,
+      "loss": 0.7856,
+      "step": 272
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001813627555373071,
+      "loss": 0.8244,
+      "step": 273
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018120963108585872,
+      "loss": 0.8663,
+      "step": 274
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001810559453651954,
+      "loss": 0.7725,
+      "step": 275
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 0.8087,
+      "step": 276
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00018074689436880644,
+      "loss": 0.8206,
+      "step": 277
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018059153122904442,
+      "loss": 0.8277,
+      "step": 278
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018043561109197967,
+      "loss": 0.8269,
+      "step": 279
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018027913503523287,
+      "loss": 0.7978,
+      "step": 280
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00018012210414026676,
+      "loss": 0.817,
+      "step": 281
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001799645194923788,
+      "loss": 0.804,
+      "step": 282
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00017980638218069366,
+      "loss": 0.8329,
+      "step": 283
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00017964769329815568,
+      "loss": 0.8,
+      "step": 284
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00017948845394152123,
+      "loss": 0.81,
+      "step": 285
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001793286652113513,
+      "loss": 0.8142,
+      "step": 286
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00017916832821200375,
+      "loss": 0.8112,
+      "step": 287
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00017900744405162578,
+      "loss": 0.8561,
+      "step": 288
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00017884601384214607,
+      "loss": 0.8109,
+      "step": 289
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00017868403869926748,
+      "loss": 0.8078,
+      "step": 290
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00017852151974245882,
+      "loss": 0.8065,
+      "step": 291
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00017835845809494768,
+      "loss": 0.7977,
+      "step": 292
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001781948548837122,
+      "loss": 0.8435,
+      "step": 293
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001780307112394735,
+      "loss": 0.8257,
+      "step": 294
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00017786602829668782,
+      "loss": 0.82,
+      "step": 295
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00017770080719353874,
+      "loss": 0.8379,
+      "step": 296
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00017753504907192923,
+      "loss": 0.8252,
+      "step": 297
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00017736875507747379,
+      "loss": 0.8051,
+      "step": 298
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001772019263594905,
+      "loss": 0.824,
+      "step": 299
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001770345640709932,
+      "loss": 0.7962,
+      "step": 300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001768666693686833,
+      "loss": 0.8196,
+      "step": 301
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00017669824341294202,
+      "loss": 0.7996,
+      "step": 302
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001765292873678222,
+      "loss": 0.8398,
+      "step": 303
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001763598024010404,
+      "loss": 0.8654,
+      "step": 304
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00017618978968396865,
+      "loss": 0.8273,
+      "step": 305
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00017601925039162648,
+      "loss": 0.8495,
+      "step": 306
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00017584818570267284,
+      "loss": 0.8057,
+      "step": 307
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00017567659679939787,
+      "loss": 0.787,
+      "step": 308
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00017550448486771466,
+      "loss": 0.8078,
+      "step": 309
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001753318510971512,
+      "loss": 0.8258,
+      "step": 310
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00017515869668084208,
+      "loss": 0.8385,
+      "step": 311
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001749850228155203,
+      "loss": 0.8271,
+      "step": 312
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00017481083070150895,
+      "loss": 0.8354,
+      "step": 313
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001746361215427129,
+      "loss": 0.8115,
+      "step": 314
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00017446089654661048,
+      "loss": 0.8236,
+      "step": 315
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001742851569242452,
+      "loss": 0.8361,
+      "step": 316
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00017410890389021736,
+      "loss": 0.8145,
+      "step": 317
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001739321386626756,
+      "loss": 0.82,
+      "step": 318
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00017375486246330848,
+      "loss": 0.8335,
+      "step": 319
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00017357707651733618,
+      "loss": 0.8257,
+      "step": 320
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00017339878205350177,
+      "loss": 0.8041,
+      "step": 321
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.000173219980304063,
+      "loss": 0.8074,
+      "step": 322
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00017304067250478364,
+      "loss": 0.8282,
+      "step": 323
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001728608598949249,
+      "loss": 0.78,
+      "step": 324
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00017268054371723693,
+      "loss": 0.8572,
+      "step": 325
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00017249972521795023,
+      "loss": 0.8519,
+      "step": 326
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001723184056467671,
+      "loss": 0.826,
+      "step": 327
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00017213658625685275,
+      "loss": 0.7963,
+      "step": 328
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00017195426830482705,
+      "loss": 0.8229,
+      "step": 329
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001717714530507554,
+      "loss": 0.8098,
+      "step": 330
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00017158814175814033,
+      "loss": 0.7983,
+      "step": 331
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00017140433569391275,
+      "loss": 0.8109,
+      "step": 332
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00017122003612842296,
+      "loss": 0.8054,
+      "step": 333
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001710352443354321,
+      "loss": 0.8151,
+      "step": 334
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00017084996159210332,
+      "loss": 0.8103,
+      "step": 335
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00017066418917899284,
+      "loss": 0.8166,
+      "step": 336
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001704779283800412,
+      "loss": 0.837,
+      "step": 337
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00017029118048256435,
+      "loss": 0.8088,
+      "step": 338
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001701039467772447,
+      "loss": 0.8175,
+      "step": 339
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001699162285581223,
+      "loss": 0.8159,
+      "step": 340
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00016972802712258585,
+      "loss": 0.7862,
+      "step": 341
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00016953934377136377,
+      "loss": 0.8378,
+      "step": 342
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00016935017980851508,
+      "loss": 0.803,
+      "step": 343
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001691605365414206,
+      "loss": 0.834,
+      "step": 344
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00016897041528077372,
+      "loss": 0.8252,
+      "step": 345
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00016877981734057137,
+      "loss": 0.7935,
+      "step": 346
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00016858874403810509,
+      "loss": 0.8249,
+      "step": 347
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00016839719669395174,
+      "loss": 0.7923,
+      "step": 348
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00016820517663196455,
+      "loss": 0.8313,
+      "step": 349
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00016801268517926377,
+      "loss": 0.814,
+      "step": 350
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00016781972366622765,
+      "loss": 0.797,
+      "step": 351
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00016762629342648318,
+      "loss": 0.7787,
+      "step": 352
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00016743239579689699,
+      "loss": 0.8169,
+      "step": 353
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001672380321175658,
+      "loss": 0.8121,
+      "step": 354
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00016704320373180754,
+      "loss": 0.82,
+      "step": 355
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001668479119861519,
+      "loss": 0.8347,
+      "step": 356
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001666521582303309,
+      "loss": 0.8612,
+      "step": 357
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00016645594381726979,
+      "loss": 0.8683,
+      "step": 358
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001662592701030775,
+      "loss": 0.7905,
+      "step": 359
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00016606213844703744,
+      "loss": 0.8197,
+      "step": 360
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00016586455021159799,
+      "loss": 0.8564,
+      "step": 361
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00016566650676236305,
+      "loss": 0.7834,
+      "step": 362
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001654680094680828,
+      "loss": 0.8129,
+      "step": 363
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00016526905970064396,
+      "loss": 0.7988,
+      "step": 364
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00016506965883506065,
+      "loss": 0.8237,
+      "step": 365
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00016486980824946445,
+      "loss": 0.8235,
+      "step": 366
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001646695093250953,
+      "loss": 0.808,
+      "step": 367
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00016446876344629172,
+      "loss": 0.8048,
+      "step": 368
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00016426757200048125,
+      "loss": 0.8149,
+      "step": 369
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00016406593637817101,
+      "loss": 0.8105,
+      "step": 370
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00016386385797293783,
+      "loss": 0.8352,
+      "step": 371
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00016366133818141893,
+      "loss": 0.8293,
+      "step": 372
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.000163458378403302,
+      "loss": 0.8103,
+      "step": 373
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00016325498004131575,
+      "loss": 0.8462,
+      "step": 374
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00016305114450121993,
+      "loss": 0.8148,
+      "step": 375
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.000162846873191796,
+      "loss": 0.8143,
+      "step": 376
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00016264216752483697,
+      "loss": 0.8549,
+      "step": 377
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.000162437028915138,
+      "loss": 0.8198,
+      "step": 378
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00016223145878048643,
+      "loss": 0.8192,
+      "step": 379
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.000162025458541652,
+      "loss": 0.8664,
+      "step": 380
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00016181902962237706,
+      "loss": 0.8031,
+      "step": 381
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001616121734493668,
+      "loss": 0.8072,
+      "step": 382
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00016140489145227918,
+      "loss": 0.8022,
+      "step": 383
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00016119718506371535,
+      "loss": 0.8285,
+      "step": 384
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00016098905571920944,
+      "loss": 0.8052,
+      "step": 385
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00016078050485721882,
+      "loss": 0.7821,
+      "step": 386
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00016057153391911422,
+      "loss": 0.7996,
+      "step": 387
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001603621443491695,
+      "loss": 0.7942,
+      "step": 388
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00016015233759455194,
+      "loss": 0.7925,
+      "step": 389
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00015994211510531216,
+      "loss": 0.7584,
+      "step": 390
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00015973147833437401,
+      "loss": 0.7814,
+      "step": 391
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001595204287375246,
+      "loss": 0.823,
+      "step": 392
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00015930896777340426,
+      "loss": 0.8318,
+      "step": 393
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001590970969034964,
+      "loss": 0.7857,
+      "step": 394
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001588848175921174,
+      "loss": 0.811,
+      "step": 395
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00015867213130640667,
+      "loss": 0.8082,
+      "step": 396
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00015845903951631623,
+      "loss": 0.7744,
+      "step": 397
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00015824554369460068,
+      "loss": 0.8362,
+      "step": 398
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00015803164531680714,
+      "loss": 0.7997,
+      "step": 399
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00015781734586126485,
+      "loss": 0.8208,
+      "step": 400
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00015760264680907506,
+      "loss": 0.8434,
+      "step": 401
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00015738754964410084,
+      "loss": 0.828,
+      "step": 402
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00015717205585295666,
+      "loss": 0.7948,
+      "step": 403
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00015695616692499833,
+      "loss": 0.7998,
+      "step": 404
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00015673988435231247,
+      "loss": 0.8106,
+      "step": 405
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001565232096297064,
+      "loss": 0.8396,
+      "step": 406
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00015630614425469775,
+      "loss": 0.8232,
+      "step": 407
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00015608868972750403,
+      "loss": 0.799,
+      "step": 408
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00015587084755103235,
+      "loss": 0.8134,
+      "step": 409
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00015565261923086902,
+      "loss": 0.8405,
+      "step": 410
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00015543400627526907,
+      "loss": 0.8387,
+      "step": 411
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00015521501019514597,
+      "loss": 0.7974,
+      "step": 412
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00015499563250406104,
+      "loss": 0.7667,
+      "step": 413
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00015477587471821304,
+      "loss": 0.8121,
+      "step": 414
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00015455573835642777,
+      "loss": 0.7685,
+      "step": 415
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00015433522494014753,
+      "loss": 0.8405,
+      "step": 416
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00015411433599342038,
+      "loss": 0.7916,
+      "step": 417
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00015389307304289007,
+      "loss": 0.8136,
+      "step": 418
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00015367143761778503,
+      "loss": 0.8073,
+      "step": 419
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001534494312499081,
+      "loss": 0.8033,
+      "step": 420
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001532270554736258,
+      "loss": 0.8292,
+      "step": 421
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00015300431182585777,
+      "loss": 0.8364,
+      "step": 422
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00015278120184606617,
+      "loss": 0.8061,
+      "step": 423
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.000152557727076245,
+      "loss": 0.8205,
+      "step": 424
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00015233388906090939,
+      "loss": 0.8161,
+      "step": 425
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001521096893470851,
+      "loss": 0.8355,
+      "step": 426
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00015188512948429765,
+      "loss": 0.8043,
+      "step": 427
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00015166021102456172,
+      "loss": 0.79,
+      "step": 428
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00015143493552237033,
+      "loss": 0.8219,
+      "step": 429
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001512093045346842,
+      "loss": 0.8151,
+      "step": 430
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00015098331962092097,
+      "loss": 0.8071,
+      "step": 431
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00015075698234294423,
+      "loss": 0.8061,
+      "step": 432
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001505302942650531,
+      "loss": 0.8338,
+      "step": 433
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00015030325695397108,
+      "loss": 0.8383,
+      "step": 434
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00015007587197883533,
+      "loss": 0.8131,
+      "step": 435
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00014984814091118594,
+      "loss": 0.8425,
+      "step": 436
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00014962006532495488,
+      "loss": 0.8186,
+      "step": 437
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00014939164679645528,
+      "loss": 0.8263,
+      "step": 438
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001491628869043704,
+      "loss": 0.775,
+      "step": 439
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00014893378722974284,
+      "loss": 0.7937,
+      "step": 440
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001487043493559636,
+      "loss": 0.8143,
+      "step": 441
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00014847457486876097,
+      "loss": 0.8209,
+      "step": 442
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001482444653561898,
+      "loss": 0.8313,
+      "step": 443
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00014801402240862034,
+      "loss": 0.819,
+      "step": 444
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001477832476187275,
+      "loss": 0.8585,
+      "step": 445
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00014755214258147943,
+      "loss": 0.8552,
+      "step": 446
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00014732070889412693,
+      "loss": 0.7882,
+      "step": 447
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00014708894815619212,
+      "loss": 0.8031,
+      "step": 448
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001468568619694576,
+      "loss": 0.809,
+      "step": 449
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00014662445193795506,
+      "loss": 0.8292,
+      "step": 450
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001463917196679546,
+      "loss": 0.7759,
+      "step": 451
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00014615866676795334,
+      "loss": 0.8542,
+      "step": 452
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001459252948486644,
+      "loss": 0.7821,
+      "step": 453
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00014569160552300574,
+      "loss": 0.7906,
+      "step": 454
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00014545760040608904,
+      "loss": 0.8272,
+      "step": 455
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00014522328111520857,
+      "loss": 0.8078,
+      "step": 456
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00014498864926982996,
+      "loss": 0.8192,
+      "step": 457
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001447537064915789,
+      "loss": 0.81,
+      "step": 458
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00014451845440423025,
+      "loss": 0.7829,
+      "step": 459
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00014428289463369644,
+      "loss": 0.7969,
+      "step": 460
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00014404702880801648,
+      "loss": 0.8164,
+      "step": 461
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00014381085855734468,
+      "loss": 0.8304,
+      "step": 462
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00014357438551393928,
+      "loss": 0.8089,
+      "step": 463
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00014333761131215126,
+      "loss": 0.8389,
+      "step": 464
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.000143100537588413,
+      "loss": 0.8531,
+      "step": 465
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.000142863165981227,
+      "loss": 0.8122,
+      "step": 466
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001426254981311545,
+      "loss": 0.8037,
+      "step": 467
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00014238753568080422,
+      "loss": 0.8358,
+      "step": 468
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.000142149280274821,
+      "loss": 0.8059,
+      "step": 469
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00014191073355987436,
+      "loss": 0.7988,
+      "step": 470
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00014167189718464714,
+      "loss": 0.7869,
+      "step": 471
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00014143277279982414,
+      "loss": 0.7896,
+      "step": 472
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00014119336205808077,
+      "loss": 0.8211,
+      "step": 473
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00014095366661407148,
+      "loss": 0.808,
+      "step": 474
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001407136881244184,
+      "loss": 0.8142,
+      "step": 475
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001404734282476999,
+      "loss": 0.7964,
+      "step": 476
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00014023288864443916,
+      "loss": 0.8306,
+      "step": 477
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001399920709770926,
+      "loss": 0.8294,
+      "step": 478
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00013975097691003852,
+      "loss": 0.821,
+      "step": 479
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00013950960810956537,
+      "loss": 0.8275,
+      "step": 480
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00013926796624386052,
+      "loss": 0.7761,
+      "step": 481
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001390260529829986,
+      "loss": 0.8043,
+      "step": 482
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001387838699989298,
+      "loss": 0.7889,
+      "step": 483
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00013854141896546863,
+      "loss": 0.8187,
+      "step": 484
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00013829870155828209,
+      "loss": 0.8142,
+      "step": 485
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001380557194548782,
+      "loss": 0.8178,
+      "step": 486
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00013781247433459449,
+      "loss": 0.7844,
+      "step": 487
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00013756896787858608,
+      "loss": 0.8219,
+      "step": 488
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001373252017698145,
+      "loss": 0.7854,
+      "step": 489
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001370811776930357,
+      "loss": 0.825,
+      "step": 490
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001368368973347886,
+      "loss": 0.8396,
+      "step": 491
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001365923623833834,
+      "loss": 0.7876,
+      "step": 492
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00013634757452888973,
+      "loss": 0.835,
+      "step": 493
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00013610253546312532,
+      "loss": 0.8247,
+      "step": 494
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00013585724687964403,
+      "loss": 0.8519,
+      "step": 495
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00013561171047372418,
+      "loss": 0.8152,
+      "step": 496
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00013536592794235696,
+      "loss": 0.8266,
+      "step": 497
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001351199009842346,
+      "loss": 0.8171,
+      "step": 498
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001348736312997386,
+      "loss": 0.8051,
+      "step": 499
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001346271205909281,
+      "loss": 0.8377,
+      "step": 500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00013438037056152802,
+      "loss": 0.8342,
+      "step": 501
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00013413338291691726,
+      "loss": 0.8443,
+      "step": 502
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00013388615936411712,
+      "loss": 0.8484,
+      "step": 503
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00013363870161177914,
+      "loss": 0.8353,
+      "step": 504
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001333910113701736,
+      "loss": 0.7922,
+      "step": 505
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00013314309035117767,
+      "loss": 0.8235,
+      "step": 506
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00013289494026826336,
+      "loss": 0.8295,
+      "step": 507
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.000132646562836486,
+      "loss": 0.7794,
+      "step": 508
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00013239795977247197,
+      "loss": 0.8474,
+      "step": 509
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001321491327944073,
+      "loss": 0.8228,
+      "step": 510
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001319000836220255,
+      "loss": 0.8063,
+      "step": 511
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00013165081397659563,
+      "loss": 0.8242,
+      "step": 512
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001314013255809107,
+      "loss": 0.8092,
+      "step": 513
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001311516201592755,
+      "loss": 0.7921,
+      "step": 514
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 0.7826,
+      "step": 515
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001306515651428612,
+      "loss": 0.818,
+      "step": 516
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001304012190041437,
+      "loss": 0.8242,
+      "step": 517
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00013015066275157518,
+      "loss": 0.7797,
+      "step": 518
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001298998981168408,
+      "loss": 0.814,
+      "step": 519
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00012964892683306592,
+      "loss": 0.8119,
+      "step": 520
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001293977506348041,
+      "loss": 0.8061,
+      "step": 521
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00012914637125802512,
+      "loss": 0.8093,
+      "step": 522
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00012889479044010308,
+      "loss": 0.8105,
+      "step": 523
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001286430099198042,
+      "loss": 0.85,
+      "step": 524
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.000128391031437275,
+      "loss": 0.7964,
+      "step": 525
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00012813885673403015,
+      "loss": 0.8182,
+      "step": 526
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00012788648755294055,
+      "loss": 0.8153,
+      "step": 527
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00012763392563822106,
+      "loss": 0.8491,
+      "step": 528
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00012738117273541875,
+      "loss": 0.8056,
+      "step": 529
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00012712823059140055,
+      "loss": 0.824,
+      "step": 530
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00012687510095434138,
+      "loss": 0.7973,
+      "step": 531
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00012662178557371198,
+      "loss": 0.8355,
+      "step": 532
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00012636828620026682,
+      "loss": 0.7912,
+      "step": 533
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.000126114604586032,
+      "loss": 0.8361,
+      "step": 534
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00012586074248429317,
+      "loss": 0.8348,
+      "step": 535
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00012560670164958338,
+      "loss": 0.821,
+      "step": 536
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.000125352483837671,
+      "loss": 0.8252,
+      "step": 537
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00012509809080554754,
+      "loss": 0.8155,
+      "step": 538
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00012484352431141552,
+      "loss": 0.8226,
+      "step": 539
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00012458878611467622,
+      "loss": 0.8188,
+      "step": 540
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00012433387797591782,
+      "loss": 0.8256,
+      "step": 541
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00012407880165690287,
+      "loss": 0.8244,
+      "step": 542
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00012382355892055636,
+      "loss": 0.7991,
+      "step": 543
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00012356815153095337,
+      "loss": 0.8106,
+      "step": 544
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00012331258125330706,
+      "loss": 0.8132,
+      "step": 545
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00012305684985395625,
+      "loss": 0.7845,
+      "step": 546
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00012280095910035342,
+      "loss": 0.8172,
+      "step": 547
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00012254491076105242,
+      "loss": 0.7964,
+      "step": 548
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00012228870660569615,
+      "loss": 0.8265,
+      "step": 549
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00012203234840500446,
+      "loss": 0.8213,
+      "step": 550
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00012177583793076184,
+      "loss": 0.8096,
+      "step": 551
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00012151917695580523,
+      "loss": 0.8188,
+      "step": 552
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00012126236725401173,
+      "loss": 0.7981,
+      "step": 553
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00012100541060028631,
+      "loss": 0.8498,
+      "step": 554
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00012074830877054963,
+      "loss": 0.8086,
+      "step": 555
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00012049106354172566,
+      "loss": 0.8206,
+      "step": 556
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00012023367669172946,
+      "loss": 0.8054,
+      "step": 557
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011997614999945499,
+      "loss": 0.7903,
+      "step": 558
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001197184852447625,
+      "loss": 0.792,
+      "step": 559
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00011946068420846666,
+      "loss": 0.8213,
+      "step": 560
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001192027486723239,
+      "loss": 0.8361,
+      "step": 561
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001189446804190203,
+      "loss": 0.7927,
+      "step": 562
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011868648123215908,
+      "loss": 0.8034,
+      "step": 563
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011842815289624853,
+      "loss": 0.8437,
+      "step": 564
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011816969719668949,
+      "loss": 0.8226,
+      "step": 565
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011791111591976298,
+      "loss": 0.8052,
+      "step": 566
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011765241085261802,
+      "loss": 0.8103,
+      "step": 567
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001173935837832592,
+      "loss": 0.8624,
+      "step": 568
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011713463650053423,
+      "loss": 0.8515,
+      "step": 569
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011687557079412171,
+      "loss": 0.8271,
+      "step": 570
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011661638845451866,
+      "loss": 0.8075,
+      "step": 571
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011635709127302829,
+      "loss": 0.8006,
+      "step": 572
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00011609768104174743,
+      "loss": 0.8263,
+      "step": 573
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00011583815955355435,
+      "loss": 0.7716,
+      "step": 574
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00011557852860209608,
+      "loss": 0.7992,
+      "step": 575
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00011531878998177635,
+      "loss": 0.8176,
+      "step": 576
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00011505894548774294,
+      "loss": 0.7893,
+      "step": 577
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00011479899691587546,
+      "loss": 0.8309,
+      "step": 578
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00011453894606277271,
+      "loss": 0.8293,
+      "step": 579
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00011427879472574048,
+      "loss": 0.8123,
+      "step": 580
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.000114018544702779,
+      "loss": 0.8088,
+      "step": 581
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00011375819779257057,
+      "loss": 0.7775,
+      "step": 582
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00011349775579446716,
+      "loss": 0.8316,
+      "step": 583
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00011323722050847782,
+      "loss": 0.7937,
+      "step": 584
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00011297659373525645,
+      "loss": 0.8224,
+      "step": 585
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00011271587727608917,
+      "loss": 0.8126,
+      "step": 586
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00011245507293288204,
+      "loss": 0.8401,
+      "step": 587
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00011219418250814841,
+      "loss": 0.8292,
+      "step": 588
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001119332078049967,
+      "loss": 0.8056,
+      "step": 589
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00011167215062711767,
+      "loss": 0.7748,
+      "step": 590
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00011141101277877222,
+      "loss": 0.8164,
+      "step": 591
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00011114979606477866,
+      "loss": 0.8309,
+      "step": 592
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00011088850229050052,
+      "loss": 0.7974,
+      "step": 593
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00011062713326183383,
+      "loss": 0.7816,
+      "step": 594
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001103656907851947,
+      "loss": 0.7743,
+      "step": 595
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00011010417666750695,
+      "loss": 0.8423,
+      "step": 596
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00010984259271618947,
+      "loss": 0.7668,
+      "step": 597
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010958094073914384,
+      "loss": 0.8168,
+      "step": 598
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001093192225447418,
+      "loss": 0.82,
+      "step": 599
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010905743994181267,
+      "loss": 0.8123,
+      "step": 600
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010879559473963099,
+      "loss": 0.8262,
+      "step": 601
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010853368874790392,
+      "loss": 0.7848,
+      "step": 602
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010827172377675876,
+      "loss": 0.8465,
+      "step": 603
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010800970163673048,
+      "loss": 0.8079,
+      "step": 604
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010774762413874904,
+      "loss": 0.7951,
+      "step": 605
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010748549309412714,
+      "loss": 0.8003,
+      "step": 606
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010722331031454748,
+      "loss": 0.8285,
+      "step": 607
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010696107761205038,
+      "loss": 0.7596,
+      "step": 608
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010669879679902115,
+      "loss": 0.8473,
+      "step": 609
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00010643646968817757,
+      "loss": 0.8309,
+      "step": 610
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010617409809255748,
+      "loss": 0.7687,
+      "step": 611
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010591168382550616,
+      "loss": 0.7753,
+      "step": 612
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010564922870066376,
+      "loss": 0.7998,
+      "step": 613
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010538673453195285,
+      "loss": 0.7973,
+      "step": 614
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010512420313356581,
+      "loss": 0.8075,
+      "step": 615
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001048616363199524,
+      "loss": 0.7626,
+      "step": 616
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010459903590580706,
+      "loss": 0.8373,
+      "step": 617
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010433640370605652,
+      "loss": 0.8016,
+      "step": 618
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010407374153584716,
+      "loss": 0.8117,
+      "step": 619
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010381105121053255,
+      "loss": 0.8437,
+      "step": 620
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010354833454566071,
+      "loss": 0.8006,
+      "step": 621
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00010328559335696188,
+      "loss": 0.7918,
+      "step": 622
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010302282946033565,
+      "loss": 0.8137,
+      "step": 623
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010276004467183868,
+      "loss": 0.8402,
+      "step": 624
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010249724080767192,
+      "loss": 0.8324,
+      "step": 625
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010223441968416818,
+      "loss": 0.8024,
+      "step": 626
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010197158311777957,
+      "loss": 0.8088,
+      "step": 627
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010170873292506495,
+      "loss": 0.8423,
+      "step": 628
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010144587092267736,
+      "loss": 0.8435,
+      "step": 629
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001011829989273514,
+      "loss": 0.7603,
+      "step": 630
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010092011875589076,
+      "loss": 0.7918,
+      "step": 631
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010065723222515566,
+      "loss": 0.799,
+      "step": 632
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010039434115205034,
+      "loss": 0.7851,
+      "step": 633
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00010013144735351023,
+      "loss": 0.8155,
+      "step": 634
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.986855264648982e-05,
+      "loss": 0.7916,
+      "step": 635
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.960565884794973e-05,
+      "loss": 0.8252,
+      "step": 636
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.934276777484436e-05,
+      "loss": 0.8079,
+      "step": 637
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.907988124410928e-05,
+      "loss": 0.793,
+      "step": 638
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.881700107264863e-05,
+      "loss": 0.7915,
+      "step": 639
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.855412907732266e-05,
+      "loss": 0.7832,
+      "step": 640
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.829126707493504e-05,
+      "loss": 0.7901,
+      "step": 641
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.802841688222043e-05,
+      "loss": 0.8127,
+      "step": 642
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.776558031583186e-05,
+      "loss": 0.791,
+      "step": 643
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.750275919232812e-05,
+      "loss": 0.807,
+      "step": 644
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.723995532816134e-05,
+      "loss": 0.7963,
+      "step": 645
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.697717053966436e-05,
+      "loss": 0.8481,
+      "step": 646
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.671440664303814e-05,
+      "loss": 0.8108,
+      "step": 647
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.645166545433931e-05,
+      "loss": 0.8321,
+      "step": 648
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.618894878946748e-05,
+      "loss": 0.8104,
+      "step": 649
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.592625846415283e-05,
+      "loss": 0.81,
+      "step": 650
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.566359629394348e-05,
+      "loss": 0.795,
+      "step": 651
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.540096409419296e-05,
+      "loss": 0.8111,
+      "step": 652
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.513836368004763e-05,
+      "loss": 0.834,
+      "step": 653
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.487579686643422e-05,
+      "loss": 0.7868,
+      "step": 654
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.461326546804719e-05,
+      "loss": 0.8383,
+      "step": 655
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.435077129933627e-05,
+      "loss": 0.8138,
+      "step": 656
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.408831617449385e-05,
+      "loss": 0.8056,
+      "step": 657
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.382590190744252e-05,
+      "loss": 0.8241,
+      "step": 658
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.356353031182244e-05,
+      "loss": 0.8276,
+      "step": 659
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.33012032009789e-05,
+      "loss": 0.804,
+      "step": 660
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.303892238794964e-05,
+      "loss": 0.7891,
+      "step": 661
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.277668968545253e-05,
+      "loss": 0.8058,
+      "step": 662
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.251450690587288e-05,
+      "loss": 0.8146,
+      "step": 663
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.225237586125099e-05,
+      "loss": 0.7977,
+      "step": 664
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.199029836326956e-05,
+      "loss": 0.8057,
+      "step": 665
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.172827622324123e-05,
+      "loss": 0.847,
+      "step": 666
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.146631125209607e-05,
+      "loss": 0.84,
+      "step": 667
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.120440526036903e-05,
+      "loss": 0.8186,
+      "step": 668
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.094256005818736e-05,
+      "loss": 0.8148,
+      "step": 669
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.068077745525823e-05,
+      "loss": 0.8082,
+      "step": 670
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.041905926085618e-05,
+      "loss": 0.7919,
+      "step": 671
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.015740728381054e-05,
+      "loss": 0.8318,
+      "step": 672
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.989582333249306e-05,
+      "loss": 0.8049,
+      "step": 673
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.963430921480533e-05,
+      "loss": 0.8111,
+      "step": 674
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.937286673816621e-05,
+      "loss": 0.8214,
+      "step": 675
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.911149770949948e-05,
+      "loss": 0.8034,
+      "step": 676
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.885020393522135e-05,
+      "loss": 0.7979,
+      "step": 677
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.858898722122782e-05,
+      "loss": 0.8224,
+      "step": 678
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.832784937288237e-05,
+      "loss": 0.8366,
+      "step": 679
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.806679219500333e-05,
+      "loss": 0.7683,
+      "step": 680
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.78058174918516e-05,
+      "loss": 0.7787,
+      "step": 681
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.754492706711798e-05,
+      "loss": 0.8001,
+      "step": 682
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.728412272391083e-05,
+      "loss": 0.8092,
+      "step": 683
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.702340626474355e-05,
+      "loss": 0.7945,
+      "step": 684
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.676277949152223e-05,
+      "loss": 0.8103,
+      "step": 685
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.650224420553288e-05,
+      "loss": 0.803,
+      "step": 686
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.624180220742946e-05,
+      "loss": 0.835,
+      "step": 687
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.598145529722104e-05,
+      "loss": 0.7927,
+      "step": 688
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.572120527425955e-05,
+      "loss": 0.8494,
+      "step": 689
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.546105393722731e-05,
+      "loss": 0.7974,
+      "step": 690
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.520100308412455e-05,
+      "loss": 0.7868,
+      "step": 691
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.494105451225704e-05,
+      "loss": 0.806,
+      "step": 692
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.46812100182237e-05,
+      "loss": 0.8497,
+      "step": 693
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.442147139790397e-05,
+      "loss": 0.7978,
+      "step": 694
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.416184044644571e-05,
+      "loss": 0.7948,
+      "step": 695
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.390231895825258e-05,
+      "loss": 0.8258,
+      "step": 696
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.364290872697173e-05,
+      "loss": 0.7766,
+      "step": 697
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.338361154548137e-05,
+      "loss": 0.8083,
+      "step": 698
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.312442920587833e-05,
+      "loss": 0.7764,
+      "step": 699
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.286536349946581e-05,
+      "loss": 0.8153,
+      "step": 700
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.260641621674082e-05,
+      "loss": 0.7889,
+      "step": 701
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.234758914738199e-05,
+      "loss": 0.8295,
+      "step": 702
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.208888408023704e-05,
+      "loss": 0.8362,
+      "step": 703
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.183030280331054e-05,
+      "loss": 0.8143,
+      "step": 704
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.157184710375149e-05,
+      "loss": 0.8415,
+      "step": 705
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.131351876784094e-05,
+      "loss": 0.7935,
+      "step": 706
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.105531958097972e-05,
+      "loss": 0.8125,
+      "step": 707
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.07972513276761e-05,
+      "loss": 0.7786,
+      "step": 708
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.053931579153333e-05,
+      "loss": 0.7615,
+      "step": 709
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.028151475523753e-05,
+      "loss": 0.8124,
+      "step": 710
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.002385000054506e-05,
+      "loss": 0.8166,
+      "step": 711
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.976632330827056e-05,
+      "loss": 0.8199,
+      "step": 712
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.950893645827437e-05,
+      "loss": 0.849,
+      "step": 713
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.925169122945039e-05,
+      "loss": 0.8452,
+      "step": 714
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.89945893997137e-05,
+      "loss": 0.7745,
+      "step": 715
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.873763274598827e-05,
+      "loss": 0.7951,
+      "step": 716
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.848082304419478e-05,
+      "loss": 0.7904,
+      "step": 717
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.822416206923816e-05,
+      "loss": 0.7961,
+      "step": 718
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.796765159499558e-05,
+      "loss": 0.7937,
+      "step": 719
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.771129339430389e-05,
+      "loss": 0.7796,
+      "step": 720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.74550892389476e-05,
+      "loss": 0.8198,
+      "step": 721
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.719904089964658e-05,
+      "loss": 0.7931,
+      "step": 722
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.694315014604376e-05,
+      "loss": 0.8085,
+      "step": 723
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.668741874669297e-05,
+      "loss": 0.7916,
+      "step": 724
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.643184846904662e-05,
+      "loss": 0.7898,
+      "step": 725
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.617644107944366e-05,
+      "loss": 0.7726,
+      "step": 726
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.592119834309715e-05,
+      "loss": 0.8049,
+      "step": 727
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.56661220240822e-05,
+      "loss": 0.8017,
+      "step": 728
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.541121388532379e-05,
+      "loss": 0.8183,
+      "step": 729
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.515647568858452e-05,
+      "loss": 0.8198,
+      "step": 730
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.490190919445248e-05,
+      "loss": 0.8118,
+      "step": 731
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.464751616232902e-05,
+      "loss": 0.7877,
+      "step": 732
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.439329835041661e-05,
+      "loss": 0.8125,
+      "step": 733
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.413925751570684e-05,
+      "loss": 0.8046,
+      "step": 734
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.388539541396802e-05,
+      "loss": 0.812,
+      "step": 735
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.363171379973322e-05,
+      "loss": 0.8285,
+      "step": 736
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.337821442628805e-05,
+      "loss": 0.8057,
+      "step": 737
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.312489904565864e-05,
+      "loss": 0.807,
+      "step": 738
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.287176940859948e-05,
+      "loss": 0.813,
+      "step": 739
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.261882726458126e-05,
+      "loss": 0.8012,
+      "step": 740
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.236607436177893e-05,
+      "loss": 0.8265,
+      "step": 741
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.211351244705946e-05,
+      "loss": 0.7669,
+      "step": 742
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.186114326596984e-05,
+      "loss": 0.8134,
+      "step": 743
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.160896856272506e-05,
+      "loss": 0.8125,
+      "step": 744
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.135699008019585e-05,
+      "loss": 0.8046,
+      "step": 745
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.110520955989697e-05,
+      "loss": 0.7902,
+      "step": 746
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.08536287419749e-05,
+      "loss": 0.775,
+      "step": 747
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.060224936519592e-05,
+      "loss": 0.8155,
+      "step": 748
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.035107316693408e-05,
+      "loss": 0.8077,
+      "step": 749
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.010010188315919e-05,
+      "loss": 0.8634,
+      "step": 750
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.984933724842481e-05,
+      "loss": 0.7546,
+      "step": 751
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.959878099585635e-05,
+      "loss": 0.814,
+      "step": 752
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.934843485713882e-05,
+      "loss": 0.8141,
+      "step": 753
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 0.8381,
+      "step": 754
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.884837984072451e-05,
+      "loss": 0.8395,
+      "step": 755
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.85986744190893e-05,
+      "loss": 0.8538,
+      "step": 756
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.834918602340438e-05,
+      "loss": 0.7836,
+      "step": 757
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.809991637797455e-05,
+      "loss": 0.8477,
+      "step": 758
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.78508672055927e-05,
+      "loss": 0.815,
+      "step": 759
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.760204022752805e-05,
+      "loss": 0.8176,
+      "step": 760
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.735343716351405e-05,
+      "loss": 0.7783,
+      "step": 761
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.710505973173664e-05,
+      "loss": 0.8209,
+      "step": 762
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.685690964882236e-05,
+      "loss": 0.809,
+      "step": 763
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.660898862982641e-05,
+      "loss": 0.7956,
+      "step": 764
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.63612983882209e-05,
+      "loss": 0.8222,
+      "step": 765
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.611384063588291e-05,
+      "loss": 0.8217,
+      "step": 766
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.586661708308272e-05,
+      "loss": 0.825,
+      "step": 767
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.5619629438472e-05,
+      "loss": 0.7832,
+      "step": 768
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.537287940907194e-05,
+      "loss": 0.8176,
+      "step": 769
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.512636870026143e-05,
+      "loss": 0.8192,
+      "step": 770
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.488009901576544e-05,
+      "loss": 0.7934,
+      "step": 771
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.463407205764305e-05,
+      "loss": 0.8246,
+      "step": 772
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.438828952627582e-05,
+      "loss": 0.7871,
+      "step": 773
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.414275312035598e-05,
+      "loss": 0.8346,
+      "step": 774
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.389746453687468e-05,
+      "loss": 0.7983,
+      "step": 775
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.365242547111027e-05,
+      "loss": 0.8078,
+      "step": 776
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.340763761661665e-05,
+      "loss": 0.7919,
+      "step": 777
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.316310266521142e-05,
+      "loss": 0.7963,
+      "step": 778
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.291882230696432e-05,
+      "loss": 0.7612,
+      "step": 779
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.267479823018553e-05,
+      "loss": 0.7916,
+      "step": 780
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.243103212141394e-05,
+      "loss": 0.8153,
+      "step": 781
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.218752566540554e-05,
+      "loss": 0.7633,
+      "step": 782
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.19442805451218e-05,
+      "loss": 0.7785,
+      "step": 783
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.170129844171792e-05,
+      "loss": 0.7678,
+      "step": 784
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.145858103453141e-05,
+      "loss": 0.8018,
+      "step": 785
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.121613000107023e-05,
+      "loss": 0.8144,
+      "step": 786
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.097394701700145e-05,
+      "loss": 0.7911,
+      "step": 787
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.073203375613949e-05,
+      "loss": 0.8215,
+      "step": 788
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.049039189043466e-05,
+      "loss": 0.8043,
+      "step": 789
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.024902308996152e-05,
+      "loss": 0.7887,
+      "step": 790
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.000792902290741e-05,
+      "loss": 0.7964,
+      "step": 791
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.976711135556086e-05,
+      "loss": 0.7591,
+      "step": 792
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.952657175230012e-05,
+      "loss": 0.8198,
+      "step": 793
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.9286311875581656e-05,
+      "loss": 0.8087,
+      "step": 794
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.904633338592857e-05,
+      "loss": 0.812,
+      "step": 795
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.880663794191925e-05,
+      "loss": 0.8213,
+      "step": 796
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.8567227200175865e-05,
+      "loss": 0.7847,
+      "step": 797
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.832810281535288e-05,
+      "loss": 0.8096,
+      "step": 798
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.8089266440125645e-05,
+      "loss": 0.8193,
+      "step": 799
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.7850719725178994e-05,
+      "loss": 0.8281,
+      "step": 800
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.761246431919576e-05,
+      "loss": 0.8248,
+      "step": 801
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.7374501868845544e-05,
+      "loss": 0.8119,
+      "step": 802
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.7136834018773047e-05,
+      "loss": 0.799,
+      "step": 803
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6899462411587027e-05,
+      "loss": 0.825,
+      "step": 804
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.666238868784875e-05,
+      "loss": 0.8279,
+      "step": 805
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.642561448606072e-05,
+      "loss": 0.8371,
+      "step": 806
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6189141442655325e-05,
+      "loss": 0.8323,
+      "step": 807
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.595297119198352e-05,
+      "loss": 0.806,
+      "step": 808
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.571710536630357e-05,
+      "loss": 0.8405,
+      "step": 809
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.548154559576979e-05,
+      "loss": 0.8206,
+      "step": 810
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.524629350842111e-05,
+      "loss": 0.8014,
+      "step": 811
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.501135073017008e-05,
+      "loss": 0.8278,
+      "step": 812
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.477671888479142e-05,
+      "loss": 0.7979,
+      "step": 813
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.454239959391095e-05,
+      "loss": 0.8028,
+      "step": 814
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.430839447699427e-05,
+      "loss": 0.8156,
+      "step": 815
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.4074705151335615e-05,
+      "loss": 0.7965,
+      "step": 816
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.3841333232046654e-05,
+      "loss": 0.7889,
+      "step": 817
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.360828033204539e-05,
+      "loss": 0.8335,
+      "step": 818
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.337554806204497e-05,
+      "loss": 0.7984,
+      "step": 819
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.3143138030542436e-05,
+      "loss": 0.8282,
+      "step": 820
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.291105184380787e-05,
+      "loss": 0.7922,
+      "step": 821
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.267929110587307e-05,
+      "loss": 0.8171,
+      "step": 822
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.2447857418520564e-05,
+      "loss": 0.8156,
+      "step": 823
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.221675238127255e-05,
+      "loss": 0.8184,
+      "step": 824
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.1985977591379664e-05,
+      "loss": 0.7958,
+      "step": 825
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.175553464381023e-05,
+      "loss": 0.8159,
+      "step": 826
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.1525425131239056e-05,
+      "loss": 0.8094,
+      "step": 827
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.1295650644036406e-05,
+      "loss": 0.7892,
+      "step": 828
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.1066212770257185e-05,
+      "loss": 0.7657,
+      "step": 829
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.083711309562963e-05,
+      "loss": 0.8209,
+      "step": 830
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.0608353203544754e-05,
+      "loss": 0.7937,
+      "step": 831
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.0379934675045145e-05,
+      "loss": 0.7819,
+      "step": 832
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.015185908881409e-05,
+      "loss": 0.8014,
+      "step": 833
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.992412802116469e-05,
+      "loss": 0.7952,
+      "step": 834
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.969674304602895e-05,
+      "loss": 0.8214,
+      "step": 835
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.9469705734946946e-05,
+      "loss": 0.7833,
+      "step": 836
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.92430176570558e-05,
+      "loss": 0.777,
+      "step": 837
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.9016680379079085e-05,
+      "loss": 0.7728,
+      "step": 838
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.879069546531581e-05,
+      "loss": 0.7876,
+      "step": 839
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.8565064477629683e-05,
+      "loss": 0.8113,
+      "step": 840
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.833978897543829e-05,
+      "loss": 0.7826,
+      "step": 841
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.811487051570235e-05,
+      "loss": 0.7929,
+      "step": 842
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.78903106529149e-05,
+      "loss": 0.8002,
+      "step": 843
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.7666110939090644e-05,
+      "loss": 0.7983,
+      "step": 844
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.744227292375505e-05,
+      "loss": 0.8178,
+      "step": 845
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.721879815393385e-05,
+      "loss": 0.7793,
+      "step": 846
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.699568817414224e-05,
+      "loss": 0.7883,
+      "step": 847
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.677294452637422e-05,
+      "loss": 0.8348,
+      "step": 848
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.655056875009192e-05,
+      "loss": 0.781,
+      "step": 849
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.632856238221498e-05,
+      "loss": 0.788,
+      "step": 850
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.610692695710994e-05,
+      "loss": 0.846,
+      "step": 851
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.5885664006579645e-05,
+      "loss": 0.7841,
+      "step": 852
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.5664775059852515e-05,
+      "loss": 0.7937,
+      "step": 853
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.5444261643572226e-05,
+      "loss": 0.795,
+      "step": 854
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.522412528178698e-05,
+      "loss": 0.8179,
+      "step": 855
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.500436749593899e-05,
+      "loss": 0.8183,
+      "step": 856
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.478498980485405e-05,
+      "loss": 0.7822,
+      "step": 857
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.456599372473094e-05,
+      "loss": 0.7643,
+      "step": 858
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.4347380769130995e-05,
+      "loss": 0.7564,
+      "step": 859
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.412915244896764e-05,
+      "loss": 0.8128,
+      "step": 860
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.3911310272495996e-05,
+      "loss": 0.7835,
+      "step": 861
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.369385574530227e-05,
+      "loss": 0.7863,
+      "step": 862
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.347679037029361e-05,
+      "loss": 0.7858,
+      "step": 863
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.326011564768755e-05,
+      "loss": 0.7923,
+      "step": 864
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.304383307500169e-05,
+      "loss": 0.792,
+      "step": 865
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.282794414704333e-05,
+      "loss": 0.7805,
+      "step": 866
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.261245035589917e-05,
+      "loss": 0.7819,
+      "step": 867
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.2397353190924926e-05,
+      "loss": 0.7702,
+      "step": 868
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.218265413873519e-05,
+      "loss": 0.8366,
+      "step": 869
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.196835468319289e-05,
+      "loss": 0.8126,
+      "step": 870
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.1754456305399336e-05,
+      "loss": 0.8017,
+      "step": 871
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.15409604836838e-05,
+      "loss": 0.7923,
+      "step": 872
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.132786869359333e-05,
+      "loss": 0.7866,
+      "step": 873
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.111518240788258e-05,
+      "loss": 0.7939,
+      "step": 874
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.090290309650361e-05,
+      "loss": 0.7811,
+      "step": 875
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.069103222659576e-05,
+      "loss": 0.8036,
+      "step": 876
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.047957126247541e-05,
+      "loss": 0.8059,
+      "step": 877
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.0268521665626e-05,
+      "loss": 0.798,
+      "step": 878
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.005788489468784e-05,
+      "loss": 0.8026,
+      "step": 879
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.984766240544805e-05,
+      "loss": 0.7973,
+      "step": 880
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.9637855650830544e-05,
+      "loss": 0.8118,
+      "step": 881
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.942846608088583e-05,
+      "loss": 0.8033,
+      "step": 882
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.921949514278119e-05,
+      "loss": 0.8063,
+      "step": 883
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.90109442807906e-05,
+      "loss": 0.8118,
+      "step": 884
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.880281493628467e-05,
+      "loss": 0.8072,
+      "step": 885
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.859510854772086e-05,
+      "loss": 0.7659,
+      "step": 886
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.838782655063325e-05,
+      "loss": 0.7933,
+      "step": 887
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.818097037762297e-05,
+      "loss": 0.8759,
+      "step": 888
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.7974541458348046e-05,
+      "loss": 0.7941,
+      "step": 889
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.776854121951361e-05,
+      "loss": 0.8474,
+      "step": 890
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.756297108486202e-05,
+      "loss": 0.7739,
+      "step": 891
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.7357832475163045e-05,
+      "loss": 0.7987,
+      "step": 892
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.715312680820402e-05,
+      "loss": 0.8013,
+      "step": 893
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.6948855498780086e-05,
+      "loss": 0.8452,
+      "step": 894
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.674501995868429e-05,
+      "loss": 0.7593,
+      "step": 895
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.654162159669802e-05,
+      "loss": 0.8122,
+      "step": 896
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.63386618185811e-05,
+      "loss": 0.8006,
+      "step": 897
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.613614202706219e-05,
+      "loss": 0.8209,
+      "step": 898
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.593406362182903e-05,
+      "loss": 0.7865,
+      "step": 899
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.573242799951876e-05,
+      "loss": 0.7869,
+      "step": 900
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5531236553708305e-05,
+      "loss": 0.7938,
+      "step": 901
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5330490674904735e-05,
+      "loss": 0.7875,
+      "step": 902
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.513019175053559e-05,
+      "loss": 0.8168,
+      "step": 903
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.49303411649394e-05,
+      "loss": 0.7877,
+      "step": 904
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.473094029935603e-05,
+      "loss": 0.8029,
+      "step": 905
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.453199053191722e-05,
+      "loss": 0.7772,
+      "step": 906
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.433349323763696e-05,
+      "loss": 0.8099,
+      "step": 907
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.4135449788402044e-05,
+      "loss": 0.8148,
+      "step": 908
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.3937861552962555e-05,
+      "loss": 0.8049,
+      "step": 909
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.3740729896922496e-05,
+      "loss": 0.7994,
+      "step": 910
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.3544056182730244e-05,
+      "loss": 0.8184,
+      "step": 911
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.334784176966912e-05,
+      "loss": 0.8077,
+      "step": 912
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.315208801384811e-05,
+      "loss": 0.788,
+      "step": 913
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.2956796268192456e-05,
+      "loss": 0.8267,
+      "step": 914
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.276196788243422e-05,
+      "loss": 0.8291,
+      "step": 915
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.256760420310303e-05,
+      "loss": 0.8217,
+      "step": 916
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.2373706573516794e-05,
+      "loss": 0.7935,
+      "step": 917
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.218027633377235e-05,
+      "loss": 0.815,
+      "step": 918
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.1987314820736257e-05,
+      "loss": 0.8171,
+      "step": 919
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.1794823368035476e-05,
+      "loss": 0.8348,
+      "step": 920
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.160280330604827e-05,
+      "loss": 0.8527,
+      "step": 921
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.141125596189494e-05,
+      "loss": 0.8236,
+      "step": 922
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.1220182659428655e-05,
+      "loss": 0.8053,
+      "step": 923
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.10295847192263e-05,
+      "loss": 0.7938,
+      "step": 924
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.083946345857939e-05,
+      "loss": 0.7825,
+      "step": 925
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.0649820191484904e-05,
+      "loss": 0.8002,
+      "step": 926
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.0460656228636254e-05,
+      "loss": 0.8148,
+      "step": 927
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.0271972877414167e-05,
+      "loss": 0.8283,
+      "step": 928
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.0083771441877728e-05,
+      "loss": 0.7986,
+      "step": 929
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.9896053222755338e-05,
+      "loss": 0.8055,
+      "step": 930
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.970881951743567e-05,
+      "loss": 0.8383,
+      "step": 931
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.952207161995879e-05,
+      "loss": 0.8118,
+      "step": 932
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.933581082100717e-05,
+      "loss": 0.7901,
+      "step": 933
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.9150038407896696e-05,
+      "loss": 0.8006,
+      "step": 934
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.8964755664567923e-05,
+      "loss": 0.817,
+      "step": 935
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.8779963871577076e-05,
+      "loss": 0.8218,
+      "step": 936
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.8595664306087312e-05,
+      "loss": 0.831,
+      "step": 937
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.84118582418597e-05,
+      "loss": 0.8062,
+      "step": 938
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.822854694924465e-05,
+      "loss": 0.7987,
+      "step": 939
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.8045731695172994e-05,
+      "loss": 0.8115,
+      "step": 940
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.786341374314725e-05,
+      "loss": 0.8073,
+      "step": 941
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.7681594353232932e-05,
+      "loss": 0.8217,
+      "step": 942
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.7500274782049762e-05,
+      "loss": 0.8077,
+      "step": 943
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.7319456282763113e-05,
+      "loss": 0.7838,
+      "step": 944
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.7139140105075134e-05,
+      "loss": 0.8148,
+      "step": 945
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.695932749521638e-05,
+      "loss": 0.7771,
+      "step": 946
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.6780019695937008e-05,
+      "loss": 0.7994,
+      "step": 947
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.660121794649826e-05,
+      "loss": 0.83,
+      "step": 948
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.6422923482663863e-05,
+      "loss": 0.8065,
+      "step": 949
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.6245137536691532e-05,
+      "loss": 0.7764,
+      "step": 950
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.606786133732442e-05,
+      "loss": 0.8078,
+      "step": 951
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5891096109782642e-05,
+      "loss": 0.7956,
+      "step": 952
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5714843075754823e-05,
+      "loss": 0.7851,
+      "step": 953
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5539103453389556e-05,
+      "loss": 0.7831,
+      "step": 954
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5363878457287138e-05,
+      "loss": 0.7725,
+      "step": 955
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5189169298491054e-05,
+      "loss": 0.7962,
+      "step": 956
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5014977184479694e-05,
+      "loss": 0.8407,
+      "step": 957
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.4841303319157925e-05,
+      "loss": 0.7742,
+      "step": 958
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.466814890284882e-05,
+      "loss": 0.8264,
+      "step": 959
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.449551513228535e-05,
+      "loss": 0.8347,
+      "step": 960
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.432340320060216e-05,
+      "loss": 0.8082,
+      "step": 961
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.4151814297327158e-05,
+      "loss": 0.8147,
+      "step": 962
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.3980749608373533e-05,
+      "loss": 0.8335,
+      "step": 963
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.381021031603139e-05,
+      "loss": 0.7961,
+      "step": 964
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.3640197598959622e-05,
+      "loss": 0.782,
+      "step": 965
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.3470712632177804e-05,
+      "loss": 0.7593,
+      "step": 966
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.3301756587057987e-05,
+      "loss": 0.765,
+      "step": 967
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.3133330631316708e-05,
+      "loss": 0.848,
+      "step": 968
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.296543592900683e-05,
+      "loss": 0.8053,
+      "step": 969
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.27980736405095e-05,
+      "loss": 0.7994,
+      "step": 970
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.263124492252623e-05,
+      "loss": 0.7858,
+      "step": 971
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.246495092807077e-05,
+      "loss": 0.7803,
+      "step": 972
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.2299192806461268e-05,
+      "loss": 0.776,
+      "step": 973
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.2133971703312194e-05,
+      "loss": 0.7904,
+      "step": 974
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.196928876052652e-05,
+      "loss": 0.816,
+      "step": 975
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.1805145116287807e-05,
+      "loss": 0.8175,
+      "step": 976
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.164154190505231e-05,
+      "loss": 0.8091,
+      "step": 977
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.1478480257541177e-05,
+      "loss": 0.795,
+      "step": 978
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.1315961300732557e-05,
+      "loss": 0.7874,
+      "step": 979
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.1153986157853923e-05,
+      "loss": 0.799,
+      "step": 980
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.099255594837424e-05,
+      "loss": 0.7883,
+      "step": 981
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.083167178799623e-05,
+      "loss": 0.7699,
+      "step": 982
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.067133478864869e-05,
+      "loss": 0.7969,
+      "step": 983
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.0511546058478793e-05,
+      "loss": 0.8028,
+      "step": 984
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.035230670184436e-05,
+      "loss": 0.8186,
+      "step": 985
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.0193617819306353e-05,
+      "loss": 0.7867,
+      "step": 986
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.0035480507621218e-05,
+      "loss": 0.7893,
+      "step": 987
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.9877895859733252e-05,
+      "loss": 0.7985,
+      "step": 988
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.972086496476716e-05,
+      "loss": 0.8733,
+      "step": 989
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.956438890802034e-05,
+      "loss": 0.831,
+      "step": 990
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.94084687709556e-05,
+      "loss": 0.8037,
+      "step": 991
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.925310563119358e-05,
+      "loss": 0.8096,
+      "step": 992
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 0.8189,
+      "step": 993
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8944054634804666e-05,
+      "loss": 0.7901,
+      "step": 994
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.879036891414129e-05,
+      "loss": 0.8087,
+      "step": 995
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8637244462692915e-05,
+      "loss": 0.7719,
+      "step": 996
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8484682338758152e-05,
+      "loss": 0.7858,
+      "step": 997
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.833268359674919e-05,
+      "loss": 0.8043,
+      "step": 998
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8181249287184466e-05,
+      "loss": 0.8041,
+      "step": 999
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8030380456681396e-05,
+      "loss": 0.7879,
+      "step": 1000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7880078147949207e-05,
+      "loss": 0.791,
+      "step": 1001
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7730343399781668e-05,
+      "loss": 0.7996,
+      "step": 1002
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.758117724704995e-05,
+      "loss": 0.8118,
+      "step": 1003
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.743258072069539e-05,
+      "loss": 0.7825,
+      "step": 1004
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.72845548477225e-05,
+      "loss": 0.8106,
+      "step": 1005
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7137100651191773e-05,
+      "loss": 0.7638,
+      "step": 1006
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.699021915021266e-05,
+      "loss": 0.7942,
+      "step": 1007
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.684391135993647e-05,
+      "loss": 0.8063,
+      "step": 1008
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6698178291549417e-05,
+      "loss": 0.7814,
+      "step": 1009
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6553020952265587e-05,
+      "loss": 0.8218,
+      "step": 1010
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.640844034532002e-05,
+      "loss": 0.8174,
+      "step": 1011
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6264437469961703e-05,
+      "loss": 0.8018,
+      "step": 1012
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6121013321446733e-05,
+      "loss": 0.8135,
+      "step": 1013
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.597816889103144e-05,
+      "loss": 0.7998,
+      "step": 1014
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.583590516596548e-05,
+      "loss": 0.7513,
+      "step": 1015
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.5694223129485075e-05,
+      "loss": 0.7904,
+      "step": 1016
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.5553123760806143e-05,
+      "loss": 0.7906,
+      "step": 1017
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.5412608035117604e-05,
+      "loss": 0.8285,
+      "step": 1018
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.5272676923574637e-05,
+      "loss": 0.7718,
+      "step": 1019
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.513333139329186e-05,
+      "loss": 0.7614,
+      "step": 1020
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4994572407336783e-05,
+      "loss": 0.7947,
+      "step": 1021
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4856400924723079e-05,
+      "loss": 0.7867,
+      "step": 1022
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4718817900404014e-05,
+      "loss": 0.8164,
+      "step": 1023
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4581824285265733e-05,
+      "loss": 0.806,
+      "step": 1024
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.444542102612082e-05,
+      "loss": 0.7663,
+      "step": 1025
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4309609065701656e-05,
+      "loss": 0.802,
+      "step": 1026
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4174389342653971e-05,
+      "loss": 0.79,
+      "step": 1027
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4039762791530343e-05,
+      "loss": 0.8284,
+      "step": 1028
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.3905730342783629e-05,
+      "loss": 0.8016,
+      "step": 1029
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.3772292922760732e-05,
+      "loss": 0.7947,
+      "step": 1030
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.363945145369606e-05,
+      "loss": 0.8162,
+      "step": 1031
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.3507206853705178e-05,
+      "loss": 0.8006,
+      "step": 1032
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.3375560036778501e-05,
+      "loss": 0.8207,
+      "step": 1033
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.3244511912774905e-05,
+      "loss": 0.7851,
+      "step": 1034
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.311406338741551e-05,
+      "loss": 0.7976,
+      "step": 1035
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.298421536227743e-05,
+      "loss": 0.8137,
+      "step": 1036
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2854968734787398e-05,
+      "loss": 0.8328,
+      "step": 1037
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2726324398215762e-05,
+      "loss": 0.7728,
+      "step": 1038
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2598283241670195e-05,
+      "loss": 0.7832,
+      "step": 1039
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2470846150089566e-05,
+      "loss": 0.7839,
+      "step": 1040
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2344014004237859e-05,
+      "loss": 0.8048,
+      "step": 1041
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.221778768069799e-05,
+      "loss": 0.796,
+      "step": 1042
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.2092168051865892e-05,
+      "loss": 0.7832,
+      "step": 1043
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1967155985944367e-05,
+      "loss": 0.8112,
+      "step": 1044
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.184275234693717e-05,
+      "loss": 0.8135,
+      "step": 1045
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1718957994642965e-05,
+      "loss": 0.8058,
+      "step": 1046
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1595773784649389e-05,
+      "loss": 0.7911,
+      "step": 1047
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1473200568327224e-05,
+      "loss": 0.8009,
+      "step": 1048
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1351239192824415e-05,
+      "loss": 0.8356,
+      "step": 1049
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1229890501060258e-05,
+      "loss": 0.7931,
+      "step": 1050
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.11091553317196e-05,
+      "loss": 0.7974,
+      "step": 1051
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.0989034519246954e-05,
+      "loss": 0.7679,
+      "step": 1052
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.0869528893840875e-05,
+      "loss": 0.7925,
+      "step": 1053
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0750639281448017e-05,
+      "loss": 0.7569,
+      "step": 1054
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0632366503757652e-05,
+      "loss": 0.8122,
+      "step": 1055
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.051471137819583e-05,
+      "loss": 0.7836,
+      "step": 1056
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0397674717919802e-05,
+      "loss": 0.8261,
+      "step": 1057
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0281257331812377e-05,
+      "loss": 0.8043,
+      "step": 1058
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0165460024476337e-05,
+      "loss": 0.7816,
+      "step": 1059
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0050283596228871e-05,
+      "loss": 0.7982,
+      "step": 1060
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.93572884309607e-06,
+      "loss": 0.8247,
+      "step": 1061
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.821796556807339e-06,
+      "loss": 0.7891,
+      "step": 1062
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.708487524790045e-06,
+      "loss": 0.7698,
+      "step": 1063
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.595802530164011e-06,
+      "loss": 0.7457,
+      "step": 1064
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.483742351736113e-06,
+      "loss": 0.7622,
+      "step": 1065
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.372307763994891e-06,
+      "loss": 0.7885,
+      "step": 1066
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 9.261499537105177e-06,
+      "loss": 0.8177,
+      "step": 1067
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 9.151318436902834e-06,
+      "loss": 0.8245,
+      "step": 1068
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 9.041765224889399e-06,
+      "loss": 0.8346,
+      "step": 1069
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.93284065822687e-06,
+      "loss": 0.823,
+      "step": 1070
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.824545489732417e-06,
+      "loss": 0.8254,
+      "step": 1071
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.716880467873234e-06,
+      "loss": 0.7924,
+      "step": 1072
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.609846336761329e-06,
+      "loss": 0.8331,
+      "step": 1073
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.503443836148416e-06,
+      "loss": 0.8041,
+      "step": 1074
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.39767370142076e-06,
+      "loss": 0.7816,
+      "step": 1075
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.29253666359413e-06,
+      "loss": 0.7861,
+      "step": 1076
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.188033449308719e-06,
+      "loss": 0.8057,
+      "step": 1077
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.084164780824178e-06,
+      "loss": 0.7928,
+      "step": 1078
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.9809313760145e-06,
+      "loss": 0.7635,
+      "step": 1079
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.878333948363203e-06,
+      "loss": 0.7887,
+      "step": 1080
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.776373206958332e-06,
+      "loss": 0.7875,
+      "step": 1081
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.675049856487549e-06,
+      "loss": 0.793,
+      "step": 1082
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.574364597233275e-06,
+      "loss": 0.7894,
+      "step": 1083
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.4743181250678475e-06,
+      "loss": 0.8915,
+      "step": 1084
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.374911131448747e-06,
+      "loss": 0.7938,
+      "step": 1085
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.276144303413757e-06,
+      "loss": 0.7834,
+      "step": 1086
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.178018323576208e-06,
+      "loss": 0.7901,
+      "step": 1087
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.080533870120376e-06,
+      "loss": 0.8224,
+      "step": 1088
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.983691616796672e-06,
+      "loss": 0.7735,
+      "step": 1089
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.887492232917025e-06,
+      "loss": 0.8153,
+      "step": 1090
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.791936383350261e-06,
+      "loss": 0.8225,
+      "step": 1091
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.6970247285175315e-06,
+      "loss": 0.8206,
+      "step": 1092
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.602757924387715e-06,
+      "loss": 0.7909,
+      "step": 1093
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.5091366224728554e-06,
+      "loss": 0.7709,
+      "step": 1094
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.416161469823734e-06,
+      "loss": 0.7875,
+      "step": 1095
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.323833109025368e-06,
+      "loss": 0.8268,
+      "step": 1096
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.23215217819253e-06,
+      "loss": 0.8078,
+      "step": 1097
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.1411193109654155e-06,
+      "loss": 0.7711,
+      "step": 1098
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.0507351365051475e-06,
+      "loss": 0.7942,
+      "step": 1099
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.961000279489559e-06,
+      "loss": 0.8448,
+      "step": 1100
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.871915360108781e-06,
+      "loss": 0.7911,
+      "step": 1101
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.7834809940610195e-06,
+      "loss": 0.8042,
+      "step": 1102
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.69569779254826e-06,
+      "loss": 0.7878,
+      "step": 1103
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.608566362272017e-06,
+      "loss": 0.8116,
+      "step": 1104
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.522087305429214e-06,
+      "loss": 0.8222,
+      "step": 1105
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.436261219707994e-06,
+      "loss": 0.8349,
+      "step": 1106
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.351088698283558e-06,
+      "loss": 0.8312,
+      "step": 1107
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.266570329814091e-06,
+      "loss": 0.8031,
+      "step": 1108
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.18270669843669e-06,
+      "loss": 0.7961,
+      "step": 1109
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.099498383763346e-06,
+      "loss": 0.7575,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.016945960876884e-06,
+      "loss": 0.8461,
+      "step": 1111
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.9350500003270465e-06,
+      "loss": 0.7653,
+      "step": 1112
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.853811068126502e-06,
+      "loss": 0.7727,
+      "step": 1113
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.773229725747008e-06,
+      "loss": 0.8009,
+      "step": 1114
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.693306530115415e-06,
+      "loss": 0.7981,
+      "step": 1115
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.614042033609945e-06,
+      "loss": 0.786,
+      "step": 1116
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.535436784056269e-06,
+      "loss": 0.8475,
+      "step": 1117
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.457491324723784e-06,
+      "loss": 0.7938,
+      "step": 1118
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.380206194321812e-06,
+      "loss": 0.8026,
+      "step": 1119
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.303581926995959e-06,
+      "loss": 0.8043,
+      "step": 1120
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.227619052324283e-06,
+      "loss": 0.8171,
+      "step": 1121
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.152318095313778e-06,
+      "loss": 0.8081,
+      "step": 1122
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.077679576396676e-06,
+      "loss": 0.8094,
+      "step": 1123
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.003704011426845e-06,
+      "loss": 0.8117,
+      "step": 1124
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.930391911676256e-06,
+      "loss": 0.793,
+      "step": 1125
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.857743783831413e-06,
+      "loss": 0.7993,
+      "step": 1126
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.785760129989868e-06,
+      "loss": 0.7828,
+      "step": 1127
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.714441447656791e-06,
+      "loss": 0.8079,
+      "step": 1128
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.6437882297414292e-06,
+      "loss": 0.8215,
+      "step": 1129
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.5738009645538195e-06,
+      "loss": 0.7831,
+      "step": 1130
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.5044801358013136e-06,
+      "loss": 0.7595,
+      "step": 1131
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.4358262225853254e-06,
+      "loss": 0.8087,
+      "step": 1132
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3678396993979343e-06,
+      "loss": 0.7865,
+      "step": 1133
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3005210361186643e-06,
+      "loss": 0.7779,
+      "step": 1134
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.23387069801121e-06,
+      "loss": 0.7769,
+      "step": 1135
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.1678891457202597e-06,
+      "loss": 0.7906,
+      "step": 1136
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.102576835268212e-06,
+      "loss": 0.8089,
+      "step": 1137
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.037934218052152e-06,
+      "loss": 0.8102,
+      "step": 1138
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.9739617408406585e-06,
+      "loss": 0.8208,
+      "step": 1139
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.9106598457707134e-06,
+      "loss": 0.8134,
+      "step": 1140
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.8480289703446737e-06,
+      "loss": 0.7775,
+      "step": 1141
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7860695474272392e-06,
+      "loss": 0.8033,
+      "step": 1142
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7247820052424657e-06,
+      "loss": 0.7912,
+      "step": 1143
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.664166767370757e-06,
+      "loss": 0.8087,
+      "step": 1144
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.6042242527460457e-06,
+      "loss": 0.7833,
+      "step": 1145
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.544954875652761e-06,
+      "loss": 0.7881,
+      "step": 1146
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.4863590457230743e-06,
+      "loss": 0.7758,
+      "step": 1147
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.4284371679340166e-06,
+      "loss": 0.7712,
+      "step": 1148
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.3711896426046987e-06,
+      "loss": 0.8,
+      "step": 1149
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.3146168653935374e-06,
+      "loss": 0.7782,
+      "step": 1150
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.2587192272954917e-06,
+      "loss": 0.794,
+      "step": 1151
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.20349711463943e-06,
+      "loss": 0.7982,
+      "step": 1152
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.148950909085401e-06,
+      "loss": 0.7854,
+      "step": 1153
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.0950809876220113e-06,
+      "loss": 0.7992,
+      "step": 1154
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.0418877225638288e-06,
+      "loss": 0.7914,
+      "step": 1155
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.989371481548807e-06,
+      "loss": 0.8131,
+      "step": 1156
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.9375326275357208e-06,
+      "loss": 0.8106,
+      "step": 1157
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.8863715188016994e-06,
+      "loss": 0.8079,
+      "step": 1158
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.8358885089397205e-06,
+      "loss": 0.7994,
+      "step": 1159
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.7860839468561874e-06,
+      "loss": 0.8128,
+      "step": 1160
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.7369581767684774e-06,
+      "loss": 0.7503,
+      "step": 1161
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6885115382026085e-06,
+      "loss": 0.8003,
+      "step": 1162
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6407443659908762e-06,
+      "loss": 0.7964,
+      "step": 1163
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.593656990269521e-06,
+      "loss": 0.7815,
+      "step": 1164
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5472497364764859e-06,
+      "loss": 0.8163,
+      "step": 1165
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5015229253491302e-06,
+      "loss": 0.7678,
+      "step": 1166
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4564768729220412e-06,
+      "loss": 0.7895,
+      "step": 1167
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4121118905248031e-06,
+      "loss": 0.7894,
+      "step": 1168
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.3684282847799213e-06,
+      "loss": 0.8079,
+      "step": 1169
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.3254263576006343e-06,
+      "loss": 0.7869,
+      "step": 1170
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.2831064061888498e-06,
+      "loss": 0.8026,
+      "step": 1171
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.2414687230331123e-06,
+      "loss": 0.82,
+      "step": 1172
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.2005135959065383e-06,
+      "loss": 0.8091,
+      "step": 1173
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.1602413078648733e-06,
+      "loss": 0.7868,
+      "step": 1174
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.1206521372445044e-06,
+      "loss": 0.7791,
+      "step": 1175
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0817463576605513e-06,
+      "loss": 0.8065,
+      "step": 1176
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0435242380049559e-06,
+      "loss": 0.827,
+      "step": 1177
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0059860424446842e-06,
+      "loss": 0.8174,
+      "step": 1178
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.691320304197837e-07,
+      "loss": 0.7883,
+      "step": 1179
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.329624566417172e-07,
+      "loss": 0.7991,
+      "step": 1180
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.974775710915207e-07,
+      "loss": 0.7715,
+      "step": 1181
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.626776190181041e-07,
+      "loss": 0.7636,
+      "step": 1182
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.285628409365753e-07,
+      "loss": 0.7824,
+      "step": 1183
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.951334726265191e-07,
+      "loss": 0.7822,
+      "step": 1184
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.623897451304318e-07,
+      "loss": 0.7775,
+      "step": 1185
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.303318847521001e-07,
+      "loss": 0.8232,
+      "step": 1186
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 6.98960113055025e-07,
+      "loss": 0.7522,
+      "step": 1187
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 6.682746468608781e-07,
+      "loss": 0.7847,
+      "step": 1188
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 6.382756982480587e-07,
+      "loss": 0.832,
+      "step": 1189
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.089634745501838e-07,
+      "loss": 0.7886,
+      "step": 1190
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.803381783546669e-07,
+      "loss": 0.7937,
+      "step": 1191
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.52400007501297e-07,
+      "loss": 0.7814,
+      "step": 1192
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.251491550809062e-07,
+      "loss": 0.8052,
+      "step": 1193
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.985858094340379e-07,
+      "loss": 0.837,
+      "step": 1194
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.727101541496026e-07,
+      "loss": 0.7821,
+      "step": 1195
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.4752236806363535e-07,
+      "loss": 0.7863,
+      "step": 1196
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.230226252580516e-07,
+      "loss": 0.804,
+      "step": 1197
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.9921109505947075e-07,
+      "loss": 0.8122,
+      "step": 1198
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.76087942038017e-07,
+      "loss": 0.7868,
+      "step": 1199
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.5365332600617586e-07,
+      "loss": 0.7883,
+      "step": 1200
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.319074020177282e-07,
+      "loss": 0.7943,
+      "step": 1201
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.108503203666402e-07,
+      "loss": 0.7858,
+      "step": 1202
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.904822265860307e-07,
+      "loss": 0.7784,
+      "step": 1203
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.7080326144719447e-07,
+      "loss": 0.8026,
+      "step": 1204
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5181356095858033e-07,
+      "loss": 0.8184,
+      "step": 1205
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.335132563649145e-07,
+      "loss": 0.7813,
+      "step": 1206
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.1590247414624566e-07,
+      "loss": 0.7606,
+      "step": 1207
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.989813360170456e-07,
+      "loss": 0.7656,
+      "step": 1208
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.8274995892546555e-07,
+      "loss": 0.8222,
+      "step": 1209
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.6720845505242555e-07,
+      "loss": 0.7894,
+      "step": 1210
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5235693181090417e-07,
+      "loss": 0.7985,
+      "step": 1211
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.3819549184516112e-07,
+      "loss": 0.7977,
+      "step": 1212
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.24724233030038e-07,
+      "loss": 0.7925,
+      "step": 1213
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.1194324847030313e-07,
+      "loss": 0.7931,
+      "step": 1214
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 9.98526264999633e-08,
+      "loss": 0.7947,
+      "step": 1215
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.84524506817086e-08,
+      "loss": 0.7849,
+      "step": 1216
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 7.774279980626853e-08,
+      "loss": 0.8017,
+      "step": 1217
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 6.772374789194569e-08,
+      "loss": 0.8175,
+      "step": 1218
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.839536418401625e-08,
+      "loss": 0.8091,
+      "step": 1219
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.975771315435252e-08,
+      "loss": 0.7687,
+      "step": 1220
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.181085450087885e-08,
+      "loss": 0.8129,
+      "step": 1221
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.4554843147216464e-08,
+      "loss": 0.7838,
+      "step": 1222
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.7989729242294814e-08,
+      "loss": 0.8235,
+      "step": 1223
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2115558159985226e-08,
+      "loss": 0.8205,
+      "step": 1224
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.693237049883445e-08,
+      "loss": 0.7926,
+      "step": 1225
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.2440202081731578e-08,
+      "loss": 0.7902,
+      "step": 1226
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.639083955663818e-09,
+      "loss": 0.7921,
+      "step": 1227
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.529042391549944e-09,
+      "loss": 0.8395,
+      "step": 1228
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.110098884007151e-09,
+      "loss": 0.7931,
+      "step": 1229
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3822701512400394e-09,
+      "loss": 0.7951,
+      "step": 1230
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4556813489627914e-10,
+      "loss": 0.8307,
+      "step": 1231
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0,
+      "loss": 0.8206,
+      "step": 1232
+    },
+    {
+      "epoch": 1.0,
+      "step": 1232,
+      "total_flos": 1826545100652544.0,
+      "train_loss": 0.8080920413523526,
+      "train_runtime": 32957.0807,
+      "train_samples_per_second": 4.785,
+      "train_steps_per_second": 0.037
+    }
+  ],
+  "max_steps": 1232,
+  "num_train_epochs": 1,
+  "total_flos": 1826545100652544.0,
+  "trial_name": null,
+  "trial_params": null
+}