diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9824 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 12.548024703460445,
+  "global_step": 160000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 2.0248,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 1.8708,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 1.8434,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 1.7994,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 1.8287,
+      "step": 500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 1.8163,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 1.8126,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 1.7973,
+      "step": 800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 1.7749,
+      "step": 900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 1.7852,
+      "step": 1000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 1.7927,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 1.7736,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 1.7792,
+      "step": 1300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 1.7559,
+      "step": 1400
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 1.7508,
+      "step": 1500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 1.7809,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 1.767,
+      "step": 1700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 1.7906,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 1.7622,
+      "step": 1900
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 1.7726,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 1.7597,
+      "step": 2100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 1.7649,
+      "step": 2200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 1.7352,
+      "step": 2300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 1.7408,
+      "step": 2400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 1.7453,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 1.7417,
+      "step": 2600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 1.7524,
+      "step": 2700
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 1.7461,
+      "step": 2800
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 1.7584,
+      "step": 2900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 1.7485,
+      "step": 3000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 1.7532,
+      "step": 3100
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 1.7221,
+      "step": 3200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 1.7351,
+      "step": 3300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 1.7286,
+      "step": 3400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 1.724,
+      "step": 3500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 1.711,
+      "step": 3600
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 1.7233,
+      "step": 3700
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 1.7075,
+      "step": 3800
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 1.7264,
+      "step": 3900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 1.7189,
+      "step": 4000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 1.7265,
+      "step": 4100
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 1.7145,
+      "step": 4200
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 1.7188,
+      "step": 4300
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 1.7173,
+      "step": 4400
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 1.7454,
+      "step": 4500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 1.7071,
+      "step": 4600
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 1.7,
+      "step": 4700
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 1.7312,
+      "step": 4800
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 1.6972,
+      "step": 4900
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 1.717,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 1.7042,
+      "step": 5100
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 1.7075,
+      "step": 5200
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 1.7037,
+      "step": 5300
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 1.689,
+      "step": 5400
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 1.7099,
+      "step": 5500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 1.7175,
+      "step": 5600
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 1.6906,
+      "step": 5700
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 1.7112,
+      "step": 5800
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 1.7183,
+      "step": 5900
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 1.7158,
+      "step": 6000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 1.6972,
+      "step": 6100
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 1.6869,
+      "step": 6200
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 1.6896,
+      "step": 6300
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 1.6895,
+      "step": 6400
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 1.7044,
+      "step": 6500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 1.7103,
+      "step": 6600
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 1.6788,
+      "step": 6700
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 1.7091,
+      "step": 6800
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 1.7046,
+      "step": 6900
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 1.6683,
+      "step": 7000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 1.7084,
+      "step": 7100
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 1.6933,
+      "step": 7200
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 1.6533,
+      "step": 7300
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 1.6701,
+      "step": 7400
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 1.6786,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 1.6886,
+      "step": 7600
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 1.6805,
+      "step": 7700
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 1.6651,
+      "step": 7800
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 1.666,
+      "step": 7900
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 1.6671,
+      "step": 8000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 1.6979,
+      "step": 8100
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 1.6823,
+      "step": 8200
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 1.6944,
+      "step": 8300
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 1.6571,
+      "step": 8400
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 1.7135,
+      "step": 8500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 1.67,
+      "step": 8600
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 1.6908,
+      "step": 8700
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 1.665,
+      "step": 8800
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 1.6862,
+      "step": 8900
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 1.6903,
+      "step": 9000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 1.6551,
+      "step": 9100
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 1.6588,
+      "step": 9200
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 1.6589,
+      "step": 9300
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 1.671,
+      "step": 9400
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 1.6702,
+      "step": 9500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 1.6641,
+      "step": 9600
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 1.6681,
+      "step": 9700
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 1.6798,
+      "step": 9800
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 1.6762,
+      "step": 9900
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 1.6781,
+      "step": 10000
+    },
+    {
+      "epoch": 0.78,
+      "eval_gen_len": 18.729231040875785,
+      "eval_loss": 1.5352756977081299,
+      "eval_rouge1": 38.2416,
+      "eval_rouge2": 16.0401,
+      "eval_rougeL": 31.3511,
+      "eval_rougeLsum": 31.3586,
+      "eval_runtime": 2357.8446,
+      "eval_samples_per_second": 4.804,
+      "eval_steps_per_second": 1.201,
+      "step": 10000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 1.6725,
+      "step": 10100
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 1.6569,
+      "step": 10200
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 1.631,
+      "step": 10300
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 1.6549,
+      "step": 10400
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 1.6714,
+      "step": 10500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 1.6827,
+      "step": 10600
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 1.6725,
+      "step": 10700
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 1.6517,
+      "step": 10800
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 1.647,
+      "step": 10900
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 1.6536,
+      "step": 11000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 1.6545,
+      "step": 11100
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 1.6697,
+      "step": 11200
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 1.6329,
+      "step": 11300
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 1.6601,
+      "step": 11400
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 1.665,
+      "step": 11500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 1.6462,
+      "step": 11600
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 1.6512,
+      "step": 11700
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 1.673,
+      "step": 11800
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 1.6818,
+      "step": 11900
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 1.6698,
+      "step": 12000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 1.6772,
+      "step": 12100
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 1.6526,
+      "step": 12200
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 1.6561,
+      "step": 12300
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 1.6429,
+      "step": 12400
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 1.6517,
+      "step": 12500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 1.6766,
+      "step": 12600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 1.6708,
+      "step": 12700
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 1.6229,
+      "step": 12800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 1.5406,
+      "step": 12900
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 1.5325,
+      "step": 13000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 1.5196,
+      "step": 13100
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 1.5283,
+      "step": 13200
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 1.5463,
+      "step": 13300
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 1.5461,
+      "step": 13400
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 1.5496,
+      "step": 13500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 1.5646,
+      "step": 13600
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 1.5527,
+      "step": 13700
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 1.551,
+      "step": 13800
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 1.5566,
+      "step": 13900
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 1.5514,
+      "step": 14000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 1.5425,
+      "step": 14100
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 1.5321,
+      "step": 14200
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 1.5552,
+      "step": 14300
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 1.5316,
+      "step": 14400
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 1.5271,
+      "step": 14500
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 1.5331,
+      "step": 14600
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 1.5428,
+      "step": 14700
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 1.5534,
+      "step": 14800
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 1.5306,
+      "step": 14900
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 1.5468,
+      "step": 15000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 1.529,
+      "step": 15100
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 1.5349,
+      "step": 15200
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 1.5596,
+      "step": 15300
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 1.5328,
+      "step": 15400
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 1.5495,
+      "step": 15500
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 1.5447,
+      "step": 15600
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 1.5442,
+      "step": 15700
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 1.5435,
+      "step": 15800
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 1.5588,
+      "step": 15900
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 1.5632,
+      "step": 16000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 1.5504,
+      "step": 16100
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 1.5148,
+      "step": 16200
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 1.5523,
+      "step": 16300
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 1.544,
+      "step": 16400
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 1.5472,
+      "step": 16500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 1.5294,
+      "step": 16600
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 1.5458,
+      "step": 16700
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 1.5579,
+      "step": 16800
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 1.5524,
+      "step": 16900
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 1.5455,
+      "step": 17000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 1.5549,
+      "step": 17100
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 1.5457,
+      "step": 17200
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 1.5378,
+      "step": 17300
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 1.5422,
+      "step": 17400
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 1.5559,
+      "step": 17500
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 1.5463,
+      "step": 17600
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 1.5385,
+      "step": 17700
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 1.5487,
+      "step": 17800
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 1.5535,
+      "step": 17900
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 1.5348,
+      "step": 18000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 1.5555,
+      "step": 18100
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 1.5378,
+      "step": 18200
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 1.5212,
+      "step": 18300
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 1.5559,
+      "step": 18400
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 1.5353,
+      "step": 18500
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 1.5461,
+      "step": 18600
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 1.5407,
+      "step": 18700
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 1.5359,
+      "step": 18800
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 1.5561,
+      "step": 18900
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 1.5274,
+      "step": 19000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 1.5544,
+      "step": 19100
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 1.5407,
+      "step": 19200
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 1.5374,
+      "step": 19300
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 1.5572,
+      "step": 19400
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 1.5177,
+      "step": 19500
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 1.5504,
+      "step": 19600
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 1.5611,
+      "step": 19700
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 1.5362,
+      "step": 19800
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 1.5411,
+      "step": 19900
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 1.5473,
+      "step": 20000
+    },
+    {
+      "epoch": 1.57,
+      "eval_gen_len": 18.753509314028427,
+      "eval_loss": 1.5065853595733643,
+      "eval_rouge1": 38.5975,
+      "eval_rouge2": 16.6862,
+      "eval_rougeL": 31.8033,
+      "eval_rougeLsum": 31.8084,
+      "eval_runtime": 2373.7069,
+      "eval_samples_per_second": 4.772,
+      "eval_steps_per_second": 1.193,
+      "step": 20000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 1.5304,
+      "step": 20100
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 1.55,
+      "step": 20200
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 1.5375,
+      "step": 20300
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 1.5366,
+      "step": 20400
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 1.5254,
+      "step": 20500
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 1.5369,
+      "step": 20600
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 1.5455,
+      "step": 20700
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 1.5337,
+      "step": 20800
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 1.5411,
+      "step": 20900
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 1.5416,
+      "step": 21000
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 1.5477,
+      "step": 21100
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 1.5276,
+      "step": 21200
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 1.5643,
+      "step": 21300
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 1.558,
+      "step": 21400
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 1.5548,
+      "step": 21500
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 1.5486,
+      "step": 21600
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 1.551,
+      "step": 21700
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 1.5494,
+      "step": 21800
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 1.5526,
+      "step": 21900
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 1.5294,
+      "step": 22000
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 1.5288,
+      "step": 22100
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 1.5093,
+      "step": 22200
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 1.5276,
+      "step": 22300
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 1.5321,
+      "step": 22400
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 1.5655,
+      "step": 22500
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 1.5438,
+      "step": 22600
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 1.5507,
+      "step": 22700
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 1.5506,
+      "step": 22800
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 1.5377,
+      "step": 22900
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 1.5447,
+      "step": 23000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 1.5361,
+      "step": 23100
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 1.5568,
+      "step": 23200
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 1.5498,
+      "step": 23300
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 1.5405,
+      "step": 23400
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 1.5198,
+      "step": 23500
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 1.5227,
+      "step": 23600
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 1.5163,
+      "step": 23700
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 1.5811,
+      "step": 23800
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 1.5303,
+      "step": 23900
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 1.5416,
+      "step": 24000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 1.5296,
+      "step": 24100
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 1.5521,
+      "step": 24200
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 1.5387,
+      "step": 24300
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 1.5311,
+      "step": 24400
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 1.5539,
+      "step": 24500
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 1.53,
+      "step": 24600
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 1.5588,
+      "step": 24700
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 1.5402,
+      "step": 24800
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 1.5269,
+      "step": 24900
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 1.5407,
+      "step": 25000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 1.5141,
+      "step": 25100
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 1.523,
+      "step": 25200
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 1.5112,
+      "step": 25300
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 1.5314,
+      "step": 25400
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 1.5345,
+      "step": 25500
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 1.4314,
+      "step": 25600
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 1.445,
+      "step": 25700
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 1.4215,
+      "step": 25800
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 1.4116,
+      "step": 25900
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 1.4264,
+      "step": 26000
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 1.4238,
+      "step": 26100
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 1.4172,
+      "step": 26200
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 1.4319,
+      "step": 26300
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 1.4174,
+      "step": 26400
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 1.4293,
+      "step": 26500
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 1.4393,
+      "step": 26600
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 1.4474,
+      "step": 26700
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 1.4227,
+      "step": 26800
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 1.4232,
+      "step": 26900
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 1.4294,
+      "step": 27000
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 1.4118,
+      "step": 27100
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 1.4277,
+      "step": 27200
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 1.4171,
+      "step": 27300
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 1.4413,
+      "step": 27400
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 1.4119,
+      "step": 27500
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 1.3893,
+      "step": 27600
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 1.4255,
+      "step": 27700
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 1.4221,
+      "step": 27800
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 1.4339,
+      "step": 27900
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 1.4394,
+      "step": 28000
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 1.4249,
+      "step": 28100
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 1.4458,
+      "step": 28200
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 1.4323,
+      "step": 28300
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 1.4167,
+      "step": 28400
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 1.431,
+      "step": 28500
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 1.4568,
+      "step": 28600
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 1.4716,
+      "step": 28700
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 1.4147,
+      "step": 28800
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 1.427,
+      "step": 28900
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 1.4643,
+      "step": 29000
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 1.4436,
+      "step": 29100
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 1.4321,
+      "step": 29200
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 1.4464,
+      "step": 29300
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 1.4052,
+      "step": 29400
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 1.4278,
+      "step": 29500
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 1.401,
+      "step": 29600
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 1.4458,
+      "step": 29700
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 1.4517,
+      "step": 29800
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 1.4541,
+      "step": 29900
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 1.4334,
+      "step": 30000
+    },
+    {
+      "epoch": 2.35,
+      "eval_gen_len": 18.717400900503222,
+      "eval_loss": 1.4981415271759033,
+      "eval_rouge1": 39.0553,
+      "eval_rouge2": 17.2294,
+      "eval_rougeL": 32.3778,
+      "eval_rougeLsum": 32.3837,
+      "eval_runtime": 2349.3062,
+      "eval_samples_per_second": 4.821,
+      "eval_steps_per_second": 1.205,
+      "step": 30000
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 1.4209,
+      "step": 30100
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 1.401,
+      "step": 30200
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5e-05,
+      "loss": 1.4375,
+      "step": 30300
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5e-05,
+      "loss": 1.4383,
+      "step": 30400
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 1.4378,
+      "step": 30500
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 1.444,
+      "step": 30600
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 1.432,
+      "step": 30700
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 1.4397,
+      "step": 30800
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 1.4196,
+      "step": 30900
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 1.4123,
+      "step": 31000
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 1.4194,
+      "step": 31100
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 1.4466,
+      "step": 31200
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 1.4358,
+      "step": 31300
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 1.4641,
+      "step": 31400
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 1.4346,
+      "step": 31500
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 1.4372,
+      "step": 31600
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 1.4128,
+      "step": 31700
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 1.4267,
+      "step": 31800
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 1.4233,
+      "step": 31900
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 1.4357,
+      "step": 32000
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 1.4127,
+      "step": 32100
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 1.4324,
+      "step": 32200
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 1.4501,
+      "step": 32300
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 1.426,
+      "step": 32400
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 1.445,
+      "step": 32500
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 1.4226,
+      "step": 32600
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 1.4569,
+      "step": 32700
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 1.4501,
+      "step": 32800
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 1.4113,
+      "step": 32900
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 1.4629,
+      "step": 33000
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 1.4288,
+      "step": 33100
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 1.4311,
+      "step": 33200
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 1.4409,
+      "step": 33300
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 1.4597,
+      "step": 33400
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 1.4401,
+      "step": 33500
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 1.4155,
+      "step": 33600
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 1.438,
+      "step": 33700
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 1.4159,
+      "step": 33800
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5e-05,
+      "loss": 1.4377,
+      "step": 33900
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 1.4474,
+      "step": 34000
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 1.4338,
+      "step": 34100
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 1.4747,
+      "step": 34200
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 1.475,
+      "step": 34300
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 1.4491,
+      "step": 34400
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5e-05,
+      "loss": 1.4372,
+      "step": 34500
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5e-05,
+      "loss": 1.455,
+      "step": 34600
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5e-05,
+      "loss": 1.4535,
+      "step": 34700
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5e-05,
+      "loss": 1.4649,
+      "step": 34800
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5e-05,
+      "loss": 1.457,
+      "step": 34900
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5e-05,
+      "loss": 1.4422,
+      "step": 35000
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5e-05,
+      "loss": 1.431,
+      "step": 35100
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 5e-05,
+      "loss": 1.4279,
+      "step": 35200
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 5e-05,
+      "loss": 1.4487,
+      "step": 35300
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5e-05,
+      "loss": 1.456,
+      "step": 35400
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5e-05,
+      "loss": 1.4208,
+      "step": 35500
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5e-05,
+      "loss": 1.4449,
+      "step": 35600
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 5e-05,
+      "loss": 1.4324,
+      "step": 35700
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5e-05,
+      "loss": 1.4399,
+      "step": 35800
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5e-05,
+      "loss": 1.4257,
+      "step": 35900
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5e-05,
+      "loss": 1.457,
+      "step": 36000
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5e-05,
+      "loss": 1.4549,
+      "step": 36100
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5e-05,
+      "loss": 1.4525,
+      "step": 36200
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5e-05,
+      "loss": 1.4587,
+      "step": 36300
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5e-05,
+      "loss": 1.4319,
+      "step": 36400
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5e-05,
+      "loss": 1.4267,
+      "step": 36500
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5e-05,
+      "loss": 1.4644,
+      "step": 36600
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5e-05,
+      "loss": 1.4477,
+      "step": 36700
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5e-05,
+      "loss": 1.4342,
+      "step": 36800
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5e-05,
+      "loss": 1.4348,
+      "step": 36900
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5e-05,
+      "loss": 1.4384,
+      "step": 37000
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5e-05,
+      "loss": 1.4432,
+      "step": 37100
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5e-05,
+      "loss": 1.4382,
+      "step": 37200
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5e-05,
+      "loss": 1.4454,
+      "step": 37300
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5e-05,
+      "loss": 1.4196,
+      "step": 37400
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5e-05,
+      "loss": 1.4476,
+      "step": 37500
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5e-05,
+      "loss": 1.4554,
+      "step": 37600
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5e-05,
+      "loss": 1.4315,
+      "step": 37700
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5e-05,
+      "loss": 1.4471,
+      "step": 37800
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5e-05,
+      "loss": 1.4523,
+      "step": 37900
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5e-05,
+      "loss": 1.4562,
+      "step": 38000
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5e-05,
+      "loss": 1.4413,
+      "step": 38100
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5e-05,
+      "loss": 1.4608,
+      "step": 38200
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5e-05,
+      "loss": 1.4121,
+      "step": 38300
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 5e-05,
+      "loss": 1.3319,
+      "step": 38400
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 5e-05,
+      "loss": 1.3408,
+      "step": 38500
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 5e-05,
+      "loss": 1.3191,
+      "step": 38600
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 5e-05,
+      "loss": 1.326,
+      "step": 38700
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 5e-05,
+      "loss": 1.3219,
+      "step": 38800
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 5e-05,
+      "loss": 1.333,
+      "step": 38900
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 5e-05,
+      "loss": 1.3286,
+      "step": 39000
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 5e-05,
+      "loss": 1.3439,
+      "step": 39100
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 5e-05,
+      "loss": 1.3256,
+      "step": 39200
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 5e-05,
+      "loss": 1.3363,
+      "step": 39300
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 5e-05,
+      "loss": 1.3229,
+      "step": 39400
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 5e-05,
+      "loss": 1.3285,
+      "step": 39500
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 5e-05,
+      "loss": 1.3242,
+      "step": 39600
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 5e-05,
+      "loss": 1.329,
+      "step": 39700
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 5e-05,
+      "loss": 1.3486,
+      "step": 39800
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 5e-05,
+      "loss": 1.3239,
+      "step": 39900
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 5e-05,
+      "loss": 1.3582,
+      "step": 40000
+    },
+    {
+      "epoch": 3.14,
+      "eval_gen_len": 18.757482122362497,
+      "eval_loss": 1.4984853267669678,
+      "eval_rouge1": 39.0472,
+      "eval_rouge2": 17.2043,
+      "eval_rougeL": 32.2958,
+      "eval_rougeLsum": 32.2919,
+      "eval_runtime": 2351.4068,
+      "eval_samples_per_second": 4.817,
+      "eval_steps_per_second": 1.204,
+      "step": 40000
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 5e-05,
+      "loss": 1.348,
+      "step": 40100
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 5e-05,
+      "loss": 1.3358,
+      "step": 40200
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 5e-05,
+      "loss": 1.3307,
+      "step": 40300
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 5e-05,
+      "loss": 1.3387,
+      "step": 40400
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 5e-05,
+      "loss": 1.3412,
+      "step": 40500
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 5e-05,
+      "loss": 1.3391,
+      "step": 40600
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 5e-05,
+      "loss": 1.367,
+      "step": 40700
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 5e-05,
+      "loss": 1.3217,
+      "step": 40800
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 5e-05,
+      "loss": 1.3407,
+      "step": 40900
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 5e-05,
+      "loss": 1.3321,
+      "step": 41000
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 5e-05,
+      "loss": 1.3565,
+      "step": 41100
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 5e-05,
+      "loss": 1.33,
+      "step": 41200
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 5e-05,
+      "loss": 1.3402,
+      "step": 41300
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 5e-05,
+      "loss": 1.3641,
+      "step": 41400
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 5e-05,
+      "loss": 1.3589,
+      "step": 41500
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 5e-05,
+      "loss": 1.329,
+      "step": 41600
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 5e-05,
+      "loss": 1.3262,
+      "step": 41700
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 5e-05,
+      "loss": 1.3384,
+      "step": 41800
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 5e-05,
+      "loss": 1.3475,
+      "step": 41900
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 5e-05,
+      "loss": 1.3532,
+      "step": 42000
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 5e-05,
+      "loss": 1.3395,
+      "step": 42100
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 5e-05,
+      "loss": 1.3357,
+      "step": 42200
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 5e-05,
+      "loss": 1.3364,
+      "step": 42300
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 5e-05,
+      "loss": 1.351,
+      "step": 42400
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 5e-05,
+      "loss": 1.3656,
+      "step": 42500
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 5e-05,
+      "loss": 1.3651,
+      "step": 42600
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 5e-05,
+      "loss": 1.3317,
+      "step": 42700
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 5e-05,
+      "loss": 1.3492,
+      "step": 42800
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 5e-05,
+      "loss": 1.3497,
+      "step": 42900
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 5e-05,
+      "loss": 1.3523,
+      "step": 43000
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5e-05,
+      "loss": 1.3484,
+      "step": 43100
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 5e-05,
+      "loss": 1.3495,
+      "step": 43200
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 5e-05,
+      "loss": 1.3633,
+      "step": 43300
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 5e-05,
+      "loss": 1.3578,
+      "step": 43400
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 5e-05,
+      "loss": 1.334,
+      "step": 43500
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 5e-05,
+      "loss": 1.3591,
+      "step": 43600
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 5e-05,
+      "loss": 1.334,
+      "step": 43700
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 5e-05,
+      "loss": 1.3484,
+      "step": 43800
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 5e-05,
+      "loss": 1.3557,
+      "step": 43900
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 5e-05,
+      "loss": 1.3453,
+      "step": 44000
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 5e-05,
+      "loss": 1.342,
+      "step": 44100
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 5e-05,
+      "loss": 1.3449,
+      "step": 44200
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 5e-05,
+      "loss": 1.3392,
+      "step": 44300
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 5e-05,
+      "loss": 1.3421,
+      "step": 44400
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 5e-05,
+      "loss": 1.3444,
+      "step": 44500
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 5e-05,
+      "loss": 1.3492,
+      "step": 44600
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 5e-05,
+      "loss": 1.3575,
+      "step": 44700
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 5e-05,
+      "loss": 1.3487,
+      "step": 44800
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 5e-05,
+      "loss": 1.3432,
+      "step": 44900
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5e-05,
+      "loss": 1.3119,
+      "step": 45000
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 5e-05,
+      "loss": 1.3624,
+      "step": 45100
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 5e-05,
+      "loss": 1.3454,
+      "step": 45200
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 5e-05,
+      "loss": 1.3507,
+      "step": 45300
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 5e-05,
+      "loss": 1.3481,
+      "step": 45400
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 5e-05,
+      "loss": 1.3822,
+      "step": 45500
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5e-05,
+      "loss": 1.3479,
+      "step": 45600
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5e-05,
+      "loss": 1.3556,
+      "step": 45700
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5e-05,
+      "loss": 1.3584,
+      "step": 45800
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 5e-05,
+      "loss": 1.3583,
+      "step": 45900
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 5e-05,
+      "loss": 1.3644,
+      "step": 46000
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 5e-05,
+      "loss": 1.3376,
+      "step": 46100
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 5e-05,
+      "loss": 1.3345,
+      "step": 46200
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 5e-05,
+      "loss": 1.3796,
+      "step": 46300
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 5e-05,
+      "loss": 1.3425,
+      "step": 46400
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 5e-05,
+      "loss": 1.3434,
+      "step": 46500
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 5e-05,
+      "loss": 1.382,
+      "step": 46600
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 5e-05,
+      "loss": 1.3829,
+      "step": 46700
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 5e-05,
+      "loss": 1.3671,
+      "step": 46800
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 5e-05,
+      "loss": 1.3561,
+      "step": 46900
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 5e-05,
+      "loss": 1.3528,
+      "step": 47000
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 5e-05,
+      "loss": 1.3359,
+      "step": 47100
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 5e-05,
+      "loss": 1.3458,
+      "step": 47200
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 5e-05,
+      "loss": 1.3547,
+      "step": 47300
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 5e-05,
+      "loss": 1.3579,
+      "step": 47400
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 5e-05,
+      "loss": 1.3383,
+      "step": 47500
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 5e-05,
+      "loss": 1.3555,
+      "step": 47600
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 5e-05,
+      "loss": 1.3702,
+      "step": 47700
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 5e-05,
+      "loss": 1.3472,
+      "step": 47800
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 5e-05,
+      "loss": 1.3567,
+      "step": 47900
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 5e-05,
+      "loss": 1.3464,
+      "step": 48000
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 5e-05,
+      "loss": 1.3709,
+      "step": 48100
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 5e-05,
+      "loss": 1.3677,
+      "step": 48200
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 5e-05,
+      "loss": 1.3653,
+      "step": 48300
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 5e-05,
+      "loss": 1.3726,
+      "step": 48400
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 5e-05,
+      "loss": 1.3684,
+      "step": 48500
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 5e-05,
+      "loss": 1.3459,
+      "step": 48600
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5e-05,
+      "loss": 1.3812,
+      "step": 48700
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 5e-05,
+      "loss": 1.3432,
+      "step": 48800
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 5e-05,
+      "loss": 1.3573,
+      "step": 48900
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 5e-05,
+      "loss": 1.363,
+      "step": 49000
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 5e-05,
+      "loss": 1.3744,
+      "step": 49100
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 5e-05,
+      "loss": 1.3751,
+      "step": 49200
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 5e-05,
+      "loss": 1.3629,
+      "step": 49300
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 5e-05,
+      "loss": 1.3549,
+      "step": 49400
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5e-05,
+      "loss": 1.3704,
+      "step": 49500
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 5e-05,
+      "loss": 1.3743,
+      "step": 49600
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 5e-05,
+      "loss": 1.3361,
+      "step": 49700
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 5e-05,
+      "loss": 1.3684,
+      "step": 49800
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 5e-05,
+      "loss": 1.3545,
+      "step": 49900
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 5e-05,
+      "loss": 1.3716,
+      "step": 50000
+    },
+    {
+      "epoch": 3.92,
+      "eval_gen_len": 18.721197139578,
+      "eval_loss": 1.4834269285202026,
+      "eval_rouge1": 39.5208,
+      "eval_rouge2": 17.6414,
+      "eval_rougeL": 32.8518,
+      "eval_rougeLsum": 32.8448,
+      "eval_runtime": 2347.5327,
+      "eval_samples_per_second": 4.825,
+      "eval_steps_per_second": 1.206,
+      "step": 50000
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 5e-05,
+      "loss": 1.355,
+      "step": 50100
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 5e-05,
+      "loss": 1.3659,
+      "step": 50200
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 5e-05,
+      "loss": 1.3819,
+      "step": 50300
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 5e-05,
+      "loss": 1.3587,
+      "step": 50400
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 5e-05,
+      "loss": 1.3802,
+      "step": 50500
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 5e-05,
+      "loss": 1.3681,
+      "step": 50600
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 5e-05,
+      "loss": 1.3567,
+      "step": 50700
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 5e-05,
+      "loss": 1.3517,
+      "step": 50800
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 5e-05,
+      "loss": 1.3769,
+      "step": 50900
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 5e-05,
+      "loss": 1.3688,
+      "step": 51000
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 5e-05,
+      "loss": 1.2497,
+      "step": 51100
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 5e-05,
+      "loss": 1.2667,
+      "step": 51200
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 5e-05,
+      "loss": 1.2817,
+      "step": 51300
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5e-05,
+      "loss": 1.2461,
+      "step": 51400
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5e-05,
+      "loss": 1.2406,
+      "step": 51500
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 5e-05,
+      "loss": 1.2409,
+      "step": 51600
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 5e-05,
+      "loss": 1.2674,
+      "step": 51700
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 5e-05,
+      "loss": 1.2651,
+      "step": 51800
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 5e-05,
+      "loss": 1.2241,
+      "step": 51900
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 5e-05,
+      "loss": 1.2928,
+      "step": 52000
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 5e-05,
+      "loss": 1.2599,
+      "step": 52100
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 5e-05,
+      "loss": 1.2408,
+      "step": 52200
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 5e-05,
+      "loss": 1.2664,
+      "step": 52300
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 5e-05,
+      "loss": 1.2434,
+      "step": 52400
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 5e-05,
+      "loss": 1.2654,
+      "step": 52500
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 5e-05,
+      "loss": 1.2663,
+      "step": 52600
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 5e-05,
+      "loss": 1.2511,
+      "step": 52700
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 5e-05,
+      "loss": 1.2763,
+      "step": 52800
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 5e-05,
+      "loss": 1.2623,
+      "step": 52900
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 5e-05,
+      "loss": 1.2667,
+      "step": 53000
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 5e-05,
+      "loss": 1.2627,
+      "step": 53100
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 5e-05,
+      "loss": 1.2648,
+      "step": 53200
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 5e-05,
+      "loss": 1.2609,
+      "step": 53300
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 5e-05,
+      "loss": 1.2636,
+      "step": 53400
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 5e-05,
+      "loss": 1.2592,
+      "step": 53500
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 5e-05,
+      "loss": 1.2517,
+      "step": 53600
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 5e-05,
+      "loss": 1.2881,
+      "step": 53700
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 5e-05,
+      "loss": 1.254,
+      "step": 53800
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 5e-05,
+      "loss": 1.2749,
+      "step": 53900
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 5e-05,
+      "loss": 1.2544,
+      "step": 54000
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 5e-05,
+      "loss": 1.2681,
+      "step": 54100
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 5e-05,
+      "loss": 1.2929,
+      "step": 54200
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 5e-05,
+      "loss": 1.273,
+      "step": 54300
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 5e-05,
+      "loss": 1.2709,
+      "step": 54400
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 5e-05,
+      "loss": 1.2504,
+      "step": 54500
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 5e-05,
+      "loss": 1.2867,
+      "step": 54600
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 5e-05,
+      "loss": 1.2771,
+      "step": 54700
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 5e-05,
+      "loss": 1.257,
+      "step": 54800
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 5e-05,
+      "loss": 1.287,
+      "step": 54900
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 5e-05,
+      "loss": 1.2629,
+      "step": 55000
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 5e-05,
+      "loss": 1.2612,
+      "step": 55100
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 5e-05,
+      "loss": 1.2433,
+      "step": 55200
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 5e-05,
+      "loss": 1.2655,
+      "step": 55300
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 5e-05,
+      "loss": 1.2762,
+      "step": 55400
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 5e-05,
+      "loss": 1.2714,
+      "step": 55500
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 5e-05,
+      "loss": 1.2821,
+      "step": 55600
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 5e-05,
+      "loss": 1.265,
+      "step": 55700
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 5e-05,
+      "loss": 1.2547,
+      "step": 55800
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 5e-05,
+      "loss": 1.2555,
+      "step": 55900
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 5e-05,
+      "loss": 1.2784,
+      "step": 56000
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 5e-05,
+      "loss": 1.2729,
+      "step": 56100
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 5e-05,
+      "loss": 1.2978,
+      "step": 56200
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 5e-05,
+      "loss": 1.2682,
+      "step": 56300
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 5e-05,
+      "loss": 1.2884,
+      "step": 56400
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 5e-05,
+      "loss": 1.2571,
+      "step": 56500
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 5e-05,
+      "loss": 1.2852,
+      "step": 56600
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 5e-05,
+      "loss": 1.2858,
+      "step": 56700
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 5e-05,
+      "loss": 1.2736,
+      "step": 56800
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 5e-05,
+      "loss": 1.2969,
+      "step": 56900
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 5e-05,
+      "loss": 1.2767,
+      "step": 57000
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 5e-05,
+      "loss": 1.2777,
+      "step": 57100
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 5e-05,
+      "loss": 1.2589,
+      "step": 57200
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 5e-05,
+      "loss": 1.2571,
+      "step": 57300
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 5e-05,
+      "loss": 1.2631,
+      "step": 57400
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 5e-05,
+      "loss": 1.2696,
+      "step": 57500
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 5e-05,
+      "loss": 1.2654,
+      "step": 57600
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 5e-05,
+      "loss": 1.2738,
+      "step": 57700
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 5e-05,
+      "loss": 1.2798,
+      "step": 57800
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 5e-05,
+      "loss": 1.301,
+      "step": 57900
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 5e-05,
+      "loss": 1.2678,
+      "step": 58000
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 5e-05,
+      "loss": 1.2659,
+      "step": 58100
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 5e-05,
+      "loss": 1.2779,
+      "step": 58200
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 5e-05,
+      "loss": 1.2936,
+      "step": 58300
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 5e-05,
+      "loss": 1.2843,
+      "step": 58400
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 5e-05,
+      "loss": 1.2633,
+      "step": 58500
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 5e-05,
+      "loss": 1.2969,
+      "step": 58600
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 5e-05,
+      "loss": 1.2624,
+      "step": 58700
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 5e-05,
+      "loss": 1.2952,
+      "step": 58800
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 5e-05,
+      "loss": 1.2845,
+      "step": 58900
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 5e-05,
+      "loss": 1.2759,
+      "step": 59000
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 5e-05,
+      "loss": 1.2948,
+      "step": 59100
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 5e-05,
+      "loss": 1.2547,
+      "step": 59200
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 5e-05,
+      "loss": 1.2721,
+      "step": 59300
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 5e-05,
+      "loss": 1.2678,
+      "step": 59400
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 5e-05,
+      "loss": 1.2821,
+      "step": 59500
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 5e-05,
+      "loss": 1.2667,
+      "step": 59600
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 5e-05,
+      "loss": 1.2791,
+      "step": 59700
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 5e-05,
+      "loss": 1.284,
+      "step": 59800
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 5e-05,
+      "loss": 1.2932,
+      "step": 59900
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 5e-05,
+      "loss": 1.3006,
+      "step": 60000
+    },
+    {
+      "epoch": 4.71,
+      "eval_gen_len": 18.713604661428445,
+      "eval_loss": 1.4940598011016846,
+      "eval_rouge1": 39.3973,
+      "eval_rouge2": 17.7088,
+      "eval_rougeL": 32.7938,
+      "eval_rougeLsum": 32.7949,
+      "eval_runtime": 2600.1572,
+      "eval_samples_per_second": 4.356,
+      "eval_steps_per_second": 1.089,
+      "step": 60000
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 5e-05,
+      "loss": 1.2774,
+      "step": 60100
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 5e-05,
+      "loss": 1.296,
+      "step": 60200
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 5e-05,
+      "loss": 1.3068,
+      "step": 60300
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 5e-05,
+      "loss": 1.2759,
+      "step": 60400
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 5e-05,
+      "loss": 1.2824,
+      "step": 60500
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 5e-05,
+      "loss": 1.3045,
+      "step": 60600
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 5e-05,
+      "loss": 1.2766,
+      "step": 60700
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 5e-05,
+      "loss": 1.2674,
+      "step": 60800
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 5e-05,
+      "loss": 1.2793,
+      "step": 60900
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 5e-05,
+      "loss": 1.284,
+      "step": 61000
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 5e-05,
+      "loss": 1.2861,
+      "step": 61100
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 5e-05,
+      "loss": 1.2924,
+      "step": 61200
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 5e-05,
+      "loss": 1.2949,
+      "step": 61300
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 5e-05,
+      "loss": 1.2945,
+      "step": 61400
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 5e-05,
+      "loss": 1.2832,
+      "step": 61500
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 5e-05,
+      "loss": 1.2674,
+      "step": 61600
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 5e-05,
+      "loss": 1.2703,
+      "step": 61700
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 5e-05,
+      "loss": 1.2977,
+      "step": 61800
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 5e-05,
+      "loss": 1.2994,
+      "step": 61900
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 5e-05,
+      "loss": 1.2934,
+      "step": 62000
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 5e-05,
+      "loss": 1.2885,
+      "step": 62100
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 5e-05,
+      "loss": 1.2849,
+      "step": 62200
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 5e-05,
+      "loss": 1.3009,
+      "step": 62300
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 5e-05,
+      "loss": 1.284,
+      "step": 62400
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 5e-05,
+      "loss": 1.2807,
+      "step": 62500
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 5e-05,
+      "loss": 1.2911,
+      "step": 62600
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 5e-05,
+      "loss": 1.2881,
+      "step": 62700
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 5e-05,
+      "loss": 1.2812,
+      "step": 62800
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 5e-05,
+      "loss": 1.2934,
+      "step": 62900
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 5e-05,
+      "loss": 1.284,
+      "step": 63000
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 5e-05,
+      "loss": 1.3022,
+      "step": 63100
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 5e-05,
+      "loss": 1.2916,
+      "step": 63200
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 5e-05,
+      "loss": 1.3006,
+      "step": 63300
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 5e-05,
+      "loss": 1.3157,
+      "step": 63400
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 5e-05,
+      "loss": 1.2868,
+      "step": 63500
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 5e-05,
+      "loss": 1.2879,
+      "step": 63600
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 5e-05,
+      "loss": 1.2873,
+      "step": 63700
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 5e-05,
+      "loss": 1.2625,
+      "step": 63800
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 5e-05,
+      "loss": 1.1801,
+      "step": 63900
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 5e-05,
+      "loss": 1.178,
+      "step": 64000
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 5e-05,
+      "loss": 1.1779,
+      "step": 64100
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 5e-05,
+      "loss": 1.1858,
+      "step": 64200
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 5e-05,
+      "loss": 1.1579,
+      "step": 64300
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 5e-05,
+      "loss": 1.1974,
+      "step": 64400
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 5e-05,
+      "loss": 1.1938,
+      "step": 64500
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 5e-05,
+      "loss": 1.1938,
+      "step": 64600
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 5e-05,
+      "loss": 1.2015,
+      "step": 64700
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 5e-05,
+      "loss": 1.1884,
+      "step": 64800
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 5e-05,
+      "loss": 1.1764,
+      "step": 64900
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 5e-05,
+      "loss": 1.1796,
+      "step": 65000
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 5e-05,
+      "loss": 1.1815,
+      "step": 65100
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 5e-05,
+      "loss": 1.2015,
+      "step": 65200
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 5e-05,
+      "loss": 1.1862,
+      "step": 65300
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 5e-05,
+      "loss": 1.2127,
+      "step": 65400
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 5e-05,
+      "loss": 1.1809,
+      "step": 65500
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 5e-05,
+      "loss": 1.1978,
+      "step": 65600
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 5e-05,
+      "loss": 1.1866,
+      "step": 65700
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 5e-05,
+      "loss": 1.1646,
+      "step": 65800
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 5e-05,
+      "loss": 1.2152,
+      "step": 65900
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 5e-05,
+      "loss": 1.1896,
+      "step": 66000
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 5e-05,
+      "loss": 1.1779,
+      "step": 66100
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 5e-05,
+      "loss": 1.1951,
+      "step": 66200
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 5e-05,
+      "loss": 1.2209,
+      "step": 66300
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 5e-05,
+      "loss": 1.2072,
+      "step": 66400
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 5e-05,
+      "loss": 1.1887,
+      "step": 66500
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 5e-05,
+      "loss": 1.2142,
+      "step": 66600
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 5e-05,
+      "loss": 1.157,
+      "step": 66700
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 5e-05,
+      "loss": 1.2097,
+      "step": 66800
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 5e-05,
+      "loss": 1.2041,
+      "step": 66900
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 5e-05,
+      "loss": 1.1877,
+      "step": 67000
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 5e-05,
+      "loss": 1.1838,
+      "step": 67100
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 5e-05,
+      "loss": 1.1936,
+      "step": 67200
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 5e-05,
+      "loss": 1.2084,
+      "step": 67300
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 5e-05,
+      "loss": 1.2174,
+      "step": 67400
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 5e-05,
+      "loss": 1.2014,
+      "step": 67500
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 5e-05,
+      "loss": 1.2259,
+      "step": 67600
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 5e-05,
+      "loss": 1.2031,
+      "step": 67700
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 5e-05,
+      "loss": 1.2046,
+      "step": 67800
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 5e-05,
+      "loss": 1.2045,
+      "step": 67900
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 5e-05,
+      "loss": 1.2016,
+      "step": 68000
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 5e-05,
+      "loss": 1.1902,
+      "step": 68100
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 5e-05,
+      "loss": 1.1769,
+      "step": 68200
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 5e-05,
+      "loss": 1.1939,
+      "step": 68300
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 5e-05,
+      "loss": 1.1984,
+      "step": 68400
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 5e-05,
+      "loss": 1.1975,
+      "step": 68500
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 5e-05,
+      "loss": 1.2162,
+      "step": 68600
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 5e-05,
+      "loss": 1.219,
+      "step": 68700
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 5e-05,
+      "loss": 1.1927,
+      "step": 68800
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 5e-05,
+      "loss": 1.2096,
+      "step": 68900
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 5e-05,
+      "loss": 1.2227,
+      "step": 69000
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 5e-05,
+      "loss": 1.2265,
+      "step": 69100
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 5e-05,
+      "loss": 1.2238,
+      "step": 69200
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 5e-05,
+      "loss": 1.2066,
+      "step": 69300
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 5e-05,
+      "loss": 1.2013,
+      "step": 69400
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 5e-05,
+      "loss": 1.1938,
+      "step": 69500
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 5e-05,
+      "loss": 1.2008,
+      "step": 69600
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 5e-05,
+      "loss": 1.2239,
+      "step": 69700
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 5e-05,
+      "loss": 1.2272,
+      "step": 69800
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 5e-05,
+      "loss": 1.2162,
+      "step": 69900
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 5e-05,
+      "loss": 1.2137,
+      "step": 70000
+    },
+    {
+      "epoch": 5.49,
+      "eval_gen_len": 18.72711220976428,
+      "eval_loss": 1.511597752571106,
+      "eval_rouge1": 39.4402,
+      "eval_rouge2": 17.7009,
+      "eval_rougeL": 32.8481,
+      "eval_rougeLsum": 32.8525,
+      "eval_runtime": 2553.8215,
+      "eval_samples_per_second": 4.435,
+      "eval_steps_per_second": 1.109,
+      "step": 70000
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 5e-05,
+      "loss": 1.2279,
+      "step": 70100
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 5e-05,
+      "loss": 1.2261,
+      "step": 70200
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 5e-05,
+      "loss": 1.2069,
+      "step": 70300
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 5e-05,
+      "loss": 1.2134,
+      "step": 70400
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 5e-05,
+      "loss": 1.215,
+      "step": 70500
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 5e-05,
+      "loss": 1.22,
+      "step": 70600
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 5e-05,
+      "loss": 1.2227,
+      "step": 70700
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 5e-05,
+      "loss": 1.2098,
+      "step": 70800
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 5e-05,
+      "loss": 1.2228,
+      "step": 70900
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 5e-05,
+      "loss": 1.2188,
+      "step": 71000
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 5e-05,
+      "loss": 1.2287,
+      "step": 71100
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 5e-05,
+      "loss": 1.2088,
+      "step": 71200
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 5e-05,
+      "loss": 1.2171,
+      "step": 71300
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 5e-05,
+      "loss": 1.2172,
+      "step": 71400
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 5e-05,
+      "loss": 1.2379,
+      "step": 71500
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 5e-05,
+      "loss": 1.2229,
+      "step": 71600
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 5e-05,
+      "loss": 1.2165,
+      "step": 71700
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 5e-05,
+      "loss": 1.2165,
+      "step": 71800
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 5e-05,
+      "loss": 1.2346,
+      "step": 71900
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 5e-05,
+      "loss": 1.224,
+      "step": 72000
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 5e-05,
+      "loss": 1.2332,
+      "step": 72100
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 5e-05,
+      "loss": 1.2139,
+      "step": 72200
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 5e-05,
+      "loss": 1.2102,
+      "step": 72300
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 5e-05,
+      "loss": 1.202,
+      "step": 72400
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 5e-05,
+      "loss": 1.2048,
+      "step": 72500
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 5e-05,
+      "loss": 1.2215,
+      "step": 72600
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 5e-05,
+      "loss": 1.2297,
+      "step": 72700
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 5e-05,
+      "loss": 1.2171,
+      "step": 72800
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 5e-05,
+      "loss": 1.2257,
+      "step": 72900
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 5e-05,
+      "loss": 1.2091,
+      "step": 73000
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 5e-05,
+      "loss": 1.2063,
+      "step": 73100
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 5e-05,
+      "loss": 1.1986,
+      "step": 73200
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 5e-05,
+      "loss": 1.222,
+      "step": 73300
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 5e-05,
+      "loss": 1.2118,
+      "step": 73400
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 5e-05,
+      "loss": 1.2252,
+      "step": 73500
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 5e-05,
+      "loss": 1.2296,
+      "step": 73600
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 5e-05,
+      "loss": 1.2421,
+      "step": 73700
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 5e-05,
+      "loss": 1.24,
+      "step": 73800
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 5e-05,
+      "loss": 1.2032,
+      "step": 73900
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 5e-05,
+      "loss": 1.2204,
+      "step": 74000
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 5e-05,
+      "loss": 1.2147,
+      "step": 74100
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 5e-05,
+      "loss": 1.2317,
+      "step": 74200
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 5e-05,
+      "loss": 1.2055,
+      "step": 74300
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 5e-05,
+      "loss": 1.2386,
+      "step": 74400
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 5e-05,
+      "loss": 1.227,
+      "step": 74500
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 5e-05,
+      "loss": 1.2341,
+      "step": 74600
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 5e-05,
+      "loss": 1.2191,
+      "step": 74700
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 5e-05,
+      "loss": 1.2396,
+      "step": 74800
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 5e-05,
+      "loss": 1.2343,
+      "step": 74900
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 5e-05,
+      "loss": 1.2257,
+      "step": 75000
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 5e-05,
+      "loss": 1.2295,
+      "step": 75100
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 5e-05,
+      "loss": 1.2128,
+      "step": 75200
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 5e-05,
+      "loss": 1.2317,
+      "step": 75300
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 5e-05,
+      "loss": 1.2195,
+      "step": 75400
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 5e-05,
+      "loss": 1.2275,
+      "step": 75500
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 5e-05,
+      "loss": 1.2202,
+      "step": 75600
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 5e-05,
+      "loss": 1.2161,
+      "step": 75700
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 5e-05,
+      "loss": 1.2215,
+      "step": 75800
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 5e-05,
+      "loss": 1.2285,
+      "step": 75900
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 5e-05,
+      "loss": 1.228,
+      "step": 76000
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 5e-05,
+      "loss": 1.2368,
+      "step": 76100
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 5e-05,
+      "loss": 1.2231,
+      "step": 76200
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 5e-05,
+      "loss": 1.213,
+      "step": 76300
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 5e-05,
+      "loss": 1.233,
+      "step": 76400
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 5e-05,
+      "loss": 1.2155,
+      "step": 76500
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 5e-05,
+      "loss": 1.1235,
+      "step": 76600
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 5e-05,
+      "loss": 1.1247,
+      "step": 76700
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 5e-05,
+      "loss": 1.1164,
+      "step": 76800
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 5e-05,
+      "loss": 1.1156,
+      "step": 76900
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 5e-05,
+      "loss": 1.1105,
+      "step": 77000
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 5e-05,
+      "loss": 1.107,
+      "step": 77100
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 5e-05,
+      "loss": 1.1293,
+      "step": 77200
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 5e-05,
+      "loss": 1.1294,
+      "step": 77300
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 5e-05,
+      "loss": 1.1274,
+      "step": 77400
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 5e-05,
+      "loss": 1.1346,
+      "step": 77500
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 5e-05,
+      "loss": 1.1089,
+      "step": 77600
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 5e-05,
+      "loss": 1.1235,
+      "step": 77700
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 5e-05,
+      "loss": 1.1154,
+      "step": 77800
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 5e-05,
+      "loss": 1.1244,
+      "step": 77900
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 5e-05,
+      "loss": 1.1424,
+      "step": 78000
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 5e-05,
+      "loss": 1.1417,
+      "step": 78100
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 5e-05,
+      "loss": 1.1393,
+      "step": 78200
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 5e-05,
+      "loss": 1.1147,
+      "step": 78300
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 5e-05,
+      "loss": 1.1396,
+      "step": 78400
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 5e-05,
+      "loss": 1.1564,
+      "step": 78500
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 5e-05,
+      "loss": 1.1457,
+      "step": 78600
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 5e-05,
+      "loss": 1.1349,
+      "step": 78700
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 5e-05,
+      "loss": 1.1724,
+      "step": 78800
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 5e-05,
+      "loss": 1.1388,
+      "step": 78900
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 5e-05,
+      "loss": 1.1393,
+      "step": 79000
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 5e-05,
+      "loss": 1.1325,
+      "step": 79100
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 5e-05,
+      "loss": 1.1315,
+      "step": 79200
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 5e-05,
+      "loss": 1.1428,
+      "step": 79300
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 5e-05,
+      "loss": 1.148,
+      "step": 79400
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 5e-05,
+      "loss": 1.1534,
+      "step": 79500
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 5e-05,
+      "loss": 1.1454,
+      "step": 79600
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 5e-05,
+      "loss": 1.1383,
+      "step": 79700
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 5e-05,
+      "loss": 1.1266,
+      "step": 79800
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 5e-05,
+      "loss": 1.1382,
+      "step": 79900
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 5e-05,
+      "loss": 1.1169,
+      "step": 80000
+    },
+    {
+      "epoch": 6.27,
+      "eval_gen_len": 18.758806391807187,
+      "eval_loss": 1.5334253311157227,
+      "eval_rouge1": 39.579,
+      "eval_rouge2": 17.841,
+      "eval_rougeL": 32.8832,
+      "eval_rougeLsum": 32.8889,
+      "eval_runtime": 2444.6143,
+      "eval_samples_per_second": 4.633,
+      "eval_steps_per_second": 1.158,
+      "step": 80000
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 5e-05,
+      "loss": 1.1758,
+      "step": 80100
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 5e-05,
+      "loss": 1.1429,
+      "step": 80200
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 5e-05,
+      "loss": 1.1347,
+      "step": 80300
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 5e-05,
+      "loss": 1.149,
+      "step": 80400
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 5e-05,
+      "loss": 1.1293,
+      "step": 80500
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 5e-05,
+      "loss": 1.1366,
+      "step": 80600
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 5e-05,
+      "loss": 1.1534,
+      "step": 80700
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 5e-05,
+      "loss": 1.1561,
+      "step": 80800
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 5e-05,
+      "loss": 1.1476,
+      "step": 80900
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 5e-05,
+      "loss": 1.1491,
+      "step": 81000
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 5e-05,
+      "loss": 1.136,
+      "step": 81100
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 5e-05,
+      "loss": 1.1254,
+      "step": 81200
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 5e-05,
+      "loss": 1.1536,
+      "step": 81300
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 5e-05,
+      "loss": 1.13,
+      "step": 81400
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 5e-05,
+      "loss": 1.1626,
+      "step": 81500
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 5e-05,
+      "loss": 1.1309,
+      "step": 81600
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 5e-05,
+      "loss": 1.1467,
+      "step": 81700
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 5e-05,
+      "loss": 1.1761,
+      "step": 81800
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 5e-05,
+      "loss": 1.1415,
+      "step": 81900
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 5e-05,
+      "loss": 1.1515,
+      "step": 82000
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 5e-05,
+      "loss": 1.1396,
+      "step": 82100
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 5e-05,
+      "loss": 1.1501,
+      "step": 82200
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 5e-05,
+      "loss": 1.1594,
+      "step": 82300
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 5e-05,
+      "loss": 1.1486,
+      "step": 82400
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 5e-05,
+      "loss": 1.1459,
+      "step": 82500
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 5e-05,
+      "loss": 1.1682,
+      "step": 82600
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 5e-05,
+      "loss": 1.1647,
+      "step": 82700
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 5e-05,
+      "loss": 1.1347,
+      "step": 82800
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 5e-05,
+      "loss": 1.1628,
+      "step": 82900
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 5e-05,
+      "loss": 1.129,
+      "step": 83000
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 5e-05,
+      "loss": 1.1791,
+      "step": 83100
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 5e-05,
+      "loss": 1.1582,
+      "step": 83200
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 5e-05,
+      "loss": 1.157,
+      "step": 83300
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 5e-05,
+      "loss": 1.1459,
+      "step": 83400
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 5e-05,
+      "loss": 1.15,
+      "step": 83500
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 5e-05,
+      "loss": 1.1235,
+      "step": 83600
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 5e-05,
+      "loss": 1.142,
+      "step": 83700
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 5e-05,
+      "loss": 1.1587,
+      "step": 83800
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 5e-05,
+      "loss": 1.1564,
+      "step": 83900
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 5e-05,
+      "loss": 1.1457,
+      "step": 84000
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 5e-05,
+      "loss": 1.1434,
+      "step": 84100
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 5e-05,
+      "loss": 1.1512,
+      "step": 84200
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 5e-05,
+      "loss": 1.1666,
+      "step": 84300
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 5e-05,
+      "loss": 1.1624,
+      "step": 84400
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 5e-05,
+      "loss": 1.1629,
+      "step": 84500
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 5e-05,
+      "loss": 1.1617,
+      "step": 84600
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 5e-05,
+      "loss": 1.1573,
+      "step": 84700
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 5e-05,
+      "loss": 1.1666,
+      "step": 84800
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 5e-05,
+      "loss": 1.1675,
+      "step": 84900
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 5e-05,
+      "loss": 1.1446,
+      "step": 85000
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 5e-05,
+      "loss": 1.1584,
+      "step": 85100
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 5e-05,
+      "loss": 1.1524,
+      "step": 85200
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 5e-05,
+      "loss": 1.1391,
+      "step": 85300
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 5e-05,
+      "loss": 1.1488,
+      "step": 85400
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 5e-05,
+      "loss": 1.1671,
+      "step": 85500
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 5e-05,
+      "loss": 1.1627,
+      "step": 85600
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 5e-05,
+      "loss": 1.1586,
+      "step": 85700
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 5e-05,
+      "loss": 1.1468,
+      "step": 85800
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 5e-05,
+      "loss": 1.1746,
+      "step": 85900
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 5e-05,
+      "loss": 1.1674,
+      "step": 86000
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 5e-05,
+      "loss": 1.148,
+      "step": 86100
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 5e-05,
+      "loss": 1.1649,
+      "step": 86200
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 5e-05,
+      "loss": 1.1567,
+      "step": 86300
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 5e-05,
+      "loss": 1.1611,
+      "step": 86400
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 5e-05,
+      "loss": 1.1541,
+      "step": 86500
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 5e-05,
+      "loss": 1.1589,
+      "step": 86600
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 5e-05,
+      "loss": 1.1942,
+      "step": 86700
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 5e-05,
+      "loss": 1.1697,
+      "step": 86800
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 5e-05,
+      "loss": 1.1645,
+      "step": 86900
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 5e-05,
+      "loss": 1.1638,
+      "step": 87000
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 5e-05,
+      "loss": 1.1851,
+      "step": 87100
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 5e-05,
+      "loss": 1.1837,
+      "step": 87200
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 5e-05,
+      "loss": 1.1693,
+      "step": 87300
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 5e-05,
+      "loss": 1.1633,
+      "step": 87400
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 5e-05,
+      "loss": 1.1601,
+      "step": 87500
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 5e-05,
+      "loss": 1.1706,
+      "step": 87600
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 5e-05,
+      "loss": 1.1805,
+      "step": 87700
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 5e-05,
+      "loss": 1.1802,
+      "step": 87800
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 5e-05,
+      "loss": 1.1563,
+      "step": 87900
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 5e-05,
+      "loss": 1.182,
+      "step": 88000
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 5e-05,
+      "loss": 1.1532,
+      "step": 88100
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 5e-05,
+      "loss": 1.1903,
+      "step": 88200
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 5e-05,
+      "loss": 1.163,
+      "step": 88300
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 5e-05,
+      "loss": 1.1692,
+      "step": 88400
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 5e-05,
+      "loss": 1.1884,
+      "step": 88500
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 5e-05,
+      "loss": 1.1697,
+      "step": 88600
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 5e-05,
+      "loss": 1.183,
+      "step": 88700
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 5e-05,
+      "loss": 1.1736,
+      "step": 88800
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 5e-05,
+      "loss": 1.1652,
+      "step": 88900
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 5e-05,
+      "loss": 1.1804,
+      "step": 89000
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 5e-05,
+      "loss": 1.1731,
+      "step": 89100
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 5e-05,
+      "loss": 1.1745,
+      "step": 89200
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 5e-05,
+      "loss": 1.1464,
+      "step": 89300
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 5e-05,
+      "loss": 1.0755,
+      "step": 89400
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 5e-05,
+      "loss": 1.0766,
+      "step": 89500
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 5e-05,
+      "loss": 1.068,
+      "step": 89600
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 5e-05,
+      "loss": 1.0792,
+      "step": 89700
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 5e-05,
+      "loss": 1.0746,
+      "step": 89800
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 5e-05,
+      "loss": 1.0775,
+      "step": 89900
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 5e-05,
+      "loss": 1.0757,
+      "step": 90000
+    },
+    {
+      "epoch": 7.06,
+      "eval_gen_len": 18.725523086430652,
+      "eval_loss": 1.5569835901260376,
+      "eval_rouge1": 39.4694,
+      "eval_rouge2": 17.8419,
+      "eval_rougeL": 32.87,
+      "eval_rougeLsum": 32.864,
+      "eval_runtime": 2362.234,
+      "eval_samples_per_second": 4.795,
+      "eval_steps_per_second": 1.199,
+      "step": 90000
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 5e-05,
+      "loss": 1.0726,
+      "step": 90100
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 5e-05,
+      "loss": 1.0736,
+      "step": 90200
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 5e-05,
+      "loss": 1.0776,
+      "step": 90300
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 5e-05,
+      "loss": 1.0533,
+      "step": 90400
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 5e-05,
+      "loss": 1.0583,
+      "step": 90500
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 5e-05,
+      "loss": 1.0816,
+      "step": 90600
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 5e-05,
+      "loss": 1.0684,
+      "step": 90700
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 5e-05,
+      "loss": 1.0656,
+      "step": 90800
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 5e-05,
+      "loss": 1.0685,
+      "step": 90900
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 5e-05,
+      "loss": 1.1002,
+      "step": 91000
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 5e-05,
+      "loss": 1.0789,
+      "step": 91100
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 5e-05,
+      "loss": 1.0738,
+      "step": 91200
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 5e-05,
+      "loss": 1.0764,
+      "step": 91300
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 5e-05,
+      "loss": 1.0733,
+      "step": 91400
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 5e-05,
+      "loss": 1.0984,
+      "step": 91500
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 5e-05,
+      "loss": 1.0767,
+      "step": 91600
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 5e-05,
+      "loss": 1.0991,
+      "step": 91700
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 5e-05,
+      "loss": 1.0882,
+      "step": 91800
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 5e-05,
+      "loss": 1.0868,
+      "step": 91900
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 5e-05,
+      "loss": 1.0789,
+      "step": 92000
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 5e-05,
+      "loss": 1.0887,
+      "step": 92100
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 5e-05,
+      "loss": 1.0777,
+      "step": 92200
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 5e-05,
+      "loss": 1.0868,
+      "step": 92300
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 5e-05,
+      "loss": 1.0971,
+      "step": 92400
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 5e-05,
+      "loss": 1.0849,
+      "step": 92500
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 5e-05,
+      "loss": 1.0796,
+      "step": 92600
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 5e-05,
+      "loss": 1.0923,
+      "step": 92700
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 5e-05,
+      "loss": 1.0929,
+      "step": 92800
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 5e-05,
+      "loss": 1.0868,
+      "step": 92900
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 5e-05,
+      "loss": 1.0987,
+      "step": 93000
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 5e-05,
+      "loss": 1.0891,
+      "step": 93100
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 5e-05,
+      "loss": 1.0812,
+      "step": 93200
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 5e-05,
+      "loss": 1.087,
+      "step": 93300
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 5e-05,
+      "loss": 1.0809,
+      "step": 93400
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 5e-05,
+      "loss": 1.0906,
+      "step": 93500
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 5e-05,
+      "loss": 1.104,
+      "step": 93600
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 5e-05,
+      "loss": 1.0869,
+      "step": 93700
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 5e-05,
+      "loss": 1.074,
+      "step": 93800
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 5e-05,
+      "loss": 1.091,
+      "step": 93900
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 5e-05,
+      "loss": 1.0797,
+      "step": 94000
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 5e-05,
+      "loss": 1.0966,
+      "step": 94100
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 5e-05,
+      "loss": 1.0721,
+      "step": 94200
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 5e-05,
+      "loss": 1.0829,
+      "step": 94300
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 5e-05,
+      "loss": 1.0853,
+      "step": 94400
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 5e-05,
+      "loss": 1.0875,
+      "step": 94500
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 5e-05,
+      "loss": 1.0935,
+      "step": 94600
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 5e-05,
+      "loss": 1.1017,
+      "step": 94700
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 5e-05,
+      "loss": 1.0981,
+      "step": 94800
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 5e-05,
+      "loss": 1.0926,
+      "step": 94900
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 5e-05,
+      "loss": 1.0902,
+      "step": 95000
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 5e-05,
+      "loss": 1.1018,
+      "step": 95100
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 5e-05,
+      "loss": 1.0786,
+      "step": 95200
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 5e-05,
+      "loss": 1.1005,
+      "step": 95300
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 5e-05,
+      "loss": 1.094,
+      "step": 95400
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 5e-05,
+      "loss": 1.1042,
+      "step": 95500
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 5e-05,
+      "loss": 1.0926,
+      "step": 95600
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 5e-05,
+      "loss": 1.1012,
+      "step": 95700
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 5e-05,
+      "loss": 1.1069,
+      "step": 95800
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 5e-05,
+      "loss": 1.0873,
+      "step": 95900
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 5e-05,
+      "loss": 1.0865,
+      "step": 96000
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 5e-05,
+      "loss": 1.1074,
+      "step": 96100
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 5e-05,
+      "loss": 1.0932,
+      "step": 96200
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 5e-05,
+      "loss": 1.061,
+      "step": 96300
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 5e-05,
+      "loss": 1.1158,
+      "step": 96400
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 5e-05,
+      "loss": 1.0995,
+      "step": 96500
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 5e-05,
+      "loss": 1.0997,
+      "step": 96600
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 5e-05,
+      "loss": 1.1066,
+      "step": 96700
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 5e-05,
+      "loss": 1.089,
+      "step": 96800
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 5e-05,
+      "loss": 1.0987,
+      "step": 96900
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 5e-05,
+      "loss": 1.0963,
+      "step": 97000
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 5e-05,
+      "loss": 1.0996,
+      "step": 97100
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 5e-05,
+      "loss": 1.0909,
+      "step": 97200
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 5e-05,
+      "loss": 1.1201,
+      "step": 97300
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 5e-05,
+      "loss": 1.1071,
+      "step": 97400
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 5e-05,
+      "loss": 1.1108,
+      "step": 97500
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 5e-05,
+      "loss": 1.0933,
+      "step": 97600
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 5e-05,
+      "loss": 1.1054,
+      "step": 97700
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 5e-05,
+      "loss": 1.1041,
+      "step": 97800
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 5e-05,
+      "loss": 1.1092,
+      "step": 97900
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 5e-05,
+      "loss": 1.1093,
+      "step": 98000
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 5e-05,
+      "loss": 1.1123,
+      "step": 98100
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 5e-05,
+      "loss": 1.1099,
+      "step": 98200
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 5e-05,
+      "loss": 1.1093,
+      "step": 98300
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 5e-05,
+      "loss": 1.1016,
+      "step": 98400
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 5e-05,
+      "loss": 1.1115,
+      "step": 98500
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 5e-05,
+      "loss": 1.0945,
+      "step": 98600
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 5e-05,
+      "loss": 1.1069,
+      "step": 98700
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 5e-05,
+      "loss": 1.1253,
+      "step": 98800
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 5e-05,
+      "loss": 1.1174,
+      "step": 98900
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 5e-05,
+      "loss": 1.1218,
+      "step": 99000
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 5e-05,
+      "loss": 1.1059,
+      "step": 99100
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 5e-05,
+      "loss": 1.1036,
+      "step": 99200
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 5e-05,
+      "loss": 1.1115,
+      "step": 99300
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 5e-05,
+      "loss": 1.1155,
+      "step": 99400
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 5e-05,
+      "loss": 1.1088,
+      "step": 99500
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 5e-05,
+      "loss": 1.1035,
+      "step": 99600
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 5e-05,
+      "loss": 1.1217,
+      "step": 99700
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 5e-05,
+      "loss": 1.114,
+      "step": 99800
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 5e-05,
+      "loss": 1.1107,
+      "step": 99900
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 5e-05,
+      "loss": 1.0901,
+      "step": 100000
+    },
+    {
+      "epoch": 7.84,
+      "eval_gen_len": 18.753509314028427,
+      "eval_loss": 1.5419243574142456,
+      "eval_rouge1": 39.5689,
+      "eval_rouge2": 17.9809,
+      "eval_rougeL": 33.0274,
+      "eval_rougeLsum": 33.0285,
+      "eval_runtime": 2388.4938,
+      "eval_samples_per_second": 4.742,
+      "eval_steps_per_second": 1.186,
+      "step": 100000
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 5e-05,
+      "loss": 1.1357,
+      "step": 100100
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 5e-05,
+      "loss": 1.1203,
+      "step": 100200
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 5e-05,
+      "loss": 1.1088,
+      "step": 100300
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 5e-05,
+      "loss": 1.1058,
+      "step": 100400
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 5e-05,
+      "loss": 1.1042,
+      "step": 100500
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 5e-05,
+      "loss": 1.1145,
+      "step": 100600
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 5e-05,
+      "loss": 1.1221,
+      "step": 100700
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 5e-05,
+      "loss": 1.0963,
+      "step": 100800
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 5e-05,
+      "loss": 1.1021,
+      "step": 100900
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 5e-05,
+      "loss": 1.1095,
+      "step": 101000
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 5e-05,
+      "loss": 1.1184,
+      "step": 101100
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 5e-05,
+      "loss": 1.1181,
+      "step": 101200
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 5e-05,
+      "loss": 1.1212,
+      "step": 101300
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 5e-05,
+      "loss": 1.1323,
+      "step": 101400
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 5e-05,
+      "loss": 1.1228,
+      "step": 101500
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 5e-05,
+      "loss": 1.123,
+      "step": 101600
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 5e-05,
+      "loss": 1.1024,
+      "step": 101700
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 5e-05,
+      "loss": 1.1225,
+      "step": 101800
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 5e-05,
+      "loss": 1.0943,
+      "step": 101900
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 5e-05,
+      "loss": 1.1048,
+      "step": 102000
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 5e-05,
+      "loss": 1.0375,
+      "step": 102100
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 5e-05,
+      "loss": 1.0024,
+      "step": 102200
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 5e-05,
+      "loss": 1.0088,
+      "step": 102300
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 5e-05,
+      "loss": 0.9942,
+      "step": 102400
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 5e-05,
+      "loss": 1.0172,
+      "step": 102500
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 5e-05,
+      "loss": 1.0192,
+      "step": 102600
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 5e-05,
+      "loss": 1.016,
+      "step": 102700
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 5e-05,
+      "loss": 1.0227,
+      "step": 102800
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 5e-05,
+      "loss": 1.0317,
+      "step": 102900
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 5e-05,
+      "loss": 1.0192,
+      "step": 103000
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 5e-05,
+      "loss": 1.0271,
+      "step": 103100
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 5e-05,
+      "loss": 1.0337,
+      "step": 103200
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 5e-05,
+      "loss": 1.017,
+      "step": 103300
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 5e-05,
+      "loss": 1.0143,
+      "step": 103400
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 5e-05,
+      "loss": 1.0128,
+      "step": 103500
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 5e-05,
+      "loss": 1.0181,
+      "step": 103600
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 5e-05,
+      "loss": 1.0114,
+      "step": 103700
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 5e-05,
+      "loss": 1.0279,
+      "step": 103800
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 5e-05,
+      "loss": 1.0109,
+      "step": 103900
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 5e-05,
+      "loss": 1.0197,
+      "step": 104000
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 5e-05,
+      "loss": 1.0286,
+      "step": 104100
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 5e-05,
+      "loss": 1.037,
+      "step": 104200
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 5e-05,
+      "loss": 1.021,
+      "step": 104300
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 5e-05,
+      "loss": 1.0114,
+      "step": 104400
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 5e-05,
+      "loss": 1.0398,
+      "step": 104500
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 5e-05,
+      "loss": 1.0266,
+      "step": 104600
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 5e-05,
+      "loss": 1.0313,
+      "step": 104700
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 5e-05,
+      "loss": 1.0181,
+      "step": 104800
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 5e-05,
+      "loss": 1.0293,
+      "step": 104900
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 5e-05,
+      "loss": 1.0336,
+      "step": 105000
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 5e-05,
+      "loss": 1.0231,
+      "step": 105100
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 5e-05,
+      "loss": 1.039,
+      "step": 105200
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 5e-05,
+      "loss": 1.0291,
+      "step": 105300
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 5e-05,
+      "loss": 1.0375,
+      "step": 105400
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 5e-05,
+      "loss": 1.0385,
+      "step": 105500
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 5e-05,
+      "loss": 1.0182,
+      "step": 105600
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 5e-05,
+      "loss": 1.0498,
+      "step": 105700
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 5e-05,
+      "loss": 1.0368,
+      "step": 105800
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 5e-05,
+      "loss": 1.0403,
+      "step": 105900
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 5e-05,
+      "loss": 1.0292,
+      "step": 106000
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 5e-05,
+      "loss": 1.0306,
+      "step": 106100
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 5e-05,
+      "loss": 1.038,
+      "step": 106200
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 5e-05,
+      "loss": 1.0445,
+      "step": 106300
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 5e-05,
+      "loss": 1.0541,
+      "step": 106400
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 5e-05,
+      "loss": 1.0281,
+      "step": 106500
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 5e-05,
+      "loss": 1.0506,
+      "step": 106600
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 5e-05,
+      "loss": 1.0351,
+      "step": 106700
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 5e-05,
+      "loss": 1.0396,
+      "step": 106800
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 5e-05,
+      "loss": 1.0616,
+      "step": 106900
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 5e-05,
+      "loss": 1.0317,
+      "step": 107000
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 5e-05,
+      "loss": 1.0462,
+      "step": 107100
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 5e-05,
+      "loss": 1.0388,
+      "step": 107200
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 5e-05,
+      "loss": 1.0529,
+      "step": 107300
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 5e-05,
+      "loss": 1.0231,
+      "step": 107400
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 5e-05,
+      "loss": 1.0487,
+      "step": 107500
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 5e-05,
+      "loss": 1.0489,
+      "step": 107600
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 5e-05,
+      "loss": 1.033,
+      "step": 107700
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 5e-05,
+      "loss": 1.0567,
+      "step": 107800
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 5e-05,
+      "loss": 1.0471,
+      "step": 107900
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 5e-05,
+      "loss": 1.0462,
+      "step": 108000
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 5e-05,
+      "loss": 1.0405,
+      "step": 108100
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 5e-05,
+      "loss": 1.0485,
+      "step": 108200
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 5e-05,
+      "loss": 1.0537,
+      "step": 108300
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 5e-05,
+      "loss": 1.0363,
+      "step": 108400
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 5e-05,
+      "loss": 1.0451,
+      "step": 108500
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 5e-05,
+      "loss": 1.0405,
+      "step": 108600
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 5e-05,
+      "loss": 1.0504,
+      "step": 108700
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 5e-05,
+      "loss": 1.0418,
+      "step": 108800
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 5e-05,
+      "loss": 1.0605,
+      "step": 108900
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 5e-05,
+      "loss": 1.0347,
+      "step": 109000
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 5e-05,
+      "loss": 1.0526,
+      "step": 109100
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 5e-05,
+      "loss": 1.0434,
+      "step": 109200
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 5e-05,
+      "loss": 1.0608,
+      "step": 109300
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 5e-05,
+      "loss": 1.0426,
+      "step": 109400
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 5e-05,
+      "loss": 1.0462,
+      "step": 109500
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 5e-05,
+      "loss": 1.0283,
+      "step": 109600
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 5e-05,
+      "loss": 1.0405,
+      "step": 109700
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 5e-05,
+      "loss": 1.0557,
+      "step": 109800
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 5e-05,
+      "loss": 1.042,
+      "step": 109900
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 5e-05,
+      "loss": 1.0506,
+      "step": 110000
+    },
+    {
+      "epoch": 8.63,
+      "eval_gen_len": 18.76313233865984,
+      "eval_loss": 1.561972737312317,
+      "eval_rouge1": 39.6409,
+      "eval_rouge2": 18.0941,
+      "eval_rougeL": 33.1012,
+      "eval_rougeLsum": 33.0983,
+      "eval_runtime": 2370.4984,
+      "eval_samples_per_second": 4.778,
+      "eval_steps_per_second": 1.195,
+      "step": 110000
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 5e-05,
+      "loss": 1.0483,
+      "step": 110100
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 5e-05,
+      "loss": 1.0626,
+      "step": 110200
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 5e-05,
+      "loss": 1.0567,
+      "step": 110300
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 5e-05,
+      "loss": 1.0455,
+      "step": 110400
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 5e-05,
+      "loss": 1.0634,
+      "step": 110500
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 5e-05,
+      "loss": 1.0537,
+      "step": 110600
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 5e-05,
+      "loss": 1.0522,
+      "step": 110700
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 5e-05,
+      "loss": 1.0724,
+      "step": 110800
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 5e-05,
+      "loss": 1.0534,
+      "step": 110900
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 5e-05,
+      "loss": 1.0407,
+      "step": 111000
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 5e-05,
+      "loss": 1.0354,
+      "step": 111100
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 5e-05,
+      "loss": 1.0423,
+      "step": 111200
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 5e-05,
+      "loss": 1.0671,
+      "step": 111300
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 5e-05,
+      "loss": 1.0798,
+      "step": 111400
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 5e-05,
+      "loss": 1.0678,
+      "step": 111500
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 5e-05,
+      "loss": 1.046,
+      "step": 111600
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 5e-05,
+      "loss": 1.0487,
+      "step": 111700
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 5e-05,
+      "loss": 1.0628,
+      "step": 111800
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 5e-05,
+      "loss": 1.0606,
+      "step": 111900
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 5e-05,
+      "loss": 1.0695,
+      "step": 112000
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 5e-05,
+      "loss": 1.0646,
+      "step": 112100
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 5e-05,
+      "loss": 1.0613,
+      "step": 112200
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 5e-05,
+      "loss": 1.0652,
+      "step": 112300
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 5e-05,
+      "loss": 1.059,
+      "step": 112400
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 5e-05,
+      "loss": 1.0661,
+      "step": 112500
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 5e-05,
+      "loss": 1.0529,
+      "step": 112600
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 5e-05,
+      "loss": 1.0732,
+      "step": 112700
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 5e-05,
+      "loss": 1.0477,
+      "step": 112800
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 5e-05,
+      "loss": 1.0621,
+      "step": 112900
+    },
+    {
+      "epoch": 8.86,
+      "learning_rate": 5e-05,
+      "loss": 1.0757,
+      "step": 113000
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 5e-05,
+      "loss": 1.0802,
+      "step": 113100
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 5e-05,
+      "loss": 1.0634,
+      "step": 113200
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 5e-05,
+      "loss": 1.0512,
+      "step": 113300
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 5e-05,
+      "loss": 1.0457,
+      "step": 113400
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 5e-05,
+      "loss": 1.0774,
+      "step": 113500
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 5e-05,
+      "loss": 1.084,
+      "step": 113600
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 5e-05,
+      "loss": 1.0792,
+      "step": 113700
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 5e-05,
+      "loss": 1.0755,
+      "step": 113800
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 5e-05,
+      "loss": 1.0546,
+      "step": 113900
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 5e-05,
+      "loss": 1.0596,
+      "step": 114000
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 5e-05,
+      "loss": 1.0675,
+      "step": 114100
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 5e-05,
+      "loss": 1.0627,
+      "step": 114200
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 5e-05,
+      "loss": 1.0663,
+      "step": 114300
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 5e-05,
+      "loss": 1.0533,
+      "step": 114400
+    },
+    {
+      "epoch": 8.98,
+      "learning_rate": 5e-05,
+      "loss": 1.0769,
+      "step": 114500
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 5e-05,
+      "loss": 1.0566,
+      "step": 114600
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 5e-05,
+      "loss": 1.0622,
+      "step": 114700
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 5e-05,
+      "loss": 1.027,
+      "step": 114800
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 5e-05,
+      "loss": 0.9537,
+      "step": 114900
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 5e-05,
+      "loss": 0.9737,
+      "step": 115000
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 5e-05,
+      "loss": 0.9674,
+      "step": 115100
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 5e-05,
+      "loss": 0.9575,
+      "step": 115200
+    },
+    {
+      "epoch": 9.04,
+      "learning_rate": 5e-05,
+      "loss": 0.973,
+      "step": 115300
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 5e-05,
+      "loss": 0.9645,
+      "step": 115400
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 5e-05,
+      "loss": 0.9557,
+      "step": 115500
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 5e-05,
+      "loss": 0.9579,
+      "step": 115600
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 5e-05,
+      "loss": 0.9655,
+      "step": 115700
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 5e-05,
+      "loss": 0.9783,
+      "step": 115800
+    },
+    {
+      "epoch": 9.09,
+      "learning_rate": 5e-05,
+      "loss": 0.9927,
+      "step": 115900
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 5e-05,
+      "loss": 0.9748,
+      "step": 116000
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 5e-05,
+      "loss": 0.9799,
+      "step": 116100
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 5e-05,
+      "loss": 0.9669,
+      "step": 116200
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 5e-05,
+      "loss": 1.0048,
+      "step": 116300
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 5e-05,
+      "loss": 0.9806,
+      "step": 116400
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 5e-05,
+      "loss": 0.9617,
+      "step": 116500
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 5e-05,
+      "loss": 0.9767,
+      "step": 116600
+    },
+    {
+      "epoch": 9.15,
+      "learning_rate": 5e-05,
+      "loss": 0.9616,
+      "step": 116700
+    },
+    {
+      "epoch": 9.16,
+      "learning_rate": 5e-05,
+      "loss": 0.9755,
+      "step": 116800
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 5e-05,
+      "loss": 0.9715,
+      "step": 116900
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 5e-05,
+      "loss": 0.9821,
+      "step": 117000
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 5e-05,
+      "loss": 0.9752,
+      "step": 117100
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 5e-05,
+      "loss": 0.982,
+      "step": 117200
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 5e-05,
+      "loss": 0.969,
+      "step": 117300
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 5e-05,
+      "loss": 0.9874,
+      "step": 117400
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 5e-05,
+      "loss": 0.9779,
+      "step": 117500
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 5e-05,
+      "loss": 0.9752,
+      "step": 117600
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 5e-05,
+      "loss": 0.9965,
+      "step": 117700
+    },
+    {
+      "epoch": 9.24,
+      "learning_rate": 5e-05,
+      "loss": 0.9936,
+      "step": 117800
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 5e-05,
+      "loss": 0.9964,
+      "step": 117900
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 5e-05,
+      "loss": 0.9793,
+      "step": 118000
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 5e-05,
+      "loss": 0.9735,
+      "step": 118100
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 5e-05,
+      "loss": 0.9839,
+      "step": 118200
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 5e-05,
+      "loss": 0.9986,
+      "step": 118300
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 5e-05,
+      "loss": 0.9789,
+      "step": 118400
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 5e-05,
+      "loss": 0.9874,
+      "step": 118500
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 5e-05,
+      "loss": 0.9839,
+      "step": 118600
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 5e-05,
+      "loss": 0.9716,
+      "step": 118700
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 5e-05,
+      "loss": 0.9764,
+      "step": 118800
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 5e-05,
+      "loss": 0.974,
+      "step": 118900
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 5e-05,
+      "loss": 0.9817,
+      "step": 119000
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 5e-05,
+      "loss": 0.9874,
+      "step": 119100
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 5e-05,
+      "loss": 0.9955,
+      "step": 119200
+    },
+    {
+      "epoch": 9.36,
+      "learning_rate": 5e-05,
+      "loss": 0.9879,
+      "step": 119300
+    },
+    {
+      "epoch": 9.36,
+      "learning_rate": 5e-05,
+      "loss": 1.0022,
+      "step": 119400
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 5e-05,
+      "loss": 0.9908,
+      "step": 119500
+    },
+    {
+      "epoch": 9.38,
+      "learning_rate": 5e-05,
+      "loss": 0.9849,
+      "step": 119600
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 5e-05,
+      "loss": 0.9749,
+      "step": 119700
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 5e-05,
+      "loss": 0.9932,
+      "step": 119800
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 5e-05,
+      "loss": 0.9971,
+      "step": 119900
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 5e-05,
+      "loss": 0.9846,
+      "step": 120000
+    },
+    {
+      "epoch": 9.41,
+      "eval_gen_len": 18.769135693475764,
+      "eval_loss": 1.6027214527130127,
+      "eval_rouge1": 39.3134,
+      "eval_rouge2": 17.9425,
+      "eval_rougeL": 32.8563,
+      "eval_rougeLsum": 32.8559,
+      "eval_runtime": 2347.8962,
+      "eval_samples_per_second": 4.824,
+      "eval_steps_per_second": 1.206,
+      "step": 120000
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 5e-05,
+      "loss": 0.9855,
+      "step": 120100
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 5e-05,
+      "loss": 0.9865,
+      "step": 120200
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 5e-05,
+      "loss": 1.0009,
+      "step": 120300
+    },
+    {
+      "epoch": 9.44,
+      "learning_rate": 5e-05,
+      "loss": 0.9908,
+      "step": 120400
+    },
+    {
+      "epoch": 9.45,
+      "learning_rate": 5e-05,
+      "loss": 1.0213,
+      "step": 120500
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 5e-05,
+      "loss": 1.0078,
+      "step": 120600
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 5e-05,
+      "loss": 1.0009,
+      "step": 120700
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 5e-05,
+      "loss": 0.987,
+      "step": 120800
+    },
+    {
+      "epoch": 9.48,
+      "learning_rate": 5e-05,
+      "loss": 0.9886,
+      "step": 120900
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 5e-05,
+      "loss": 1.0088,
+      "step": 121000
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 5e-05,
+      "loss": 0.9998,
+      "step": 121100
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 5e-05,
+      "loss": 1.0107,
+      "step": 121200
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 5e-05,
+      "loss": 0.9858,
+      "step": 121300
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 5e-05,
+      "loss": 0.9944,
+      "step": 121400
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 5e-05,
+      "loss": 0.9938,
+      "step": 121500
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 5e-05,
+      "loss": 0.9932,
+      "step": 121600
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 5e-05,
+      "loss": 1.0013,
+      "step": 121700
+    },
+    {
+      "epoch": 9.55,
+      "learning_rate": 5e-05,
+      "loss": 1.0077,
+      "step": 121800
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 5e-05,
+      "loss": 1.0019,
+      "step": 121900
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 5e-05,
+      "loss": 1.0033,
+      "step": 122000
+    },
+    {
+      "epoch": 9.58,
+      "learning_rate": 5e-05,
+      "loss": 1.0015,
+      "step": 122100
+    },
+    {
+      "epoch": 9.58,
+      "learning_rate": 5e-05,
+      "loss": 1.0139,
+      "step": 122200
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 5e-05,
+      "loss": 1.0096,
+      "step": 122300
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 5e-05,
+      "loss": 0.9902,
+      "step": 122400
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 5e-05,
+      "loss": 1.0116,
+      "step": 122500
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 5e-05,
+      "loss": 1.0116,
+      "step": 122600
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 5e-05,
+      "loss": 1.0046,
+      "step": 122700
+    },
+    {
+      "epoch": 9.63,
+      "learning_rate": 5e-05,
+      "loss": 1.014,
+      "step": 122800
+    },
+    {
+      "epoch": 9.64,
+      "learning_rate": 5e-05,
+      "loss": 1.0187,
+      "step": 122900
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 5e-05,
+      "loss": 1.006,
+      "step": 123000
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 5e-05,
+      "loss": 1.0128,
+      "step": 123100
+    },
+    {
+      "epoch": 9.66,
+      "learning_rate": 5e-05,
+      "loss": 0.9984,
+      "step": 123200
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 5e-05,
+      "loss": 0.995,
+      "step": 123300
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 5e-05,
+      "loss": 1.0211,
+      "step": 123400
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 5e-05,
+      "loss": 1.028,
+      "step": 123500
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 5e-05,
+      "loss": 1.0029,
+      "step": 123600
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 5e-05,
+      "loss": 1.0003,
+      "step": 123700
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 5e-05,
+      "loss": 0.991,
+      "step": 123800
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 5e-05,
+      "loss": 1.0048,
+      "step": 123900
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 5e-05,
+      "loss": 1.0188,
+      "step": 124000
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 5e-05,
+      "loss": 1.0033,
+      "step": 124100
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 5e-05,
+      "loss": 1.0226,
+      "step": 124200
+    },
+    {
+      "epoch": 9.75,
+      "learning_rate": 5e-05,
+      "loss": 1.0104,
+      "step": 124300
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 5e-05,
+      "loss": 0.9867,
+      "step": 124400
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 5e-05,
+      "loss": 1.0037,
+      "step": 124500
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 5e-05,
+      "loss": 1.0097,
+      "step": 124600
+    },
+    {
+      "epoch": 9.78,
+      "learning_rate": 5e-05,
+      "loss": 1.0213,
+      "step": 124700
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 5e-05,
+      "loss": 1.0441,
+      "step": 124800
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 5e-05,
+      "loss": 0.9998,
+      "step": 124900
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 5e-05,
+      "loss": 1.0209,
+      "step": 125000
+    },
+    {
+      "epoch": 9.81,
+      "learning_rate": 5e-05,
+      "loss": 1.012,
+      "step": 125100
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 5e-05,
+      "loss": 1.0061,
+      "step": 125200
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 5e-05,
+      "loss": 1.0079,
+      "step": 125300
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 5e-05,
+      "loss": 1.0104,
+      "step": 125400
+    },
+    {
+      "epoch": 9.84,
+      "learning_rate": 5e-05,
+      "loss": 1.0232,
+      "step": 125500
+    },
+    {
+      "epoch": 9.85,
+      "learning_rate": 5e-05,
+      "loss": 0.9992,
+      "step": 125600
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 5e-05,
+      "loss": 0.9957,
+      "step": 125700
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 5e-05,
+      "loss": 1.0075,
+      "step": 125800
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 5e-05,
+      "loss": 1.0014,
+      "step": 125900
+    },
+    {
+      "epoch": 9.88,
+      "learning_rate": 5e-05,
+      "loss": 1.0205,
+      "step": 126000
+    },
+    {
+      "epoch": 9.89,
+      "learning_rate": 5e-05,
+      "loss": 1.0097,
+      "step": 126100
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 5e-05,
+      "loss": 1.0078,
+      "step": 126200
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 5e-05,
+      "loss": 1.0099,
+      "step": 126300
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 5e-05,
+      "loss": 1.0139,
+      "step": 126400
+    },
+    {
+      "epoch": 9.92,
+      "learning_rate": 5e-05,
+      "loss": 1.0171,
+      "step": 126500
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 5e-05,
+      "loss": 1.0505,
+      "step": 126600
+    },
+    {
+      "epoch": 9.94,
+      "learning_rate": 5e-05,
+      "loss": 1.0137,
+      "step": 126700
+    },
+    {
+      "epoch": 9.94,
+      "learning_rate": 5e-05,
+      "loss": 1.0241,
+      "step": 126800
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 5e-05,
+      "loss": 1.0221,
+      "step": 126900
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 5e-05,
+      "loss": 1.0107,
+      "step": 127000
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 5e-05,
+      "loss": 1.022,
+      "step": 127100
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 5e-05,
+      "loss": 1.0076,
+      "step": 127200
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 5e-05,
+      "loss": 1.0194,
+      "step": 127300
+    },
+    {
+      "epoch": 9.99,
+      "learning_rate": 5e-05,
+      "loss": 1.011,
+      "step": 127400
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 5e-05,
+      "loss": 1.0128,
+      "step": 127500
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 5e-05,
+      "loss": 0.9376,
+      "step": 127600
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 5e-05,
+      "loss": 0.9231,
+      "step": 127700
+    },
+    {
+      "epoch": 10.02,
+      "learning_rate": 5e-05,
+      "loss": 0.9108,
+      "step": 127800
+    },
+    {
+      "epoch": 10.03,
+      "learning_rate": 5e-05,
+      "loss": 0.9122,
+      "step": 127900
+    },
+    {
+      "epoch": 10.04,
+      "learning_rate": 5e-05,
+      "loss": 0.918,
+      "step": 128000
+    },
+    {
+      "epoch": 10.05,
+      "learning_rate": 5e-05,
+      "loss": 0.9231,
+      "step": 128100
+    },
+    {
+      "epoch": 10.05,
+      "learning_rate": 5e-05,
+      "loss": 0.9267,
+      "step": 128200
+    },
+    {
+      "epoch": 10.06,
+      "learning_rate": 5e-05,
+      "loss": 0.9139,
+      "step": 128300
+    },
+    {
+      "epoch": 10.07,
+      "learning_rate": 5e-05,
+      "loss": 0.9155,
+      "step": 128400
+    },
+    {
+      "epoch": 10.08,
+      "learning_rate": 5e-05,
+      "loss": 0.94,
+      "step": 128500
+    },
+    {
+      "epoch": 10.09,
+      "learning_rate": 5e-05,
+      "loss": 0.938,
+      "step": 128600
+    },
+    {
+      "epoch": 10.09,
+      "learning_rate": 5e-05,
+      "loss": 0.9136,
+      "step": 128700
+    },
+    {
+      "epoch": 10.1,
+      "learning_rate": 5e-05,
+      "loss": 0.9205,
+      "step": 128800
+    },
+    {
+      "epoch": 10.11,
+      "learning_rate": 5e-05,
+      "loss": 0.9269,
+      "step": 128900
+    },
+    {
+      "epoch": 10.12,
+      "learning_rate": 5e-05,
+      "loss": 0.9215,
+      "step": 129000
+    },
+    {
+      "epoch": 10.12,
+      "learning_rate": 5e-05,
+      "loss": 0.9153,
+      "step": 129100
+    },
+    {
+      "epoch": 10.13,
+      "learning_rate": 5e-05,
+      "loss": 0.9522,
+      "step": 129200
+    },
+    {
+      "epoch": 10.14,
+      "learning_rate": 5e-05,
+      "loss": 0.9342,
+      "step": 129300
+    },
+    {
+      "epoch": 10.15,
+      "learning_rate": 5e-05,
+      "loss": 0.9349,
+      "step": 129400
+    },
+    {
+      "epoch": 10.16,
+      "learning_rate": 5e-05,
+      "loss": 0.9316,
+      "step": 129500
+    },
+    {
+      "epoch": 10.16,
+      "learning_rate": 5e-05,
+      "loss": 0.9214,
+      "step": 129600
+    },
+    {
+      "epoch": 10.17,
+      "learning_rate": 5e-05,
+      "loss": 0.931,
+      "step": 129700
+    },
+    {
+      "epoch": 10.18,
+      "learning_rate": 5e-05,
+      "loss": 0.9418,
+      "step": 129800
+    },
+    {
+      "epoch": 10.19,
+      "learning_rate": 5e-05,
+      "loss": 0.9534,
+      "step": 129900
+    },
+    {
+      "epoch": 10.2,
+      "learning_rate": 5e-05,
+      "loss": 0.9217,
+      "step": 130000
+    },
+    {
+      "epoch": 10.2,
+      "eval_gen_len": 18.793502251258055,
+      "eval_loss": 1.6331114768981934,
+      "eval_rouge1": 39.2465,
+      "eval_rouge2": 17.7996,
+      "eval_rougeL": 32.7625,
+      "eval_rougeLsum": 32.7679,
+      "eval_runtime": 2377.4093,
+      "eval_samples_per_second": 4.764,
+      "eval_steps_per_second": 1.191,
+      "step": 130000
+    },
+    {
+      "epoch": 10.2,
+      "learning_rate": 5e-05,
+      "loss": 0.9296,
+      "step": 130100
+    },
+    {
+      "epoch": 10.21,
+      "learning_rate": 5e-05,
+      "loss": 0.9386,
+      "step": 130200
+    },
+    {
+      "epoch": 10.22,
+      "learning_rate": 5e-05,
+      "loss": 0.9402,
+      "step": 130300
+    },
+    {
+      "epoch": 10.23,
+      "learning_rate": 5e-05,
+      "loss": 0.9317,
+      "step": 130400
+    },
+    {
+      "epoch": 10.23,
+      "learning_rate": 5e-05,
+      "loss": 0.9484,
+      "step": 130500
+    },
+    {
+      "epoch": 10.24,
+      "learning_rate": 5e-05,
+      "loss": 0.9338,
+      "step": 130600
+    },
+    {
+      "epoch": 10.25,
+      "learning_rate": 5e-05,
+      "loss": 0.9273,
+      "step": 130700
+    },
+    {
+      "epoch": 10.26,
+      "learning_rate": 5e-05,
+      "loss": 0.9402,
+      "step": 130800
+    },
+    {
+      "epoch": 10.27,
+      "learning_rate": 5e-05,
+      "loss": 0.9333,
+      "step": 130900
+    },
+    {
+      "epoch": 10.27,
+      "learning_rate": 5e-05,
+      "loss": 0.9379,
+      "step": 131000
+    },
+    {
+      "epoch": 10.28,
+      "learning_rate": 5e-05,
+      "loss": 0.9531,
+      "step": 131100
+    },
+    {
+      "epoch": 10.29,
+      "learning_rate": 5e-05,
+      "loss": 0.933,
+      "step": 131200
+    },
+    {
+      "epoch": 10.3,
+      "learning_rate": 5e-05,
+      "loss": 0.9423,
+      "step": 131300
+    },
+    {
+      "epoch": 10.31,
+      "learning_rate": 5e-05,
+      "loss": 0.9395,
+      "step": 131400
+    },
+    {
+      "epoch": 10.31,
+      "learning_rate": 5e-05,
+      "loss": 0.9337,
+      "step": 131500
+    },
+    {
+      "epoch": 10.32,
+      "learning_rate": 5e-05,
+      "loss": 0.9532,
+      "step": 131600
+    },
+    {
+      "epoch": 10.33,
+      "learning_rate": 5e-05,
+      "loss": 0.9587,
+      "step": 131700
+    },
+    {
+      "epoch": 10.34,
+      "learning_rate": 5e-05,
+      "loss": 0.9321,
+      "step": 131800
+    },
+    {
+      "epoch": 10.34,
+      "learning_rate": 5e-05,
+      "loss": 0.9541,
+      "step": 131900
+    },
+    {
+      "epoch": 10.35,
+      "learning_rate": 5e-05,
+      "loss": 0.951,
+      "step": 132000
+    },
+    {
+      "epoch": 10.36,
+      "learning_rate": 5e-05,
+      "loss": 0.9324,
+      "step": 132100
+    },
+    {
+      "epoch": 10.37,
+      "learning_rate": 5e-05,
+      "loss": 0.9521,
+      "step": 132200
+    },
+    {
+      "epoch": 10.38,
+      "learning_rate": 5e-05,
+      "loss": 0.9568,
+      "step": 132300
+    },
+    {
+      "epoch": 10.38,
+      "learning_rate": 5e-05,
+      "loss": 0.9492,
+      "step": 132400
+    },
+    {
+      "epoch": 10.39,
+      "learning_rate": 5e-05,
+      "loss": 0.9577,
+      "step": 132500
+    },
+    {
+      "epoch": 10.4,
+      "learning_rate": 5e-05,
+      "loss": 0.9533,
+      "step": 132600
+    },
+    {
+      "epoch": 10.41,
+      "learning_rate": 5e-05,
+      "loss": 0.9551,
+      "step": 132700
+    },
+    {
+      "epoch": 10.41,
+      "learning_rate": 5e-05,
+      "loss": 0.9283,
+      "step": 132800
+    },
+    {
+      "epoch": 10.42,
+      "learning_rate": 5e-05,
+      "loss": 0.9644,
+      "step": 132900
+    },
+    {
+      "epoch": 10.43,
+      "learning_rate": 5e-05,
+      "loss": 0.967,
+      "step": 133000
+    },
+    {
+      "epoch": 10.44,
+      "learning_rate": 5e-05,
+      "loss": 0.947,
+      "step": 133100
+    },
+    {
+      "epoch": 10.45,
+      "learning_rate": 5e-05,
+      "loss": 0.9688,
+      "step": 133200
+    },
+    {
+      "epoch": 10.45,
+      "learning_rate": 5e-05,
+      "loss": 0.9266,
+      "step": 133300
+    },
+    {
+      "epoch": 10.46,
+      "learning_rate": 5e-05,
+      "loss": 0.9544,
+      "step": 133400
+    },
+    {
+      "epoch": 10.47,
+      "learning_rate": 5e-05,
+      "loss": 0.9429,
+      "step": 133500
+    },
+    {
+      "epoch": 10.48,
+      "learning_rate": 5e-05,
+      "loss": 0.9553,
+      "step": 133600
+    },
+    {
+      "epoch": 10.49,
+      "learning_rate": 5e-05,
+      "loss": 0.9385,
+      "step": 133700
+    },
+    {
+      "epoch": 10.49,
+      "learning_rate": 5e-05,
+      "loss": 0.9545,
+      "step": 133800
+    },
+    {
+      "epoch": 10.5,
+      "learning_rate": 5e-05,
+      "loss": 0.9577,
+      "step": 133900
+    },
+    {
+      "epoch": 10.51,
+      "learning_rate": 5e-05,
+      "loss": 0.9532,
+      "step": 134000
+    },
+    {
+      "epoch": 10.52,
+      "learning_rate": 5e-05,
+      "loss": 0.9503,
+      "step": 134100
+    },
+    {
+      "epoch": 10.52,
+      "learning_rate": 5e-05,
+      "loss": 0.9504,
+      "step": 134200
+    },
+    {
+      "epoch": 10.53,
+      "learning_rate": 5e-05,
+      "loss": 0.9556,
+      "step": 134300
+    },
+    {
+      "epoch": 10.54,
+      "learning_rate": 5e-05,
+      "loss": 0.963,
+      "step": 134400
+    },
+    {
+      "epoch": 10.55,
+      "learning_rate": 5e-05,
+      "loss": 0.9663,
+      "step": 134500
+    },
+    {
+      "epoch": 10.56,
+      "learning_rate": 5e-05,
+      "loss": 0.9645,
+      "step": 134600
+    },
+    {
+      "epoch": 10.56,
+      "learning_rate": 5e-05,
+      "loss": 0.9531,
+      "step": 134700
+    },
+    {
+      "epoch": 10.57,
+      "learning_rate": 5e-05,
+      "loss": 0.9566,
+      "step": 134800
+    },
+    {
+      "epoch": 10.58,
+      "learning_rate": 5e-05,
+      "loss": 0.9443,
+      "step": 134900
+    },
+    {
+      "epoch": 10.59,
+      "learning_rate": 5e-05,
+      "loss": 0.9575,
+      "step": 135000
+    },
+    {
+      "epoch": 10.6,
+      "learning_rate": 5e-05,
+      "loss": 0.9522,
+      "step": 135100
+    },
+    {
+      "epoch": 10.6,
+      "learning_rate": 5e-05,
+      "loss": 0.9417,
+      "step": 135200
+    },
+    {
+      "epoch": 10.61,
+      "learning_rate": 5e-05,
+      "loss": 0.9595,
+      "step": 135300
+    },
+    {
+      "epoch": 10.62,
+      "learning_rate": 5e-05,
+      "loss": 0.9746,
+      "step": 135400
+    },
+    {
+      "epoch": 10.63,
+      "learning_rate": 5e-05,
+      "loss": 0.9531,
+      "step": 135500
+    },
+    {
+      "epoch": 10.63,
+      "learning_rate": 5e-05,
+      "loss": 0.9666,
+      "step": 135600
+    },
+    {
+      "epoch": 10.64,
+      "learning_rate": 5e-05,
+      "loss": 0.9508,
+      "step": 135700
+    },
+    {
+      "epoch": 10.65,
+      "learning_rate": 5e-05,
+      "loss": 0.9707,
+      "step": 135800
+    },
+    {
+      "epoch": 10.66,
+      "learning_rate": 5e-05,
+      "loss": 0.9508,
+      "step": 135900
+    },
+    {
+      "epoch": 10.67,
+      "learning_rate": 5e-05,
+      "loss": 0.9637,
+      "step": 136000
+    },
+    {
+      "epoch": 10.67,
+      "learning_rate": 5e-05,
+      "loss": 0.9641,
+      "step": 136100
+    },
+    {
+      "epoch": 10.68,
+      "learning_rate": 5e-05,
+      "loss": 0.9605,
+      "step": 136200
+    },
+    {
+      "epoch": 10.69,
+      "learning_rate": 5e-05,
+      "loss": 0.9672,
+      "step": 136300
+    },
+    {
+      "epoch": 10.7,
+      "learning_rate": 5e-05,
+      "loss": 0.9683,
+      "step": 136400
+    },
+    {
+      "epoch": 10.71,
+      "learning_rate": 5e-05,
+      "loss": 0.9686,
+      "step": 136500
+    },
+    {
+      "epoch": 10.71,
+      "learning_rate": 5e-05,
+      "loss": 0.9655,
+      "step": 136600
+    },
+    {
+      "epoch": 10.72,
+      "learning_rate": 5e-05,
+      "loss": 0.9457,
+      "step": 136700
+    },
+    {
+      "epoch": 10.73,
+      "learning_rate": 5e-05,
+      "loss": 0.9543,
+      "step": 136800
+    },
+    {
+      "epoch": 10.74,
+      "learning_rate": 5e-05,
+      "loss": 0.9712,
+      "step": 136900
+    },
+    {
+      "epoch": 10.74,
+      "learning_rate": 5e-05,
+      "loss": 0.9574,
+      "step": 137000
+    },
+    {
+      "epoch": 10.75,
+      "learning_rate": 5e-05,
+      "loss": 0.9726,
+      "step": 137100
+    },
+    {
+      "epoch": 10.76,
+      "learning_rate": 5e-05,
+      "loss": 0.9752,
+      "step": 137200
+    },
+    {
+      "epoch": 10.77,
+      "learning_rate": 5e-05,
+      "loss": 0.9676,
+      "step": 137300
+    },
+    {
+      "epoch": 10.78,
+      "learning_rate": 5e-05,
+      "loss": 0.9694,
+      "step": 137400
+    },
+    {
+      "epoch": 10.78,
+      "learning_rate": 5e-05,
+      "loss": 0.9805,
+      "step": 137500
+    },
+    {
+      "epoch": 10.79,
+      "learning_rate": 5e-05,
+      "loss": 0.9632,
+      "step": 137600
+    },
+    {
+      "epoch": 10.8,
+      "learning_rate": 5e-05,
+      "loss": 0.9553,
+      "step": 137700
+    },
+    {
+      "epoch": 10.81,
+      "learning_rate": 5e-05,
+      "loss": 0.9712,
+      "step": 137800
+    },
+    {
+      "epoch": 10.81,
+      "learning_rate": 5e-05,
+      "loss": 0.9857,
+      "step": 137900
+    },
+    {
+      "epoch": 10.82,
+      "learning_rate": 5e-05,
+      "loss": 0.9676,
+      "step": 138000
+    },
+    {
+      "epoch": 10.83,
+      "learning_rate": 5e-05,
+      "loss": 0.9511,
+      "step": 138100
+    },
+    {
+      "epoch": 10.84,
+      "learning_rate": 5e-05,
+      "loss": 0.9619,
+      "step": 138200
+    },
+    {
+      "epoch": 10.85,
+      "learning_rate": 5e-05,
+      "loss": 0.9716,
+      "step": 138300
+    },
+    {
+      "epoch": 10.85,
+      "learning_rate": 5e-05,
+      "loss": 0.966,
+      "step": 138400
+    },
+    {
+      "epoch": 10.86,
+      "learning_rate": 5e-05,
+      "loss": 0.9637,
+      "step": 138500
+    },
+    {
+      "epoch": 10.87,
+      "learning_rate": 5e-05,
+      "loss": 0.9726,
+      "step": 138600
+    },
+    {
+      "epoch": 10.88,
+      "learning_rate": 5e-05,
+      "loss": 0.9753,
+      "step": 138700
+    },
+    {
+      "epoch": 10.89,
+      "learning_rate": 5e-05,
+      "loss": 0.9845,
+      "step": 138800
+    },
+    {
+      "epoch": 10.89,
+      "learning_rate": 5e-05,
+      "loss": 0.9738,
+      "step": 138900
+    },
+    {
+      "epoch": 10.9,
+      "learning_rate": 5e-05,
+      "loss": 0.9612,
+      "step": 139000
+    },
+    {
+      "epoch": 10.91,
+      "learning_rate": 5e-05,
+      "loss": 0.9791,
+      "step": 139100
+    },
+    {
+      "epoch": 10.92,
+      "learning_rate": 5e-05,
+      "loss": 0.9772,
+      "step": 139200
+    },
+    {
+      "epoch": 10.92,
+      "learning_rate": 5e-05,
+      "loss": 0.9645,
+      "step": 139300
+    },
+    {
+      "epoch": 10.93,
+      "learning_rate": 5e-05,
+      "loss": 0.9761,
+      "step": 139400
+    },
+    {
+      "epoch": 10.94,
+      "learning_rate": 5e-05,
+      "loss": 0.982,
+      "step": 139500
+    },
+    {
+      "epoch": 10.95,
+      "learning_rate": 5e-05,
+      "loss": 0.9827,
+      "step": 139600
+    },
+    {
+      "epoch": 10.96,
+      "learning_rate": 5e-05,
+      "loss": 0.9782,
+      "step": 139700
+    },
+    {
+      "epoch": 10.96,
+      "learning_rate": 5e-05,
+      "loss": 0.9825,
+      "step": 139800
+    },
+    {
+      "epoch": 10.97,
+      "learning_rate": 5e-05,
+      "loss": 0.9786,
+      "step": 139900
+    },
+    {
+      "epoch": 10.98,
+      "learning_rate": 5e-05,
+      "loss": 0.9657,
+      "step": 140000
+    },
+    {
+      "epoch": 10.98,
+      "eval_gen_len": 18.765427739030635,
+      "eval_loss": 1.6096522808074951,
+      "eval_rouge1": 39.4889,
+      "eval_rouge2": 17.9974,
+      "eval_rougeL": 33.0222,
+      "eval_rougeLsum": 33.0106,
+      "eval_runtime": 2438.6189,
+      "eval_samples_per_second": 4.645,
+      "eval_steps_per_second": 1.161,
+      "step": 140000
+    },
+    {
+      "epoch": 10.99,
+      "learning_rate": 5e-05,
+      "loss": 0.9702,
+      "step": 140100
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 5e-05,
+      "loss": 0.9868,
+      "step": 140200
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 5e-05,
+      "loss": 0.9282,
+      "step": 140300
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 5e-05,
+      "loss": 0.8992,
+      "step": 140400
+    },
+    {
+      "epoch": 11.02,
+      "learning_rate": 5e-05,
+      "loss": 0.8785,
+      "step": 140500
+    },
+    {
+      "epoch": 11.03,
+      "learning_rate": 5e-05,
+      "loss": 0.8643,
+      "step": 140600
+    },
+    {
+      "epoch": 11.03,
+      "learning_rate": 5e-05,
+      "loss": 0.8896,
+      "step": 140700
+    },
+    {
+      "epoch": 11.04,
+      "learning_rate": 5e-05,
+      "loss": 0.8829,
+      "step": 140800
+    },
+    {
+      "epoch": 11.05,
+      "learning_rate": 5e-05,
+      "loss": 0.8793,
+      "step": 140900
+    },
+    {
+      "epoch": 11.06,
+      "learning_rate": 5e-05,
+      "loss": 0.8632,
+      "step": 141000
+    },
+    {
+      "epoch": 11.07,
+      "learning_rate": 5e-05,
+      "loss": 0.8656,
+      "step": 141100
+    },
+    {
+      "epoch": 11.07,
+      "learning_rate": 5e-05,
+      "loss": 0.887,
+      "step": 141200
+    },
+    {
+      "epoch": 11.08,
+      "learning_rate": 5e-05,
+      "loss": 0.8705,
+      "step": 141300
+    },
+    {
+      "epoch": 11.09,
+      "learning_rate": 5e-05,
+      "loss": 0.8785,
+      "step": 141400
+    },
+    {
+      "epoch": 11.1,
+      "learning_rate": 5e-05,
+      "loss": 0.8813,
+      "step": 141500
+    },
+    {
+      "epoch": 11.11,
+      "learning_rate": 5e-05,
+      "loss": 0.8884,
+      "step": 141600
+    },
+    {
+      "epoch": 11.11,
+      "learning_rate": 5e-05,
+      "loss": 0.8907,
+      "step": 141700
+    },
+    {
+      "epoch": 11.12,
+      "learning_rate": 5e-05,
+      "loss": 0.8836,
+      "step": 141800
+    },
+    {
+      "epoch": 11.13,
+      "learning_rate": 5e-05,
+      "loss": 0.8849,
+      "step": 141900
+    },
+    {
+      "epoch": 11.14,
+      "learning_rate": 5e-05,
+      "loss": 0.8802,
+      "step": 142000
+    },
+    {
+      "epoch": 11.14,
+      "learning_rate": 5e-05,
+      "loss": 0.8824,
+      "step": 142100
+    },
+    {
+      "epoch": 11.15,
+      "learning_rate": 5e-05,
+      "loss": 0.8882,
+      "step": 142200
+    },
+    {
+      "epoch": 11.16,
+      "learning_rate": 5e-05,
+      "loss": 0.8845,
+      "step": 142300
+    },
+    {
+      "epoch": 11.17,
+      "learning_rate": 5e-05,
+      "loss": 0.8946,
+      "step": 142400
+    },
+    {
+      "epoch": 11.18,
+      "learning_rate": 5e-05,
+      "loss": 0.8921,
+      "step": 142500
+    },
+    {
+      "epoch": 11.18,
+      "learning_rate": 5e-05,
+      "loss": 0.8864,
+      "step": 142600
+    },
+    {
+      "epoch": 11.19,
+      "learning_rate": 5e-05,
+      "loss": 0.8896,
+      "step": 142700
+    },
+    {
+      "epoch": 11.2,
+      "learning_rate": 5e-05,
+      "loss": 0.8809,
+      "step": 142800
+    },
+    {
+      "epoch": 11.21,
+      "learning_rate": 5e-05,
+      "loss": 0.9066,
+      "step": 142900
+    },
+    {
+      "epoch": 11.21,
+      "learning_rate": 5e-05,
+      "loss": 0.9023,
+      "step": 143000
+    },
+    {
+      "epoch": 11.22,
+      "learning_rate": 5e-05,
+      "loss": 0.9192,
+      "step": 143100
+    },
+    {
+      "epoch": 11.23,
+      "learning_rate": 5e-05,
+      "loss": 0.9137,
+      "step": 143200
+    },
+    {
+      "epoch": 11.24,
+      "learning_rate": 5e-05,
+      "loss": 0.8832,
+      "step": 143300
+    },
+    {
+      "epoch": 11.25,
+      "learning_rate": 5e-05,
+      "loss": 0.8887,
+      "step": 143400
+    },
+    {
+      "epoch": 11.25,
+      "learning_rate": 5e-05,
+      "loss": 0.9082,
+      "step": 143500
+    },
+    {
+      "epoch": 11.26,
+      "learning_rate": 5e-05,
+      "loss": 0.8935,
+      "step": 143600
+    },
+    {
+      "epoch": 11.27,
+      "learning_rate": 5e-05,
+      "loss": 0.8923,
+      "step": 143700
+    },
+    {
+      "epoch": 11.28,
+      "learning_rate": 5e-05,
+      "loss": 0.9177,
+      "step": 143800
+    },
+    {
+      "epoch": 11.29,
+      "learning_rate": 5e-05,
+      "loss": 0.8995,
+      "step": 143900
+    },
+    {
+      "epoch": 11.29,
+      "learning_rate": 5e-05,
+      "loss": 0.8996,
+      "step": 144000
+    },
+    {
+      "epoch": 11.3,
+      "learning_rate": 5e-05,
+      "loss": 0.9046,
+      "step": 144100
+    },
+    {
+      "epoch": 11.31,
+      "learning_rate": 5e-05,
+      "loss": 0.9144,
+      "step": 144200
+    },
+    {
+      "epoch": 11.32,
+      "learning_rate": 5e-05,
+      "loss": 0.8997,
+      "step": 144300
+    },
+    {
+      "epoch": 11.32,
+      "learning_rate": 5e-05,
+      "loss": 0.8818,
+      "step": 144400
+    },
+    {
+      "epoch": 11.33,
+      "learning_rate": 5e-05,
+      "loss": 0.8952,
+      "step": 144500
+    },
+    {
+      "epoch": 11.34,
+      "learning_rate": 5e-05,
+      "loss": 0.9112,
+      "step": 144600
+    },
+    {
+      "epoch": 11.35,
+      "learning_rate": 5e-05,
+      "loss": 0.8976,
+      "step": 144700
+    },
+    {
+      "epoch": 11.36,
+      "learning_rate": 5e-05,
+      "loss": 0.8983,
+      "step": 144800
+    },
+    {
+      "epoch": 11.36,
+      "learning_rate": 5e-05,
+      "loss": 0.9207,
+      "step": 144900
+    },
+    {
+      "epoch": 11.37,
+      "learning_rate": 5e-05,
+      "loss": 0.9011,
+      "step": 145000
+    },
+    {
+      "epoch": 11.38,
+      "learning_rate": 5e-05,
+      "loss": 0.9068,
+      "step": 145100
+    },
+    {
+      "epoch": 11.39,
+      "learning_rate": 5e-05,
+      "loss": 0.8939,
+      "step": 145200
+    },
+    {
+      "epoch": 11.4,
+      "learning_rate": 5e-05,
+      "loss": 0.9102,
+      "step": 145300
+    },
+    {
+      "epoch": 11.4,
+      "learning_rate": 5e-05,
+      "loss": 0.914,
+      "step": 145400
+    },
+    {
+      "epoch": 11.41,
+      "learning_rate": 5e-05,
+      "loss": 0.8868,
+      "step": 145500
+    },
+    {
+      "epoch": 11.42,
+      "learning_rate": 5e-05,
+      "loss": 0.9171,
+      "step": 145600
+    },
+    {
+      "epoch": 11.43,
+      "learning_rate": 5e-05,
+      "loss": 0.9036,
+      "step": 145700
+    },
+    {
+      "epoch": 11.43,
+      "learning_rate": 5e-05,
+      "loss": 0.9035,
+      "step": 145800
+    },
+    {
+      "epoch": 11.44,
+      "learning_rate": 5e-05,
+      "loss": 0.916,
+      "step": 145900
+    },
+    {
+      "epoch": 11.45,
+      "learning_rate": 5e-05,
+      "loss": 0.9135,
+      "step": 146000
+    },
+    {
+      "epoch": 11.46,
+      "learning_rate": 5e-05,
+      "loss": 0.9106,
+      "step": 146100
+    },
+    {
+      "epoch": 11.47,
+      "learning_rate": 5e-05,
+      "loss": 0.9052,
+      "step": 146200
+    },
+    {
+      "epoch": 11.47,
+      "learning_rate": 5e-05,
+      "loss": 0.9017,
+      "step": 146300
+    },
+    {
+      "epoch": 11.48,
+      "learning_rate": 5e-05,
+      "loss": 0.8923,
+      "step": 146400
+    },
+    {
+      "epoch": 11.49,
+      "learning_rate": 5e-05,
+      "loss": 0.9145,
+      "step": 146500
+    },
+    {
+      "epoch": 11.5,
+      "learning_rate": 5e-05,
+      "loss": 0.917,
+      "step": 146600
+    },
+    {
+      "epoch": 11.5,
+      "learning_rate": 5e-05,
+      "loss": 0.9145,
+      "step": 146700
+    },
+    {
+      "epoch": 11.51,
+      "learning_rate": 5e-05,
+      "loss": 0.9204,
+      "step": 146800
+    },
+    {
+      "epoch": 11.52,
+      "learning_rate": 5e-05,
+      "loss": 0.9049,
+      "step": 146900
+    },
+    {
+      "epoch": 11.53,
+      "learning_rate": 5e-05,
+      "loss": 0.9188,
+      "step": 147000
+    },
+    {
+      "epoch": 11.54,
+      "learning_rate": 5e-05,
+      "loss": 0.9211,
+      "step": 147100
+    },
+    {
+      "epoch": 11.54,
+      "learning_rate": 5e-05,
+      "loss": 0.8907,
+      "step": 147200
+    },
+    {
+      "epoch": 11.55,
+      "learning_rate": 5e-05,
+      "loss": 0.9142,
+      "step": 147300
+    },
+    {
+      "epoch": 11.56,
+      "learning_rate": 5e-05,
+      "loss": 0.9182,
+      "step": 147400
+    },
+    {
+      "epoch": 11.57,
+      "learning_rate": 5e-05,
+      "loss": 0.9251,
+      "step": 147500
+    },
+    {
+      "epoch": 11.58,
+      "learning_rate": 5e-05,
+      "loss": 0.9156,
+      "step": 147600
+    },
+    {
+      "epoch": 11.58,
+      "learning_rate": 5e-05,
+      "loss": 0.9317,
+      "step": 147700
+    },
+    {
+      "epoch": 11.59,
+      "learning_rate": 5e-05,
+      "loss": 0.9117,
+      "step": 147800
+    },
+    {
+      "epoch": 11.6,
+      "learning_rate": 5e-05,
+      "loss": 0.9183,
+      "step": 147900
+    },
+    {
+      "epoch": 11.61,
+      "learning_rate": 5e-05,
+      "loss": 0.9069,
+      "step": 148000
+    },
+    {
+      "epoch": 11.61,
+      "learning_rate": 5e-05,
+      "loss": 0.9165,
+      "step": 148100
+    },
+    {
+      "epoch": 11.62,
+      "learning_rate": 5e-05,
+      "loss": 0.9078,
+      "step": 148200
+    },
+    {
+      "epoch": 11.63,
+      "learning_rate": 5e-05,
+      "loss": 0.9342,
+      "step": 148300
+    },
+    {
+      "epoch": 11.64,
+      "learning_rate": 5e-05,
+      "loss": 0.9206,
+      "step": 148400
+    },
+    {
+      "epoch": 11.65,
+      "learning_rate": 5e-05,
+      "loss": 0.9127,
+      "step": 148500
+    },
+    {
+      "epoch": 11.65,
+      "learning_rate": 5e-05,
+      "loss": 0.921,
+      "step": 148600
+    },
+    {
+      "epoch": 11.66,
+      "learning_rate": 5e-05,
+      "loss": 0.9207,
+      "step": 148700
+    },
+    {
+      "epoch": 11.67,
+      "learning_rate": 5e-05,
+      "loss": 0.9156,
+      "step": 148800
+    },
+    {
+      "epoch": 11.68,
+      "learning_rate": 5e-05,
+      "loss": 0.9028,
+      "step": 148900
+    },
+    {
+      "epoch": 11.69,
+      "learning_rate": 5e-05,
+      "loss": 0.9092,
+      "step": 149000
+    },
+    {
+      "epoch": 11.69,
+      "learning_rate": 5e-05,
+      "loss": 0.9215,
+      "step": 149100
+    },
+    {
+      "epoch": 11.7,
+      "learning_rate": 5e-05,
+      "loss": 0.9208,
+      "step": 149200
+    },
+    {
+      "epoch": 11.71,
+      "learning_rate": 5e-05,
+      "loss": 0.9308,
+      "step": 149300
+    },
+    {
+      "epoch": 11.72,
+      "learning_rate": 5e-05,
+      "loss": 0.9075,
+      "step": 149400
+    },
+    {
+      "epoch": 11.72,
+      "learning_rate": 5e-05,
+      "loss": 0.9103,
+      "step": 149500
+    },
+    {
+      "epoch": 11.73,
+      "learning_rate": 5e-05,
+      "loss": 0.9306,
+      "step": 149600
+    },
+    {
+      "epoch": 11.74,
+      "learning_rate": 5e-05,
+      "loss": 0.924,
+      "step": 149700
+    },
+    {
+      "epoch": 11.75,
+      "learning_rate": 5e-05,
+      "loss": 0.9214,
+      "step": 149800
+    },
+    {
+      "epoch": 11.76,
+      "learning_rate": 5e-05,
+      "loss": 0.9252,
+      "step": 149900
+    },
+    {
+      "epoch": 11.76,
+      "learning_rate": 5e-05,
+      "loss": 0.9206,
+      "step": 150000
+    },
+    {
+      "epoch": 11.76,
+      "eval_gen_len": 18.793767105146994,
+      "eval_loss": 1.6424586772918701,
+      "eval_rouge1": 39.3295,
+      "eval_rouge2": 17.9689,
+      "eval_rougeL": 32.8956,
+      "eval_rougeLsum": 32.889,
+      "eval_runtime": 2509.3575,
+      "eval_samples_per_second": 4.514,
+      "eval_steps_per_second": 1.129,
+      "step": 150000
+    },
+    {
+      "epoch": 11.77,
+      "learning_rate": 5e-05,
+      "loss": 0.9152,
+      "step": 150100
+    },
+    {
+      "epoch": 11.78,
+      "learning_rate": 5e-05,
+      "loss": 0.9187,
+      "step": 150200
+    },
+    {
+      "epoch": 11.79,
+      "learning_rate": 5e-05,
+      "loss": 0.9092,
+      "step": 150300
+    },
+    {
+      "epoch": 11.8,
+      "learning_rate": 5e-05,
+      "loss": 0.9096,
+      "step": 150400
+    },
+    {
+      "epoch": 11.8,
+      "learning_rate": 5e-05,
+      "loss": 0.9297,
+      "step": 150500
+    },
+    {
+      "epoch": 11.81,
+      "learning_rate": 5e-05,
+      "loss": 0.9272,
+      "step": 150600
+    },
+    {
+      "epoch": 11.82,
+      "learning_rate": 5e-05,
+      "loss": 0.9477,
+      "step": 150700
+    },
+    {
+      "epoch": 11.83,
+      "learning_rate": 5e-05,
+      "loss": 0.9237,
+      "step": 150800
+    },
+    {
+      "epoch": 11.83,
+      "learning_rate": 5e-05,
+      "loss": 0.9298,
+      "step": 150900
+    },
+    {
+      "epoch": 11.84,
+      "learning_rate": 5e-05,
+      "loss": 0.9283,
+      "step": 151000
+    },
+    {
+      "epoch": 11.85,
+      "learning_rate": 5e-05,
+      "loss": 0.9449,
+      "step": 151100
+    },
+    {
+      "epoch": 11.86,
+      "learning_rate": 5e-05,
+      "loss": 0.9287,
+      "step": 151200
+    },
+    {
+      "epoch": 11.87,
+      "learning_rate": 5e-05,
+      "loss": 0.9181,
+      "step": 151300
+    },
+    {
+      "epoch": 11.87,
+      "learning_rate": 5e-05,
+      "loss": 0.9318,
+      "step": 151400
+    },
+    {
+      "epoch": 11.88,
+      "learning_rate": 5e-05,
+      "loss": 0.9127,
+      "step": 151500
+    },
+    {
+      "epoch": 11.89,
+      "learning_rate": 5e-05,
+      "loss": 0.9318,
+      "step": 151600
+    },
+    {
+      "epoch": 11.9,
+      "learning_rate": 5e-05,
+      "loss": 0.9263,
+      "step": 151700
+    },
+    {
+      "epoch": 11.9,
+      "learning_rate": 5e-05,
+      "loss": 0.9319,
+      "step": 151800
+    },
+    {
+      "epoch": 11.91,
+      "learning_rate": 5e-05,
+      "loss": 0.9412,
+      "step": 151900
+    },
+    {
+      "epoch": 11.92,
+      "learning_rate": 5e-05,
+      "loss": 0.9246,
+      "step": 152000
+    },
+    {
+      "epoch": 11.93,
+      "learning_rate": 5e-05,
+      "loss": 0.935,
+      "step": 152100
+    },
+    {
+      "epoch": 11.94,
+      "learning_rate": 5e-05,
+      "loss": 0.9311,
+      "step": 152200
+    },
+    {
+      "epoch": 11.94,
+      "learning_rate": 5e-05,
+      "loss": 0.9228,
+      "step": 152300
+    },
+    {
+      "epoch": 11.95,
+      "learning_rate": 5e-05,
+      "loss": 0.9253,
+      "step": 152400
+    },
+    {
+      "epoch": 11.96,
+      "learning_rate": 5e-05,
+      "loss": 0.9347,
+      "step": 152500
+    },
+    {
+      "epoch": 11.97,
+      "learning_rate": 5e-05,
+      "loss": 0.9357,
+      "step": 152600
+    },
+    {
+      "epoch": 11.98,
+      "learning_rate": 5e-05,
+      "loss": 0.9402,
+      "step": 152700
+    },
+    {
+      "epoch": 11.98,
+      "learning_rate": 5e-05,
+      "loss": 0.9422,
+      "step": 152800
+    },
+    {
+      "epoch": 11.99,
+      "learning_rate": 5e-05,
+      "loss": 0.9344,
+      "step": 152900
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 5e-05,
+      "loss": 0.9344,
+      "step": 153000
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 5e-05,
+      "loss": 0.8553,
+      "step": 153100
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 5e-05,
+      "loss": 0.8309,
+      "step": 153200
+    },
+    {
+      "epoch": 12.02,
+      "learning_rate": 5e-05,
+      "loss": 0.8392,
+      "step": 153300
+    },
+    {
+      "epoch": 12.03,
+      "learning_rate": 5e-05,
+      "loss": 0.839,
+      "step": 153400
+    },
+    {
+      "epoch": 12.04,
+      "learning_rate": 5e-05,
+      "loss": 0.832,
+      "step": 153500
+    },
+    {
+      "epoch": 12.05,
+      "learning_rate": 5e-05,
+      "loss": 0.8413,
+      "step": 153600
+    },
+    {
+      "epoch": 12.05,
+      "learning_rate": 5e-05,
+      "loss": 0.8306,
+      "step": 153700
+    },
+    {
+      "epoch": 12.06,
+      "learning_rate": 5e-05,
+      "loss": 0.8318,
+      "step": 153800
+    },
+    {
+      "epoch": 12.07,
+      "learning_rate": 5e-05,
+      "loss": 0.8422,
+      "step": 153900
+    },
+    {
+      "epoch": 12.08,
+      "learning_rate": 5e-05,
+      "loss": 0.8423,
+      "step": 154000
+    },
+    {
+      "epoch": 12.09,
+      "learning_rate": 5e-05,
+      "loss": 0.8485,
+      "step": 154100
+    },
+    {
+      "epoch": 12.09,
+      "learning_rate": 5e-05,
+      "loss": 0.8339,
+      "step": 154200
+    },
+    {
+      "epoch": 12.1,
+      "learning_rate": 5e-05,
+      "loss": 0.8444,
+      "step": 154300
+    },
+    {
+      "epoch": 12.11,
+      "learning_rate": 5e-05,
+      "loss": 0.8435,
+      "step": 154400
+    },
+    {
+      "epoch": 12.12,
+      "learning_rate": 5e-05,
+      "loss": 0.8419,
+      "step": 154500
+    },
+    {
+      "epoch": 12.12,
+      "learning_rate": 5e-05,
+      "loss": 0.8566,
+      "step": 154600
+    },
+    {
+      "epoch": 12.13,
+      "learning_rate": 5e-05,
+      "loss": 0.8516,
+      "step": 154700
+    },
+    {
+      "epoch": 12.14,
+      "learning_rate": 5e-05,
+      "loss": 0.8737,
+      "step": 154800
+    },
+    {
+      "epoch": 12.15,
+      "learning_rate": 5e-05,
+      "loss": 0.8527,
+      "step": 154900
+    },
+    {
+      "epoch": 12.16,
+      "learning_rate": 5e-05,
+      "loss": 0.8388,
+      "step": 155000
+    },
+    {
+      "epoch": 12.16,
+      "learning_rate": 5e-05,
+      "loss": 0.8528,
+      "step": 155100
+    },
+    {
+      "epoch": 12.17,
+      "learning_rate": 5e-05,
+      "loss": 0.8548,
+      "step": 155200
+    },
+    {
+      "epoch": 12.18,
+      "learning_rate": 5e-05,
+      "loss": 0.866,
+      "step": 155300
+    },
+    {
+      "epoch": 12.19,
+      "learning_rate": 5e-05,
+      "loss": 0.8604,
+      "step": 155400
+    },
+    {
+      "epoch": 12.2,
+      "learning_rate": 5e-05,
+      "loss": 0.8596,
+      "step": 155500
+    },
+    {
+      "epoch": 12.2,
+      "learning_rate": 5e-05,
+      "loss": 0.8483,
+      "step": 155600
+    },
+    {
+      "epoch": 12.21,
+      "learning_rate": 5e-05,
+      "loss": 0.8523,
+      "step": 155700
+    },
+    {
+      "epoch": 12.22,
+      "learning_rate": 5e-05,
+      "loss": 0.8588,
+      "step": 155800
+    },
+    {
+      "epoch": 12.23,
+      "learning_rate": 5e-05,
+      "loss": 0.8571,
+      "step": 155900
+    },
+    {
+      "epoch": 12.23,
+      "learning_rate": 5e-05,
+      "loss": 0.8554,
+      "step": 156000
+    },
+    {
+      "epoch": 12.24,
+      "learning_rate": 5e-05,
+      "loss": 0.8586,
+      "step": 156100
+    },
+    {
+      "epoch": 12.25,
+      "learning_rate": 5e-05,
+      "loss": 0.8577,
+      "step": 156200
+    },
+    {
+      "epoch": 12.26,
+      "learning_rate": 5e-05,
+      "loss": 0.8751,
+      "step": 156300
+    },
+    {
+      "epoch": 12.27,
+      "learning_rate": 5e-05,
+      "loss": 0.8783,
+      "step": 156400
+    },
+    {
+      "epoch": 12.27,
+      "learning_rate": 5e-05,
+      "loss": 0.8635,
+      "step": 156500
+    },
+    {
+      "epoch": 12.28,
+      "learning_rate": 5e-05,
+      "loss": 0.8546,
+      "step": 156600
+    },
+    {
+      "epoch": 12.29,
+      "learning_rate": 5e-05,
+      "loss": 0.8548,
+      "step": 156700
+    },
+    {
+      "epoch": 12.3,
+      "learning_rate": 5e-05,
+      "loss": 0.8657,
+      "step": 156800
+    },
+    {
+      "epoch": 12.3,
+      "learning_rate": 5e-05,
+      "loss": 0.8578,
+      "step": 156900
+    },
+    {
+      "epoch": 12.31,
+      "learning_rate": 5e-05,
+      "loss": 0.8626,
+      "step": 157000
+    },
+    {
+      "epoch": 12.32,
+      "learning_rate": 5e-05,
+      "loss": 0.869,
+      "step": 157100
+    },
+    {
+      "epoch": 12.33,
+      "learning_rate": 5e-05,
+      "loss": 0.864,
+      "step": 157200
+    },
+    {
+      "epoch": 12.34,
+      "learning_rate": 5e-05,
+      "loss": 0.8677,
+      "step": 157300
+    },
+    {
+      "epoch": 12.34,
+      "learning_rate": 5e-05,
+      "loss": 0.8673,
+      "step": 157400
+    },
+    {
+      "epoch": 12.35,
+      "learning_rate": 5e-05,
+      "loss": 0.8685,
+      "step": 157500
+    },
+    {
+      "epoch": 12.36,
+      "learning_rate": 5e-05,
+      "loss": 0.8777,
+      "step": 157600
+    },
+    {
+      "epoch": 12.37,
+      "learning_rate": 5e-05,
+      "loss": 0.8707,
+      "step": 157700
+    },
+    {
+      "epoch": 12.38,
+      "learning_rate": 5e-05,
+      "loss": 0.8495,
+      "step": 157800
+    },
+    {
+      "epoch": 12.38,
+      "learning_rate": 5e-05,
+      "loss": 0.8891,
+      "step": 157900
+    },
+    {
+      "epoch": 12.39,
+      "learning_rate": 5e-05,
+      "loss": 0.8652,
+      "step": 158000
+    },
+    {
+      "epoch": 12.4,
+      "learning_rate": 5e-05,
+      "loss": 0.8688,
+      "step": 158100
+    },
+    {
+      "epoch": 12.41,
+      "learning_rate": 5e-05,
+      "loss": 0.8519,
+      "step": 158200
+    },
+    {
+      "epoch": 12.41,
+      "learning_rate": 5e-05,
+      "loss": 0.8686,
+      "step": 158300
+    },
+    {
+      "epoch": 12.42,
+      "learning_rate": 5e-05,
+      "loss": 0.859,
+      "step": 158400
+    },
+    {
+      "epoch": 12.43,
+      "learning_rate": 5e-05,
+      "loss": 0.8559,
+      "step": 158500
+    },
+    {
+      "epoch": 12.44,
+      "learning_rate": 5e-05,
+      "loss": 0.8713,
+      "step": 158600
+    },
+    {
+      "epoch": 12.45,
+      "learning_rate": 5e-05,
+      "loss": 0.8868,
+      "step": 158700
+    },
+    {
+      "epoch": 12.45,
+      "learning_rate": 5e-05,
+      "loss": 0.8699,
+      "step": 158800
+    },
+    {
+      "epoch": 12.46,
+      "learning_rate": 5e-05,
+      "loss": 0.8768,
+      "step": 158900
+    },
+    {
+      "epoch": 12.47,
+      "learning_rate": 5e-05,
+      "loss": 0.8746,
+      "step": 159000
+    },
+    {
+      "epoch": 12.48,
+      "learning_rate": 5e-05,
+      "loss": 0.873,
+      "step": 159100
+    },
+    {
+      "epoch": 12.49,
+      "learning_rate": 5e-05,
+      "loss": 0.8646,
+      "step": 159200
+    },
+    {
+      "epoch": 12.49,
+      "learning_rate": 5e-05,
+      "loss": 0.86,
+      "step": 159300
+    },
+    {
+      "epoch": 12.5,
+      "learning_rate": 5e-05,
+      "loss": 0.8631,
+      "step": 159400
+    },
+    {
+      "epoch": 12.51,
+      "learning_rate": 5e-05,
+      "loss": 0.8907,
+      "step": 159500
+    },
+    {
+      "epoch": 12.52,
+      "learning_rate": 5e-05,
+      "loss": 0.8596,
+      "step": 159600
+    },
+    {
+      "epoch": 12.52,
+      "learning_rate": 5e-05,
+      "loss": 0.8781,
+      "step": 159700
+    },
+    {
+      "epoch": 12.53,
+      "learning_rate": 5e-05,
+      "loss": 0.8728,
+      "step": 159800
+    },
+    {
+      "epoch": 12.54,
+      "learning_rate": 5e-05,
+      "loss": 0.8767,
+      "step": 159900
+    },
+    {
+      "epoch": 12.55,
+      "learning_rate": 5e-05,
+      "loss": 0.8635,
+      "step": 160000
+    },
+    {
+      "epoch": 12.55,
+      "eval_gen_len": 18.793943674406286,
+      "eval_loss": 1.6820098161697388,
+      "eval_rouge1": 39.4633,
+      "eval_rouge2": 18.0301,
+      "eval_rougeL": 32.9134,
+      "eval_rougeLsum": 32.919,
+      "eval_runtime": 2390.8284,
+      "eval_samples_per_second": 4.738,
+      "eval_steps_per_second": 1.185,
+      "step": 160000
+    }
+  ],
+  "max_steps": 637550,
+  "num_train_epochs": 50,
+  "total_flos": 8.741750241043968e+18,
+  "trial_name": null,
+  "trial_params": null
+}