diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9895 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 823291,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 0.7037,
+      "step": 500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 0.6481,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 0.6272,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 0.6114,
+      "step": 2000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 0.5996,
+      "step": 2500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 0.5917,
+      "step": 3000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 0.5829,
+      "step": 3500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 0.5755,
+      "step": 4000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.572,
+      "step": 4500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.5664,
+      "step": 5000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.5611,
+      "step": 5500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.5561,
+      "step": 6000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.5526,
+      "step": 6500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.5506,
+      "step": 7000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.5485,
+      "step": 7500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.5415,
+      "step": 8000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.5384,
+      "step": 8500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.537,
+      "step": 9000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.5359,
+      "step": 9500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.5303,
+      "step": 10000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.5296,
+      "step": 10500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.528,
+      "step": 11000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.5258,
+      "step": 11500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.5231,
+      "step": 12000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.5208,
+      "step": 12500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.5199,
+      "step": 13000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.52,
+      "step": 13500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.517,
+      "step": 14000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.5142,
+      "step": 14500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.5132,
+      "step": 15000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.513,
+      "step": 15500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.5111,
+      "step": 16000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.5093,
+      "step": 16500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.5077,
+      "step": 17000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.5064,
+      "step": 17500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.5039,
+      "step": 18000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.5021,
+      "step": 18500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.5018,
+      "step": 19000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.4999,
+      "step": 19500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.499,
+      "step": 20000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.4995,
+      "step": 20500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.4969,
+      "step": 21000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.4977,
+      "step": 21500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.4941,
+      "step": 22000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.495,
+      "step": 22500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.4936,
+      "step": 23000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.492,
+      "step": 23500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.4906,
+      "step": 24000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.4904,
+      "step": 24500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.4891,
+      "step": 25000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.4885,
+      "step": 25500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.4871,
+      "step": 26000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.4868,
+      "step": 26500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.485,
+      "step": 27000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.4862,
+      "step": 27500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.4857,
+      "step": 28000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.482,
+      "step": 28500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.4816,
+      "step": 29000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.481,
+      "step": 29500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.4813,
+      "step": 30000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.48,
+      "step": 30500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.4783,
+      "step": 31000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.4784,
+      "step": 31500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.4784,
+      "step": 32000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.4777,
+      "step": 32500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.4773,
+      "step": 33000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.4758,
+      "step": 33500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.4753,
+      "step": 34000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.4742,
+      "step": 34500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.4762,
+      "step": 35000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.4728,
+      "step": 35500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.4725,
+      "step": 36000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.4731,
+      "step": 36500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.4721,
+      "step": 37000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.4693,
+      "step": 37500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.4711,
+      "step": 38000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.4702,
+      "step": 38500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.468,
+      "step": 39000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.4659,
+      "step": 39500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.4684,
+      "step": 40000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.4683,
+      "step": 40500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.4673,
+      "step": 41000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.4659,
+      "step": 41500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.4648,
+      "step": 42000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.4655,
+      "step": 42500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.466,
+      "step": 43000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.4646,
+      "step": 43500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.4643,
+      "step": 44000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.4632,
+      "step": 44500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.4635,
+      "step": 45000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.4624,
+      "step": 45500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.4647,
+      "step": 46000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.4604,
+      "step": 46500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.4614,
+      "step": 47000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.4596,
+      "step": 47500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.4598,
+      "step": 48000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.4601,
+      "step": 48500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.4586,
+      "step": 49000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.459,
+      "step": 49500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.4591,
+      "step": 50000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.4591,
+      "step": 50500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.4598,
+      "step": 51000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.4567,
+      "step": 51500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.4559,
+      "step": 52000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.4565,
+      "step": 52500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.4559,
+      "step": 53000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.4561,
+      "step": 53500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4557,
+      "step": 54000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4551,
+      "step": 54500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.455,
+      "step": 55000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4538,
+      "step": 55500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4531,
+      "step": 56000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4526,
+      "step": 56500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4517,
+      "step": 57000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4525,
+      "step": 57500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4498,
+      "step": 58000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4534,
+      "step": 58500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4509,
+      "step": 59000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4512,
+      "step": 59500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4499,
+      "step": 60000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4501,
+      "step": 60500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4521,
+      "step": 61000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4489,
+      "step": 61500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.448,
+      "step": 62000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4487,
+      "step": 62500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4471,
+      "step": 63000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4485,
+      "step": 63500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4483,
+      "step": 64000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4471,
+      "step": 64500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4473,
+      "step": 65000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4463,
+      "step": 65500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4464,
+      "step": 66000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4461,
+      "step": 66500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4467,
+      "step": 67000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4458,
+      "step": 67500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4458,
+      "step": 68000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4443,
+      "step": 68500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4455,
+      "step": 69000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4452,
+      "step": 69500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.445,
+      "step": 70000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4438,
+      "step": 70500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4434,
+      "step": 71000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4423,
+      "step": 71500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4435,
+      "step": 72000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4423,
+      "step": 72500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4443,
+      "step": 73000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4431,
+      "step": 73500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4428,
+      "step": 74000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4435,
+      "step": 74500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.442,
+      "step": 75000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4406,
+      "step": 75500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4418,
+      "step": 76000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4418,
+      "step": 76500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4413,
+      "step": 77000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4402,
+      "step": 77500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4387,
+      "step": 78000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.4401,
+      "step": 78500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.4387,
+      "step": 79000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.4369,
+      "step": 79500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.439,
+      "step": 80000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.4389,
+      "step": 80500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.4377,
+      "step": 81000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.4382,
+      "step": 81500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.4369,
+      "step": 82000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.4368,
+      "step": 82500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.437,
+      "step": 83000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.4365,
+      "step": 83500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.4373,
+      "step": 84000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.4359,
+      "step": 84500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.4362,
+      "step": 85000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.4354,
+      "step": 85500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.4344,
+      "step": 86000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.4364,
+      "step": 86500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.4357,
+      "step": 87000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.4352,
+      "step": 87500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.4347,
+      "step": 88000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.435,
+      "step": 88500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.4338,
+      "step": 89000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.4348,
+      "step": 89500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.4351,
+      "step": 90000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.4334,
+      "step": 90500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.4329,
+      "step": 91000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.4332,
+      "step": 91500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.4335,
+      "step": 92000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.4329,
+      "step": 92500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.4324,
+      "step": 93000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.432,
+      "step": 93500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.4322,
+      "step": 94000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.4307,
+      "step": 94500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.4316,
+      "step": 95000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.4316,
+      "step": 95500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.4308,
+      "step": 96000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.4318,
+      "step": 96500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.4315,
+      "step": 97000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.4311,
+      "step": 97500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.4315,
+      "step": 98000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.4291,
+      "step": 98500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.4296,
+      "step": 99000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.4305,
+      "step": 99500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.4313,
+      "step": 100000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.4299,
+      "step": 100500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.4298,
+      "step": 101000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.427,
+      "step": 101500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.4283,
+      "step": 102000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.4284,
+      "step": 102500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.429,
+      "step": 103000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.4269,
+      "step": 103500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.4284,
+      "step": 104000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.4292,
+      "step": 104500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.4276,
+      "step": 105000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.4271,
+      "step": 105500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.4278,
+      "step": 106000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.4273,
+      "step": 106500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.4274,
+      "step": 107000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.4271,
+      "step": 107500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.4271,
+      "step": 108000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.4274,
+      "step": 108500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.4268,
+      "step": 109000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.4269,
+      "step": 109500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.4265,
+      "step": 110000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.4264,
+      "step": 110500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.4253,
+      "step": 111000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.4238,
+      "step": 111500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.4258,
+      "step": 112000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.4238,
+      "step": 112500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.4244,
+      "step": 113000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.4224,
+      "step": 113500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.4234,
+      "step": 114000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.4236,
+      "step": 114500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.4232,
+      "step": 115000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.4238,
+      "step": 115500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.4248,
+      "step": 116000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.4227,
+      "step": 116500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.4237,
+      "step": 117000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.4227,
+      "step": 117500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.4233,
+      "step": 118000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.4236,
+      "step": 118500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.4233,
+      "step": 119000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.422,
+      "step": 119500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.4216,
+      "step": 120000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.4223,
+      "step": 120500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.4214,
+      "step": 121000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.4212,
+      "step": 121500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.4199,
+      "step": 122000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.4225,
+      "step": 122500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.4218,
+      "step": 123000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.421,
+      "step": 123500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.4218,
+      "step": 124000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.4213,
+      "step": 124500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.4212,
+      "step": 125000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.4203,
+      "step": 125500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.4208,
+      "step": 126000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.4207,
+      "step": 126500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.421,
+      "step": 127000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.4202,
+      "step": 127500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.4191,
+      "step": 128000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.4197,
+      "step": 128500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.4196,
+      "step": 129000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.4188,
+      "step": 129500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.4207,
+      "step": 130000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.4194,
+      "step": 130500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.4176,
+      "step": 131000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.4185,
+      "step": 131500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.4174,
+      "step": 132000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.4186,
+      "step": 132500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.4193,
+      "step": 133000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.4187,
+      "step": 133500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.4182,
+      "step": 134000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.4181,
+      "step": 134500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.4194,
+      "step": 135000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.4183,
+      "step": 135500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.4191,
+      "step": 136000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.4166,
+      "step": 136500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.4172,
+      "step": 137000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.4166,
+      "step": 137500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.4158,
+      "step": 138000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.4166,
+      "step": 138500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.4163,
+      "step": 139000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.4161,
+      "step": 139500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.4158,
+      "step": 140000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.4161,
+      "step": 140500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.4154,
+      "step": 141000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.4155,
+      "step": 141500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.4149,
+      "step": 142000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.416,
+      "step": 142500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.4151,
+      "step": 143000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.4157,
+      "step": 143500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.4147,
+      "step": 144000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.4152,
+      "step": 144500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.4136,
+      "step": 145000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.414,
+      "step": 145500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.4144,
+      "step": 146000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.4156,
+      "step": 146500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.4143,
+      "step": 147000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.4145,
+      "step": 147500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.4148,
+      "step": 148000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.4149,
+      "step": 148500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.4142,
+      "step": 149000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.4123,
+      "step": 149500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.4132,
+      "step": 150000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.4137,
+      "step": 150500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.4124,
+      "step": 151000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.4131,
+      "step": 151500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.4126,
+      "step": 152000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.413,
+      "step": 152500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.413,
+      "step": 153000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.4143,
+      "step": 153500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.4129,
+      "step": 154000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.4121,
+      "step": 154500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.4117,
+      "step": 155000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.4106,
+      "step": 155500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.411,
+      "step": 156000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.4109,
+      "step": 156500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.4101,
+      "step": 157000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.4121,
+      "step": 157500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.4113,
+      "step": 158000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.4136,
+      "step": 158500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.4101,
+      "step": 159000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.4108,
+      "step": 159500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.4117,
+      "step": 160000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.4105,
+      "step": 160500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.4112,
+      "step": 161000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.4096,
+      "step": 161500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.4103,
+      "step": 162000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.4101,
+      "step": 162500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.4103,
+      "step": 163000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.4114,
+      "step": 163500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.4106,
+      "step": 164000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.4095,
+      "step": 164500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.4095,
+      "step": 165000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.4101,
+      "step": 165500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.4079,
+      "step": 166000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.4094,
+      "step": 166500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.4094,
+      "step": 167000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.4101,
+      "step": 167500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.4096,
+      "step": 168000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.409,
+      "step": 168500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.41,
+      "step": 169000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.4079,
+      "step": 169500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.4096,
+      "step": 170000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.4072,
+      "step": 170500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.408,
+      "step": 171000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.408,
+      "step": 171500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.4086,
+      "step": 172000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.4092,
+      "step": 172500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.4078,
+      "step": 173000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.4086,
+      "step": 173500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.4074,
+      "step": 174000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.4089,
+      "step": 174500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.4081,
+      "step": 175000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.4086,
+      "step": 175500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.407,
+      "step": 176000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.4082,
+      "step": 176500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.4075,
+      "step": 177000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.4069,
+      "step": 177500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.4077,
+      "step": 178000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.406,
+      "step": 178500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.4059,
+      "step": 179000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.4067,
+      "step": 179500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.4079,
+      "step": 180000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.4066,
+      "step": 180500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.4073,
+      "step": 181000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.4059,
+      "step": 181500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.4066,
+      "step": 182000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.4073,
+      "step": 182500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.4076,
+      "step": 183000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.4046,
+      "step": 183500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.4072,
+      "step": 184000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.4069,
+      "step": 184500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.4054,
+      "step": 185000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.405,
+      "step": 185500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.4065,
+      "step": 186000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.4054,
+      "step": 186500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.4056,
+      "step": 187000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.4049,
+      "step": 187500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.4048,
+      "step": 188000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.4049,
+      "step": 188500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.4058,
+      "step": 189000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.4041,
+      "step": 189500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.4077,
+      "step": 190000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.4054,
+      "step": 190500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.4032,
+      "step": 191000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.4048,
+      "step": 191500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.4049,
+      "step": 192000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.4038,
+      "step": 192500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.4048,
+      "step": 193000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.4053,
+      "step": 193500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.4039,
+      "step": 194000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.4033,
+      "step": 194500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.4035,
+      "step": 195000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.404,
+      "step": 195500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.4046,
+      "step": 196000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.4041,
+      "step": 196500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.4032,
+      "step": 197000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.404,
+      "step": 197500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.4045,
+      "step": 198000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.4027,
+      "step": 198500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.4027,
+      "step": 199000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.4031,
+      "step": 199500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.4035,
+      "step": 200000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.402,
+      "step": 200500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.4034,
+      "step": 201000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.4013,
+      "step": 201500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.4031,
+      "step": 202000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.4028,
+      "step": 202500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.4032,
+      "step": 203000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.4028,
+      "step": 203500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.4027,
+      "step": 204000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.4018,
+      "step": 204500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.4017,
+      "step": 205000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.4013,
+      "step": 205500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.4008,
+      "step": 206000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.402,
+      "step": 206500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.4014,
+      "step": 207000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.4024,
+      "step": 207500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.4003,
+      "step": 208000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.4006,
+      "step": 208500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.4013,
+      "step": 209000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.4012,
+      "step": 209500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.4005,
+      "step": 210000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.4016,
+      "step": 210500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.4019,
+      "step": 211000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.4015,
+      "step": 211500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.4012,
+      "step": 212000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.4005,
+      "step": 212500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.4014,
+      "step": 213000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.4012,
+      "step": 213500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.3999,
+      "step": 214000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.3998,
+      "step": 214500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.401,
+      "step": 215000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.4003,
+      "step": 215500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.3995,
+      "step": 216000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.4008,
+      "step": 216500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.4009,
+      "step": 217000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.4017,
+      "step": 217500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.4017,
+      "step": 218000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.3991,
+      "step": 218500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.4009,
+      "step": 219000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.3988,
+      "step": 219500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.4012,
+      "step": 220000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.3979,
+      "step": 220500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.3995,
+      "step": 221000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.3987,
+      "step": 221500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.398,
+      "step": 222000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.3995,
+      "step": 222500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.3991,
+      "step": 223000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.4001,
+      "step": 223500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.3976,
+      "step": 224000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.3984,
+      "step": 224500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.3995,
+      "step": 225000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.398,
+      "step": 225500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.3987,
+      "step": 226000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.398,
+      "step": 226500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.3972,
+      "step": 227000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.399,
+      "step": 227500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.3986,
+      "step": 228000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.3974,
+      "step": 228500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.3983,
+      "step": 229000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.3991,
+      "step": 229500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.3983,
+      "step": 230000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.3981,
+      "step": 230500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.3988,
+      "step": 231000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.3979,
+      "step": 231500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.3974,
+      "step": 232000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.399,
+      "step": 232500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.397,
+      "step": 233000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.3967,
+      "step": 233500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.3978,
+      "step": 234000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.3975,
+      "step": 234500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.3968,
+      "step": 235000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.3965,
+      "step": 235500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.3952,
+      "step": 236000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.3963,
+      "step": 236500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.3975,
+      "step": 237000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.3961,
+      "step": 237500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.3965,
+      "step": 238000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.3955,
+      "step": 238500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.3971,
+      "step": 239000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.3959,
+      "step": 239500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.3966,
+      "step": 240000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.3967,
+      "step": 240500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.3959,
+      "step": 241000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.3966,
+      "step": 241500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.396,
+      "step": 242000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.3966,
+      "step": 242500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.3956,
+      "step": 243000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.3966,
+      "step": 243500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.3957,
+      "step": 244000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.3967,
+      "step": 244500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.3962,
+      "step": 245000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.3942,
+      "step": 245500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.3963,
+      "step": 246000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.395,
+      "step": 246500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.3958,
+      "step": 247000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.3956,
+      "step": 247500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.3951,
+      "step": 248000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.3958,
+      "step": 248500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.3954,
+      "step": 249000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.3942,
+      "step": 249500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.3946,
+      "step": 250000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.3949,
+      "step": 250500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.3939,
+      "step": 251000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.395,
+      "step": 251500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3966,
+      "step": 252000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3936,
+      "step": 252500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3956,
+      "step": 253000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3946,
+      "step": 253500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3938,
+      "step": 254000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3935,
+      "step": 254500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3938,
+      "step": 255000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3957,
+      "step": 255500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3937,
+      "step": 256000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3944,
+      "step": 256500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3942,
+      "step": 257000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3929,
+      "step": 257500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3935,
+      "step": 258000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3934,
+      "step": 258500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3934,
+      "step": 259000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3929,
+      "step": 259500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3928,
+      "step": 260000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3928,
+      "step": 260500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3932,
+      "step": 261000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3937,
+      "step": 261500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3926,
+      "step": 262000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3939,
+      "step": 262500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3931,
+      "step": 263000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3939,
+      "step": 263500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3939,
+      "step": 264000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3925,
+      "step": 264500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3928,
+      "step": 265000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3925,
+      "step": 265500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3928,
+      "step": 266000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3927,
+      "step": 266500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3926,
+      "step": 267000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3914,
+      "step": 267500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.3918,
+      "step": 268000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.3925,
+      "step": 268500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.3917,
+      "step": 269000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.3938,
+      "step": 269500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.3926,
+      "step": 270000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.3934,
+      "step": 270500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.3923,
+      "step": 271000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.3916,
+      "step": 271500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.3922,
+      "step": 272000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.3918,
+      "step": 272500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.3907,
+      "step": 273000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.391,
+      "step": 273500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.3921,
+      "step": 274000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.3913,
+      "step": 274500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.3908,
+      "step": 275000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.3924,
+      "step": 275500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.3905,
+      "step": 276000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.3913,
+      "step": 276500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.391,
+      "step": 277000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.391,
+      "step": 277500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.3912,
+      "step": 278000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.3915,
+      "step": 278500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.3917,
+      "step": 279000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.3906,
+      "step": 279500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.3919,
+      "step": 280000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.3912,
+      "step": 280500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.3899,
+      "step": 281000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.3918,
+      "step": 281500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.3917,
+      "step": 282000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.3894,
+      "step": 282500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.3906,
+      "step": 283000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.391,
+      "step": 283500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.3902,
+      "step": 284000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.3902,
+      "step": 284500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.3908,
+      "step": 285000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.3889,
+      "step": 285500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.3908,
+      "step": 286000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.3904,
+      "step": 286500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.3898,
+      "step": 287000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.3893,
+      "step": 287500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.3897,
+      "step": 288000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.3916,
+      "step": 288500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.388,
+      "step": 289000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.3896,
+      "step": 289500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.39,
+      "step": 290000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.3908,
+      "step": 290500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.3885,
+      "step": 291000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.3895,
+      "step": 291500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.3886,
+      "step": 292000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.3907,
+      "step": 292500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.389,
+      "step": 293000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.3903,
+      "step": 293500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.3903,
+      "step": 294000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.3879,
+      "step": 294500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.3902,
+      "step": 295000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.3905,
+      "step": 295500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.3897,
+      "step": 296000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.3885,
+      "step": 296500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.3883,
+      "step": 297000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.3885,
+      "step": 297500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.3893,
+      "step": 298000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.3897,
+      "step": 298500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.3882,
+      "step": 299000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.388,
+      "step": 299500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.3898,
+      "step": 300000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.3877,
+      "step": 300500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.3887,
+      "step": 301000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.3877,
+      "step": 301500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.3891,
+      "step": 302000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.3874,
+      "step": 302500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.3871,
+      "step": 303000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.3886,
+      "step": 303500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.3864,
+      "step": 304000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.3886,
+      "step": 304500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.3885,
+      "step": 305000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.3866,
+      "step": 305500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.3877,
+      "step": 306000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.3877,
+      "step": 306500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.3883,
+      "step": 307000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.3876,
+      "step": 307500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.3872,
+      "step": 308000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.3876,
+      "step": 308500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3872,
+      "step": 309000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3866,
+      "step": 309500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3871,
+      "step": 310000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3882,
+      "step": 310500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3865,
+      "step": 311000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3868,
+      "step": 311500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3868,
+      "step": 312000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3876,
+      "step": 312500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3862,
+      "step": 313000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3849,
+      "step": 313500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3864,
+      "step": 314000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3874,
+      "step": 314500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3867,
+      "step": 315000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3871,
+      "step": 315500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3868,
+      "step": 316000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3872,
+      "step": 316500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3878,
+      "step": 317000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3874,
+      "step": 317500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3869,
+      "step": 318000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3862,
+      "step": 318500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3878,
+      "step": 319000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3864,
+      "step": 319500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3873,
+      "step": 320000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3857,
+      "step": 320500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3863,
+      "step": 321000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3869,
+      "step": 321500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3867,
+      "step": 322000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3843,
+      "step": 322500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3862,
+      "step": 323000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3861,
+      "step": 323500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3865,
+      "step": 324000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3862,
+      "step": 324500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.3859,
+      "step": 325000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.3855,
+      "step": 325500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.3863,
+      "step": 326000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.3849,
+      "step": 326500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.3852,
+      "step": 327000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.3851,
+      "step": 327500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.384,
+      "step": 328000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.3856,
+      "step": 328500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.3849,
+      "step": 329000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.3847,
+      "step": 329500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.3856,
+      "step": 330000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.3854,
+      "step": 330500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.3846,
+      "step": 331000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.3854,
+      "step": 331500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.3856,
+      "step": 332000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.3849,
+      "step": 332500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.3845,
+      "step": 333000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3846,
+      "step": 333500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3856,
+      "step": 334000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3849,
+      "step": 334500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3857,
+      "step": 335000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3846,
+      "step": 335500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3845,
+      "step": 336000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3839,
+      "step": 336500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3845,
+      "step": 337000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3843,
+      "step": 337500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3841,
+      "step": 338000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3848,
+      "step": 338500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3848,
+      "step": 339000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3854,
+      "step": 339500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3827,
+      "step": 340000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3847,
+      "step": 340500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3843,
+      "step": 341000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.3856,
+      "step": 341500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.3839,
+      "step": 342000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.3828,
+      "step": 342500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.3843,
+      "step": 343000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.3837,
+      "step": 343500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.3845,
+      "step": 344000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.384,
+      "step": 344500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.3829,
+      "step": 345000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.3835,
+      "step": 345500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.3842,
+      "step": 346000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.3829,
+      "step": 346500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.3835,
+      "step": 347000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.3838,
+      "step": 347500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.3829,
+      "step": 348000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.3837,
+      "step": 348500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.3841,
+      "step": 349000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.383,
+      "step": 349500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3826,
+      "step": 350000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3826,
+      "step": 350500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3824,
+      "step": 351000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3827,
+      "step": 351500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3819,
+      "step": 352000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3828,
+      "step": 352500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3828,
+      "step": 353000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3817,
+      "step": 353500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3819,
+      "step": 354000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3823,
+      "step": 354500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3842,
+      "step": 355000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3818,
+      "step": 355500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3829,
+      "step": 356000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3828,
+      "step": 356500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3824,
+      "step": 357000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3824,
+      "step": 357500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.3826,
+      "step": 358000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.3831,
+      "step": 358500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.3833,
+      "step": 359000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.3822,
+      "step": 359500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.3828,
+      "step": 360000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.3821,
+      "step": 360500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.3828,
+      "step": 361000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.383,
+      "step": 361500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.3822,
+      "step": 362000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.383,
+      "step": 362500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.3817,
+      "step": 363000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.382,
+      "step": 363500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.3825,
+      "step": 364000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.3821,
+      "step": 364500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.3812,
+      "step": 365000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.3806,
+      "step": 365500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.3817,
+      "step": 366000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.3816,
+      "step": 366500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.3818,
+      "step": 367000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.3814,
+      "step": 367500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.3815,
+      "step": 368000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.3815,
+      "step": 368500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.3809,
+      "step": 369000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.3809,
+      "step": 369500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.3812,
+      "step": 370000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.3807,
+      "step": 370500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.3815,
+      "step": 371000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.3822,
+      "step": 371500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.381,
+      "step": 372000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.382,
+      "step": 372500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.3815,
+      "step": 373000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.3807,
+      "step": 373500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.3824,
+      "step": 374000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.3811,
+      "step": 374500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.3809,
+      "step": 375000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.3805,
+      "step": 375500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.3795,
+      "step": 376000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.3806,
+      "step": 376500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.3812,
+      "step": 377000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.3811,
+      "step": 377500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.3803,
+      "step": 378000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.3804,
+      "step": 378500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.381,
+      "step": 379000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.3819,
+      "step": 379500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.3811,
+      "step": 380000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.3818,
+      "step": 380500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.3796,
+      "step": 381000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.3807,
+      "step": 381500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.3804,
+      "step": 382000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.38,
+      "step": 382500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3812,
+      "step": 383000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3801,
+      "step": 383500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3808,
+      "step": 384000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3801,
+      "step": 384500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3813,
+      "step": 385000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3798,
+      "step": 385500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3797,
+      "step": 386000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3799,
+      "step": 386500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3798,
+      "step": 387000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3802,
+      "step": 387500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3782,
+      "step": 388000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3797,
+      "step": 388500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3796,
+      "step": 389000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3805,
+      "step": 389500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3796,
+      "step": 390000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3809,
+      "step": 390500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.3795,
+      "step": 391000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.3802,
+      "step": 391500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.3783,
+      "step": 392000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.379,
+      "step": 392500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.3784,
+      "step": 393000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.3792,
+      "step": 393500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.3802,
+      "step": 394000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.3786,
+      "step": 394500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.379,
+      "step": 395000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.3798,
+      "step": 395500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.3788,
+      "step": 396000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.3794,
+      "step": 396500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.3801,
+      "step": 397000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.3777,
+      "step": 397500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.378,
+      "step": 398000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.3795,
+      "step": 398500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.3784,
+      "step": 399000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.3786,
+      "step": 399500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.3785,
+      "step": 400000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.378,
+      "step": 400500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.3784,
+      "step": 401000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.3775,
+      "step": 401500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.3774,
+      "step": 402000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.3783,
+      "step": 402500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.3792,
+      "step": 403000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.3792,
+      "step": 403500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.3781,
+      "step": 404000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.3792,
+      "step": 404500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.379,
+      "step": 405000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.378,
+      "step": 405500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.3797,
+      "step": 406000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.379,
+      "step": 406500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.3781,
+      "step": 407000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.3772,
+      "step": 407500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.3779,
+      "step": 408000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.3779,
+      "step": 408500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.3775,
+      "step": 409000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.3767,
+      "step": 409500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.3789,
+      "step": 410000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.3772,
+      "step": 410500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.3778,
+      "step": 411000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.3777,
+      "step": 411500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.3768,
+      "step": 412000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.3773,
+      "step": 412500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.377,
+      "step": 413000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.3774,
+      "step": 413500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.3769,
+      "step": 414000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.3778,
+      "step": 414500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.3766,
+      "step": 415000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.3777,
+      "step": 415500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.3778,
+      "step": 416000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.3773,
+      "step": 416500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.3777,
+      "step": 417000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.3772,
+      "step": 417500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.3778,
+      "step": 418000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.3771,
+      "step": 418500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.3782,
+      "step": 419000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.3786,
+      "step": 419500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.376,
+      "step": 420000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.376,
+      "step": 420500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.3773,
+      "step": 421000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.3768,
+      "step": 421500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.377,
+      "step": 422000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.3751,
+      "step": 422500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.377,
+      "step": 423000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.3777,
+      "step": 423500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.3749,
+      "step": 424000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.377,
+      "step": 424500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.3761,
+      "step": 425000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.3776,
+      "step": 425500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.3761,
+      "step": 426000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.3764,
+      "step": 426500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.3762,
+      "step": 427000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.3764,
+      "step": 427500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.3767,
+      "step": 428000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.3762,
+      "step": 428500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.3752,
+      "step": 429000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.3763,
+      "step": 429500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.3758,
+      "step": 430000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.3767,
+      "step": 430500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.3765,
+      "step": 431000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.376,
+      "step": 431500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.3776,
+      "step": 432000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.3764,
+      "step": 432500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.3764,
+      "step": 433000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.3753,
+      "step": 433500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.3762,
+      "step": 434000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.3767,
+      "step": 434500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.3771,
+      "step": 435000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.3744,
+      "step": 435500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.376,
+      "step": 436000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.3773,
+      "step": 436500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.3763,
+      "step": 437000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.3755,
+      "step": 437500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.3764,
+      "step": 438000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.3756,
+      "step": 438500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.3751,
+      "step": 439000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.3761,
+      "step": 439500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.3757,
+      "step": 440000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.3752,
+      "step": 440500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.3765,
+      "step": 441000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.3757,
+      "step": 441500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.3761,
+      "step": 442000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.376,
+      "step": 442500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.3756,
+      "step": 443000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.376,
+      "step": 443500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.3748,
+      "step": 444000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.375,
+      "step": 444500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.3759,
+      "step": 445000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.3747,
+      "step": 445500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.3761,
+      "step": 446000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.3759,
+      "step": 446500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.375,
+      "step": 447000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.3755,
+      "step": 447500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.3749,
+      "step": 448000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.3753,
+      "step": 448500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.3745,
+      "step": 449000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.3759,
+      "step": 449500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.3756,
+      "step": 450000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.375,
+      "step": 450500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.3744,
+      "step": 451000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.3748,
+      "step": 451500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.3745,
+      "step": 452000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.3751,
+      "step": 452500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.3761,
+      "step": 453000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.375,
+      "step": 453500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.3744,
+      "step": 454000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.3745,
+      "step": 454500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.3754,
+      "step": 455000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.3746,
+      "step": 455500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.3743,
+      "step": 456000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.3738,
+      "step": 456500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.3753,
+      "step": 457000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.375,
+      "step": 457500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.3741,
+      "step": 458000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.3745,
+      "step": 458500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.3761,
+      "step": 459000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.3747,
+      "step": 459500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.373,
+      "step": 460000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.3749,
+      "step": 460500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.3732,
+      "step": 461000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.3746,
+      "step": 461500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.3747,
+      "step": 462000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.3747,
+      "step": 462500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.374,
+      "step": 463000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.3741,
+      "step": 463500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.3747,
+      "step": 464000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.3745,
+      "step": 464500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.3742,
+      "step": 465000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.3734,
+      "step": 465500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.374,
+      "step": 466000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.3744,
+      "step": 466500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.373,
+      "step": 467000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.3725,
+      "step": 467500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.3738,
+      "step": 468000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.3728,
+      "step": 468500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.3734,
+      "step": 469000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.3734,
+      "step": 469500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.3733,
+      "step": 470000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.3733,
+      "step": 470500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.3728,
+      "step": 471000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.3724,
+      "step": 471500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.3731,
+      "step": 472000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.3731,
+      "step": 472500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.3737,
+      "step": 473000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.3736,
+      "step": 473500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.3734,
+      "step": 474000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.3726,
+      "step": 474500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.3726,
+      "step": 475000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.3731,
+      "step": 475500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.3732,
+      "step": 476000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.3727,
+      "step": 476500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.3727,
+      "step": 477000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.373,
+      "step": 477500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.3731,
+      "step": 478000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.374,
+      "step": 478500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.3726,
+      "step": 479000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.3732,
+      "step": 479500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.3728,
+      "step": 480000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.3729,
+      "step": 480500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.3732,
+      "step": 481000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.3727,
+      "step": 481500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.3727,
+      "step": 482000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.3722,
+      "step": 482500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.3731,
+      "step": 483000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.3731,
+      "step": 483500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.3713,
+      "step": 484000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.3711,
+      "step": 484500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.3718,
+      "step": 485000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.3725,
+      "step": 485500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.3722,
+      "step": 486000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.3721,
+      "step": 486500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.372,
+      "step": 487000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.3714,
+      "step": 487500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.3733,
+      "step": 488000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.3727,
+      "step": 488500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.3712,
+      "step": 489000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.3713,
+      "step": 489500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.3719,
+      "step": 490000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.3728,
+      "step": 490500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.372,
+      "step": 491000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.3721,
+      "step": 491500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.3724,
+      "step": 492000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.3728,
+      "step": 492500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.3721,
+      "step": 493000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.3717,
+      "step": 493500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.3731,
+      "step": 494000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.3717,
+      "step": 494500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.3713,
+      "step": 495000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.3717,
+      "step": 495500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.3718,
+      "step": 496000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.3726,
+      "step": 496500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.3722,
+      "step": 497000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.3714,
+      "step": 497500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.3713,
+      "step": 498000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.3706,
+      "step": 498500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.3717,
+      "step": 499000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.3717,
+      "step": 499500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.3712,
+      "step": 500000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.3721,
+      "step": 500500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.3716,
+      "step": 501000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.3721,
+      "step": 501500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.3714,
+      "step": 502000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.3717,
+      "step": 502500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.3711,
+      "step": 503000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.3719,
+      "step": 503500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.3713,
+      "step": 504000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.3717,
+      "step": 504500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.3708,
+      "step": 505000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.3699,
+      "step": 505500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.3704,
+      "step": 506000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.371,
+      "step": 506500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.3705,
+      "step": 507000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.37,
+      "step": 507500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.3719,
+      "step": 508000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.371,
+      "step": 508500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.3709,
+      "step": 509000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.3708,
+      "step": 509500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.3701,
+      "step": 510000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.3708,
+      "step": 510500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.3704,
+      "step": 511000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.3709,
+      "step": 511500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.3719,
+      "step": 512000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.372,
+      "step": 512500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.3698,
+      "step": 513000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.3704,
+      "step": 513500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.3706,
+      "step": 514000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.3702,
+      "step": 514500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.3705,
+      "step": 515000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.3704,
+      "step": 515500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.3699,
+      "step": 516000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.371,
+      "step": 516500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.3702,
+      "step": 517000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.3704,
+      "step": 517500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.3706,
+      "step": 518000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.3693,
+      "step": 518500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.3698,
+      "step": 519000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.3694,
+      "step": 519500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.3699,
+      "step": 520000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.3715,
+      "step": 520500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.3681,
+      "step": 521000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.3695,
+      "step": 521500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.37,
+      "step": 522000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.3694,
+      "step": 522500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.368,
+      "step": 523000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.3691,
+      "step": 523500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.3706,
+      "step": 524000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.37,
+      "step": 524500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.3701,
+      "step": 525000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.3706,
+      "step": 525500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.3689,
+      "step": 526000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.3702,
+      "step": 526500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.3692,
+      "step": 527000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.3688,
+      "step": 527500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.3704,
+      "step": 528000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.3695,
+      "step": 528500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.3696,
+      "step": 529000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.3697,
+      "step": 529500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.3707,
+      "step": 530000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.3703,
+      "step": 530500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.3688,
+      "step": 531000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.3696,
+      "step": 531500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.3687,
+      "step": 532000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.3697,
+      "step": 532500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.3698,
+      "step": 533000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.3692,
+      "step": 533500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.3685,
+      "step": 534000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.3685,
+      "step": 534500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.3701,
+      "step": 535000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.3697,
+      "step": 535500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.3692,
+      "step": 536000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.3692,
+      "step": 536500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.3699,
+      "step": 537000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.3685,
+      "step": 537500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.3685,
+      "step": 538000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.369,
+      "step": 538500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.3699,
+      "step": 539000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.3681,
+      "step": 539500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.369,
+      "step": 540000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.3688,
+      "step": 540500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.3685,
+      "step": 541000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.3683,
+      "step": 541500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.3693,
+      "step": 542000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.37,
+      "step": 542500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.3692,
+      "step": 543000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.368,
+      "step": 543500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.3687,
+      "step": 544000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.3694,
+      "step": 544500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.3704,
+      "step": 545000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.3676,
+      "step": 545500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.3679,
+      "step": 546000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.3685,
+      "step": 546500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.3689,
+      "step": 547000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.3695,
+      "step": 547500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.3695,
+      "step": 548000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.3681,
+      "step": 548500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.369,
+      "step": 549000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.3683,
+      "step": 549500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.3691,
+      "step": 550000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.3695,
+      "step": 550500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.3685,
+      "step": 551000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.3683,
+      "step": 551500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.3677,
+      "step": 552000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.3681,
+      "step": 552500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.3677,
+      "step": 553000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.3669,
+      "step": 553500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.3686,
+      "step": 554000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.3679,
+      "step": 554500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.3676,
+      "step": 555000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.3689,
+      "step": 555500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.3688,
+      "step": 556000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.3684,
+      "step": 556500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.3674,
+      "step": 557000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.3691,
+      "step": 557500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.3693,
+      "step": 558000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.369,
+      "step": 558500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.3686,
+      "step": 559000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.3669,
+      "step": 559500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.3682,
+      "step": 560000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.3681,
+      "step": 560500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.3673,
+      "step": 561000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.3693,
+      "step": 561500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.3678,
+      "step": 562000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.3684,
+      "step": 562500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.3675,
+      "step": 563000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.3672,
+      "step": 563500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.3676,
+      "step": 564000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.3672,
+      "step": 564500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.3669,
+      "step": 565000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.3674,
+      "step": 565500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.366,
+      "step": 566000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.367,
+      "step": 566500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.3666,
+      "step": 567000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.3693,
+      "step": 567500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.3675,
+      "step": 568000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.3672,
+      "step": 568500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.3677,
+      "step": 569000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.3681,
+      "step": 569500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.3668,
+      "step": 570000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.368,
+      "step": 570500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.368,
+      "step": 571000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.3676,
+      "step": 571500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.367,
+      "step": 572000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.3669,
+      "step": 572500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.3674,
+      "step": 573000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.3663,
+      "step": 573500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.3671,
+      "step": 574000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.3679,
+      "step": 574500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.3662,
+      "step": 575000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.3678,
+      "step": 575500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.3671,
+      "step": 576000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.3655,
+      "step": 576500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.3665,
+      "step": 577000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.3671,
+      "step": 577500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.3671,
+      "step": 578000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.3655,
+      "step": 578500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.3668,
+      "step": 579000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.3672,
+      "step": 579500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.3661,
+      "step": 580000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.3664,
+      "step": 580500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.3672,
+      "step": 581000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.3672,
+      "step": 581500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.3656,
+      "step": 582000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.3671,
+      "step": 582500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.3669,
+      "step": 583000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.3672,
+      "step": 583500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.3682,
+      "step": 584000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.3661,
+      "step": 584500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.3659,
+      "step": 585000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.3679,
+      "step": 585500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.3664,
+      "step": 586000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.3661,
+      "step": 586500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.366,
+      "step": 587000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.3665,
+      "step": 587500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.3658,
+      "step": 588000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.3661,
+      "step": 588500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.3664,
+      "step": 589000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.3651,
+      "step": 589500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.3681,
+      "step": 590000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.3666,
+      "step": 590500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.3651,
+      "step": 591000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.3656,
+      "step": 591500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.3661,
+      "step": 592000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.3663,
+      "step": 592500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.3665,
+      "step": 593000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.3656,
+      "step": 593500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.3661,
+      "step": 594000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.367,
+      "step": 594500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.366,
+      "step": 595000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.3648,
+      "step": 595500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.3663,
+      "step": 596000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.3661,
+      "step": 596500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.3653,
+      "step": 597000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.3654,
+      "step": 597500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.3659,
+      "step": 598000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.3651,
+      "step": 598500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.3657,
+      "step": 599000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.3654,
+      "step": 599500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.3655,
+      "step": 600000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.3671,
+      "step": 600500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.3663,
+      "step": 601000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.3664,
+      "step": 601500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.3657,
+      "step": 602000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.3646,
+      "step": 602500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.3652,
+      "step": 603000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.3661,
+      "step": 603500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.3662,
+      "step": 604000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.3659,
+      "step": 604500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.366,
+      "step": 605000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.3652,
+      "step": 605500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.3657,
+      "step": 606000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.3653,
+      "step": 606500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.3644,
+      "step": 607000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.3653,
+      "step": 607500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.3662,
+      "step": 608000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.3665,
+      "step": 608500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.3655,
+      "step": 609000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.3664,
+      "step": 609500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.3656,
+      "step": 610000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.3644,
+      "step": 610500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.3637,
+      "step": 611000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.3653,
+      "step": 611500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.3652,
+      "step": 612000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.3656,
+      "step": 612500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.3648,
+      "step": 613000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.3652,
+      "step": 613500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.3649,
+      "step": 614000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.3642,
+      "step": 614500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.3641,
+      "step": 615000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.3644,
+      "step": 615500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.3642,
+      "step": 616000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.3648,
+      "step": 616500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.366,
+      "step": 617000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.3648,
+      "step": 617500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.3649,
+      "step": 618000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.3656,
+      "step": 618500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.365,
+      "step": 619000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.3635,
+      "step": 619500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.3639,
+      "step": 620000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.3645,
+      "step": 620500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.3643,
+      "step": 621000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.3627,
+      "step": 621500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.3645,
+      "step": 622000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.3635,
+      "step": 622500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.3654,
+      "step": 623000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.3641,
+      "step": 623500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.3637,
+      "step": 624000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.3636,
+      "step": 624500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.3652,
+      "step": 625000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.3648,
+      "step": 625500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.3647,
+      "step": 626000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.3638,
+      "step": 626500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.3645,
+      "step": 627000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.3648,
+      "step": 627500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.3642,
+      "step": 628000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.3642,
+      "step": 628500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.3641,
+      "step": 629000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.3651,
+      "step": 629500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.365,
+      "step": 630000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.364,
+      "step": 630500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.3638,
+      "step": 631000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.3638,
+      "step": 631500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.3659,
+      "step": 632000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.3642,
+      "step": 632500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.3644,
+      "step": 633000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.3639,
+      "step": 633500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.3639,
+      "step": 634000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.3639,
+      "step": 634500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.3646,
+      "step": 635000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.3636,
+      "step": 635500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.3629,
+      "step": 636000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.3629,
+      "step": 636500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.3639,
+      "step": 637000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.3638,
+      "step": 637500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.3653,
+      "step": 638000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.3628,
+      "step": 638500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.3645,
+      "step": 639000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.365,
+      "step": 639500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.3653,
+      "step": 640000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.363,
+      "step": 640500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.3634,
+      "step": 641000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.3637,
+      "step": 641500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.3639,
+      "step": 642000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.364,
+      "step": 642500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.3636,
+      "step": 643000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.3641,
+      "step": 643500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.3631,
+      "step": 644000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.3633,
+      "step": 644500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.3633,
+      "step": 645000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.362,
+      "step": 645500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.3629,
+      "step": 646000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3637,
+      "step": 646500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3642,
+      "step": 647000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3637,
+      "step": 647500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3639,
+      "step": 648000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3639,
+      "step": 648500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3632,
+      "step": 649000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3631,
+      "step": 649500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3634,
+      "step": 650000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3641,
+      "step": 650500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3645,
+      "step": 651000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3636,
+      "step": 651500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3645,
+      "step": 652000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3631,
+      "step": 652500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3625,
+      "step": 653000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3627,
+      "step": 653500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.363,
+      "step": 654000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3631,
+      "step": 654500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.3631,
+      "step": 655000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.3625,
+      "step": 655500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.3638,
+      "step": 656000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.3609,
+      "step": 656500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.3625,
+      "step": 657000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.3627,
+      "step": 657500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.3631,
+      "step": 658000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.3621,
+      "step": 658500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.3629,
+      "step": 659000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.3637,
+      "step": 659500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.3627,
+      "step": 660000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.3633,
+      "step": 660500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.3625,
+      "step": 661000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.3633,
+      "step": 661500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.3636,
+      "step": 662000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.3612,
+      "step": 662500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.3624,
+      "step": 663000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.3623,
+      "step": 663500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.3629,
+      "step": 664000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.363,
+      "step": 664500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.3623,
+      "step": 665000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.3636,
+      "step": 665500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.3628,
+      "step": 666000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.3628,
+      "step": 666500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.3626,
+      "step": 667000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.3632,
+      "step": 667500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.3621,
+      "step": 668000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.3624,
+      "step": 668500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.3616,
+      "step": 669000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.3631,
+      "step": 669500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.3633,
+      "step": 670000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.3624,
+      "step": 670500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.3603,
+      "step": 671000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.3614,
+      "step": 671500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.3618,
+      "step": 672000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.3612,
+      "step": 672500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.3615,
+      "step": 673000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.363,
+      "step": 673500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.3615,
+      "step": 674000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.3629,
+      "step": 674500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.3618,
+      "step": 675000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.3623,
+      "step": 675500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.3619,
+      "step": 676000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.362,
+      "step": 676500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.3624,
+      "step": 677000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.3625,
+      "step": 677500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.3616,
+      "step": 678000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.3617,
+      "step": 678500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.3622,
+      "step": 679000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.3607,
+      "step": 679500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.3625,
+      "step": 680000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.3633,
+      "step": 680500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.3625,
+      "step": 681000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.3618,
+      "step": 681500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.363,
+      "step": 682000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.3612,
+      "step": 682500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.3606,
+      "step": 683000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.3611,
+      "step": 683500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.3618,
+      "step": 684000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.3615,
+      "step": 684500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.3617,
+      "step": 685000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.3623,
+      "step": 685500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.3611,
+      "step": 686000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.3616,
+      "step": 686500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.362,
+      "step": 687000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.3613,
+      "step": 687500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.3607,
+      "step": 688000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.3607,
+      "step": 688500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.3622,
+      "step": 689000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.3623,
+      "step": 689500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.361,
+      "step": 690000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.3599,
+      "step": 690500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.3619,
+      "step": 691000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.3615,
+      "step": 691500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.3613,
+      "step": 692000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.3608,
+      "step": 692500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.3609,
+      "step": 693000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.3616,
+      "step": 693500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.3614,
+      "step": 694000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.3607,
+      "step": 694500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.3624,
+      "step": 695000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.3614,
+      "step": 695500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.3609,
+      "step": 696000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.3615,
+      "step": 696500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.3608,
+      "step": 697000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.3607,
+      "step": 697500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.3609,
+      "step": 698000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.3606,
+      "step": 698500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.3613,
+      "step": 699000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.362,
+      "step": 699500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.3618,
+      "step": 700000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.3612,
+      "step": 700500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.3608,
+      "step": 701000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.3609,
+      "step": 701500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.36,
+      "step": 702000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.3611,
+      "step": 702500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.3624,
+      "step": 703000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.3597,
+      "step": 703500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.3607,
+      "step": 704000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.3608,
+      "step": 704500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.361,
+      "step": 705000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.3596,
+      "step": 705500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.3602,
+      "step": 706000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.3606,
+      "step": 706500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.3603,
+      "step": 707000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.3608,
+      "step": 707500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.3608,
+      "step": 708000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.3603,
+      "step": 708500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.3605,
+      "step": 709000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.3611,
+      "step": 709500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.3594,
+      "step": 710000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.3599,
+      "step": 710500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.361,
+      "step": 711000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.3594,
+      "step": 711500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.3609,
+      "step": 712000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.3603,
+      "step": 712500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.3606,
+      "step": 713000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.36,
+      "step": 713500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.3597,
+      "step": 714000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.3625,
+      "step": 714500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.3602,
+      "step": 715000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.3604,
+      "step": 715500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.3603,
+      "step": 716000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.36,
+      "step": 716500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.3602,
+      "step": 717000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.36,
+      "step": 717500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.3593,
+      "step": 718000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.3606,
+      "step": 718500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.3612,
+      "step": 719000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.3595,
+      "step": 719500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.36,
+      "step": 720000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.3591,
+      "step": 720500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.3601,
+      "step": 721000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.3602,
+      "step": 721500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.3603,
+      "step": 722000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.3604,
+      "step": 722500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.3598,
+      "step": 723000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.3587,
+      "step": 723500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.3597,
+      "step": 724000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.3598,
+      "step": 724500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.3594,
+      "step": 725000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.3612,
+      "step": 725500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.3592,
+      "step": 726000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.3591,
+      "step": 726500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.36,
+      "step": 727000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.3594,
+      "step": 727500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.3596,
+      "step": 728000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.3604,
+      "step": 728500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.3591,
+      "step": 729000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.3601,
+      "step": 729500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.3602,
+      "step": 730000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.3594,
+      "step": 730500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.3593,
+      "step": 731000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.3602,
+      "step": 731500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.3591,
+      "step": 732000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.3597,
+      "step": 732500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.3597,
+      "step": 733000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.3587,
+      "step": 733500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.3603,
+      "step": 734000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.3601,
+      "step": 734500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.3601,
+      "step": 735000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.3597,
+      "step": 735500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.3598,
+      "step": 736000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.3593,
+      "step": 736500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.3602,
+      "step": 737000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.359,
+      "step": 737500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.3588,
+      "step": 738000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.3595,
+      "step": 738500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.3599,
+      "step": 739000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.359,
+      "step": 739500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.3589,
+      "step": 740000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.3587,
+      "step": 740500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.3591,
+      "step": 741000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.3604,
+      "step": 741500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.3594,
+      "step": 742000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.3599,
+      "step": 742500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.3594,
+      "step": 743000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.3581,
+      "step": 743500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.3598,
+      "step": 744000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.3581,
+      "step": 744500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.3595,
+      "step": 745000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.3579,
+      "step": 745500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.3582,
+      "step": 746000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.3575,
+      "step": 746500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.3602,
+      "step": 747000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.3585,
+      "step": 747500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.3591,
+      "step": 748000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.3581,
+      "step": 748500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.3586,
+      "step": 749000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.3589,
+      "step": 749500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.3585,
+      "step": 750000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.3584,
+      "step": 750500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.3594,
+      "step": 751000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.36,
+      "step": 751500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.3589,
+      "step": 752000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.3584,
+      "step": 752500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.3587,
+      "step": 753000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.3573,
+      "step": 753500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.3588,
+      "step": 754000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.3588,
+      "step": 754500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.3581,
+      "step": 755000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.3585,
+      "step": 755500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.3571,
+      "step": 756000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.3571,
+      "step": 756500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.3592,
+      "step": 757000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.3594,
+      "step": 757500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.3588,
+      "step": 758000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.3584,
+      "step": 758500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.358,
+      "step": 759000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.3583,
+      "step": 759500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.3591,
+      "step": 760000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.3588,
+      "step": 760500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.3592,
+      "step": 761000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.3579,
+      "step": 761500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.3573,
+      "step": 762000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.3587,
+      "step": 762500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.3588,
+      "step": 763000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.3581,
+      "step": 763500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.3584,
+      "step": 764000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.3581,
+      "step": 764500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.3587,
+      "step": 765000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.3583,
+      "step": 765500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.3579,
+      "step": 766000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.3585,
+      "step": 766500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.359,
+      "step": 767000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.3587,
+      "step": 767500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.358,
+      "step": 768000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.3579,
+      "step": 768500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.3583,
+      "step": 769000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.3583,
+      "step": 769500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3576,
+      "step": 770000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3583,
+      "step": 770500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3581,
+      "step": 771000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3576,
+      "step": 771500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3578,
+      "step": 772000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3579,
+      "step": 772500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3571,
+      "step": 773000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3589,
+      "step": 773500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.358,
+      "step": 774000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3576,
+      "step": 774500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3572,
+      "step": 775000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3572,
+      "step": 775500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3584,
+      "step": 776000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3583,
+      "step": 776500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3578,
+      "step": 777000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3589,
+      "step": 777500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.3579,
+      "step": 778000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.357,
+      "step": 778500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.3574,
+      "step": 779000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.3569,
+      "step": 779500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.3573,
+      "step": 780000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.3576,
+      "step": 780500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.3576,
+      "step": 781000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.3571,
+      "step": 781500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.3562,
+      "step": 782000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.3576,
+      "step": 782500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.3573,
+      "step": 783000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.3573,
+      "step": 783500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.3575,
+      "step": 784000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.3587,
+      "step": 784500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.358,
+      "step": 785000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.3578,
+      "step": 785500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.3569,
+      "step": 786000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.3571,
+      "step": 786500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.3571,
+      "step": 787000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.3573,
+      "step": 787500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.3573,
+      "step": 788000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.3578,
+      "step": 788500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.3578,
+      "step": 789000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.3564,
+      "step": 789500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.3564,
+      "step": 790000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.3574,
+      "step": 790500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.3582,
+      "step": 791000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.3572,
+      "step": 791500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.357,
+      "step": 792000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.3577,
+      "step": 792500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.3581,
+      "step": 793000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.3569,
+      "step": 793500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.3566,
+      "step": 794000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.3578,
+      "step": 794500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.3557,
+      "step": 795000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.3577,
+      "step": 795500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.3566,
+      "step": 796000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.3572,
+      "step": 796500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.3564,
+      "step": 797000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.3559,
+      "step": 797500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.3569,
+      "step": 798000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.357,
+      "step": 798500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.3575,
+      "step": 799000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.3568,
+      "step": 799500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.3576,
+      "step": 800000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.3572,
+      "step": 800500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.357,
+      "step": 801000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.3566,
+      "step": 801500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.3576,
+      "step": 802000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.3576,
+      "step": 802500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.3569,
+      "step": 803000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.3577,
+      "step": 803500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.3572,
+      "step": 804000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.3558,
+      "step": 804500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.3563,
+      "step": 805000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.3557,
+      "step": 805500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.3566,
+      "step": 806000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.357,
+      "step": 806500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.3569,
+      "step": 807000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.3575,
+      "step": 807500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.3561,
+      "step": 808000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.3569,
+      "step": 808500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.3574,
+      "step": 809000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.3571,
+      "step": 809500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.3558,
+      "step": 810000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.3567,
+      "step": 810500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.3556,
+      "step": 811000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.3566,
+      "step": 811500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.3577,
+      "step": 812000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.3567,
+      "step": 812500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.3559,
+      "step": 813000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.3561,
+      "step": 813500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.3564,
+      "step": 814000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.3566,
+      "step": 814500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.357,
+      "step": 815000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.3568,
+      "step": 815500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.3559,
+      "step": 816000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.3568,
+      "step": 816500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.3564,
+      "step": 817000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.3558,
+      "step": 817500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.3554,
+      "step": 818000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.3567,
+      "step": 818500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.3561,
+      "step": 819000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.3563,
+      "step": 819500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.3565,
+      "step": 820000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.3569,
+      "step": 820500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.3553,
+      "step": 821000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.3554,
+      "step": 821500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.3578,
+      "step": 822000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.3546,
+      "step": 822500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.3564,
+      "step": 823000
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 823291,
+  "num_train_epochs": 1,
+  "save_steps": 823291,
+  "total_flos": 6.419933083302101e+20,
+  "trial_name": null,
+  "trial_params": null
+}