diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4794 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "global_step": 795,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.25e-07,
+      "loss": 0.856,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.25e-06,
+      "loss": 0.792,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 0.835,
+      "step": 3
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5e-06,
+      "loss": 0.8315,
+      "step": 4
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.125e-06,
+      "loss": 0.8115,
+      "step": 5
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.8032,
+      "step": 6
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 0.7256,
+      "step": 7
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-06,
+      "loss": 0.7007,
+      "step": 8
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.625e-06,
+      "loss": 0.647,
+      "step": 9
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 6.25e-06,
+      "loss": 0.6431,
+      "step": 10
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 6.875e-06,
+      "loss": 0.6455,
+      "step": 11
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.5786,
+      "step": 12
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 8.125000000000001e-06,
+      "loss": 0.6245,
+      "step": 13
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.5962,
+      "step": 14
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 0.5854,
+      "step": 15
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.5693,
+      "step": 16
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.0625e-05,
+      "loss": 0.5615,
+      "step": 17
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.125e-05,
+      "loss": 0.5354,
+      "step": 18
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.1875e-05,
+      "loss": 0.563,
+      "step": 19
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.25e-05,
+      "loss": 0.5596,
+      "step": 20
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.3125e-05,
+      "loss": 0.543,
+      "step": 21
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.375e-05,
+      "loss": 0.5403,
+      "step": 22
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.4375e-05,
+      "loss": 0.5437,
+      "step": 23
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.5286,
+      "step": 24
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.5625e-05,
+      "loss": 0.5417,
+      "step": 25
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.6250000000000002e-05,
+      "loss": 0.5151,
+      "step": 26
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.6875e-05,
+      "loss": 0.5154,
+      "step": 27
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 0.5271,
+      "step": 28
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.8125e-05,
+      "loss": 0.5283,
+      "step": 29
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.5127,
+      "step": 30
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9375e-05,
+      "loss": 0.5125,
+      "step": 31
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 0.5166,
+      "step": 32
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9999915234318064e-05,
+      "loss": 0.5288,
+      "step": 33
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.999966093870929e-05,
+      "loss": 0.5264,
+      "step": 34
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.99992371174848e-05,
+      "loss": 0.5229,
+      "step": 35
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9998643777829674e-05,
+      "loss": 0.5239,
+      "step": 36
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9997880929802895e-05,
+      "loss": 0.4995,
+      "step": 37
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9996948586337127e-05,
+      "loss": 0.4934,
+      "step": 38
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9995846763238514e-05,
+      "loss": 0.5022,
+      "step": 39
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.999457547918641e-05,
+      "loss": 0.4658,
+      "step": 40
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9993134755733075e-05,
+      "loss": 0.4968,
+      "step": 41
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9991524617303282e-05,
+      "loss": 0.5078,
+      "step": 42
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.998974509119393e-05,
+      "loss": 0.5183,
+      "step": 43
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9987796207573573e-05,
+      "loss": 0.4844,
+      "step": 44
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9985677999481898e-05,
+      "loss": 0.5,
+      "step": 45
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9983390502829168e-05,
+      "loss": 0.488,
+      "step": 46
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9980933756395635e-05,
+      "loss": 0.4744,
+      "step": 47
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9978307801830855e-05,
+      "loss": 0.5098,
+      "step": 48
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9975512683652985e-05,
+      "loss": 0.4851,
+      "step": 49
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.997254844924806e-05,
+      "loss": 0.4861,
+      "step": 50
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.996941514886914e-05,
+      "loss": 0.5112,
+      "step": 51
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9966112835635493e-05,
+      "loss": 0.4678,
+      "step": 52
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9962641565531694e-05,
+      "loss": 0.4897,
+      "step": 53
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.995900139740665e-05,
+      "loss": 0.4563,
+      "step": 54
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9955192392972628e-05,
+      "loss": 0.488,
+      "step": 55
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9951214616804203e-05,
+      "loss": 0.4812,
+      "step": 56
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.994706813633716e-05,
+      "loss": 0.4617,
+      "step": 57
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.994275302186734e-05,
+      "loss": 0.4836,
+      "step": 58
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9938269346549473e-05,
+      "loss": 0.4846,
+      "step": 59
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9933617186395917e-05,
+      "loss": 0.4749,
+      "step": 60
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9928796620275377e-05,
+      "loss": 0.4788,
+      "step": 61
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9923807729911567e-05,
+      "loss": 0.4802,
+      "step": 62
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9918650599881828e-05,
+      "loss": 0.468,
+      "step": 63
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9913325317615684e-05,
+      "loss": 0.4854,
+      "step": 64
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9907831973393377e-05,
+      "loss": 0.4731,
+      "step": 65
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9902170660344323e-05,
+      "loss": 0.5017,
+      "step": 66
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9896341474445526e-05,
+      "loss": 0.4915,
+      "step": 67
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9890344514519974e-05,
+      "loss": 0.4868,
+      "step": 68
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9884179882234946e-05,
+      "loss": 0.457,
+      "step": 69
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9877847682100294e-05,
+      "loss": 0.48,
+      "step": 70
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9871348021466673e-05,
+      "loss": 0.4707,
+      "step": 71
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.986468101052371e-05,
+      "loss": 0.4912,
+      "step": 72
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9857846762298157e-05,
+      "loss": 0.4768,
+      "step": 73
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.985084539265195e-05,
+      "loss": 0.4492,
+      "step": 74
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.984367702028027e-05,
+      "loss": 0.4761,
+      "step": 75
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.983634176670951e-05,
+      "loss": 0.437,
+      "step": 76
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9828839756295223e-05,
+      "loss": 0.4619,
+      "step": 77
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.982117111622001e-05,
+      "loss": 0.4719,
+      "step": 78
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9813335976491387e-05,
+      "loss": 0.4753,
+      "step": 79
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.980533446993953e-05,
+      "loss": 0.4697,
+      "step": 80
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9797166732215078e-05,
+      "loss": 0.4648,
+      "step": 81
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.97888329017868e-05,
+      "loss": 0.4534,
+      "step": 82
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9780333119939264e-05,
+      "loss": 0.4553,
+      "step": 83
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9771667530770427e-05,
+      "loss": 0.4714,
+      "step": 84
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9762836281189207e-05,
+      "loss": 0.4385,
+      "step": 85
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9753839520912984e-05,
+      "loss": 0.4697,
+      "step": 86
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9744677402465053e-05,
+      "loss": 0.4775,
+      "step": 87
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.973535008117207e-05,
+      "loss": 0.4287,
+      "step": 88
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9725857715161375e-05,
+      "loss": 0.4473,
+      "step": 89
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9716200465358352e-05,
+      "loss": 0.4531,
+      "step": 90
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9706378495483664e-05,
+      "loss": 0.4524,
+      "step": 91
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9696391972050516e-05,
+      "loss": 0.4756,
+      "step": 92
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9686241064361792e-05,
+      "loss": 0.436,
+      "step": 93
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9675925944507226e-05,
+      "loss": 0.4265,
+      "step": 94
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9665446787360444e-05,
+      "loss": 0.4766,
+      "step": 95
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.965480377057603e-05,
+      "loss": 0.4683,
+      "step": 96
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.964399707458649e-05,
+      "loss": 0.4602,
+      "step": 97
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9633026882599228e-05,
+      "loss": 0.4497,
+      "step": 98
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9621893380593398e-05,
+      "loss": 0.4358,
+      "step": 99
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.961059675731678e-05,
+      "loss": 0.4641,
+      "step": 100
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9599137204282566e-05,
+      "loss": 0.4429,
+      "step": 101
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9587514915766124e-05,
+      "loss": 0.4343,
+      "step": 102
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9575730088801696e-05,
+      "loss": 0.4553,
+      "step": 103
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9563782923179063e-05,
+      "loss": 0.4294,
+      "step": 104
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.955167362144015e-05,
+      "loss": 0.4282,
+      "step": 105
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9539402388875598e-05,
+      "loss": 0.4333,
+      "step": 106
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9526969433521298e-05,
+      "loss": 0.4631,
+      "step": 107
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9514374966154826e-05,
+      "loss": 0.4358,
+      "step": 108
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.950161920029191e-05,
+      "loss": 0.4507,
+      "step": 109
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.948870235218279e-05,
+      "loss": 0.4512,
+      "step": 110
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9475624640808542e-05,
+      "loss": 0.4495,
+      "step": 111
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.946238628787741e-05,
+      "loss": 0.4546,
+      "step": 112
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9448987517820982e-05,
+      "loss": 0.4451,
+      "step": 113
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.943542855779044e-05,
+      "loss": 0.4504,
+      "step": 114
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9421709637652683e-05,
+      "loss": 0.4504,
+      "step": 115
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.940783098998643e-05,
+      "loss": 0.4219,
+      "step": 116
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9393792850078294e-05,
+      "loss": 0.4253,
+      "step": 117
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9379595455918773e-05,
+      "loss": 0.4309,
+      "step": 118
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9365239048198227e-05,
+      "loss": 0.4587,
+      "step": 119
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.935072387030279e-05,
+      "loss": 0.4404,
+      "step": 120
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.933605016831026e-05,
+      "loss": 0.4568,
+      "step": 121
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9321218190985906e-05,
+      "loss": 0.4399,
+      "step": 122
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9306228189778255e-05,
+      "loss": 0.429,
+      "step": 123
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9291080418814852e-05,
+      "loss": 0.4341,
+      "step": 124
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.927577513489792e-05,
+      "loss": 0.4377,
+      "step": 125
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.926031259750002e-05,
+      "loss": 0.4258,
+      "step": 126
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9244693068759668e-05,
+      "loss": 0.4233,
+      "step": 127
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9228916813476855e-05,
+      "loss": 0.4243,
+      "step": 128
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9212984099108594e-05,
+      "loss": 0.467,
+      "step": 129
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9196895195764363e-05,
+      "loss": 0.4299,
+      "step": 130
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9180650376201536e-05,
+      "loss": 0.4329,
+      "step": 131
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9164249915820753e-05,
+      "loss": 0.4695,
+      "step": 132
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9147694092661254e-05,
+      "loss": 0.4431,
+      "step": 133
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9130983187396174e-05,
+      "loss": 0.3972,
+      "step": 134
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.911411748332776e-05,
+      "loss": 0.4617,
+      "step": 135
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9097097266382598e-05,
+      "loss": 0.4087,
+      "step": 136
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.907992282510675e-05,
+      "loss": 0.4275,
+      "step": 137
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.9062594450660857e-05,
+      "loss": 0.4285,
+      "step": 138
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.9045112436815217e-05,
+      "loss": 0.4099,
+      "step": 139
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.902747707994479e-05,
+      "loss": 0.4299,
+      "step": 140
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 0.4216,
+      "step": 141
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8991747535622607e-05,
+      "loss": 0.426,
+      "step": 142
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.897365395389869e-05,
+      "loss": 0.449,
+      "step": 143
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8955408240595396e-05,
+      "loss": 0.4387,
+      "step": 144
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8937010705034788e-05,
+      "loss": 0.4146,
+      "step": 145
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8918461659112805e-05,
+      "loss": 0.4326,
+      "step": 146
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8899761417293944e-05,
+      "loss": 0.4514,
+      "step": 147
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8880910296605956e-05,
+      "loss": 0.4436,
+      "step": 148
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8861908616634465e-05,
+      "loss": 0.4062,
+      "step": 149
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.884275669951754e-05,
+      "loss": 0.4158,
+      "step": 150
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8823454869940243e-05,
+      "loss": 0.3972,
+      "step": 151
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.880400345512913e-05,
+      "loss": 0.408,
+      "step": 152
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8784402784846683e-05,
+      "loss": 0.4277,
+      "step": 153
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8764653191385737e-05,
+      "loss": 0.4163,
+      "step": 154
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.874475500956385e-05,
+      "loss": 0.4158,
+      "step": 155
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8724708576717607e-05,
+      "loss": 0.4512,
+      "step": 156
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.870451423269692e-05,
+      "loss": 0.4094,
+      "step": 157
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8684172319859258e-05,
+      "loss": 0.408,
+      "step": 158
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8663683183063846e-05,
+      "loss": 0.4189,
+      "step": 159
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.864304716966581e-05,
+      "loss": 0.4333,
+      "step": 160
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.86222646295103e-05,
+      "loss": 0.4329,
+      "step": 161
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8601335914926558e-05,
+      "loss": 0.437,
+      "step": 162
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8580261380721932e-05,
+      "loss": 0.407,
+      "step": 163
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8559041384175874e-05,
+      "loss": 0.4077,
+      "step": 164
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8537676285033886e-05,
+      "loss": 0.4346,
+      "step": 165
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8516166445501405e-05,
+      "loss": 0.4246,
+      "step": 166
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.849451223023767e-05,
+      "loss": 0.4456,
+      "step": 167
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8472714006349554e-05,
+      "loss": 0.4209,
+      "step": 168
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.845077214338531e-05,
+      "loss": 0.4275,
+      "step": 169
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8428687013328338e-05,
+      "loss": 0.4268,
+      "step": 170
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.840645899059086e-05,
+      "loss": 0.4041,
+      "step": 171
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.838408845200758e-05,
+      "loss": 0.3972,
+      "step": 172
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.836157577682928e-05,
+      "loss": 0.4041,
+      "step": 173
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8338921346716426e-05,
+      "loss": 0.3882,
+      "step": 174
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8316125545732653e-05,
+      "loss": 0.3889,
+      "step": 175
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8293188760338285e-05,
+      "loss": 0.3784,
+      "step": 176
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8270111379383773e-05,
+      "loss": 0.386,
+      "step": 177
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8246893794103113e-05,
+      "loss": 0.3862,
+      "step": 178
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8223536398107177e-05,
+      "loss": 0.3838,
+      "step": 179
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.82000395873771e-05,
+      "loss": 0.3848,
+      "step": 180
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.817640376025751e-05,
+      "loss": 0.4333,
+      "step": 181
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.8152629317449814e-05,
+      "loss": 0.3643,
+      "step": 182
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.8128716662005385e-05,
+      "loss": 0.3896,
+      "step": 183
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.810466619931872e-05,
+      "loss": 0.4285,
+      "step": 184
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.8080478337120604e-05,
+      "loss": 0.4031,
+      "step": 185
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.8056153485471167e-05,
+      "loss": 0.4028,
+      "step": 186
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.8031692056752926e-05,
+      "loss": 0.4155,
+      "step": 187
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.8007094465663823e-05,
+      "loss": 0.3867,
+      "step": 188
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7982361129210172e-05,
+      "loss": 0.4058,
+      "step": 189
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7957492466699606e-05,
+      "loss": 0.394,
+      "step": 190
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.793248889973395e-05,
+      "loss": 0.4062,
+      "step": 191
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7907350852202078e-05,
+      "loss": 0.3909,
+      "step": 192
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.788207875027274e-05,
+      "loss": 0.394,
+      "step": 193
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.785667302238734e-05,
+      "loss": 0.4043,
+      "step": 194
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7831134099252633e-05,
+      "loss": 0.4062,
+      "step": 195
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.780546241383347e-05,
+      "loss": 0.3811,
+      "step": 196
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7779658401345437e-05,
+      "loss": 0.3818,
+      "step": 197
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.775372249924748e-05,
+      "loss": 0.3811,
+      "step": 198
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.772765514723448e-05,
+      "loss": 0.3726,
+      "step": 199
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7701456787229805e-05,
+      "loss": 0.3767,
+      "step": 200
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.767512786337784e-05,
+      "loss": 0.3813,
+      "step": 201
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.764866882203641e-05,
+      "loss": 0.3943,
+      "step": 202
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7622080111769257e-05,
+      "loss": 0.376,
+      "step": 203
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.759536218333841e-05,
+      "loss": 0.3779,
+      "step": 204
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7568515489696558e-05,
+      "loss": 0.3838,
+      "step": 205
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7541540485979357e-05,
+      "loss": 0.3579,
+      "step": 206
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.751443762949772e-05,
+      "loss": 0.4128,
+      "step": 207
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7487207379730078e-05,
+      "loss": 0.3806,
+      "step": 208
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7459850198314562e-05,
+      "loss": 0.3726,
+      "step": 209
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7432366549041203e-05,
+      "loss": 0.3682,
+      "step": 210
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7404756897844054e-05,
+      "loss": 0.3828,
+      "step": 211
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.73770217127933e-05,
+      "loss": 0.3622,
+      "step": 212
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7349161464087312e-05,
+      "loss": 0.4253,
+      "step": 213
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.732117662404469e-05,
+      "loss": 0.3638,
+      "step": 214
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.729306766709624e-05,
+      "loss": 0.4026,
+      "step": 215
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7264835069776945e-05,
+      "loss": 0.4021,
+      "step": 216
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7236479310717878e-05,
+      "loss": 0.3625,
+      "step": 217
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7208000870638094e-05,
+      "loss": 0.3806,
+      "step": 218
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.7179400232336462e-05,
+      "loss": 0.3914,
+      "step": 219
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.7150677880683515e-05,
+      "loss": 0.3923,
+      "step": 220
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.712183430261319e-05,
+      "loss": 0.3843,
+      "step": 221
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.70928699871146e-05,
+      "loss": 0.3945,
+      "step": 222
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.706378542522374e-05,
+      "loss": 0.3948,
+      "step": 223
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.7034581110015156e-05,
+      "loss": 0.3804,
+      "step": 224
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.7005257536593577e-05,
+      "loss": 0.3738,
+      "step": 225
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6975815202085556e-05,
+      "loss": 0.3582,
+      "step": 226
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6946254605630995e-05,
+      "loss": 0.3545,
+      "step": 227
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.691657624837472e-05,
+      "loss": 0.3586,
+      "step": 228
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6886780633457975e-05,
+      "loss": 0.3726,
+      "step": 229
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6856868266009874e-05,
+      "loss": 0.3818,
+      "step": 230
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6826839653138872e-05,
+      "loss": 0.3713,
+      "step": 231
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.679669530392413e-05,
+      "loss": 0.3711,
+      "step": 232
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6766435729406913e-05,
+      "loss": 0.3855,
+      "step": 233
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6736061442581922e-05,
+      "loss": 0.3943,
+      "step": 234
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6705572958388576e-05,
+      "loss": 0.3867,
+      "step": 235
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.667497079370231e-05,
+      "loss": 0.3679,
+      "step": 236
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6644255467325793e-05,
+      "loss": 0.3945,
+      "step": 237
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6613427499980143e-05,
+      "loss": 0.3906,
+      "step": 238
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6582487414296097e-05,
+      "loss": 0.3706,
+      "step": 239
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.655143573480515e-05,
+      "loss": 0.3794,
+      "step": 240
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6520272987930652e-05,
+      "loss": 0.376,
+      "step": 241
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6488999701978905e-05,
+      "loss": 0.3486,
+      "step": 242
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.645761640713019e-05,
+      "loss": 0.3806,
+      "step": 243
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6426123635429787e-05,
+      "loss": 0.3521,
+      "step": 244
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.639452192077895e-05,
+      "loss": 0.3962,
+      "step": 245
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6362811798925852e-05,
+      "loss": 0.3372,
+      "step": 246
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.633099380745652e-05,
+      "loss": 0.3545,
+      "step": 247
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.62990684857857e-05,
+      "loss": 0.3774,
+      "step": 248
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6267036375147728e-05,
+      "loss": 0.3672,
+      "step": 249
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.3418,
+      "step": 250
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6202653960950474e-05,
+      "loss": 0.3789,
+      "step": 251
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.617030474887505e-05,
+      "loss": 0.3457,
+      "step": 252
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.613785093078166e-05,
+      "loss": 0.3298,
+      "step": 253
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.6105293056864314e-05,
+      "loss": 0.3906,
+      "step": 254
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.607263167908109e-05,
+      "loss": 0.3518,
+      "step": 255
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.6039867351144778e-05,
+      "loss": 0.3496,
+      "step": 256
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.6007000628513498e-05,
+      "loss": 0.3816,
+      "step": 257
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.597403206838128e-05,
+      "loss": 0.385,
+      "step": 258
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5940962229668625e-05,
+      "loss": 0.3774,
+      "step": 259
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5907791673013016e-05,
+      "loss": 0.3354,
+      "step": 260
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5874520960759423e-05,
+      "loss": 0.3489,
+      "step": 261
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.584115065695077e-05,
+      "loss": 0.3491,
+      "step": 262
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5807681327318372e-05,
+      "loss": 0.3892,
+      "step": 263
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5774113539272332e-05,
+      "loss": 0.3838,
+      "step": 264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5740447861891946e-05,
+      "loss": 0.3872,
+      "step": 265
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5706684865916025e-05,
+      "loss": 0.324,
+      "step": 266
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5672825123733257e-05,
+      "loss": 0.3145,
+      "step": 267
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.563886920937246e-05,
+      "loss": 0.2955,
+      "step": 268
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5604817698492886e-05,
+      "loss": 0.3337,
+      "step": 269
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.557067116837444e-05,
+      "loss": 0.3225,
+      "step": 270
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5536430197907904e-05,
+      "loss": 0.3074,
+      "step": 271
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5502095367585124e-05,
+      "loss": 0.2902,
+      "step": 272
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5467667259489157e-05,
+      "loss": 0.3196,
+      "step": 273
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.543314645728442e-05,
+      "loss": 0.3165,
+      "step": 274
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.539853354620678e-05,
+      "loss": 0.3162,
+      "step": 275
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5363829113053633e-05,
+      "loss": 0.307,
+      "step": 276
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5329033746173975e-05,
+      "loss": 0.3289,
+      "step": 277
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5294148035458406e-05,
+      "loss": 0.3269,
+      "step": 278
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5259172572329132e-05,
+      "loss": 0.3296,
+      "step": 279
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5224107949729952e-05,
+      "loss": 0.3134,
+      "step": 280
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5188954762116197e-05,
+      "loss": 0.3328,
+      "step": 281
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.515371360544465e-05,
+      "loss": 0.2966,
+      "step": 282
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.5118385077163446e-05,
+      "loss": 0.3091,
+      "step": 283
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.5082969776201948e-05,
+      "loss": 0.3149,
+      "step": 284
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.5047468302960577e-05,
+      "loss": 0.2905,
+      "step": 285
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.5011881259300654e-05,
+      "loss": 0.3082,
+      "step": 286
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4976209248534183e-05,
+      "loss": 0.3025,
+      "step": 287
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4940452875413627e-05,
+      "loss": 0.3101,
+      "step": 288
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4904612746121657e-05,
+      "loss": 0.304,
+      "step": 289
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4868689468260876e-05,
+      "loss": 0.3428,
+      "step": 290
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.483268365084351e-05,
+      "loss": 0.2863,
+      "step": 291
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.479659590428109e-05,
+      "loss": 0.3323,
+      "step": 292
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.476042684037411e-05,
+      "loss": 0.3051,
+      "step": 293
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4724177072301642e-05,
+      "loss": 0.2874,
+      "step": 294
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4687847214610944e-05,
+      "loss": 0.3176,
+      "step": 295
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4651437883207056e-05,
+      "loss": 0.3197,
+      "step": 296
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4614949695342335e-05,
+      "loss": 0.3259,
+      "step": 297
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4578383269606004e-05,
+      "loss": 0.2832,
+      "step": 298
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4541739225913669e-05,
+      "loss": 0.3047,
+      "step": 299
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4505018185496802e-05,
+      "loss": 0.318,
+      "step": 300
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4468220770892208e-05,
+      "loss": 0.2836,
+      "step": 301
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.443134760593147e-05,
+      "loss": 0.275,
+      "step": 302
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4394399315730389e-05,
+      "loss": 0.3,
+      "step": 303
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4357376526678368e-05,
+      "loss": 0.2927,
+      "step": 304
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4320279866427798e-05,
+      "loss": 0.3098,
+      "step": 305
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.428310996388342e-05,
+      "loss": 0.2933,
+      "step": 306
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.424586744919166e-05,
+      "loss": 0.2913,
+      "step": 307
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4208552953729949e-05,
+      "loss": 0.2844,
+      "step": 308
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.4171167110096017e-05,
+      "loss": 0.2993,
+      "step": 309
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.4133710552097175e-05,
+      "loss": 0.2803,
+      "step": 310
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.4096183914739554e-05,
+      "loss": 0.3308,
+      "step": 311
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.4058587834217356e-05,
+      "loss": 0.3281,
+      "step": 312
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.4020922947902067e-05,
+      "loss": 0.3112,
+      "step": 313
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3983189894331636e-05,
+      "loss": 0.3196,
+      "step": 314
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3945389313199669e-05,
+      "loss": 0.313,
+      "step": 315
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3907521845344571e-05,
+      "loss": 0.2998,
+      "step": 316
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3869588132738696e-05,
+      "loss": 0.317,
+      "step": 317
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3831588818477437e-05,
+      "loss": 0.3042,
+      "step": 318
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3793524546768358e-05,
+      "loss": 0.3219,
+      "step": 319
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.375539596292025e-05,
+      "loss": 0.3298,
+      "step": 320
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.371720371333219e-05,
+      "loss": 0.2942,
+      "step": 321
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3678948445482598e-05,
+      "loss": 0.2849,
+      "step": 322
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3640630807918246e-05,
+      "loss": 0.2966,
+      "step": 323
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3602251450243273e-05,
+      "loss": 0.2968,
+      "step": 324
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3563811023108157e-05,
+      "loss": 0.2969,
+      "step": 325
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3525310178198707e-05,
+      "loss": 0.3016,
+      "step": 326
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3486749568225002e-05,
+      "loss": 0.2727,
+      "step": 327
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3448129846910312e-05,
+      "loss": 0.2738,
+      "step": 328
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3409451668980047e-05,
+      "loss": 0.2787,
+      "step": 329
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3370715690150631e-05,
+      "loss": 0.2789,
+      "step": 330
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3331922567118394e-05,
+      "loss": 0.285,
+      "step": 331
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3293072957548443e-05,
+      "loss": 0.2745,
+      "step": 332
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.325416752006351e-05,
+      "loss": 0.3013,
+      "step": 333
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.321520691423278e-05,
+      "loss": 0.2968,
+      "step": 334
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.317619180056072e-05,
+      "loss": 0.2772,
+      "step": 335
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.313712284047587e-05,
+      "loss": 0.2599,
+      "step": 336
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.3098000696319642e-05,
+      "loss": 0.283,
+      "step": 337
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.305882603133508e-05,
+      "loss": 0.2819,
+      "step": 338
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.301959950965562e-05,
+      "loss": 0.2982,
+      "step": 339
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2980321796293838e-05,
+      "loss": 0.2552,
+      "step": 340
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2940993557130166e-05,
+      "loss": 0.2743,
+      "step": 341
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2901615458901602e-05,
+      "loss": 0.2858,
+      "step": 342
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2862188169190419e-05,
+      "loss": 0.2767,
+      "step": 343
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.282271235641284e-05,
+      "loss": 0.3038,
+      "step": 344
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2783188689807697e-05,
+      "loss": 0.2587,
+      "step": 345
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.27436178394251e-05,
+      "loss": 0.2593,
+      "step": 346
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2704000476115079e-05,
+      "loss": 0.2947,
+      "step": 347
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2664337271516194e-05,
+      "loss": 0.3076,
+      "step": 348
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.262462889804416e-05,
+      "loss": 0.2772,
+      "step": 349
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2584876028880455e-05,
+      "loss": 0.2466,
+      "step": 350
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2545079337960883e-05,
+      "loss": 0.2823,
+      "step": 351
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2505239499964179e-05,
+      "loss": 0.3049,
+      "step": 352
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.246535719030055e-05,
+      "loss": 0.2764,
+      "step": 353
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2425433085100224e-05,
+      "loss": 0.2744,
+      "step": 354
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.238546786120201e-05,
+      "loss": 0.2751,
+      "step": 355
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2345462196141797e-05,
+      "loss": 0.2932,
+      "step": 356
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2305416768141082e-05,
+      "loss": 0.2756,
+      "step": 357
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2265332256095463e-05,
+      "loss": 0.2816,
+      "step": 358
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.3037,
+      "step": 359
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2185048698753403e-05,
+      "loss": 0.2765,
+      "step": 360
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2144851014515055e-05,
+      "loss": 0.274,
+      "step": 361
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.2104616968324928e-05,
+      "loss": 0.2786,
+      "step": 362
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.2064347242276293e-05,
+      "loss": 0.2832,
+      "step": 363
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.202404251906731e-05,
+      "loss": 0.2985,
+      "step": 364
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1983703481989443e-05,
+      "loss": 0.2792,
+      "step": 365
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1943330814915897e-05,
+      "loss": 0.3042,
+      "step": 366
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1902925202289997e-05,
+      "loss": 0.2667,
+      "step": 367
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1862487329113606e-05,
+      "loss": 0.2931,
+      "step": 368
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1822017880935507e-05,
+      "loss": 0.2621,
+      "step": 369
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.178151754383977e-05,
+      "loss": 0.2799,
+      "step": 370
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1740987004434137e-05,
+      "loss": 0.302,
+      "step": 371
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1700426949838364e-05,
+      "loss": 0.2913,
+      "step": 372
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.16598380676726e-05,
+      "loss": 0.276,
+      "step": 373
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1619221046045688e-05,
+      "loss": 0.2812,
+      "step": 374
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1578576573543541e-05,
+      "loss": 0.291,
+      "step": 375
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1537905339217448e-05,
+      "loss": 0.3162,
+      "step": 376
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1497208032572385e-05,
+      "loss": 0.2717,
+      "step": 377
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1456485343555344e-05,
+      "loss": 0.2635,
+      "step": 378
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.141573796254363e-05,
+      "loss": 0.2666,
+      "step": 379
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1374966580333147e-05,
+      "loss": 0.2942,
+      "step": 380
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1334171888126698e-05,
+      "loss": 0.2633,
+      "step": 381
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1293354577522264e-05,
+      "loss": 0.2731,
+      "step": 382
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1252515340501282e-05,
+      "loss": 0.238,
+      "step": 383
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1211654869416901e-05,
+      "loss": 0.2789,
+      "step": 384
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1170773856982268e-05,
+      "loss": 0.2583,
+      "step": 385
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.1129872996258757e-05,
+      "loss": 0.2806,
+      "step": 386
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.1088952980644242e-05,
+      "loss": 0.2756,
+      "step": 387
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.1048014503861321e-05,
+      "loss": 0.2637,
+      "step": 388
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.1007058259945584e-05,
+      "loss": 0.25,
+      "step": 389
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0966084943233818e-05,
+      "loss": 0.2723,
+      "step": 390
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.092509524835224e-05,
+      "loss": 0.2329,
+      "step": 391
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0884089870204751e-05,
+      "loss": 0.2628,
+      "step": 392
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0843069503961112e-05,
+      "loss": 0.2418,
+      "step": 393
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0802034845045189e-05,
+      "loss": 0.2628,
+      "step": 394
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0760986589123145e-05,
+      "loss": 0.2655,
+      "step": 395
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0719925432091671e-05,
+      "loss": 0.2441,
+      "step": 396
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.067885207006616e-05,
+      "loss": 0.2496,
+      "step": 397
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0637767199368911e-05,
+      "loss": 0.2797,
+      "step": 398
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0596671516517356e-05,
+      "loss": 0.2412,
+      "step": 399
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0555565718212198e-05,
+      "loss": 0.2534,
+      "step": 400
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0514450501325646e-05,
+      "loss": 0.2886,
+      "step": 401
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0473326562889583e-05,
+      "loss": 0.2683,
+      "step": 402
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.043219460008374e-05,
+      "loss": 0.2771,
+      "step": 403
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0391055310223899e-05,
+      "loss": 0.2427,
+      "step": 404
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0349909390750046e-05,
+      "loss": 0.2482,
+      "step": 405
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0308757539214573e-05,
+      "loss": 0.2629,
+      "step": 406
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0267600453270422e-05,
+      "loss": 0.2706,
+      "step": 407
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.022643883065929e-05,
+      "loss": 0.2333,
+      "step": 408
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0185273369199781e-05,
+      "loss": 0.2584,
+      "step": 409
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.0144104766775574e-05,
+      "loss": 0.2981,
+      "step": 410
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.01029337213236e-05,
+      "loss": 0.267,
+      "step": 411
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.006176093082221e-05,
+      "loss": 0.2473,
+      "step": 412
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.0020587093279339e-05,
+      "loss": 0.2633,
+      "step": 413
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.979412906720663e-06,
+      "loss": 0.2687,
+      "step": 414
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.938239069177792e-06,
+      "loss": 0.2538,
+      "step": 415
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.897066278676405e-06,
+      "loss": 0.2701,
+      "step": 416
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.855895233224431e-06,
+      "loss": 0.2322,
+      "step": 417
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.81472663080022e-06,
+      "loss": 0.2654,
+      "step": 418
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.773561169340711e-06,
+      "loss": 0.2854,
+      "step": 419
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.732399546729578e-06,
+      "loss": 0.2571,
+      "step": 420
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.691242460785433e-06,
+      "loss": 0.2462,
+      "step": 421
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.650090609249957e-06,
+      "loss": 0.2491,
+      "step": 422
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.608944689776104e-06,
+      "loss": 0.2666,
+      "step": 423
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.56780539991626e-06,
+      "loss": 0.2628,
+      "step": 424
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.52667343711042e-06,
+      "loss": 0.2853,
+      "step": 425
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.485549498674357e-06,
+      "loss": 0.2388,
+      "step": 426
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.444434281787806e-06,
+      "loss": 0.2714,
+      "step": 427
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.403328483482647e-06,
+      "loss": 0.2351,
+      "step": 428
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.362232800631087e-06,
+      "loss": 0.2605,
+      "step": 429
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.321147929933847e-06,
+      "loss": 0.2599,
+      "step": 430
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.28007456790833e-06,
+      "loss": 0.2548,
+      "step": 431
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.239013410876856e-06,
+      "loss": 0.2678,
+      "step": 432
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.197965154954815e-06,
+      "loss": 0.2654,
+      "step": 433
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.15693049603889e-06,
+      "loss": 0.2583,
+      "step": 434
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.115910129795252e-06,
+      "loss": 0.264,
+      "step": 435
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.074904751647762e-06,
+      "loss": 0.2699,
+      "step": 436
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.033915056766187e-06,
+      "loss": 0.2476,
+      "step": 437
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.992941740054418e-06,
+      "loss": 0.2108,
+      "step": 438
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.951985496138679e-06,
+      "loss": 0.2362,
+      "step": 439
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.911047019355763e-06,
+      "loss": 0.2617,
+      "step": 440
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.870127003741245e-06,
+      "loss": 0.2454,
+      "step": 441
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.829226143017735e-06,
+      "loss": 0.2444,
+      "step": 442
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.788345130583099e-06,
+      "loss": 0.2322,
+      "step": 443
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.747484659498725e-06,
+      "loss": 0.2427,
+      "step": 444
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.706645422477739e-06,
+      "loss": 0.2528,
+      "step": 445
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.665828111873307e-06,
+      "loss": 0.2742,
+      "step": 446
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.625033419666858e-06,
+      "loss": 0.2197,
+      "step": 447
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.584262037456374e-06,
+      "loss": 0.244,
+      "step": 448
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.54351465644466e-06,
+      "loss": 0.2678,
+      "step": 449
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.50279196742762e-06,
+      "loss": 0.2372,
+      "step": 450
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.462094660782555e-06,
+      "loss": 0.2557,
+      "step": 451
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.42142342645646e-06,
+      "loss": 0.2335,
+      "step": 452
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.380778953954314e-06,
+      "loss": 0.2389,
+      "step": 453
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.340161932327405e-06,
+      "loss": 0.2683,
+      "step": 454
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.299573050161637e-06,
+      "loss": 0.2378,
+      "step": 455
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.259012995565868e-06,
+      "loss": 0.2603,
+      "step": 456
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.21848245616023e-06,
+      "loss": 0.2318,
+      "step": 457
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.177982119064497e-06,
+      "loss": 0.2061,
+      "step": 458
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.137512670886397e-06,
+      "loss": 0.2288,
+      "step": 459
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.097074797710007e-06,
+      "loss": 0.2244,
+      "step": 460
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.056669185084108e-06,
+      "loss": 0.2323,
+      "step": 461
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.016296518010558e-06,
+      "loss": 0.2546,
+      "step": 462
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.975957480932695e-06,
+      "loss": 0.2544,
+      "step": 463
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.935652757723709e-06,
+      "loss": 0.2522,
+      "step": 464
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.895383031675074e-06,
+      "loss": 0.2426,
+      "step": 465
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.855148985484946e-06,
+      "loss": 0.225,
+      "step": 466
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.814951301246598e-06,
+      "loss": 0.2477,
+      "step": 467
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.2391,
+      "step": 468
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.73466774390454e-06,
+      "loss": 0.2243,
+      "step": 469
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.694583231858921e-06,
+      "loss": 0.2365,
+      "step": 470
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.654537803858205e-06,
+      "loss": 0.2336,
+      "step": 471
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.614532138797994e-06,
+      "loss": 0.2428,
+      "step": 472
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.574566914899779e-06,
+      "loss": 0.2498,
+      "step": 473
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.534642809699455e-06,
+      "loss": 0.2444,
+      "step": 474
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.494760500035824e-06,
+      "loss": 0.2194,
+      "step": 475
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.454920662039118e-06,
+      "loss": 0.2544,
+      "step": 476
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.415123971119549e-06,
+      "loss": 0.2506,
+      "step": 477
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.375371101955842e-06,
+      "loss": 0.2413,
+      "step": 478
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.335662728483808e-06,
+      "loss": 0.2262,
+      "step": 479
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.295999523884921e-06,
+      "loss": 0.2096,
+      "step": 480
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.256382160574902e-06,
+      "loss": 0.2446,
+      "step": 481
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.2168113101923085e-06,
+      "loss": 0.2427,
+      "step": 482
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.177287643587164e-06,
+      "loss": 0.2389,
+      "step": 483
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.1378118308095835e-06,
+      "loss": 0.2334,
+      "step": 484
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.0983845410984e-06,
+      "loss": 0.2528,
+      "step": 485
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.05900644286984e-06,
+      "loss": 0.2317,
+      "step": 486
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.019678203706164e-06,
+      "loss": 0.2352,
+      "step": 487
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.980400490344383e-06,
+      "loss": 0.2289,
+      "step": 488
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.941173968664923e-06,
+      "loss": 0.2273,
+      "step": 489
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.901999303680359e-06,
+      "loss": 0.1824,
+      "step": 490
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.862877159524133e-06,
+      "loss": 0.2325,
+      "step": 491
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.8238081994392836e-06,
+      "loss": 0.2399,
+      "step": 492
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.7847930857672205e-06,
+      "loss": 0.2269,
+      "step": 493
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.745832479936492e-06,
+      "loss": 0.2161,
+      "step": 494
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.706927042451561e-06,
+      "loss": 0.2291,
+      "step": 495
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.66807743288161e-06,
+      "loss": 0.2207,
+      "step": 496
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.629284309849373e-06,
+      "loss": 0.2136,
+      "step": 497
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.590548331019957e-06,
+      "loss": 0.2098,
+      "step": 498
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.55187015308969e-06,
+      "loss": 0.2432,
+      "step": 499
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.513250431775003e-06,
+      "loss": 0.2526,
+      "step": 500
+    },
+    {
+      "epoch": 1.89,
+      "eval_loss": 0.2714967727661133,
+      "eval_runtime": 477.7572,
+      "eval_samples_per_second": 1.448,
+      "eval_steps_per_second": 0.362,
+      "step": 500
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.474689821801295e-06,
+      "loss": 0.2352,
+      "step": 501
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.436188976891846e-06,
+      "loss": 0.2379,
+      "step": 502
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.39774854975673e-06,
+      "loss": 0.217,
+      "step": 503
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.359369192081756e-06,
+      "loss": 0.2463,
+      "step": 504
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.321051554517406e-06,
+      "loss": 0.241,
+      "step": 505
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.282796286667814e-06,
+      "loss": 0.2594,
+      "step": 506
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.244604037079754e-06,
+      "loss": 0.2195,
+      "step": 507
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.206475453231644e-06,
+      "loss": 0.2463,
+      "step": 508
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.168411181522569e-06,
+      "loss": 0.239,
+      "step": 509
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.13041186726131e-06,
+      "loss": 0.2115,
+      "step": 510
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.092478154655431e-06,
+      "loss": 0.1979,
+      "step": 511
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.054610686800333e-06,
+      "loss": 0.215,
+      "step": 512
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 6.016810105668365e-06,
+      "loss": 0.2428,
+      "step": 513
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.979077052097936e-06,
+      "loss": 0.2042,
+      "step": 514
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.941412165782645e-06,
+      "loss": 0.2126,
+      "step": 515
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.903816085260447e-06,
+      "loss": 0.223,
+      "step": 516
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.866289447902829e-06,
+      "loss": 0.2163,
+      "step": 517
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.828832889903983e-06,
+      "loss": 0.2302,
+      "step": 518
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.791447046270055e-06,
+      "loss": 0.2329,
+      "step": 519
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.754132550808345e-06,
+      "loss": 0.1956,
+      "step": 520
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.716890036116582e-06,
+      "loss": 0.2402,
+      "step": 521
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.6797201335722064e-06,
+      "loss": 0.2158,
+      "step": 522
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.642623473321638e-06,
+      "loss": 0.2141,
+      "step": 523
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.6056006842696145e-06,
+      "loss": 0.2346,
+      "step": 524
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.568652394068532e-06,
+      "loss": 0.1821,
+      "step": 525
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.531779229107797e-06,
+      "loss": 0.21,
+      "step": 526
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.494981814503199e-06,
+      "loss": 0.1902,
+      "step": 527
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.458260774086332e-06,
+      "loss": 0.1962,
+      "step": 528
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.421616730394e-06,
+      "loss": 0.2277,
+      "step": 529
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.38505030465767e-06,
+      "loss": 0.2076,
+      "step": 530
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.348562116792946e-06,
+      "loss": 0.2212,
+      "step": 531
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.312152785389056e-06,
+      "loss": 0.1982,
+      "step": 532
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.275822927698362e-06,
+      "loss": 0.2003,
+      "step": 533
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.2395731596258925e-06,
+      "loss": 0.184,
+      "step": 534
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.20340409571891e-06,
+      "loss": 0.1873,
+      "step": 535
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.167316349156495e-06,
+      "loss": 0.1909,
+      "step": 536
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.131310531739129e-06,
+      "loss": 0.1779,
+      "step": 537
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.095387253878346e-06,
+      "loss": 0.1965,
+      "step": 538
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.0595471245863745e-06,
+      "loss": 0.184,
+      "step": 539
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.023790751465818e-06,
+      "loss": 0.1933,
+      "step": 540
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.9881187406993455e-06,
+      "loss": 0.158,
+      "step": 541
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.952531697039424e-06,
+      "loss": 0.1793,
+      "step": 542
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.917030223798057e-06,
+      "loss": 0.1794,
+      "step": 543
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.881614922836555e-06,
+      "loss": 0.1873,
+      "step": 544
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.846286394555352e-06,
+      "loss": 0.1944,
+      "step": 545
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.811045237883803e-06,
+      "loss": 0.1942,
+      "step": 546
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.775892050270051e-06,
+      "loss": 0.1942,
+      "step": 547
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.740827427670871e-06,
+      "loss": 0.183,
+      "step": 548
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.7058519645416004e-06,
+      "loss": 0.1706,
+      "step": 549
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.670966253826027e-06,
+      "loss": 0.1782,
+      "step": 550
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.636170886946371e-06,
+      "loss": 0.215,
+      "step": 551
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.601466453793224e-06,
+      "loss": 0.1782,
+      "step": 552
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.5668535427155816e-06,
+      "loss": 0.1744,
+      "step": 553
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.532332740510842e-06,
+      "loss": 0.1895,
+      "step": 554
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.497904632414879e-06,
+      "loss": 0.1639,
+      "step": 555
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.4635698020921016e-06,
+      "loss": 0.1809,
+      "step": 556
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.429328831625565e-06,
+      "loss": 0.183,
+      "step": 557
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.3951823015071186e-06,
+      "loss": 0.1825,
+      "step": 558
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.361130790627541e-06,
+      "loss": 0.2015,
+      "step": 559
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.327174876266743e-06,
+      "loss": 0.1918,
+      "step": 560
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.293315134083975e-06,
+      "loss": 0.1882,
+      "step": 561
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.259552138108061e-06,
+      "loss": 0.215,
+      "step": 562
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.225886460727671e-06,
+      "loss": 0.1973,
+      "step": 563
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.192318672681631e-06,
+      "loss": 0.1719,
+      "step": 564
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.158849343049233e-06,
+      "loss": 0.1976,
+      "step": 565
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.12547903924058e-06,
+      "loss": 0.2042,
+      "step": 566
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.092208326986986e-06,
+      "loss": 0.1572,
+      "step": 567
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.059037770331379e-06,
+      "loss": 0.1841,
+      "step": 568
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 4.025967931618722e-06,
+      "loss": 0.1952,
+      "step": 569
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.992999371486508e-06,
+      "loss": 0.1924,
+      "step": 570
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.960132648855226e-06,
+      "loss": 0.1885,
+      "step": 571
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.9273683209189115e-06,
+      "loss": 0.1849,
+      "step": 572
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.894706943135686e-06,
+      "loss": 0.1721,
+      "step": 573
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.862149069218343e-06,
+      "loss": 0.181,
+      "step": 574
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.829695251124953e-06,
+      "loss": 0.2257,
+      "step": 575
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.797346039049529e-06,
+      "loss": 0.1974,
+      "step": 576
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 0.1716,
+      "step": 577
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.732963624852275e-06,
+      "loss": 0.1873,
+      "step": 578
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.7009315142143033e-06,
+      "loss": 0.1727,
+      "step": 579
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.6690061925434817e-06,
+      "loss": 0.2046,
+      "step": 580
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.637188201074149e-06,
+      "loss": 0.1947,
+      "step": 581
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.6054780792210542e-06,
+      "loss": 0.1555,
+      "step": 582
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.5738763645702145e-06,
+      "loss": 0.1697,
+      "step": 583
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.5423835928698126e-06,
+      "loss": 0.2059,
+      "step": 584
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.511000298021098e-06,
+      "loss": 0.1851,
+      "step": 585
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.479727012069349e-06,
+      "loss": 0.1702,
+      "step": 586
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.4485642651948516e-06,
+      "loss": 0.1691,
+      "step": 587
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.4175125857039027e-06,
+      "loss": 0.1613,
+      "step": 588
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.38657250001986e-06,
+      "loss": 0.1934,
+      "step": 589
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.355744532674211e-06,
+      "loss": 0.1686,
+      "step": 590
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.325029206297694e-06,
+      "loss": 0.1669,
+      "step": 591
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.2944270416114256e-06,
+      "loss": 0.1832,
+      "step": 592
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.2639385574180825e-06,
+      "loss": 0.1459,
+      "step": 593
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.2335642705930883e-06,
+      "loss": 0.1718,
+      "step": 594
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.2033046960758763e-06,
+      "loss": 0.1847,
+      "step": 595
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.173160346861134e-06,
+      "loss": 0.1511,
+      "step": 596
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.1431317339901267e-06,
+      "loss": 0.18,
+      "step": 597
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.1132193665420306e-06,
+      "loss": 0.1697,
+      "step": 598
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0834237516252817e-06,
+      "loss": 0.1908,
+      "step": 599
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0537453943690076e-06,
+      "loss": 0.166,
+      "step": 600
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 3.024184797914449e-06,
+      "loss": 0.174,
+      "step": 601
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.994742463406427e-06,
+      "loss": 0.1539,
+      "step": 602
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.96541888998485e-06,
+      "loss": 0.1866,
+      "step": 603
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.9362145747762626e-06,
+      "loss": 0.1911,
+      "step": 604
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.9071300128854007e-06,
+      "loss": 0.1734,
+      "step": 605
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.878165697386812e-06,
+      "loss": 0.1763,
+      "step": 606
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.8493221193164886e-06,
+      "loss": 0.1606,
+      "step": 607
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.82059976766354e-06,
+      "loss": 0.1938,
+      "step": 608
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.79199912936191e-06,
+      "loss": 0.1426,
+      "step": 609
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.763520689282122e-06,
+      "loss": 0.1766,
+      "step": 610
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.7351649302230553e-06,
+      "loss": 0.1729,
+      "step": 611
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.7069323329037632e-06,
+      "loss": 0.1669,
+      "step": 612
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.678823375955314e-06,
+      "loss": 0.1874,
+      "step": 613
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.650838535912692e-06,
+      "loss": 0.1873,
+      "step": 614
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.6229782872067042e-06,
+      "loss": 0.1581,
+      "step": 615
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.595243102155951e-06,
+      "loss": 0.1828,
+      "step": 616
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.567633450958801e-06,
+      "loss": 0.1967,
+      "step": 617
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.540149801685441e-06,
+      "loss": 0.1593,
+      "step": 618
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.512792620269924e-06,
+      "loss": 0.1614,
+      "step": 619
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.485562370502279e-06,
+      "loss": 0.1665,
+      "step": 620
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.4584595140206457e-06,
+      "loss": 0.1723,
+      "step": 621
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.4314845103034456e-06,
+      "loss": 0.1884,
+      "step": 622
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.40463781666159e-06,
+      "loss": 0.202,
+      "step": 623
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3779198882307443e-06,
+      "loss": 0.1616,
+      "step": 624
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.3513311779635904e-06,
+      "loss": 0.2026,
+      "step": 625
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.324872136622164e-06,
+      "loss": 0.1823,
+      "step": 626
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2985432127701945e-06,
+      "loss": 0.1833,
+      "step": 627
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2723448527655267e-06,
+      "loss": 0.1729,
+      "step": 628
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.246277500752524e-06,
+      "loss": 0.1671,
+      "step": 629
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.220341598654565e-06,
+      "loss": 0.1648,
+      "step": 630
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.194537586166532e-06,
+      "loss": 0.1545,
+      "step": 631
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.16886590074737e-06,
+      "loss": 0.1763,
+      "step": 632
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.143326977612662e-06,
+      "loss": 0.1793,
+      "step": 633
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.1179212497272582e-06,
+      "loss": 0.1722,
+      "step": 634
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0926491477979272e-06,
+      "loss": 0.1847,
+      "step": 635
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0675111002660566e-06,
+      "loss": 0.1674,
+      "step": 636
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.042507533300395e-06,
+      "loss": 0.1815,
+      "step": 637
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.0176388707898274e-06,
+      "loss": 0.1676,
+      "step": 638
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.992905534336178e-06,
+      "loss": 0.1753,
+      "step": 639
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9683079432470774e-06,
+      "loss": 0.1696,
+      "step": 640
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9438465145288377e-06,
+      "loss": 0.189,
+      "step": 641
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9195216628793956e-06,
+      "loss": 0.1935,
+      "step": 642
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8953338006812805e-06,
+      "loss": 0.1814,
+      "step": 643
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8712833379946217e-06,
+      "loss": 0.1982,
+      "step": 644
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.847370682550187e-06,
+      "loss": 0.182,
+      "step": 645
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.82359623974249e-06,
+      "loss": 0.1696,
+      "step": 646
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7999604126229043e-06,
+      "loss": 0.1931,
+      "step": 647
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7764636018928249e-06,
+      "loss": 0.1597,
+      "step": 648
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.753106205896895e-06,
+      "loss": 0.1479,
+      "step": 649
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.729888620616228e-06,
+      "loss": 0.148,
+      "step": 650
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.7068112396617164e-06,
+      "loss": 0.1808,
+      "step": 651
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6838744542673492e-06,
+      "loss": 0.1599,
+      "step": 652
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6610786532835776e-06,
+      "loss": 0.1671,
+      "step": 653
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.6384242231707203e-06,
+      "loss": 0.1345,
+      "step": 654
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.6159115479924259e-06,
+      "loss": 0.1743,
+      "step": 655
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.593541009409143e-06,
+      "loss": 0.1677,
+      "step": 656
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5713129866716647e-06,
+      "loss": 0.1418,
+      "step": 657
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5492278566146945e-06,
+      "loss": 0.1627,
+      "step": 658
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.5272859936504513e-06,
+      "loss": 0.1649,
+      "step": 659
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.5054877697623305e-06,
+      "loss": 0.1418,
+      "step": 660
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4838335544985982e-06,
+      "loss": 0.1819,
+      "step": 661
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.462323714966114e-06,
+      "loss": 0.1536,
+      "step": 662
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4409586158241272e-06,
+      "loss": 0.1654,
+      "step": 663
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.4197386192780715e-06,
+      "loss": 0.16,
+      "step": 664
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3986640850734444e-06,
+      "loss": 0.1573,
+      "step": 665
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3777353704897002e-06,
+      "loss": 0.1831,
+      "step": 666
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3569528303341927e-06,
+      "loss": 0.1633,
+      "step": 667
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3363168169361574e-06,
+      "loss": 0.1652,
+      "step": 668
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3158276801407432e-06,
+      "loss": 0.1839,
+      "step": 669
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2954857673030807e-06,
+      "loss": 0.1727,
+      "step": 670
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2752914232823942e-06,
+      "loss": 0.1725,
+      "step": 671
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.255244990436153e-06,
+      "loss": 0.201,
+      "step": 672
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2353468086142639e-06,
+      "loss": 0.1599,
+      "step": 673
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2155972151533225e-06,
+      "loss": 0.1528,
+      "step": 674
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1959965448708731e-06,
+      "loss": 0.2054,
+      "step": 675
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1765451300597574e-06,
+      "loss": 0.1611,
+      "step": 676
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1572433004824635e-06,
+      "loss": 0.162,
+      "step": 677
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1380913833655383e-06,
+      "loss": 0.1448,
+      "step": 678
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1190897033940461e-06,
+      "loss": 0.1481,
+      "step": 679
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.1002385827060602e-06,
+      "loss": 0.1831,
+      "step": 680
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0815383408871983e-06,
+      "loss": 0.1454,
+      "step": 681
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0629892949652133e-06,
+      "loss": 0.1602,
+      "step": 682
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0445917594046073e-06,
+      "loss": 0.1445,
+      "step": 683
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.026346046101312e-06,
+      "loss": 0.1743,
+      "step": 684
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0082524643773916e-06,
+      "loss": 0.1913,
+      "step": 685
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 0.1639,
+      "step": 686
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.725229200552123e-07,
+      "loss": 0.1642,
+      "step": 687
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.548875631847875e-07,
+      "loss": 0.1589,
+      "step": 688
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.374055493391455e-07,
+      "loss": 0.1874,
+      "step": 689
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.200771748932513e-07,
+      "loss": 0.1727,
+      "step": 690
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 9.029027336174023e-07,
+      "loss": 0.1633,
+      "step": 691
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.858825166722418e-07,
+      "loss": 0.1641,
+      "step": 692
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.690168126038301e-07,
+      "loss": 0.1526,
+      "step": 693
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.523059073387474e-07,
+      "loss": 0.1612,
+      "step": 694
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.35750084179251e-07,
+      "loss": 0.1459,
+      "step": 695
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 8.193496237984677e-07,
+      "loss": 0.179,
+      "step": 696
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 8.031048042356393e-07,
+      "loss": 0.2097,
+      "step": 697
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.870159008914069e-07,
+      "loss": 0.1656,
+      "step": 698
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.710831865231461e-07,
+      "loss": 0.1652,
+      "step": 699
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.55306931240335e-07,
+      "loss": 0.1738,
+      "step": 700
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.396874024999811e-07,
+      "loss": 0.1501,
+      "step": 701
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.242248651020845e-07,
+      "loss": 0.1627,
+      "step": 702
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.089195811851502e-07,
+      "loss": 0.1574,
+      "step": 703
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.937718102217461e-07,
+      "loss": 0.1866,
+      "step": 704
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.787818090140985e-07,
+      "loss": 0.1519,
+      "step": 705
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.639498316897419e-07,
+      "loss": 0.1476,
+      "step": 706
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.492761296972117e-07,
+      "loss": 0.1718,
+      "step": 707
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.347609518017761e-07,
+      "loss": 0.1735,
+      "step": 708
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 6.2040454408123e-07,
+      "loss": 0.1581,
+      "step": 709
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 6.062071499217081e-07,
+      "loss": 0.1473,
+      "step": 710
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.921690100135713e-07,
+      "loss": 0.1606,
+      "step": 711
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.782903623473202e-07,
+      "loss": 0.1654,
+      "step": 712
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.645714422095627e-07,
+      "loss": 0.1842,
+      "step": 713
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.510124821790208e-07,
+      "loss": 0.178,
+      "step": 714
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.376137121225933e-07,
+      "loss": 0.1729,
+      "step": 715
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.24375359191458e-07,
+      "loss": 0.1461,
+      "step": 716
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5.11297647817216e-07,
+      "loss": 0.1383,
+      "step": 717
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.983807997080925e-07,
+      "loss": 0.1639,
+      "step": 718
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.856250338451763e-07,
+      "loss": 0.1607,
+      "step": 719
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.7303056647870605e-07,
+      "loss": 0.1604,
+      "step": 720
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.605976111244015e-07,
+      "loss": 0.1653,
+      "step": 721
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.483263785598524e-07,
+      "loss": 0.1677,
+      "step": 722
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.3621707682094063e-07,
+      "loss": 0.1464,
+      "step": 723
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.242699111983051e-07,
+      "loss": 0.1339,
+      "step": 724
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.124850842338779e-07,
+      "loss": 0.1628,
+      "step": 725
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.0086279571743715e-07,
+      "loss": 0.1604,
+      "step": 726
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.8940324268322285e-07,
+      "loss": 0.1587,
+      "step": 727
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.781066194066052e-07,
+      "loss": 0.1534,
+      "step": 728
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.66973117400774e-07,
+      "loss": 0.1775,
+      "step": 729
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.5600292541351e-07,
+      "loss": 0.1565,
+      "step": 730
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.451962294239741e-07,
+      "loss": 0.1609,
+      "step": 731
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.345532126395579e-07,
+      "loss": 0.1426,
+      "step": 732
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.2407405549277683e-07,
+      "loss": 0.1815,
+      "step": 733
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.137589356382076e-07,
+      "loss": 0.1801,
+      "step": 734
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.0360802794948687e-07,
+      "loss": 0.1709,
+      "step": 735
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.936215045163371e-07,
+      "loss": 0.1714,
+      "step": 736
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.8379953464165334e-07,
+      "loss": 0.1487,
+      "step": 737
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.741422848386266e-07,
+      "loss": 0.1533,
+      "step": 738
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.646499188279328e-07,
+      "loss": 0.1635,
+      "step": 739
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.5532259753494825e-07,
+      "loss": 0.1524,
+      "step": 740
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.461604790870209e-07,
+      "loss": 0.1764,
+      "step": 741
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.3716371881079558e-07,
+      "loss": 0.1636,
+      "step": 742
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2833246922957408e-07,
+      "loss": 0.1606,
+      "step": 743
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.196668800607382e-07,
+      "loss": 0.1533,
+      "step": 744
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.111670982132008e-07,
+      "loss": 0.1442,
+      "step": 745
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.028332677849254e-07,
+      "loss": 0.1649,
+      "step": 746
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.9466553006047383e-07,
+      "loss": 0.1686,
+      "step": 747
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.8666402350861701e-07,
+      "loss": 0.1672,
+      "step": 748
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.7882888377998787e-07,
+      "loss": 0.1764,
+      "step": 749
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.711602437047788e-07,
+      "loss": 0.142,
+      "step": 750
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.6365823329049124e-07,
+      "loss": 0.1835,
+      "step": 751
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.5632297971972966e-07,
+      "loss": 0.1624,
+      "step": 752
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.49154607348051e-07,
+      "loss": 0.1537,
+      "step": 753
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.4215323770184642e-07,
+      "loss": 0.1417,
+      "step": 754
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.3531898947629296e-07,
+      "loss": 0.1468,
+      "step": 755
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.2865197853333179e-07,
+      "loss": 0.1514,
+      "step": 756
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.221523178997075e-07,
+      "loss": 0.149,
+      "step": 757
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.158201177650553e-07,
+      "loss": 0.1629,
+      "step": 758
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.0965548548002802e-07,
+      "loss": 0.1525,
+      "step": 759
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.0365852555447642e-07,
+      "loss": 0.16,
+      "step": 760
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.782933965567953e-08,
+      "loss": 0.1482,
+      "step": 761
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 9.216802660662161e-08,
+      "loss": 0.1664,
+      "step": 762
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.667468238431453e-08,
+      "loss": 0.1436,
+      "step": 763
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.13494001181725e-08,
+      "loss": 0.1717,
+      "step": 764
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.619227008843322e-08,
+      "loss": 0.1703,
+      "step": 765
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.120337972462365e-08,
+      "loss": 0.1708,
+      "step": 766
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.638281360408339e-08,
+      "loss": 0.1758,
+      "step": 767
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.173065345052687e-08,
+      "loss": 0.1775,
+      "step": 768
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.7246978132659e-08,
+      "loss": 0.1589,
+      "step": 769
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.2931863662841755e-08,
+      "loss": 0.1775,
+      "step": 770
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.878538319579629e-08,
+      "loss": 0.1509,
+      "step": 771
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.480760702737286e-08,
+      "loss": 0.1893,
+      "step": 772
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 4.099860259335287e-08,
+      "loss": 0.1917,
+      "step": 773
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.735843446830867e-08,
+      "loss": 0.1607,
+      "step": 774
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.3887164364506676e-08,
+      "loss": 0.1372,
+      "step": 775
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 3.0584851130861516e-08,
+      "loss": 0.1535,
+      "step": 776
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.7451550751941282e-08,
+      "loss": 0.1685,
+      "step": 777
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.4487316347013845e-08,
+      "loss": 0.1531,
+      "step": 778
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.169219816914869e-08,
+      "loss": 0.1663,
+      "step": 779
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.9066243604367595e-08,
+      "loss": 0.1647,
+      "step": 780
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.6609497170834154e-08,
+      "loss": 0.1571,
+      "step": 781
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.4322000518106616e-08,
+      "loss": 0.1685,
+      "step": 782
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.220379242642844e-08,
+      "loss": 0.1361,
+      "step": 783
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.0254908806068831e-08,
+      "loss": 0.1469,
+      "step": 784
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 8.47538269671988e-09,
+      "loss": 0.1719,
+      "step": 785
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 6.8652442669281394e-09,
+      "loss": 0.1594,
+      "step": 786
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.4245208135905725e-09,
+      "loss": 0.1523,
+      "step": 787
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.153236761488266e-09,
+      "loss": 0.1614,
+      "step": 788
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.0514136628745363e-09,
+      "loss": 0.1537,
+      "step": 789
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.1190701971052218e-09,
+      "loss": 0.1628,
+      "step": 790
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.3562221703267153e-09,
+      "loss": 0.1766,
+      "step": 791
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 7.628825152050656e-10,
+      "loss": 0.1805,
+      "step": 792
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 3.390612907094859e-10,
+      "loss": 0.1617,
+      "step": 793
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.476568193804824e-11,
+      "loss": 0.1549,
+      "step": 794
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0,
+      "loss": 0.1619,
+      "step": 795
+    }
+  ],
+  "max_steps": 795,
+  "num_train_epochs": 3,
+  "total_flos": 1006074152878080.0,
+  "trial_name": null,
+  "trial_params": null
+}