diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7631 @@
+{
+  "best_metric": 3.5567543506622314,
+  "best_model_checkpoint": "/home/hpcpudu1/rds/hpc-work/data/pretrain-mds/led_pretrain/ver2/gen_model/Centrum_pretrain_base_batch_16_21-7-22.1/checkpoint-96500",
+  "epoch": 9.269546276127357,
+  "global_step": 100000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.0000000000000004e-09,
+      "loss": 5.1238,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.97e-07,
+      "loss": 4.8301,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.970000000000001e-07,
+      "loss": 4.4653,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.97e-07,
+      "loss": 4.2831,
+      "step": 300
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.197e-06,
+      "loss": 4.2033,
+      "step": 400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.497e-06,
+      "loss": 4.1628,
+      "step": 500
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 4.073188304901123,
+      "eval_runtime": 23.8328,
+      "eval_samples_per_second": 105.821,
+      "eval_steps_per_second": 6.63,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.7970000000000001e-06,
+      "loss": 4.1047,
+      "step": 600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.097e-06,
+      "loss": 4.1061,
+      "step": 700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.397e-06,
+      "loss": 4.0666,
+      "step": 800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.6969999999999998e-06,
+      "loss": 4.05,
+      "step": 900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.9970000000000003e-06,
+      "loss": 4.0278,
+      "step": 1000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 3.9799606800079346,
+      "eval_runtime": 23.808,
+      "eval_samples_per_second": 105.931,
+      "eval_steps_per_second": 6.636,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.297e-06,
+      "loss": 4.0257,
+      "step": 1100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3.5970000000000005e-06,
+      "loss": 4.0151,
+      "step": 1200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.897e-06,
+      "loss": 3.9934,
+      "step": 1300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.197e-06,
+      "loss": 4.0018,
+      "step": 1400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.497e-06,
+      "loss": 4.0008,
+      "step": 1500
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 3.9282920360565186,
+      "eval_runtime": 23.8216,
+      "eval_samples_per_second": 105.87,
+      "eval_steps_per_second": 6.633,
+      "step": 1500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.797e-06,
+      "loss": 3.98,
+      "step": 1600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.097e-06,
+      "loss": 3.9486,
+      "step": 1700
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.397e-06,
+      "loss": 3.9664,
+      "step": 1800
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.697000000000001e-06,
+      "loss": 3.9469,
+      "step": 1900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.997e-06,
+      "loss": 3.9564,
+      "step": 2000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 3.8940863609313965,
+      "eval_runtime": 23.785,
+      "eval_samples_per_second": 106.033,
+      "eval_steps_per_second": 6.643,
+      "step": 2000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 6.297e-06,
+      "loss": 3.9533,
+      "step": 2100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 6.5970000000000005e-06,
+      "loss": 3.9263,
+      "step": 2200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 6.897e-06,
+      "loss": 3.9351,
+      "step": 2300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 7.197e-06,
+      "loss": 3.9339,
+      "step": 2400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 7.497e-06,
+      "loss": 3.9193,
+      "step": 2500
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 3.878038167953491,
+      "eval_runtime": 23.806,
+      "eval_samples_per_second": 105.94,
+      "eval_steps_per_second": 6.637,
+      "step": 2500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 7.797e-06,
+      "loss": 3.9323,
+      "step": 2600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.096999999999999e-06,
+      "loss": 3.9102,
+      "step": 2700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.397e-06,
+      "loss": 3.891,
+      "step": 2800
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.694e-06,
+      "loss": 3.9072,
+      "step": 2900
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.994e-06,
+      "loss": 3.9185,
+      "step": 3000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 3.8500914573669434,
+      "eval_runtime": 23.7916,
+      "eval_samples_per_second": 106.004,
+      "eval_steps_per_second": 6.641,
+      "step": 3000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.294000000000001e-06,
+      "loss": 3.8927,
+      "step": 3100
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.594e-06,
+      "loss": 3.8977,
+      "step": 3200
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.894e-06,
+      "loss": 3.8906,
+      "step": 3300
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.0194e-05,
+      "loss": 3.8869,
+      "step": 3400
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.0494e-05,
+      "loss": 3.8881,
+      "step": 3500
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 3.833404779434204,
+      "eval_runtime": 23.8004,
+      "eval_samples_per_second": 105.965,
+      "eval_steps_per_second": 6.639,
+      "step": 3500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.0794e-05,
+      "loss": 3.8722,
+      "step": 3600
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.1094e-05,
+      "loss": 3.8633,
+      "step": 3700
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.1394000000000001e-05,
+      "loss": 3.8719,
+      "step": 3800
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.1694e-05,
+      "loss": 3.8499,
+      "step": 3900
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.1994e-05,
+      "loss": 3.8869,
+      "step": 4000
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 3.8210906982421875,
+      "eval_runtime": 23.7934,
+      "eval_samples_per_second": 105.996,
+      "eval_steps_per_second": 6.64,
+      "step": 4000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.2294e-05,
+      "loss": 3.8583,
+      "step": 4100
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.2594e-05,
+      "loss": 3.8594,
+      "step": 4200
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.2894e-05,
+      "loss": 3.8824,
+      "step": 4300
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3194000000000001e-05,
+      "loss": 3.8901,
+      "step": 4400
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.3494e-05,
+      "loss": 3.876,
+      "step": 4500
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 3.8056981563568115,
+      "eval_runtime": 23.8047,
+      "eval_samples_per_second": 105.946,
+      "eval_steps_per_second": 6.637,
+      "step": 4500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3794e-05,
+      "loss": 3.8473,
+      "step": 4600
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.4094000000000001e-05,
+      "loss": 3.8718,
+      "step": 4700
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.4394e-05,
+      "loss": 3.8671,
+      "step": 4800
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.4694e-05,
+      "loss": 3.8396,
+      "step": 4900
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.4994e-05,
+      "loss": 3.8552,
+      "step": 5000
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 3.795402765274048,
+      "eval_runtime": 23.8098,
+      "eval_samples_per_second": 105.923,
+      "eval_steps_per_second": 6.636,
+      "step": 5000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5294000000000003e-05,
+      "loss": 3.838,
+      "step": 5100
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5594e-05,
+      "loss": 3.8231,
+      "step": 5200
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.5894e-05,
+      "loss": 3.8139,
+      "step": 5300
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.6193999999999998e-05,
+      "loss": 3.8249,
+      "step": 5400
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.6493999999999998e-05,
+      "loss": 3.8198,
+      "step": 5500
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 3.7860653400421143,
+      "eval_runtime": 23.7811,
+      "eval_samples_per_second": 106.051,
+      "eval_steps_per_second": 6.644,
+      "step": 5500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.6794e-05,
+      "loss": 3.8332,
+      "step": 5600
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.7094e-05,
+      "loss": 3.833,
+      "step": 5700
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.7394e-05,
+      "loss": 3.8316,
+      "step": 5800
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.7694e-05,
+      "loss": 3.8073,
+      "step": 5900
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.7994e-05,
+      "loss": 3.8016,
+      "step": 6000
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 3.7749924659729004,
+      "eval_runtime": 23.805,
+      "eval_samples_per_second": 105.944,
+      "eval_steps_per_second": 6.637,
+      "step": 6000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8294e-05,
+      "loss": 3.8169,
+      "step": 6100
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8594000000000002e-05,
+      "loss": 3.8168,
+      "step": 6200
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8894000000000002e-05,
+      "loss": 3.8178,
+      "step": 6300
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9194000000000003e-05,
+      "loss": 3.7942,
+      "step": 6400
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.9494000000000003e-05,
+      "loss": 3.8033,
+      "step": 6500
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 3.765125036239624,
+      "eval_runtime": 23.776,
+      "eval_samples_per_second": 106.074,
+      "eval_steps_per_second": 6.645,
+      "step": 6500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9794e-05,
+      "loss": 3.7867,
+      "step": 6600
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.0093999999999998e-05,
+      "loss": 3.7787,
+      "step": 6700
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.0393999999999998e-05,
+      "loss": 3.8017,
+      "step": 6800
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.0694e-05,
+      "loss": 3.7995,
+      "step": 6900
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.0994e-05,
+      "loss": 3.7927,
+      "step": 7000
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 3.7527832984924316,
+      "eval_runtime": 23.795,
+      "eval_samples_per_second": 105.988,
+      "eval_steps_per_second": 6.64,
+      "step": 7000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.1294e-05,
+      "loss": 3.788,
+      "step": 7100
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.1594e-05,
+      "loss": 3.8054,
+      "step": 7200
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.1894e-05,
+      "loss": 3.8069,
+      "step": 7300
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.2194e-05,
+      "loss": 3.7761,
+      "step": 7400
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.2494000000000002e-05,
+      "loss": 3.7978,
+      "step": 7500
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 3.742873191833496,
+      "eval_runtime": 23.8023,
+      "eval_samples_per_second": 105.956,
+      "eval_steps_per_second": 6.638,
+      "step": 7500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.2794000000000002e-05,
+      "loss": 3.7738,
+      "step": 7600
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.3094000000000003e-05,
+      "loss": 3.7788,
+      "step": 7700
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.3394000000000003e-05,
+      "loss": 3.7823,
+      "step": 7800
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.3694e-05,
+      "loss": 3.7879,
+      "step": 7900
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.3993999999999998e-05,
+      "loss": 3.7727,
+      "step": 8000
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 3.7367067337036133,
+      "eval_runtime": 23.7892,
+      "eval_samples_per_second": 106.014,
+      "eval_steps_per_second": 6.642,
+      "step": 8000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.4293999999999998e-05,
+      "loss": 3.7747,
+      "step": 8100
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.4594e-05,
+      "loss": 3.7716,
+      "step": 8200
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.4894e-05,
+      "loss": 3.765,
+      "step": 8300
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5191e-05,
+      "loss": 3.7819,
+      "step": 8400
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.5491e-05,
+      "loss": 3.7634,
+      "step": 8500
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 3.7275290489196777,
+      "eval_runtime": 23.7785,
+      "eval_samples_per_second": 106.062,
+      "eval_steps_per_second": 6.645,
+      "step": 8500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.5791e-05,
+      "loss": 3.7757,
+      "step": 8600
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.6091e-05,
+      "loss": 3.7559,
+      "step": 8700
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.6391000000000002e-05,
+      "loss": 3.7709,
+      "step": 8800
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.6688e-05,
+      "loss": 3.7626,
+      "step": 8900
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.6985e-05,
+      "loss": 3.7395,
+      "step": 9000
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 3.715789318084717,
+      "eval_runtime": 23.7885,
+      "eval_samples_per_second": 106.018,
+      "eval_steps_per_second": 6.642,
+      "step": 9000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.7285e-05,
+      "loss": 3.7773,
+      "step": 9100
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.7585e-05,
+      "loss": 3.7273,
+      "step": 9200
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.7885e-05,
+      "loss": 3.7598,
+      "step": 9300
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.8185000000000002e-05,
+      "loss": 3.7531,
+      "step": 9400
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.8485000000000003e-05,
+      "loss": 3.7432,
+      "step": 9500
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 3.706634283065796,
+      "eval_runtime": 23.7692,
+      "eval_samples_per_second": 106.103,
+      "eval_steps_per_second": 6.647,
+      "step": 9500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.8785e-05,
+      "loss": 3.7297,
+      "step": 9600
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.9085e-05,
+      "loss": 3.7366,
+      "step": 9700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.9385e-05,
+      "loss": 3.7179,
+      "step": 9800
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.9685e-05,
+      "loss": 3.7166,
+      "step": 9900
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.9985000000000002e-05,
+      "loss": 3.7623,
+      "step": 10000
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 3.7039339542388916,
+      "eval_runtime": 23.7947,
+      "eval_samples_per_second": 105.99,
+      "eval_steps_per_second": 6.64,
+      "step": 10000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.9968333333333332e-05,
+      "loss": 3.7351,
+      "step": 10100
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.9935e-05,
+      "loss": 3.7235,
+      "step": 10200
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.9901666666666665e-05,
+      "loss": 3.7207,
+      "step": 10300
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.9868333333333333e-05,
+      "loss": 3.7299,
+      "step": 10400
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.9835e-05,
+      "loss": 3.7182,
+      "step": 10500
+    },
+    {
+      "epoch": 0.97,
+      "eval_loss": 3.6904149055480957,
+      "eval_runtime": 23.8525,
+      "eval_samples_per_second": 105.733,
+      "eval_steps_per_second": 6.624,
+      "step": 10500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.9801666666666666e-05,
+      "loss": 3.7165,
+      "step": 10600
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.9768333333333334e-05,
+      "loss": 3.7112,
+      "step": 10700
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9735e-05,
+      "loss": 3.7438,
+      "step": 10800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9701666666666667e-05,
+      "loss": 3.6832,
+      "step": 10900
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9668333333333332e-05,
+      "loss": 3.7146,
+      "step": 11000
+    },
+    {
+      "epoch": 1.02,
+      "eval_loss": 3.6881186962127686,
+      "eval_runtime": 23.7931,
+      "eval_samples_per_second": 105.997,
+      "eval_steps_per_second": 6.641,
+      "step": 11000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9635e-05,
+      "loss": 3.6748,
+      "step": 11100
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.9601666666666665e-05,
+      "loss": 3.6982,
+      "step": 11200
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9568333333333333e-05,
+      "loss": 3.7015,
+      "step": 11300
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.9535e-05,
+      "loss": 3.6715,
+      "step": 11400
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.9501666666666666e-05,
+      "loss": 3.681,
+      "step": 11500
+    },
+    {
+      "epoch": 1.07,
+      "eval_loss": 3.679689407348633,
+      "eval_runtime": 23.7803,
+      "eval_samples_per_second": 106.054,
+      "eval_steps_per_second": 6.644,
+      "step": 11500
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.9468333333333334e-05,
+      "loss": 3.6938,
+      "step": 11600
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.9435e-05,
+      "loss": 3.6871,
+      "step": 11700
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.9401666666666667e-05,
+      "loss": 3.6833,
+      "step": 11800
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.936833333333333e-05,
+      "loss": 3.6986,
+      "step": 11900
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.9335e-05,
+      "loss": 3.6745,
+      "step": 12000
+    },
+    {
+      "epoch": 1.11,
+      "eval_loss": 3.674957275390625,
+      "eval_runtime": 23.7927,
+      "eval_samples_per_second": 105.999,
+      "eval_steps_per_second": 6.641,
+      "step": 12000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.9301666666666668e-05,
+      "loss": 3.666,
+      "step": 12100
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.9268333333333332e-05,
+      "loss": 3.6803,
+      "step": 12200
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.9235e-05,
+      "loss": 3.6671,
+      "step": 12300
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.9201666666666665e-05,
+      "loss": 3.6783,
+      "step": 12400
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.9168333333333333e-05,
+      "loss": 3.6794,
+      "step": 12500
+    },
+    {
+      "epoch": 1.16,
+      "eval_loss": 3.6748008728027344,
+      "eval_runtime": 23.8003,
+      "eval_samples_per_second": 105.965,
+      "eval_steps_per_second": 6.639,
+      "step": 12500
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.9134999999999998e-05,
+      "loss": 3.69,
+      "step": 12600
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.9101666666666666e-05,
+      "loss": 3.6674,
+      "step": 12700
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.9068333333333334e-05,
+      "loss": 3.647,
+      "step": 12800
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.9035e-05,
+      "loss": 3.6818,
+      "step": 12900
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.9001666666666667e-05,
+      "loss": 3.6802,
+      "step": 13000
+    },
+    {
+      "epoch": 1.21,
+      "eval_loss": 3.6695618629455566,
+      "eval_runtime": 23.7884,
+      "eval_samples_per_second": 106.018,
+      "eval_steps_per_second": 6.642,
+      "step": 13000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.896866666666667e-05,
+      "loss": 3.6639,
+      "step": 13100
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.8935333333333334e-05,
+      "loss": 3.6644,
+      "step": 13200
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.8902000000000002e-05,
+      "loss": 3.6877,
+      "step": 13300
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.8868666666666667e-05,
+      "loss": 3.6686,
+      "step": 13400
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.8835333333333335e-05,
+      "loss": 3.665,
+      "step": 13500
+    },
+    {
+      "epoch": 1.25,
+      "eval_loss": 3.6609461307525635,
+      "eval_runtime": 23.7865,
+      "eval_samples_per_second": 106.027,
+      "eval_steps_per_second": 6.642,
+      "step": 13500
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.8802e-05,
+      "loss": 3.6411,
+      "step": 13600
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.8768666666666668e-05,
+      "loss": 3.6634,
+      "step": 13700
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.8735333333333336e-05,
+      "loss": 3.6865,
+      "step": 13800
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.8702e-05,
+      "loss": 3.6675,
+      "step": 13900
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.866866666666667e-05,
+      "loss": 3.6516,
+      "step": 14000
+    },
+    {
+      "epoch": 1.3,
+      "eval_loss": 3.663334608078003,
+      "eval_runtime": 23.7962,
+      "eval_samples_per_second": 105.983,
+      "eval_steps_per_second": 6.64,
+      "step": 14000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.8635333333333333e-05,
+      "loss": 3.6612,
+      "step": 14100
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.8602e-05,
+      "loss": 3.68,
+      "step": 14200
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.8568666666666666e-05,
+      "loss": 3.6674,
+      "step": 14300
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.8535333333333334e-05,
+      "loss": 3.6875,
+      "step": 14400
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.8502e-05,
+      "loss": 3.6577,
+      "step": 14500
+    },
+    {
+      "epoch": 1.34,
+      "eval_loss": 3.657306432723999,
+      "eval_runtime": 23.7877,
+      "eval_samples_per_second": 106.021,
+      "eval_steps_per_second": 6.642,
+      "step": 14500
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.8468666666666667e-05,
+      "loss": 3.6649,
+      "step": 14600
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.8435333333333335e-05,
+      "loss": 3.6539,
+      "step": 14700
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.8402e-05,
+      "loss": 3.6726,
+      "step": 14800
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.8368666666666668e-05,
+      "loss": 3.6542,
+      "step": 14900
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.8335333333333333e-05,
+      "loss": 3.6409,
+      "step": 15000
+    },
+    {
+      "epoch": 1.39,
+      "eval_loss": 3.6518640518188477,
+      "eval_runtime": 23.7734,
+      "eval_samples_per_second": 106.085,
+      "eval_steps_per_second": 6.646,
+      "step": 15000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.8302e-05,
+      "loss": 3.6629,
+      "step": 15100
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.8268666666666666e-05,
+      "loss": 3.6467,
+      "step": 15200
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.8235333333333334e-05,
+      "loss": 3.6499,
+      "step": 15300
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.8202000000000002e-05,
+      "loss": 3.6623,
+      "step": 15400
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.8168666666666667e-05,
+      "loss": 3.6691,
+      "step": 15500
+    },
+    {
+      "epoch": 1.44,
+      "eval_loss": 3.6490323543548584,
+      "eval_runtime": 23.8031,
+      "eval_samples_per_second": 105.953,
+      "eval_steps_per_second": 6.638,
+      "step": 15500
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.813566666666667e-05,
+      "loss": 3.6585,
+      "step": 15600
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.8102333333333337e-05,
+      "loss": 3.6582,
+      "step": 15700
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.8069000000000002e-05,
+      "loss": 3.648,
+      "step": 15800
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.803566666666667e-05,
+      "loss": 3.6484,
+      "step": 15900
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.8002333333333335e-05,
+      "loss": 3.6521,
+      "step": 16000
+    },
+    {
+      "epoch": 1.48,
+      "eval_loss": 3.647477865219116,
+      "eval_runtime": 23.7875,
+      "eval_samples_per_second": 106.022,
+      "eval_steps_per_second": 6.642,
+      "step": 16000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.7969000000000003e-05,
+      "loss": 3.6635,
+      "step": 16100
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.7935666666666667e-05,
+      "loss": 3.6423,
+      "step": 16200
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.7902333333333336e-05,
+      "loss": 3.6403,
+      "step": 16300
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.7869000000000004e-05,
+      "loss": 3.6515,
+      "step": 16400
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.783566666666667e-05,
+      "loss": 3.6435,
+      "step": 16500
+    },
+    {
+      "epoch": 1.53,
+      "eval_loss": 3.6465208530426025,
+      "eval_runtime": 23.7862,
+      "eval_samples_per_second": 106.028,
+      "eval_steps_per_second": 6.643,
+      "step": 16500
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.7802333333333337e-05,
+      "loss": 3.667,
+      "step": 16600
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.7769e-05,
+      "loss": 3.6313,
+      "step": 16700
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.773566666666667e-05,
+      "loss": 3.6229,
+      "step": 16800
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.7702333333333334e-05,
+      "loss": 3.6347,
+      "step": 16900
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.7669000000000002e-05,
+      "loss": 3.6466,
+      "step": 17000
+    },
+    {
+      "epoch": 1.58,
+      "eval_loss": 3.6391589641571045,
+      "eval_runtime": 23.7988,
+      "eval_samples_per_second": 105.972,
+      "eval_steps_per_second": 6.639,
+      "step": 17000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.7635666666666667e-05,
+      "loss": 3.6402,
+      "step": 17100
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.7602333333333335e-05,
+      "loss": 3.6356,
+      "step": 17200
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.7569000000000003e-05,
+      "loss": 3.6462,
+      "step": 17300
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.7535666666666668e-05,
+      "loss": 3.6613,
+      "step": 17400
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.7502333333333336e-05,
+      "loss": 3.644,
+      "step": 17500
+    },
+    {
+      "epoch": 1.62,
+      "eval_loss": 3.641881227493286,
+      "eval_runtime": 23.9969,
+      "eval_samples_per_second": 105.097,
+      "eval_steps_per_second": 6.584,
+      "step": 17500
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.7469e-05,
+      "loss": 3.641,
+      "step": 17600
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.743566666666667e-05,
+      "loss": 3.6439,
+      "step": 17700
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.7402333333333334e-05,
+      "loss": 3.642,
+      "step": 17800
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.7369000000000002e-05,
+      "loss": 3.6524,
+      "step": 17900
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.733566666666667e-05,
+      "loss": 3.6347,
+      "step": 18000
+    },
+    {
+      "epoch": 1.67,
+      "eval_loss": 3.6347177028656006,
+      "eval_runtime": 23.7827,
+      "eval_samples_per_second": 106.044,
+      "eval_steps_per_second": 6.643,
+      "step": 18000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.7302333333333335e-05,
+      "loss": 3.6091,
+      "step": 18100
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.7269000000000003e-05,
+      "loss": 3.6392,
+      "step": 18200
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.7235666666666667e-05,
+      "loss": 3.6385,
+      "step": 18300
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.7202333333333336e-05,
+      "loss": 3.6259,
+      "step": 18400
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.7169e-05,
+      "loss": 3.6205,
+      "step": 18500
+    },
+    {
+      "epoch": 1.71,
+      "eval_loss": 3.6328201293945312,
+      "eval_runtime": 23.7927,
+      "eval_samples_per_second": 105.999,
+      "eval_steps_per_second": 6.641,
+      "step": 18500
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.713566666666667e-05,
+      "loss": 3.6171,
+      "step": 18600
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.7102333333333336e-05,
+      "loss": 3.62,
+      "step": 18700
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.7069e-05,
+      "loss": 3.6423,
+      "step": 18800
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.703566666666667e-05,
+      "loss": 3.6225,
+      "step": 18900
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.7002333333333334e-05,
+      "loss": 3.6451,
+      "step": 19000
+    },
+    {
+      "epoch": 1.76,
+      "eval_loss": 3.630978584289551,
+      "eval_runtime": 23.7828,
+      "eval_samples_per_second": 106.043,
+      "eval_steps_per_second": 6.643,
+      "step": 19000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.6969000000000002e-05,
+      "loss": 3.6364,
+      "step": 19100
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.6936e-05,
+      "loss": 3.6305,
+      "step": 19200
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.6902666666666666e-05,
+      "loss": 3.6389,
+      "step": 19300
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.6869333333333334e-05,
+      "loss": 3.6359,
+      "step": 19400
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.6836e-05,
+      "loss": 3.6327,
+      "step": 19500
+    },
+    {
+      "epoch": 1.81,
+      "eval_loss": 3.6284360885620117,
+      "eval_runtime": 23.7713,
+      "eval_samples_per_second": 106.094,
+      "eval_steps_per_second": 6.647,
+      "step": 19500
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.6802666666666667e-05,
+      "loss": 3.6089,
+      "step": 19600
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.6769333333333335e-05,
+      "loss": 3.6254,
+      "step": 19700
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.6736e-05,
+      "loss": 3.6378,
+      "step": 19800
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.6702666666666668e-05,
+      "loss": 3.6374,
+      "step": 19900
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.6669333333333332e-05,
+      "loss": 3.6166,
+      "step": 20000
+    },
+    {
+      "epoch": 1.85,
+      "eval_loss": 3.6267168521881104,
+      "eval_runtime": 23.7269,
+      "eval_samples_per_second": 106.293,
+      "eval_steps_per_second": 6.659,
+      "step": 20000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.6636e-05,
+      "loss": 3.6233,
+      "step": 20100
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.6602666666666665e-05,
+      "loss": 3.6276,
+      "step": 20200
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.6569333333333333e-05,
+      "loss": 3.64,
+      "step": 20300
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2.6535999999999998e-05,
+      "loss": 3.6328,
+      "step": 20400
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2.6502666666666666e-05,
+      "loss": 3.622,
+      "step": 20500
+    },
+    {
+      "epoch": 1.9,
+      "eval_loss": 3.621175765991211,
+      "eval_runtime": 23.806,
+      "eval_samples_per_second": 105.94,
+      "eval_steps_per_second": 6.637,
+      "step": 20500
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.6469333333333334e-05,
+      "loss": 3.6092,
+      "step": 20600
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.6436e-05,
+      "loss": 3.6236,
+      "step": 20700
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.6402666666666667e-05,
+      "loss": 3.6293,
+      "step": 20800
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.6369333333333332e-05,
+      "loss": 3.6257,
+      "step": 20900
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.6336e-05,
+      "loss": 3.6164,
+      "step": 21000
+    },
+    {
+      "epoch": 1.95,
+      "eval_loss": 3.619929075241089,
+      "eval_runtime": 23.9488,
+      "eval_samples_per_second": 105.308,
+      "eval_steps_per_second": 6.597,
+      "step": 21000
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.6303333333333333e-05,
+      "loss": 3.6132,
+      "step": 21100
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.627e-05,
+      "loss": 3.6215,
+      "step": 21200
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.6236666666666666e-05,
+      "loss": 3.6209,
+      "step": 21300
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.6203333333333334e-05,
+      "loss": 3.6149,
+      "step": 21400
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.6170333333333336e-05,
+      "loss": 3.6178,
+      "step": 21500
+    },
+    {
+      "epoch": 1.99,
+      "eval_loss": 3.6201303005218506,
+      "eval_runtime": 23.776,
+      "eval_samples_per_second": 106.074,
+      "eval_steps_per_second": 6.645,
+      "step": 21500
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6137e-05,
+      "loss": 3.6313,
+      "step": 21600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.610366666666667e-05,
+      "loss": 3.5876,
+      "step": 21700
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.6070333333333333e-05,
+      "loss": 3.573,
+      "step": 21800
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.6037e-05,
+      "loss": 3.5804,
+      "step": 21900
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.600366666666667e-05,
+      "loss": 3.5892,
+      "step": 22000
+    },
+    {
+      "epoch": 2.04,
+      "eval_loss": 3.620147466659546,
+      "eval_runtime": 23.7922,
+      "eval_samples_per_second": 106.001,
+      "eval_steps_per_second": 6.641,
+      "step": 22000
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.5970333333333334e-05,
+      "loss": 3.5861,
+      "step": 22100
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.5937000000000003e-05,
+      "loss": 3.5823,
+      "step": 22200
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.5903666666666667e-05,
+      "loss": 3.5576,
+      "step": 22300
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.5870333333333335e-05,
+      "loss": 3.5834,
+      "step": 22400
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.5837e-05,
+      "loss": 3.5855,
+      "step": 22500
+    },
+    {
+      "epoch": 2.09,
+      "eval_loss": 3.6220951080322266,
+      "eval_runtime": 23.7972,
+      "eval_samples_per_second": 105.979,
+      "eval_steps_per_second": 6.639,
+      "step": 22500
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.5803666666666668e-05,
+      "loss": 3.5831,
+      "step": 22600
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.5770333333333333e-05,
+      "loss": 3.5785,
+      "step": 22700
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.5737e-05,
+      "loss": 3.5825,
+      "step": 22800
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.570366666666667e-05,
+      "loss": 3.5766,
+      "step": 22900
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.5670333333333334e-05,
+      "loss": 3.5658,
+      "step": 23000
+    },
+    {
+      "epoch": 2.13,
+      "eval_loss": 3.6193323135375977,
+      "eval_runtime": 23.7837,
+      "eval_samples_per_second": 106.039,
+      "eval_steps_per_second": 6.643,
+      "step": 23000
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.5637000000000002e-05,
+      "loss": 3.5795,
+      "step": 23100
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.5603666666666667e-05,
+      "loss": 3.5644,
+      "step": 23200
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.5570333333333335e-05,
+      "loss": 3.5725,
+      "step": 23300
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.5537e-05,
+      "loss": 3.5833,
+      "step": 23400
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.5503666666666668e-05,
+      "loss": 3.5916,
+      "step": 23500
+    },
+    {
+      "epoch": 2.18,
+      "eval_loss": 3.6143603324890137,
+      "eval_runtime": 23.7839,
+      "eval_samples_per_second": 106.038,
+      "eval_steps_per_second": 6.643,
+      "step": 23500
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.5470333333333336e-05,
+      "loss": 3.5665,
+      "step": 23600
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.5437e-05,
+      "loss": 3.577,
+      "step": 23700
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.540366666666667e-05,
+      "loss": 3.5749,
+      "step": 23800
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.5370333333333333e-05,
+      "loss": 3.5777,
+      "step": 23900
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.5337e-05,
+      "loss": 3.5767,
+      "step": 24000
+    },
+    {
+      "epoch": 2.22,
+      "eval_loss": 3.6100852489471436,
+      "eval_runtime": 23.7895,
+      "eval_samples_per_second": 106.013,
+      "eval_steps_per_second": 6.642,
+      "step": 24000
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.5303666666666666e-05,
+      "loss": 3.557,
+      "step": 24100
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.5270333333333334e-05,
+      "loss": 3.5769,
+      "step": 24200
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.5237000000000002e-05,
+      "loss": 3.5622,
+      "step": 24300
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.5203666666666667e-05,
+      "loss": 3.571,
+      "step": 24400
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.5170333333333335e-05,
+      "loss": 3.5809,
+      "step": 24500
+    },
+    {
+      "epoch": 2.27,
+      "eval_loss": 3.611521005630493,
+      "eval_runtime": 23.7847,
+      "eval_samples_per_second": 106.035,
+      "eval_steps_per_second": 6.643,
+      "step": 24500
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.5137e-05,
+      "loss": 3.583,
+      "step": 24600
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.5103666666666668e-05,
+      "loss": 3.5867,
+      "step": 24700
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.5070333333333333e-05,
+      "loss": 3.5639,
+      "step": 24800
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.5037e-05,
+      "loss": 3.5715,
+      "step": 24900
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5003666666666666e-05,
+      "loss": 3.5561,
+      "step": 25000
+    },
+    {
+      "epoch": 2.32,
+      "eval_loss": 3.611006021499634,
+      "eval_runtime": 23.7869,
+      "eval_samples_per_second": 106.025,
+      "eval_steps_per_second": 6.642,
+      "step": 25000
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.4970333333333334e-05,
+      "loss": 3.5542,
+      "step": 25100
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.4937000000000002e-05,
+      "loss": 3.589,
+      "step": 25200
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.4903666666666667e-05,
+      "loss": 3.5546,
+      "step": 25300
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.4870333333333335e-05,
+      "loss": 3.547,
+      "step": 25400
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.4837e-05,
+      "loss": 3.5831,
+      "step": 25500
+    },
+    {
+      "epoch": 2.36,
+      "eval_loss": 3.6079983711242676,
+      "eval_runtime": 23.8882,
+      "eval_samples_per_second": 105.575,
+      "eval_steps_per_second": 6.614,
+      "step": 25500
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.4803666666666668e-05,
+      "loss": 3.5712,
+      "step": 25600
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.4770333333333332e-05,
+      "loss": 3.5836,
+      "step": 25700
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.4737e-05,
+      "loss": 3.5661,
+      "step": 25800
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.470366666666667e-05,
+      "loss": 3.5508,
+      "step": 25900
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.4670666666666667e-05,
+      "loss": 3.5551,
+      "step": 26000
+    },
+    {
+      "epoch": 2.41,
+      "eval_loss": 3.612078905105591,
+      "eval_runtime": 23.7677,
+      "eval_samples_per_second": 106.11,
+      "eval_steps_per_second": 6.648,
+      "step": 26000
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.4637333333333336e-05,
+      "loss": 3.564,
+      "step": 26100
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.4604000000000004e-05,
+      "loss": 3.5561,
+      "step": 26200
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.457066666666667e-05,
+      "loss": 3.5639,
+      "step": 26300
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.4537333333333336e-05,
+      "loss": 3.5583,
+      "step": 26400
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.4504e-05,
+      "loss": 3.5588,
+      "step": 26500
+    },
+    {
+      "epoch": 2.46,
+      "eval_loss": 3.6071577072143555,
+      "eval_runtime": 23.8054,
+      "eval_samples_per_second": 105.942,
+      "eval_steps_per_second": 6.637,
+      "step": 26500
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.447066666666667e-05,
+      "loss": 3.5733,
+      "step": 26600
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.4437333333333334e-05,
+      "loss": 3.5882,
+      "step": 26700
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2.4404000000000002e-05,
+      "loss": 3.5682,
+      "step": 26800
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.437066666666667e-05,
+      "loss": 3.5627,
+      "step": 26900
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2.4337333333333335e-05,
+      "loss": 3.5645,
+      "step": 27000
+    },
+    {
+      "epoch": 2.5,
+      "eval_loss": 3.6056134700775146,
+      "eval_runtime": 23.7971,
+      "eval_samples_per_second": 105.979,
+      "eval_steps_per_second": 6.639,
+      "step": 27000
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.4304000000000003e-05,
+      "loss": 3.5442,
+      "step": 27100
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.4270666666666668e-05,
+      "loss": 3.5718,
+      "step": 27200
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2.4237333333333336e-05,
+      "loss": 3.5869,
+      "step": 27300
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2.4204e-05,
+      "loss": 3.5762,
+      "step": 27400
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.417066666666667e-05,
+      "loss": 3.5804,
+      "step": 27500
+    },
+    {
+      "epoch": 2.55,
+      "eval_loss": 3.6037800312042236,
+      "eval_runtime": 23.7846,
+      "eval_samples_per_second": 106.035,
+      "eval_steps_per_second": 6.643,
+      "step": 27500
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.4137333333333334e-05,
+      "loss": 3.567,
+      "step": 27600
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.4104e-05,
+      "loss": 3.5534,
+      "step": 27700
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.407066666666667e-05,
+      "loss": 3.568,
+      "step": 27800
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.4037333333333334e-05,
+      "loss": 3.5745,
+      "step": 27900
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.4004000000000003e-05,
+      "loss": 3.5712,
+      "step": 28000
+    },
+    {
+      "epoch": 2.6,
+      "eval_loss": 3.605182647705078,
+      "eval_runtime": 23.7889,
+      "eval_samples_per_second": 106.016,
+      "eval_steps_per_second": 6.642,
+      "step": 28000
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.3970666666666667e-05,
+      "loss": 3.5636,
+      "step": 28100
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.3937333333333335e-05,
+      "loss": 3.5593,
+      "step": 28200
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.3904e-05,
+      "loss": 3.5347,
+      "step": 28300
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.3870666666666668e-05,
+      "loss": 3.5813,
+      "step": 28400
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.3837333333333336e-05,
+      "loss": 3.5494,
+      "step": 28500
+    },
+    {
+      "epoch": 2.64,
+      "eval_loss": 3.601443290710449,
+      "eval_runtime": 23.783,
+      "eval_samples_per_second": 106.042,
+      "eval_steps_per_second": 6.643,
+      "step": 28500
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.3804e-05,
+      "loss": 3.5589,
+      "step": 28600
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.377066666666667e-05,
+      "loss": 3.583,
+      "step": 28700
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.3737666666666668e-05,
+      "loss": 3.5648,
+      "step": 28800
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.3704333333333333e-05,
+      "loss": 3.549,
+      "step": 28900
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.3671e-05,
+      "loss": 3.582,
+      "step": 29000
+    },
+    {
+      "epoch": 2.69,
+      "eval_loss": 3.599480152130127,
+      "eval_runtime": 23.7091,
+      "eval_samples_per_second": 106.373,
+      "eval_steps_per_second": 6.664,
+      "step": 29000
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.3637666666666666e-05,
+      "loss": 3.5619,
+      "step": 29100
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.3604333333333334e-05,
+      "loss": 3.5682,
+      "step": 29200
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.3571e-05,
+      "loss": 3.5896,
+      "step": 29300
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.3537666666666667e-05,
+      "loss": 3.552,
+      "step": 29400
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.3504333333333335e-05,
+      "loss": 3.5487,
+      "step": 29500
+    },
+    {
+      "epoch": 2.73,
+      "eval_loss": 3.6051015853881836,
+      "eval_runtime": 23.8293,
+      "eval_samples_per_second": 105.836,
+      "eval_steps_per_second": 6.63,
+      "step": 29500
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.3471e-05,
+      "loss": 3.5452,
+      "step": 29600
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.3437666666666668e-05,
+      "loss": 3.5542,
+      "step": 29700
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.3404333333333332e-05,
+      "loss": 3.5642,
+      "step": 29800
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.3371e-05,
+      "loss": 3.5566,
+      "step": 29900
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.3337666666666665e-05,
+      "loss": 3.5709,
+      "step": 30000
+    },
+    {
+      "epoch": 2.78,
+      "eval_loss": 3.5953872203826904,
+      "eval_runtime": 23.8304,
+      "eval_samples_per_second": 105.831,
+      "eval_steps_per_second": 6.63,
+      "step": 30000
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.3304333333333333e-05,
+      "loss": 3.5628,
+      "step": 30100
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.3270999999999998e-05,
+      "loss": 3.5695,
+      "step": 30200
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.3237666666666666e-05,
+      "loss": 3.5499,
+      "step": 30300
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.3204333333333334e-05,
+      "loss": 3.5622,
+      "step": 30400
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.3171e-05,
+      "loss": 3.5546,
+      "step": 30500
+    },
+    {
+      "epoch": 2.83,
+      "eval_loss": 3.5941429138183594,
+      "eval_runtime": 23.8009,
+      "eval_samples_per_second": 105.963,
+      "eval_steps_per_second": 6.638,
+      "step": 30500
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.3137666666666667e-05,
+      "loss": 3.5545,
+      "step": 30600
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.3104333333333332e-05,
+      "loss": 3.5677,
+      "step": 30700
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.3071e-05,
+      "loss": 3.5301,
+      "step": 30800
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.3037666666666665e-05,
+      "loss": 3.5521,
+      "step": 30900
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.3004333333333333e-05,
+      "loss": 3.5525,
+      "step": 31000
+    },
+    {
+      "epoch": 2.87,
+      "eval_loss": 3.5951895713806152,
+      "eval_runtime": 23.8131,
+      "eval_samples_per_second": 105.908,
+      "eval_steps_per_second": 6.635,
+      "step": 31000
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.2971e-05,
+      "loss": 3.5602,
+      "step": 31100
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2.2937666666666666e-05,
+      "loss": 3.5652,
+      "step": 31200
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.2904333333333334e-05,
+      "loss": 3.5511,
+      "step": 31300
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.2871333333333336e-05,
+      "loss": 3.5592,
+      "step": 31400
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2.2838e-05,
+      "loss": 3.5603,
+      "step": 31500
+    },
+    {
+      "epoch": 2.92,
+      "eval_loss": 3.597166061401367,
+      "eval_runtime": 23.7699,
+      "eval_samples_per_second": 106.1,
+      "eval_steps_per_second": 6.647,
+      "step": 31500
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.280466666666667e-05,
+      "loss": 3.5604,
+      "step": 31600
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.2771333333333333e-05,
+      "loss": 3.5652,
+      "step": 31700
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 2.2738e-05,
+      "loss": 3.5589,
+      "step": 31800
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.2704666666666666e-05,
+      "loss": 3.5469,
+      "step": 31900
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.2671666666666665e-05,
+      "loss": 3.5572,
+      "step": 32000
+    },
+    {
+      "epoch": 2.97,
+      "eval_loss": 3.5946662425994873,
+      "eval_runtime": 23.7852,
+      "eval_samples_per_second": 106.032,
+      "eval_steps_per_second": 6.643,
+      "step": 32000
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.2638333333333333e-05,
+      "loss": 3.5633,
+      "step": 32100
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.2604999999999998e-05,
+      "loss": 3.5561,
+      "step": 32200
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.2571666666666666e-05,
+      "loss": 3.5372,
+      "step": 32300
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.2538333333333334e-05,
+      "loss": 3.565,
+      "step": 32400
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.2505e-05,
+      "loss": 3.5106,
+      "step": 32500
+    },
+    {
+      "epoch": 3.01,
+      "eval_loss": 3.5952203273773193,
+      "eval_runtime": 23.7897,
+      "eval_samples_per_second": 106.012,
+      "eval_steps_per_second": 6.642,
+      "step": 32500
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.2471666666666667e-05,
+      "loss": 3.5149,
+      "step": 32600
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.243866666666667e-05,
+      "loss": 3.5158,
+      "step": 32700
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 2.2405333333333334e-05,
+      "loss": 3.5136,
+      "step": 32800
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 2.2372000000000002e-05,
+      "loss": 3.5401,
+      "step": 32900
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 2.2338666666666667e-05,
+      "loss": 3.5142,
+      "step": 33000
+    },
+    {
+      "epoch": 3.06,
+      "eval_loss": 3.5936639308929443,
+      "eval_runtime": 23.7767,
+      "eval_samples_per_second": 106.07,
+      "eval_steps_per_second": 6.645,
+      "step": 33000
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 2.2305333333333335e-05,
+      "loss": 3.5154,
+      "step": 33100
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 2.2272e-05,
+      "loss": 3.5303,
+      "step": 33200
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 2.2238666666666668e-05,
+      "loss": 3.5124,
+      "step": 33300
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 2.2205333333333336e-05,
+      "loss": 3.5242,
+      "step": 33400
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 2.2172e-05,
+      "loss": 3.506,
+      "step": 33500
+    },
+    {
+      "epoch": 3.11,
+      "eval_loss": 3.5964934825897217,
+      "eval_runtime": 23.7863,
+      "eval_samples_per_second": 106.028,
+      "eval_steps_per_second": 6.642,
+      "step": 33500
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 2.213866666666667e-05,
+      "loss": 3.5077,
+      "step": 33600
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 2.2105333333333333e-05,
+      "loss": 3.501,
+      "step": 33700
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 2.2072e-05,
+      "loss": 3.512,
+      "step": 33800
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 2.2038666666666666e-05,
+      "loss": 3.5373,
+      "step": 33900
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 2.2005333333333334e-05,
+      "loss": 3.515,
+      "step": 34000
+    },
+    {
+      "epoch": 3.15,
+      "eval_loss": 3.5931718349456787,
+      "eval_runtime": 23.7854,
+      "eval_samples_per_second": 106.031,
+      "eval_steps_per_second": 6.643,
+      "step": 34000
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 2.1972000000000002e-05,
+      "loss": 3.5225,
+      "step": 34100
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 2.1938666666666667e-05,
+      "loss": 3.4866,
+      "step": 34200
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 2.1905333333333335e-05,
+      "loss": 3.526,
+      "step": 34300
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 2.1872e-05,
+      "loss": 3.5036,
+      "step": 34400
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 2.1838666666666668e-05,
+      "loss": 3.5247,
+      "step": 34500
+    },
+    {
+      "epoch": 3.2,
+      "eval_loss": 3.5951099395751953,
+      "eval_runtime": 23.7873,
+      "eval_samples_per_second": 106.023,
+      "eval_steps_per_second": 6.642,
+      "step": 34500
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 2.1805333333333333e-05,
+      "loss": 3.528,
+      "step": 34600
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 2.1772e-05,
+      "loss": 3.5123,
+      "step": 34700
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 2.1738666666666666e-05,
+      "loss": 3.5277,
+      "step": 34800
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 2.1705333333333334e-05,
+      "loss": 3.522,
+      "step": 34900
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 2.1672000000000002e-05,
+      "loss": 3.5384,
+      "step": 35000
+    },
+    {
+      "epoch": 3.24,
+      "eval_loss": 3.591693878173828,
+      "eval_runtime": 23.7823,
+      "eval_samples_per_second": 106.045,
+      "eval_steps_per_second": 6.644,
+      "step": 35000
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 2.1638666666666667e-05,
+      "loss": 3.526,
+      "step": 35100
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 2.1605333333333335e-05,
+      "loss": 3.5137,
+      "step": 35200
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 2.1572e-05,
+      "loss": 3.525,
+      "step": 35300
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 2.1538666666666668e-05,
+      "loss": 3.519,
+      "step": 35400
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 2.1505666666666666e-05,
+      "loss": 3.5165,
+      "step": 35500
+    },
+    {
+      "epoch": 3.29,
+      "eval_loss": 3.5887227058410645,
+      "eval_runtime": 23.8133,
+      "eval_samples_per_second": 105.907,
+      "eval_steps_per_second": 6.635,
+      "step": 35500
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 2.147233333333333e-05,
+      "loss": 3.5157,
+      "step": 35600
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 2.1439e-05,
+      "loss": 3.5363,
+      "step": 35700
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 2.1405666666666664e-05,
+      "loss": 3.5146,
+      "step": 35800
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 2.1372333333333332e-05,
+      "loss": 3.5313,
+      "step": 35900
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 2.1339e-05,
+      "loss": 3.5187,
+      "step": 36000
+    },
+    {
+      "epoch": 3.34,
+      "eval_loss": 3.5865957736968994,
+      "eval_runtime": 23.7698,
+      "eval_samples_per_second": 106.101,
+      "eval_steps_per_second": 6.647,
+      "step": 36000
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 2.1305666666666668e-05,
+      "loss": 3.5181,
+      "step": 36100
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 2.1272333333333336e-05,
+      "loss": 3.5087,
+      "step": 36200
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 2.1239e-05,
+      "loss": 3.5211,
+      "step": 36300
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 2.120566666666667e-05,
+      "loss": 3.5419,
+      "step": 36400
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 2.1172333333333334e-05,
+      "loss": 3.5097,
+      "step": 36500
+    },
+    {
+      "epoch": 3.38,
+      "eval_loss": 3.5894858837127686,
+      "eval_runtime": 23.8034,
+      "eval_samples_per_second": 105.951,
+      "eval_steps_per_second": 6.638,
+      "step": 36500
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 2.1139000000000002e-05,
+      "loss": 3.525,
+      "step": 36600
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 2.110566666666667e-05,
+      "loss": 3.5083,
+      "step": 36700
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 2.1072333333333335e-05,
+      "loss": 3.4998,
+      "step": 36800
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 2.1039000000000003e-05,
+      "loss": 3.5201,
+      "step": 36900
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 2.1005666666666668e-05,
+      "loss": 3.5136,
+      "step": 37000
+    },
+    {
+      "epoch": 3.43,
+      "eval_loss": 3.587806224822998,
+      "eval_runtime": 23.7779,
+      "eval_samples_per_second": 106.065,
+      "eval_steps_per_second": 6.645,
+      "step": 37000
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 2.0972333333333336e-05,
+      "loss": 3.5331,
+      "step": 37100
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 2.0939e-05,
+      "loss": 3.5299,
+      "step": 37200
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 2.090566666666667e-05,
+      "loss": 3.5274,
+      "step": 37300
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 2.0872333333333337e-05,
+      "loss": 3.5212,
+      "step": 37400
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 2.0839e-05,
+      "loss": 3.5095,
+      "step": 37500
+    },
+    {
+      "epoch": 3.48,
+      "eval_loss": 3.5838775634765625,
+      "eval_runtime": 23.7902,
+      "eval_samples_per_second": 106.01,
+      "eval_steps_per_second": 6.641,
+      "step": 37500
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 2.080566666666667e-05,
+      "loss": 3.5125,
+      "step": 37600
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 2.0772333333333334e-05,
+      "loss": 3.4908,
+      "step": 37700
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 2.0739000000000003e-05,
+      "loss": 3.5018,
+      "step": 37800
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 2.0705666666666667e-05,
+      "loss": 3.5034,
+      "step": 37900
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 2.0672333333333335e-05,
+      "loss": 3.5226,
+      "step": 38000
+    },
+    {
+      "epoch": 3.52,
+      "eval_loss": 3.5859289169311523,
+      "eval_runtime": 23.7148,
+      "eval_samples_per_second": 106.347,
+      "eval_steps_per_second": 6.663,
+      "step": 38000
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 2.0639e-05,
+      "loss": 3.5247,
+      "step": 38100
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 2.0605666666666668e-05,
+      "loss": 3.5342,
+      "step": 38200
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 2.0572333333333336e-05,
+      "loss": 3.5144,
+      "step": 38300
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 2.0539e-05,
+      "loss": 3.5106,
+      "step": 38400
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 2.050566666666667e-05,
+      "loss": 3.5277,
+      "step": 38500
+    },
+    {
+      "epoch": 3.57,
+      "eval_loss": 3.5827441215515137,
+      "eval_runtime": 23.8131,
+      "eval_samples_per_second": 105.908,
+      "eval_steps_per_second": 6.635,
+      "step": 38500
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 2.0472333333333334e-05,
+      "loss": 3.4914,
+      "step": 38600
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 2.0439000000000002e-05,
+      "loss": 3.5394,
+      "step": 38700
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 2.0405666666666667e-05,
+      "loss": 3.5314,
+      "step": 38800
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 2.0372333333333335e-05,
+      "loss": 3.5096,
+      "step": 38900
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 2.0339000000000003e-05,
+      "loss": 3.4959,
+      "step": 39000
+    },
+    {
+      "epoch": 3.62,
+      "eval_loss": 3.5846378803253174,
+      "eval_runtime": 23.7913,
+      "eval_samples_per_second": 106.005,
+      "eval_steps_per_second": 6.641,
+      "step": 39000
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 2.0305666666666668e-05,
+      "loss": 3.51,
+      "step": 39100
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 2.0272333333333336e-05,
+      "loss": 3.507,
+      "step": 39200
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 2.0239e-05,
+      "loss": 3.5048,
+      "step": 39300
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 2.020566666666667e-05,
+      "loss": 3.525,
+      "step": 39400
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 2.0172333333333333e-05,
+      "loss": 3.5003,
+      "step": 39500
+    },
+    {
+      "epoch": 3.66,
+      "eval_loss": 3.5823278427124023,
+      "eval_runtime": 23.8021,
+      "eval_samples_per_second": 105.957,
+      "eval_steps_per_second": 6.638,
+      "step": 39500
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 2.0139e-05,
+      "loss": 3.4986,
+      "step": 39600
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 2.0105666666666666e-05,
+      "loss": 3.516,
+      "step": 39700
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 2.0072333333333334e-05,
+      "loss": 3.5114,
+      "step": 39800
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 2.0039000000000002e-05,
+      "loss": 3.5071,
+      "step": 39900
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 2.0005666666666667e-05,
+      "loss": 3.5095,
+      "step": 40000
+    },
+    {
+      "epoch": 3.71,
+      "eval_loss": 3.581997871398926,
+      "eval_runtime": 23.772,
+      "eval_samples_per_second": 106.091,
+      "eval_steps_per_second": 6.646,
+      "step": 40000
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1.9972333333333335e-05,
+      "loss": 3.5524,
+      "step": 40100
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.9939e-05,
+      "loss": 3.5158,
+      "step": 40200
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.9905666666666668e-05,
+      "loss": 3.4829,
+      "step": 40300
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.9872333333333333e-05,
+      "loss": 3.498,
+      "step": 40400
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.9839e-05,
+      "loss": 3.4814,
+      "step": 40500
+    },
+    {
+      "epoch": 3.75,
+      "eval_loss": 3.5854456424713135,
+      "eval_runtime": 23.7663,
+      "eval_samples_per_second": 106.117,
+      "eval_steps_per_second": 6.648,
+      "step": 40500
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 1.980566666666667e-05,
+      "loss": 3.5146,
+      "step": 40600
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.9772333333333334e-05,
+      "loss": 3.5109,
+      "step": 40700
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1.9739000000000002e-05,
+      "loss": 3.5271,
+      "step": 40800
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1.9705666666666667e-05,
+      "loss": 3.5158,
+      "step": 40900
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 1.9672333333333335e-05,
+      "loss": 3.5173,
+      "step": 41000
+    },
+    {
+      "epoch": 3.8,
+      "eval_loss": 3.5796046257019043,
+      "eval_runtime": 23.7829,
+      "eval_samples_per_second": 106.043,
+      "eval_steps_per_second": 6.643,
+      "step": 41000
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 1.9639e-05,
+      "loss": 3.5034,
+      "step": 41100
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 1.9605666666666668e-05,
+      "loss": 3.5118,
+      "step": 41200
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 1.9572333333333336e-05,
+      "loss": 3.5265,
+      "step": 41300
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 1.9539e-05,
+      "loss": 3.513,
+      "step": 41400
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 1.950566666666667e-05,
+      "loss": 3.4968,
+      "step": 41500
+    },
+    {
+      "epoch": 3.85,
+      "eval_loss": 3.5810136795043945,
+      "eval_runtime": 23.9055,
+      "eval_samples_per_second": 105.499,
+      "eval_steps_per_second": 6.609,
+      "step": 41500
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 1.9472333333333333e-05,
+      "loss": 3.5123,
+      "step": 41600
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1.9439e-05,
+      "loss": 3.495,
+      "step": 41700
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1.9405666666666666e-05,
+      "loss": 3.5034,
+      "step": 41800
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 1.9372333333333334e-05,
+      "loss": 3.5135,
+      "step": 41900
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1.9339e-05,
+      "loss": 3.5183,
+      "step": 42000
+    },
+    {
+      "epoch": 3.89,
+      "eval_loss": 3.578335762023926,
+      "eval_runtime": 23.7695,
+      "eval_samples_per_second": 106.102,
+      "eval_steps_per_second": 6.647,
+      "step": 42000
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 1.9305666666666667e-05,
+      "loss": 3.4925,
+      "step": 42100
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.9272333333333335e-05,
+      "loss": 3.5166,
+      "step": 42200
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1.9239e-05,
+      "loss": 3.4917,
+      "step": 42300
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 1.9205666666666668e-05,
+      "loss": 3.4986,
+      "step": 42400
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 1.9172333333333333e-05,
+      "loss": 3.512,
+      "step": 42500
+    },
+    {
+      "epoch": 3.94,
+      "eval_loss": 3.5784130096435547,
+      "eval_runtime": 23.7738,
+      "eval_samples_per_second": 106.083,
+      "eval_steps_per_second": 6.646,
+      "step": 42500
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 1.9139e-05,
+      "loss": 3.5175,
+      "step": 42600
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 1.9105666666666666e-05,
+      "loss": 3.5299,
+      "step": 42700
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1.9072333333333334e-05,
+      "loss": 3.4949,
+      "step": 42800
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 1.9039000000000002e-05,
+      "loss": 3.5017,
+      "step": 42900
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 1.9005666666666667e-05,
+      "loss": 3.5069,
+      "step": 43000
+    },
+    {
+      "epoch": 3.99,
+      "eval_loss": 3.577517032623291,
+      "eval_runtime": 23.8132,
+      "eval_samples_per_second": 105.908,
+      "eval_steps_per_second": 6.635,
+      "step": 43000
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1.8972333333333335e-05,
+      "loss": 3.5017,
+      "step": 43100
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1.8939e-05,
+      "loss": 3.517,
+      "step": 43200
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1.8905666666666668e-05,
+      "loss": 3.4586,
+      "step": 43300
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 1.8872333333333332e-05,
+      "loss": 3.4568,
+      "step": 43400
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 1.8839e-05,
+      "loss": 3.5014,
+      "step": 43500
+    },
+    {
+      "epoch": 4.03,
+      "eval_loss": 3.581881046295166,
+      "eval_runtime": 23.7711,
+      "eval_samples_per_second": 106.095,
+      "eval_steps_per_second": 6.647,
+      "step": 43500
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 1.880566666666667e-05,
+      "loss": 3.4791,
+      "step": 43600
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 1.8772333333333333e-05,
+      "loss": 3.4744,
+      "step": 43700
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 1.8739e-05,
+      "loss": 3.4972,
+      "step": 43800
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 1.8705666666666666e-05,
+      "loss": 3.4811,
+      "step": 43900
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 1.8672333333333334e-05,
+      "loss": 3.4787,
+      "step": 44000
+    },
+    {
+      "epoch": 4.08,
+      "eval_loss": 3.5835728645324707,
+      "eval_runtime": 23.7828,
+      "eval_samples_per_second": 106.043,
+      "eval_steps_per_second": 6.643,
+      "step": 44000
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 1.8639e-05,
+      "loss": 3.4686,
+      "step": 44100
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 1.8605666666666667e-05,
+      "loss": 3.4751,
+      "step": 44200
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 1.8572333333333332e-05,
+      "loss": 3.4829,
+      "step": 44300
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 1.8539e-05,
+      "loss": 3.4669,
+      "step": 44400
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 1.8505666666666668e-05,
+      "loss": 3.4625,
+      "step": 44500
+    },
+    {
+      "epoch": 4.12,
+      "eval_loss": 3.5787713527679443,
+      "eval_runtime": 23.7907,
+      "eval_samples_per_second": 106.008,
+      "eval_steps_per_second": 6.641,
+      "step": 44500
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 1.8472333333333333e-05,
+      "loss": 3.4731,
+      "step": 44600
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 1.8439e-05,
+      "loss": 3.4799,
+      "step": 44700
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.8405666666666666e-05,
+      "loss": 3.5001,
+      "step": 44800
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.8372333333333334e-05,
+      "loss": 3.4744,
+      "step": 44900
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.8339e-05,
+      "loss": 3.4902,
+      "step": 45000
+    },
+    {
+      "epoch": 4.17,
+      "eval_loss": 3.5783867835998535,
+      "eval_runtime": 23.7793,
+      "eval_samples_per_second": 106.059,
+      "eval_steps_per_second": 6.644,
+      "step": 45000
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1.8305666666666667e-05,
+      "loss": 3.4946,
+      "step": 45100
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 1.8272333333333335e-05,
+      "loss": 3.4695,
+      "step": 45200
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1.8239e-05,
+      "loss": 3.4579,
+      "step": 45300
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 1.8205666666666667e-05,
+      "loss": 3.4617,
+      "step": 45400
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 1.8172333333333332e-05,
+      "loss": 3.4927,
+      "step": 45500
+    },
+    {
+      "epoch": 4.22,
+      "eval_loss": 3.577256202697754,
+      "eval_runtime": 23.8012,
+      "eval_samples_per_second": 105.961,
+      "eval_steps_per_second": 6.638,
+      "step": 45500
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 1.8139e-05,
+      "loss": 3.4659,
+      "step": 45600
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 1.8105666666666665e-05,
+      "loss": 3.4958,
+      "step": 45700
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 1.8072333333333333e-05,
+      "loss": 3.4759,
+      "step": 45800
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 1.8038999999999998e-05,
+      "loss": 3.4523,
+      "step": 45900
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 1.8005666666666666e-05,
+      "loss": 3.4813,
+      "step": 46000
+    },
+    {
+      "epoch": 4.26,
+      "eval_loss": 3.576850652694702,
+      "eval_runtime": 23.7844,
+      "eval_samples_per_second": 106.036,
+      "eval_steps_per_second": 6.643,
+      "step": 46000
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 1.7972333333333334e-05,
+      "loss": 3.4841,
+      "step": 46100
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 1.7939e-05,
+      "loss": 3.472,
+      "step": 46200
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 1.7905666666666667e-05,
+      "loss": 3.4769,
+      "step": 46300
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 1.787233333333333e-05,
+      "loss": 3.4783,
+      "step": 46400
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 1.7839e-05,
+      "loss": 3.4637,
+      "step": 46500
+    },
+    {
+      "epoch": 4.31,
+      "eval_loss": 3.576143264770508,
+      "eval_runtime": 23.8068,
+      "eval_samples_per_second": 105.936,
+      "eval_steps_per_second": 6.637,
+      "step": 46500
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 1.7805666666666665e-05,
+      "loss": 3.4812,
+      "step": 46600
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 1.7772333333333333e-05,
+      "loss": 3.476,
+      "step": 46700
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 1.7739e-05,
+      "loss": 3.4689,
+      "step": 46800
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 1.7705666666666665e-05,
+      "loss": 3.4719,
+      "step": 46900
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 1.7672333333333334e-05,
+      "loss": 3.4731,
+      "step": 47000
+    },
+    {
+      "epoch": 4.36,
+      "eval_loss": 3.577054262161255,
+      "eval_runtime": 23.719,
+      "eval_samples_per_second": 106.328,
+      "eval_steps_per_second": 6.661,
+      "step": 47000
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 1.7639e-05,
+      "loss": 3.4923,
+      "step": 47100
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 1.7605666666666666e-05,
+      "loss": 3.4679,
+      "step": 47200
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 1.757233333333333e-05,
+      "loss": 3.4742,
+      "step": 47300
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 1.7539e-05,
+      "loss": 3.4628,
+      "step": 47400
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 1.7505666666666667e-05,
+      "loss": 3.4856,
+      "step": 47500
+    },
+    {
+      "epoch": 4.4,
+      "eval_loss": 3.578598976135254,
+      "eval_runtime": 23.7952,
+      "eval_samples_per_second": 105.988,
+      "eval_steps_per_second": 6.64,
+      "step": 47500
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 1.7472333333333332e-05,
+      "loss": 3.488,
+      "step": 47600
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 1.7439e-05,
+      "loss": 3.4711,
+      "step": 47700
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 1.7405666666666665e-05,
+      "loss": 3.4814,
+      "step": 47800
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 1.7372333333333333e-05,
+      "loss": 3.4768,
+      "step": 47900
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 1.7338999999999998e-05,
+      "loss": 3.4579,
+      "step": 48000
+    },
+    {
+      "epoch": 4.45,
+      "eval_loss": 3.5790109634399414,
+      "eval_runtime": 23.779,
+      "eval_samples_per_second": 106.06,
+      "eval_steps_per_second": 6.645,
+      "step": 48000
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 1.7305666666666666e-05,
+      "loss": 3.4752,
+      "step": 48100
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 1.727233333333333e-05,
+      "loss": 3.4836,
+      "step": 48200
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 1.7239333333333333e-05,
+      "loss": 3.4835,
+      "step": 48300
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 1.7206e-05,
+      "loss": 3.4798,
+      "step": 48400
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 1.717266666666667e-05,
+      "loss": 3.5032,
+      "step": 48500
+    },
+    {
+      "epoch": 4.5,
+      "eval_loss": 3.573821544647217,
+      "eval_runtime": 23.7995,
+      "eval_samples_per_second": 105.969,
+      "eval_steps_per_second": 6.639,
+      "step": 48500
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 1.7139333333333334e-05,
+      "loss": 3.4885,
+      "step": 48600
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 1.7106000000000002e-05,
+      "loss": 3.4805,
+      "step": 48700
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 1.7072666666666667e-05,
+      "loss": 3.4588,
+      "step": 48800
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 1.7039333333333335e-05,
+      "loss": 3.4742,
+      "step": 48900
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 1.7006e-05,
+      "loss": 3.4826,
+      "step": 49000
+    },
+    {
+      "epoch": 4.54,
+      "eval_loss": 3.574939012527466,
+      "eval_runtime": 23.7807,
+      "eval_samples_per_second": 106.052,
+      "eval_steps_per_second": 6.644,
+      "step": 49000
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 1.6972666666666668e-05,
+      "loss": 3.4866,
+      "step": 49100
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 1.6939333333333332e-05,
+      "loss": 3.4699,
+      "step": 49200
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 1.6906e-05,
+      "loss": 3.4643,
+      "step": 49300
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 1.687266666666667e-05,
+      "loss": 3.4771,
+      "step": 49400
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 1.6839333333333333e-05,
+      "loss": 3.4709,
+      "step": 49500
+    },
+    {
+      "epoch": 4.59,
+      "eval_loss": 3.574620246887207,
+      "eval_runtime": 23.7928,
+      "eval_samples_per_second": 105.999,
+      "eval_steps_per_second": 6.641,
+      "step": 49500
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1.6806e-05,
+      "loss": 3.4724,
+      "step": 49600
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 1.6772666666666666e-05,
+      "loss": 3.4675,
+      "step": 49700
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 1.6739333333333334e-05,
+      "loss": 3.4958,
+      "step": 49800
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 1.6706e-05,
+      "loss": 3.4738,
+      "step": 49900
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 1.6672666666666667e-05,
+      "loss": 3.4916,
+      "step": 50000
+    },
+    {
+      "epoch": 4.63,
+      "eval_loss": 3.574481964111328,
+      "eval_runtime": 23.7895,
+      "eval_samples_per_second": 106.013,
+      "eval_steps_per_second": 6.642,
+      "step": 50000
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 1.6639333333333335e-05,
+      "loss": 3.4786,
+      "step": 50100
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 1.6606e-05,
+      "loss": 3.4727,
+      "step": 50200
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 1.6572666666666668e-05,
+      "loss": 3.4872,
+      "step": 50300
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 1.6539333333333333e-05,
+      "loss": 3.4945,
+      "step": 50400
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 1.6506e-05,
+      "loss": 3.4715,
+      "step": 50500
+    },
+    {
+      "epoch": 4.68,
+      "eval_loss": 3.5705904960632324,
+      "eval_runtime": 24.2293,
+      "eval_samples_per_second": 104.089,
+      "eval_steps_per_second": 6.521,
+      "step": 50500
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 1.6472666666666666e-05,
+      "loss": 3.4569,
+      "step": 50600
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1.6439333333333334e-05,
+      "loss": 3.4647,
+      "step": 50700
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1.6406333333333336e-05,
+      "loss": 3.4686,
+      "step": 50800
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.6373e-05,
+      "loss": 3.4876,
+      "step": 50900
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.633966666666667e-05,
+      "loss": 3.4926,
+      "step": 51000
+    },
+    {
+      "epoch": 4.73,
+      "eval_loss": 3.572871685028076,
+      "eval_runtime": 23.7903,
+      "eval_samples_per_second": 106.009,
+      "eval_steps_per_second": 6.641,
+      "step": 51000
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.6306333333333337e-05,
+      "loss": 3.4759,
+      "step": 51100
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1.6273e-05,
+      "loss": 3.4418,
+      "step": 51200
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 1.623966666666667e-05,
+      "loss": 3.4926,
+      "step": 51300
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 1.6206333333333334e-05,
+      "loss": 3.4671,
+      "step": 51400
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 1.6173000000000003e-05,
+      "loss": 3.4974,
+      "step": 51500
+    },
+    {
+      "epoch": 4.77,
+      "eval_loss": 3.5724916458129883,
+      "eval_runtime": 23.8031,
+      "eval_samples_per_second": 105.952,
+      "eval_steps_per_second": 6.638,
+      "step": 51500
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1.6139666666666667e-05,
+      "loss": 3.4833,
+      "step": 51600
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 1.6106333333333335e-05,
+      "loss": 3.4603,
+      "step": 51700
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 1.6073e-05,
+      "loss": 3.4809,
+      "step": 51800
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 1.6039666666666668e-05,
+      "loss": 3.4697,
+      "step": 51900
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.6006333333333336e-05,
+      "loss": 3.4796,
+      "step": 52000
+    },
+    {
+      "epoch": 4.82,
+      "eval_loss": 3.568263292312622,
+      "eval_runtime": 23.7971,
+      "eval_samples_per_second": 105.979,
+      "eval_steps_per_second": 6.639,
+      "step": 52000
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.5973e-05,
+      "loss": 3.4968,
+      "step": 52100
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 1.593966666666667e-05,
+      "loss": 3.469,
+      "step": 52200
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 1.5906333333333334e-05,
+      "loss": 3.4832,
+      "step": 52300
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 1.5873000000000002e-05,
+      "loss": 3.4819,
+      "step": 52400
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 1.5839666666666667e-05,
+      "loss": 3.4817,
+      "step": 52500
+    },
+    {
+      "epoch": 4.87,
+      "eval_loss": 3.5706946849823,
+      "eval_runtime": 23.8071,
+      "eval_samples_per_second": 105.935,
+      "eval_steps_per_second": 6.637,
+      "step": 52500
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 1.5806333333333335e-05,
+      "loss": 3.4747,
+      "step": 52600
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 1.5773000000000003e-05,
+      "loss": 3.4643,
+      "step": 52700
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 1.5739666666666668e-05,
+      "loss": 3.4446,
+      "step": 52800
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 1.5706333333333336e-05,
+      "loss": 3.4721,
+      "step": 52900
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 1.5673e-05,
+      "loss": 3.4683,
+      "step": 53000
+    },
+    {
+      "epoch": 4.91,
+      "eval_loss": 3.5720579624176025,
+      "eval_runtime": 23.787,
+      "eval_samples_per_second": 106.024,
+      "eval_steps_per_second": 6.642,
+      "step": 53000
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 1.563966666666667e-05,
+      "loss": 3.4983,
+      "step": 53100
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 1.5606333333333333e-05,
+      "loss": 3.4543,
+      "step": 53200
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 1.5573e-05,
+      "loss": 3.4886,
+      "step": 53300
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 1.553966666666667e-05,
+      "loss": 3.4598,
+      "step": 53400
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 1.5506333333333334e-05,
+      "loss": 3.4986,
+      "step": 53500
+    },
+    {
+      "epoch": 4.96,
+      "eval_loss": 3.5689127445220947,
+      "eval_runtime": 23.9712,
+      "eval_samples_per_second": 105.21,
+      "eval_steps_per_second": 6.591,
+      "step": 53500
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.5473000000000002e-05,
+      "loss": 3.4927,
+      "step": 53600
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 1.5439666666666667e-05,
+      "loss": 3.4767,
+      "step": 53700
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 1.5406333333333335e-05,
+      "loss": 3.4679,
+      "step": 53800
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 1.5373333333333334e-05,
+      "loss": 3.4892,
+      "step": 53900
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 1.534e-05,
+      "loss": 3.4763,
+      "step": 54000
+    },
+    {
+      "epoch": 5.01,
+      "eval_loss": 3.571553945541382,
+      "eval_runtime": 23.7751,
+      "eval_samples_per_second": 106.077,
+      "eval_steps_per_second": 6.646,
+      "step": 54000
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 1.5306666666666667e-05,
+      "loss": 3.448,
+      "step": 54100
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 1.527333333333333e-05,
+      "loss": 3.4476,
+      "step": 54200
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 1.524e-05,
+      "loss": 3.452,
+      "step": 54300
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 1.5206666666666668e-05,
+      "loss": 3.4499,
+      "step": 54400
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 1.5173333333333334e-05,
+      "loss": 3.4668,
+      "step": 54500
+    },
+    {
+      "epoch": 5.05,
+      "eval_loss": 3.5700042247772217,
+      "eval_runtime": 23.7777,
+      "eval_samples_per_second": 106.066,
+      "eval_steps_per_second": 6.645,
+      "step": 54500
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 1.5140000000000002e-05,
+      "loss": 3.446,
+      "step": 54600
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 1.5106666666666667e-05,
+      "loss": 3.449,
+      "step": 54700
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 1.5073333333333335e-05,
+      "loss": 3.4855,
+      "step": 54800
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 1.504e-05,
+      "loss": 3.4352,
+      "step": 54900
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 1.5006666666666668e-05,
+      "loss": 3.4274,
+      "step": 55000
+    },
+    {
+      "epoch": 5.1,
+      "eval_loss": 3.5723605155944824,
+      "eval_runtime": 23.7887,
+      "eval_samples_per_second": 106.017,
+      "eval_steps_per_second": 6.642,
+      "step": 55000
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 1.4973333333333335e-05,
+      "loss": 3.4536,
+      "step": 55100
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 1.4940000000000001e-05,
+      "loss": 3.4443,
+      "step": 55200
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 1.4906666666666667e-05,
+      "loss": 3.4531,
+      "step": 55300
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 1.4873333333333334e-05,
+      "loss": 3.4414,
+      "step": 55400
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 1.484e-05,
+      "loss": 3.4499,
+      "step": 55500
+    },
+    {
+      "epoch": 5.14,
+      "eval_loss": 3.571748971939087,
+      "eval_runtime": 23.7939,
+      "eval_samples_per_second": 105.994,
+      "eval_steps_per_second": 6.64,
+      "step": 55500
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 1.4806666666666668e-05,
+      "loss": 3.4533,
+      "step": 55600
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 1.4773333333333335e-05,
+      "loss": 3.4528,
+      "step": 55700
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 1.4740000000000001e-05,
+      "loss": 3.4617,
+      "step": 55800
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 1.4706666666666668e-05,
+      "loss": 3.454,
+      "step": 55900
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 1.4673333333333334e-05,
+      "loss": 3.4507,
+      "step": 56000
+    },
+    {
+      "epoch": 5.19,
+      "eval_loss": 3.5706114768981934,
+      "eval_runtime": 23.7225,
+      "eval_samples_per_second": 106.313,
+      "eval_steps_per_second": 6.66,
+      "step": 56000
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 1.4640333333333335e-05,
+      "loss": 3.4436,
+      "step": 56100
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 1.4607000000000001e-05,
+      "loss": 3.4584,
+      "step": 56200
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 1.4573666666666667e-05,
+      "loss": 3.4517,
+      "step": 56300
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 1.4540333333333334e-05,
+      "loss": 3.4581,
+      "step": 56400
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 1.4507e-05,
+      "loss": 3.4343,
+      "step": 56500
+    },
+    {
+      "epoch": 5.24,
+      "eval_loss": 3.5696890354156494,
+      "eval_runtime": 23.8115,
+      "eval_samples_per_second": 105.915,
+      "eval_steps_per_second": 6.635,
+      "step": 56500
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 1.4473666666666668e-05,
+      "loss": 3.4306,
+      "step": 56600
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 1.4440333333333335e-05,
+      "loss": 3.4402,
+      "step": 56700
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 1.4407000000000001e-05,
+      "loss": 3.453,
+      "step": 56800
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 1.4373666666666668e-05,
+      "loss": 3.434,
+      "step": 56900
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 1.4340333333333334e-05,
+      "loss": 3.4151,
+      "step": 57000
+    },
+    {
+      "epoch": 5.28,
+      "eval_loss": 3.5709781646728516,
+      "eval_runtime": 23.7861,
+      "eval_samples_per_second": 106.028,
+      "eval_steps_per_second": 6.643,
+      "step": 57000
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 1.4307e-05,
+      "loss": 3.4446,
+      "step": 57100
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 1.4273666666666667e-05,
+      "loss": 3.4401,
+      "step": 57200
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 1.4240333333333333e-05,
+      "loss": 3.4597,
+      "step": 57300
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 1.4207000000000001e-05,
+      "loss": 3.4369,
+      "step": 57400
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 1.4173666666666668e-05,
+      "loss": 3.4469,
+      "step": 57500
+    },
+    {
+      "epoch": 5.33,
+      "eval_loss": 3.571150779724121,
+      "eval_runtime": 23.8246,
+      "eval_samples_per_second": 105.857,
+      "eval_steps_per_second": 6.632,
+      "step": 57500
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 1.4140333333333334e-05,
+      "loss": 3.4317,
+      "step": 57600
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 1.4107e-05,
+      "loss": 3.4437,
+      "step": 57700
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 1.4073666666666667e-05,
+      "loss": 3.4488,
+      "step": 57800
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 1.4040333333333333e-05,
+      "loss": 3.4404,
+      "step": 57900
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 1.4007e-05,
+      "loss": 3.458,
+      "step": 58000
+    },
+    {
+      "epoch": 5.38,
+      "eval_loss": 3.5692391395568848,
+      "eval_runtime": 23.7837,
+      "eval_samples_per_second": 106.039,
+      "eval_steps_per_second": 6.643,
+      "step": 58000
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 1.3973666666666666e-05,
+      "loss": 3.4656,
+      "step": 58100
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 1.3940333333333334e-05,
+      "loss": 3.4639,
+      "step": 58200
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 1.3907000000000001e-05,
+      "loss": 3.4502,
+      "step": 58300
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 1.3873666666666667e-05,
+      "loss": 3.4752,
+      "step": 58400
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 1.3840333333333334e-05,
+      "loss": 3.4559,
+      "step": 58500
+    },
+    {
+      "epoch": 5.42,
+      "eval_loss": 3.567958116531372,
+      "eval_runtime": 23.7984,
+      "eval_samples_per_second": 105.974,
+      "eval_steps_per_second": 6.639,
+      "step": 58500
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 1.3807e-05,
+      "loss": 3.4518,
+      "step": 58600
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 1.3774e-05,
+      "loss": 3.418,
+      "step": 58700
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 1.3740666666666667e-05,
+      "loss": 3.438,
+      "step": 58800
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 1.3707333333333333e-05,
+      "loss": 3.4315,
+      "step": 58900
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 1.3674e-05,
+      "loss": 3.4354,
+      "step": 59000
+    },
+    {
+      "epoch": 5.47,
+      "eval_loss": 3.56830096244812,
+      "eval_runtime": 23.7688,
+      "eval_samples_per_second": 106.106,
+      "eval_steps_per_second": 6.647,
+      "step": 59000
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 1.3640666666666666e-05,
+      "loss": 3.4574,
+      "step": 59100
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 1.3607333333333334e-05,
+      "loss": 3.4327,
+      "step": 59200
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 1.3574e-05,
+      "loss": 3.4542,
+      "step": 59300
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 1.3540666666666667e-05,
+      "loss": 3.4424,
+      "step": 59400
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 1.3507333333333334e-05,
+      "loss": 3.4479,
+      "step": 59500
+    },
+    {
+      "epoch": 5.52,
+      "eval_loss": 3.570251703262329,
+      "eval_runtime": 23.7862,
+      "eval_samples_per_second": 106.028,
+      "eval_steps_per_second": 6.643,
+      "step": 59500
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 1.3474e-05,
+      "loss": 3.4637,
+      "step": 59600
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 1.3440666666666667e-05,
+      "loss": 3.4584,
+      "step": 59700
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 1.3407333333333333e-05,
+      "loss": 3.4347,
+      "step": 59800
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 1.3374e-05,
+      "loss": 3.4526,
+      "step": 59900
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 1.3340666666666667e-05,
+      "loss": 3.4627,
+      "step": 60000
+    },
+    {
+      "epoch": 5.56,
+      "eval_loss": 3.5677947998046875,
+      "eval_runtime": 23.7816,
+      "eval_samples_per_second": 106.049,
+      "eval_steps_per_second": 6.644,
+      "step": 60000
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 1.3307333333333334e-05,
+      "loss": 3.4423,
+      "step": 60100
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 1.3274e-05,
+      "loss": 3.4469,
+      "step": 60200
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 1.3240666666666667e-05,
+      "loss": 3.4423,
+      "step": 60300
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 1.3207333333333333e-05,
+      "loss": 3.4322,
+      "step": 60400
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 1.3174e-05,
+      "loss": 3.4478,
+      "step": 60500
+    },
+    {
+      "epoch": 5.61,
+      "eval_loss": 3.565863847732544,
+      "eval_runtime": 23.8989,
+      "eval_samples_per_second": 105.528,
+      "eval_steps_per_second": 6.611,
+      "step": 60500
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 1.3140666666666666e-05,
+      "loss": 3.4294,
+      "step": 60600
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 1.3107333333333332e-05,
+      "loss": 3.4547,
+      "step": 60700
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 1.3074e-05,
+      "loss": 3.4605,
+      "step": 60800
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 1.3040666666666667e-05,
+      "loss": 3.4517,
+      "step": 60900
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 1.3007333333333333e-05,
+      "loss": 3.4645,
+      "step": 61000
+    },
+    {
+      "epoch": 5.65,
+      "eval_loss": 3.567469596862793,
+      "eval_runtime": 23.816,
+      "eval_samples_per_second": 105.895,
+      "eval_steps_per_second": 6.634,
+      "step": 61000
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 1.2974e-05,
+      "loss": 3.4338,
+      "step": 61100
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 1.2940666666666666e-05,
+      "loss": 3.4675,
+      "step": 61200
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 1.2907333333333333e-05,
+      "loss": 3.4542,
+      "step": 61300
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 1.2873999999999999e-05,
+      "loss": 3.46,
+      "step": 61400
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 1.2840666666666667e-05,
+      "loss": 3.4658,
+      "step": 61500
+    },
+    {
+      "epoch": 5.7,
+      "eval_loss": 3.5665736198425293,
+      "eval_runtime": 23.7752,
+      "eval_samples_per_second": 106.077,
+      "eval_steps_per_second": 6.646,
+      "step": 61500
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 1.2807333333333334e-05,
+      "loss": 3.4398,
+      "step": 61600
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 1.2774e-05,
+      "loss": 3.4504,
+      "step": 61700
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 1.2740666666666666e-05,
+      "loss": 3.4197,
+      "step": 61800
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 1.2707333333333333e-05,
+      "loss": 3.4719,
+      "step": 61900
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 1.2674e-05,
+      "loss": 3.4657,
+      "step": 62000
+    },
+    {
+      "epoch": 5.75,
+      "eval_loss": 3.5658366680145264,
+      "eval_runtime": 23.7952,
+      "eval_samples_per_second": 105.988,
+      "eval_steps_per_second": 6.64,
+      "step": 62000
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 1.2641e-05,
+      "loss": 3.4558,
+      "step": 62100
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 1.2607666666666666e-05,
+      "loss": 3.4448,
+      "step": 62200
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 1.2574333333333333e-05,
+      "loss": 3.4367,
+      "step": 62300
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 1.2540999999999999e-05,
+      "loss": 3.4363,
+      "step": 62400
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 1.2507666666666667e-05,
+      "loss": 3.4618,
+      "step": 62500
+    },
+    {
+      "epoch": 5.79,
+      "eval_loss": 3.565267324447632,
+      "eval_runtime": 23.7952,
+      "eval_samples_per_second": 105.988,
+      "eval_steps_per_second": 6.64,
+      "step": 62500
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 1.2474666666666666e-05,
+      "loss": 3.4519,
+      "step": 62600
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 1.2441333333333332e-05,
+      "loss": 3.4173,
+      "step": 62700
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 1.2408e-05,
+      "loss": 3.4429,
+      "step": 62800
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 1.2374666666666667e-05,
+      "loss": 3.4334,
+      "step": 62900
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 1.2341333333333333e-05,
+      "loss": 3.4541,
+      "step": 63000
+    },
+    {
+      "epoch": 5.84,
+      "eval_loss": 3.565286874771118,
+      "eval_runtime": 23.7849,
+      "eval_samples_per_second": 106.034,
+      "eval_steps_per_second": 6.643,
+      "step": 63000
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 1.2308e-05,
+      "loss": 3.4434,
+      "step": 63100
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 1.2274666666666666e-05,
+      "loss": 3.4478,
+      "step": 63200
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 1.2241333333333333e-05,
+      "loss": 3.4517,
+      "step": 63300
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 1.2207999999999999e-05,
+      "loss": 3.462,
+      "step": 63400
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 1.2174666666666665e-05,
+      "loss": 3.4552,
+      "step": 63500
+    },
+    {
+      "epoch": 5.89,
+      "eval_loss": 3.5648128986358643,
+      "eval_runtime": 23.8,
+      "eval_samples_per_second": 105.966,
+      "eval_steps_per_second": 6.639,
+      "step": 63500
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 1.2141333333333334e-05,
+      "loss": 3.4486,
+      "step": 63600
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 1.2108e-05,
+      "loss": 3.4296,
+      "step": 63700
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 1.2074666666666666e-05,
+      "loss": 3.4678,
+      "step": 63800
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 1.2041333333333334e-05,
+      "loss": 3.4566,
+      "step": 63900
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 1.2008000000000001e-05,
+      "loss": 3.4679,
+      "step": 64000
+    },
+    {
+      "epoch": 5.93,
+      "eval_loss": 3.5647895336151123,
+      "eval_runtime": 23.7762,
+      "eval_samples_per_second": 106.073,
+      "eval_steps_per_second": 6.645,
+      "step": 64000
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 1.1974666666666667e-05,
+      "loss": 3.4522,
+      "step": 64100
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 1.1941333333333334e-05,
+      "loss": 3.4432,
+      "step": 64200
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 1.1908000000000002e-05,
+      "loss": 3.4428,
+      "step": 64300
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 1.1874666666666668e-05,
+      "loss": 3.4462,
+      "step": 64400
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 1.1841333333333335e-05,
+      "loss": 3.4423,
+      "step": 64500
+    },
+    {
+      "epoch": 5.98,
+      "eval_loss": 3.5651543140411377,
+      "eval_runtime": 23.7835,
+      "eval_samples_per_second": 106.04,
+      "eval_steps_per_second": 6.643,
+      "step": 64500
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 1.1808000000000001e-05,
+      "loss": 3.4277,
+      "step": 64600
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 1.1774666666666668e-05,
+      "loss": 3.4397,
+      "step": 64700
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 1.1741333333333334e-05,
+      "loss": 3.4286,
+      "step": 64800
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 1.1708e-05,
+      "loss": 3.4134,
+      "step": 64900
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 1.1674666666666667e-05,
+      "loss": 3.3893,
+      "step": 65000
+    },
+    {
+      "epoch": 6.03,
+      "eval_loss": 3.564636707305908,
+      "eval_runtime": 23.8888,
+      "eval_samples_per_second": 105.572,
+      "eval_steps_per_second": 6.614,
+      "step": 65000
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 1.1641333333333335e-05,
+      "loss": 3.4141,
+      "step": 65100
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 1.1608000000000001e-05,
+      "loss": 3.4166,
+      "step": 65200
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 1.1574666666666668e-05,
+      "loss": 3.4112,
+      "step": 65300
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 1.1541333333333334e-05,
+      "loss": 3.4282,
+      "step": 65400
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 1.1508e-05,
+      "loss": 3.4239,
+      "step": 65500
+    },
+    {
+      "epoch": 6.07,
+      "eval_loss": 3.566786527633667,
+      "eval_runtime": 23.7837,
+      "eval_samples_per_second": 106.039,
+      "eval_steps_per_second": 6.643,
+      "step": 65500
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 1.1474666666666667e-05,
+      "loss": 3.4181,
+      "step": 65600
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 1.1441333333333333e-05,
+      "loss": 3.4323,
+      "step": 65700
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 1.1408e-05,
+      "loss": 3.4269,
+      "step": 65800
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 1.1374666666666668e-05,
+      "loss": 3.4389,
+      "step": 65900
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 1.1341333333333334e-05,
+      "loss": 3.4329,
+      "step": 66000
+    },
+    {
+      "epoch": 6.12,
+      "eval_loss": 3.563938856124878,
+      "eval_runtime": 23.781,
+      "eval_samples_per_second": 106.051,
+      "eval_steps_per_second": 6.644,
+      "step": 66000
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 1.1308e-05,
+      "loss": 3.4201,
+      "step": 66100
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 1.1274666666666667e-05,
+      "loss": 3.4139,
+      "step": 66200
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 1.1241333333333334e-05,
+      "loss": 3.4178,
+      "step": 66300
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 1.1208333333333334e-05,
+      "loss": 3.4203,
+      "step": 66400
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 1.1175e-05,
+      "loss": 3.4151,
+      "step": 66500
+    },
+    {
+      "epoch": 6.16,
+      "eval_loss": 3.564937114715576,
+      "eval_runtime": 23.7834,
+      "eval_samples_per_second": 106.04,
+      "eval_steps_per_second": 6.643,
+      "step": 66500
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 1.1141666666666667e-05,
+      "loss": 3.4345,
+      "step": 66600
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 1.1108333333333333e-05,
+      "loss": 3.4681,
+      "step": 66700
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 1.1075e-05,
+      "loss": 3.4163,
+      "step": 66800
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 1.1041666666666668e-05,
+      "loss": 3.422,
+      "step": 66900
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 1.1008333333333334e-05,
+      "loss": 3.4181,
+      "step": 67000
+    },
+    {
+      "epoch": 6.21,
+      "eval_loss": 3.568237066268921,
+      "eval_runtime": 23.8051,
+      "eval_samples_per_second": 105.944,
+      "eval_steps_per_second": 6.637,
+      "step": 67000
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 1.0975e-05,
+      "loss": 3.4151,
+      "step": 67100
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 1.0941666666666667e-05,
+      "loss": 3.4258,
+      "step": 67200
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 1.0908333333333334e-05,
+      "loss": 3.411,
+      "step": 67300
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 1.0875e-05,
+      "loss": 3.4326,
+      "step": 67400
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 1.0841666666666666e-05,
+      "loss": 3.4314,
+      "step": 67500
+    },
+    {
+      "epoch": 6.26,
+      "eval_loss": 3.566948413848877,
+      "eval_runtime": 23.7688,
+      "eval_samples_per_second": 106.105,
+      "eval_steps_per_second": 6.647,
+      "step": 67500
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 1.0808333333333333e-05,
+      "loss": 3.3988,
+      "step": 67600
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 1.0775000000000001e-05,
+      "loss": 3.4227,
+      "step": 67700
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 1.0741666666666667e-05,
+      "loss": 3.4358,
+      "step": 67800
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 1.0708333333333334e-05,
+      "loss": 3.4499,
+      "step": 67900
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 1.0675e-05,
+      "loss": 3.4245,
+      "step": 68000
+    },
+    {
+      "epoch": 6.3,
+      "eval_loss": 3.562889575958252,
+      "eval_runtime": 23.7724,
+      "eval_samples_per_second": 106.089,
+      "eval_steps_per_second": 6.646,
+      "step": 68000
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 1.0641666666666667e-05,
+      "loss": 3.4124,
+      "step": 68100
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 1.0608333333333333e-05,
+      "loss": 3.4106,
+      "step": 68200
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 1.0575e-05,
+      "loss": 3.4115,
+      "step": 68300
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 1.0541666666666666e-05,
+      "loss": 3.4309,
+      "step": 68400
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 1.0508333333333334e-05,
+      "loss": 3.421,
+      "step": 68500
+    },
+    {
+      "epoch": 6.35,
+      "eval_loss": 3.566340446472168,
+      "eval_runtime": 23.7726,
+      "eval_samples_per_second": 106.089,
+      "eval_steps_per_second": 6.646,
+      "step": 68500
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 1.0475e-05,
+      "loss": 3.4107,
+      "step": 68600
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 1.0441666666666667e-05,
+      "loss": 3.4152,
+      "step": 68700
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 1.0408333333333333e-05,
+      "loss": 3.4019,
+      "step": 68800
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 1.0375e-05,
+      "loss": 3.3954,
+      "step": 68900
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 1.0341666666666666e-05,
+      "loss": 3.4329,
+      "step": 69000
+    },
+    {
+      "epoch": 6.4,
+      "eval_loss": 3.565971851348877,
+      "eval_runtime": 23.7798,
+      "eval_samples_per_second": 106.056,
+      "eval_steps_per_second": 6.644,
+      "step": 69000
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 1.0308333333333333e-05,
+      "loss": 3.419,
+      "step": 69100
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 1.0275e-05,
+      "loss": 3.4272,
+      "step": 69200
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 1.0241666666666667e-05,
+      "loss": 3.4056,
+      "step": 69300
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 1.0208333333333334e-05,
+      "loss": 3.4373,
+      "step": 69400
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 1.0175333333333334e-05,
+      "loss": 3.4122,
+      "step": 69500
+    },
+    {
+      "epoch": 6.44,
+      "eval_loss": 3.5650551319122314,
+      "eval_runtime": 23.7808,
+      "eval_samples_per_second": 106.052,
+      "eval_steps_per_second": 6.644,
+      "step": 69500
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 1.0142e-05,
+      "loss": 3.4152,
+      "step": 69600
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 1.0109000000000001e-05,
+      "loss": 3.463,
+      "step": 69700
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 1.0075666666666667e-05,
+      "loss": 3.4214,
+      "step": 69800
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 1.0042333333333334e-05,
+      "loss": 3.4423,
+      "step": 69900
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 1.0009e-05,
+      "loss": 3.4362,
+      "step": 70000
+    },
+    {
+      "epoch": 6.49,
+      "eval_loss": 3.5628156661987305,
+      "eval_runtime": 23.761,
+      "eval_samples_per_second": 106.14,
+      "eval_steps_per_second": 6.65,
+      "step": 70000
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 9.975666666666667e-06,
+      "loss": 3.4393,
+      "step": 70100
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 9.942333333333333e-06,
+      "loss": 3.4448,
+      "step": 70200
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 9.909e-06,
+      "loss": 3.4162,
+      "step": 70300
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 9.875666666666666e-06,
+      "loss": 3.4291,
+      "step": 70400
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 9.842333333333334e-06,
+      "loss": 3.4497,
+      "step": 70500
+    },
+    {
+      "epoch": 6.54,
+      "eval_loss": 3.5647528171539307,
+      "eval_runtime": 23.7333,
+      "eval_samples_per_second": 106.264,
+      "eval_steps_per_second": 6.657,
+      "step": 70500
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 9.809e-06,
+      "loss": 3.4127,
+      "step": 70600
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 9.775666666666667e-06,
+      "loss": 3.4347,
+      "step": 70700
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 9.742333333333333e-06,
+      "loss": 3.4317,
+      "step": 70800
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 9.709e-06,
+      "loss": 3.4201,
+      "step": 70900
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 9.675666666666666e-06,
+      "loss": 3.431,
+      "step": 71000
+    },
+    {
+      "epoch": 6.58,
+      "eval_loss": 3.562624216079712,
+      "eval_runtime": 23.7932,
+      "eval_samples_per_second": 105.997,
+      "eval_steps_per_second": 6.641,
+      "step": 71000
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 9.642333333333333e-06,
+      "loss": 3.4172,
+      "step": 71100
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 9.608999999999999e-06,
+      "loss": 3.4387,
+      "step": 71200
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 9.575666666666667e-06,
+      "loss": 3.4288,
+      "step": 71300
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 9.542333333333334e-06,
+      "loss": 3.4287,
+      "step": 71400
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 9.509e-06,
+      "loss": 3.432,
+      "step": 71500
+    },
+    {
+      "epoch": 6.63,
+      "eval_loss": 3.5648353099823,
+      "eval_runtime": 23.9564,
+      "eval_samples_per_second": 105.274,
+      "eval_steps_per_second": 6.595,
+      "step": 71500
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 9.475666666666666e-06,
+      "loss": 3.4299,
+      "step": 71600
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 9.442333333333333e-06,
+      "loss": 3.4381,
+      "step": 71700
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 9.409e-06,
+      "loss": 3.4313,
+      "step": 71800
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 9.375666666666666e-06,
+      "loss": 3.4254,
+      "step": 71900
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 9.342333333333334e-06,
+      "loss": 3.4208,
+      "step": 72000
+    },
+    {
+      "epoch": 6.67,
+      "eval_loss": 3.563504934310913,
+      "eval_runtime": 23.7942,
+      "eval_samples_per_second": 105.992,
+      "eval_steps_per_second": 6.64,
+      "step": 72000
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 9.309e-06,
+      "loss": 3.4348,
+      "step": 72100
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 9.275666666666667e-06,
+      "loss": 3.4201,
+      "step": 72200
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 9.242333333333333e-06,
+      "loss": 3.3956,
+      "step": 72300
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 9.209e-06,
+      "loss": 3.4262,
+      "step": 72400
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 9.175666666666666e-06,
+      "loss": 3.4526,
+      "step": 72500
+    },
+    {
+      "epoch": 6.72,
+      "eval_loss": 3.564483404159546,
+      "eval_runtime": 23.8074,
+      "eval_samples_per_second": 105.934,
+      "eval_steps_per_second": 6.637,
+      "step": 72500
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 9.142333333333332e-06,
+      "loss": 3.4344,
+      "step": 72600
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 9.109e-06,
+      "loss": 3.4501,
+      "step": 72700
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 9.075666666666668e-06,
+      "loss": 3.405,
+      "step": 72800
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 9.042333333333335e-06,
+      "loss": 3.4416,
+      "step": 72900
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 9.009000000000001e-06,
+      "loss": 3.4139,
+      "step": 73000
+    },
+    {
+      "epoch": 6.77,
+      "eval_loss": 3.5620737075805664,
+      "eval_runtime": 23.7869,
+      "eval_samples_per_second": 106.025,
+      "eval_steps_per_second": 6.642,
+      "step": 73000
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 8.975666666666668e-06,
+      "loss": 3.4286,
+      "step": 73100
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 8.942333333333334e-06,
+      "loss": 3.428,
+      "step": 73200
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 8.909e-06,
+      "loss": 3.4446,
+      "step": 73300
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 8.875666666666667e-06,
+      "loss": 3.415,
+      "step": 73400
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 8.842333333333333e-06,
+      "loss": 3.4212,
+      "step": 73500
+    },
+    {
+      "epoch": 6.81,
+      "eval_loss": 3.562941789627075,
+      "eval_runtime": 23.7809,
+      "eval_samples_per_second": 106.051,
+      "eval_steps_per_second": 6.644,
+      "step": 73500
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 8.809000000000002e-06,
+      "loss": 3.4272,
+      "step": 73600
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 8.775666666666668e-06,
+      "loss": 3.4448,
+      "step": 73700
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 8.742333333333334e-06,
+      "loss": 3.4102,
+      "step": 73800
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 8.709e-06,
+      "loss": 3.4299,
+      "step": 73900
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 8.675666666666667e-06,
+      "loss": 3.4352,
+      "step": 74000
+    },
+    {
+      "epoch": 6.86,
+      "eval_loss": 3.5597243309020996,
+      "eval_runtime": 23.7901,
+      "eval_samples_per_second": 106.01,
+      "eval_steps_per_second": 6.641,
+      "step": 74000
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 8.642333333333334e-06,
+      "loss": 3.4144,
+      "step": 74100
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 8.609e-06,
+      "loss": 3.4328,
+      "step": 74200
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 8.575666666666666e-06,
+      "loss": 3.4214,
+      "step": 74300
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 8.542333333333335e-06,
+      "loss": 3.4055,
+      "step": 74400
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 8.509000000000001e-06,
+      "loss": 3.4242,
+      "step": 74500
+    },
+    {
+      "epoch": 6.91,
+      "eval_loss": 3.559704303741455,
+      "eval_runtime": 23.8077,
+      "eval_samples_per_second": 105.932,
+      "eval_steps_per_second": 6.637,
+      "step": 74500
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 8.475666666666667e-06,
+      "loss": 3.4363,
+      "step": 74600
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 8.442333333333334e-06,
+      "loss": 3.4417,
+      "step": 74700
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 8.409e-06,
+      "loss": 3.4616,
+      "step": 74800
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 8.375666666666667e-06,
+      "loss": 3.4322,
+      "step": 74900
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 8.342333333333333e-06,
+      "loss": 3.429,
+      "step": 75000
+    },
+    {
+      "epoch": 6.95,
+      "eval_loss": 3.561878204345703,
+      "eval_runtime": 23.7887,
+      "eval_samples_per_second": 106.017,
+      "eval_steps_per_second": 6.642,
+      "step": 75000
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 8.309000000000001e-06,
+      "loss": 3.4187,
+      "step": 75100
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 8.276e-06,
+      "loss": 3.419,
+      "step": 75200
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 8.242666666666666e-06,
+      "loss": 3.4123,
+      "step": 75300
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 8.209333333333335e-06,
+      "loss": 3.4338,
+      "step": 75400
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 8.176000000000001e-06,
+      "loss": 3.4133,
+      "step": 75500
+    },
+    {
+      "epoch": 7.0,
+      "eval_loss": 3.5591769218444824,
+      "eval_runtime": 23.7782,
+      "eval_samples_per_second": 106.064,
+      "eval_steps_per_second": 6.645,
+      "step": 75500
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 8.142666666666667e-06,
+      "loss": 3.4184,
+      "step": 75600
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 8.109333333333334e-06,
+      "loss": 3.3904,
+      "step": 75700
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 8.076e-06,
+      "loss": 3.3813,
+      "step": 75800
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 8.042666666666667e-06,
+      "loss": 3.4422,
+      "step": 75900
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 8.009333333333333e-06,
+      "loss": 3.4086,
+      "step": 76000
+    },
+    {
+      "epoch": 7.04,
+      "eval_loss": 3.562080144882202,
+      "eval_runtime": 23.7794,
+      "eval_samples_per_second": 106.058,
+      "eval_steps_per_second": 6.644,
+      "step": 76000
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 7.976e-06,
+      "loss": 3.3967,
+      "step": 76100
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 7.942666666666668e-06,
+      "loss": 3.3924,
+      "step": 76200
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 7.909333333333334e-06,
+      "loss": 3.4045,
+      "step": 76300
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 7.876e-06,
+      "loss": 3.4003,
+      "step": 76400
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 7.842666666666667e-06,
+      "loss": 3.4056,
+      "step": 76500
+    },
+    {
+      "epoch": 7.09,
+      "eval_loss": 3.5603981018066406,
+      "eval_runtime": 23.7637,
+      "eval_samples_per_second": 106.128,
+      "eval_steps_per_second": 6.649,
+      "step": 76500
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 7.809333333333333e-06,
+      "loss": 3.4325,
+      "step": 76600
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 7.776e-06,
+      "loss": 3.4493,
+      "step": 76700
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 7.742666666666666e-06,
+      "loss": 3.4245,
+      "step": 76800
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 7.709333333333334e-06,
+      "loss": 3.3955,
+      "step": 76900
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 7.676e-06,
+      "loss": 3.4158,
+      "step": 77000
+    },
+    {
+      "epoch": 7.14,
+      "eval_loss": 3.56292986869812,
+      "eval_runtime": 23.8347,
+      "eval_samples_per_second": 105.812,
+      "eval_steps_per_second": 6.629,
+      "step": 77000
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 7.642666666666667e-06,
+      "loss": 3.3797,
+      "step": 77100
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 7.6093333333333335e-06,
+      "loss": 3.4144,
+      "step": 77200
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 7.576e-06,
+      "loss": 3.4127,
+      "step": 77300
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 7.542666666666666e-06,
+      "loss": 3.4166,
+      "step": 77400
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 7.509333333333333e-06,
+      "loss": 3.4153,
+      "step": 77500
+    },
+    {
+      "epoch": 7.18,
+      "eval_loss": 3.5608999729156494,
+      "eval_runtime": 23.7878,
+      "eval_samples_per_second": 106.021,
+      "eval_steps_per_second": 6.642,
+      "step": 77500
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 7.476e-06,
+      "loss": 3.4062,
+      "step": 77600
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 7.4426666666666665e-06,
+      "loss": 3.4288,
+      "step": 77700
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 7.409333333333333e-06,
+      "loss": 3.4145,
+      "step": 77800
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 7.376e-06,
+      "loss": 3.4103,
+      "step": 77900
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 7.342666666666667e-06,
+      "loss": 3.4155,
+      "step": 78000
+    },
+    {
+      "epoch": 7.23,
+      "eval_loss": 3.5620954036712646,
+      "eval_runtime": 23.793,
+      "eval_samples_per_second": 105.997,
+      "eval_steps_per_second": 6.641,
+      "step": 78000
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 7.309333333333333e-06,
+      "loss": 3.4204,
+      "step": 78100
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 7.2759999999999995e-06,
+      "loss": 3.3973,
+      "step": 78200
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 7.2426666666666676e-06,
+      "loss": 3.4022,
+      "step": 78300
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 7.209333333333334e-06,
+      "loss": 3.3983,
+      "step": 78400
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 7.176e-06,
+      "loss": 3.4117,
+      "step": 78500
+    },
+    {
+      "epoch": 7.28,
+      "eval_loss": 3.5625743865966797,
+      "eval_runtime": 23.7846,
+      "eval_samples_per_second": 106.035,
+      "eval_steps_per_second": 6.643,
+      "step": 78500
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 7.142666666666667e-06,
+      "loss": 3.4003,
+      "step": 78600
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 7.109333333333334e-06,
+      "loss": 3.4077,
+      "step": 78700
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 7.0760000000000005e-06,
+      "loss": 3.4098,
+      "step": 78800
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 7.042666666666667e-06,
+      "loss": 3.3968,
+      "step": 78900
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 7.009333333333333e-06,
+      "loss": 3.407,
+      "step": 79000
+    },
+    {
+      "epoch": 7.32,
+      "eval_loss": 3.563791275024414,
+      "eval_runtime": 23.7857,
+      "eval_samples_per_second": 106.03,
+      "eval_steps_per_second": 6.643,
+      "step": 79000
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 6.976000000000001e-06,
+      "loss": 3.4117,
+      "step": 79100
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 6.942666666666667e-06,
+      "loss": 3.4233,
+      "step": 79200
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 6.9093333333333335e-06,
+      "loss": 3.3917,
+      "step": 79300
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 6.876e-06,
+      "loss": 3.3981,
+      "step": 79400
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 6.843e-06,
+      "loss": 3.3977,
+      "step": 79500
+    },
+    {
+      "epoch": 7.37,
+      "eval_loss": 3.5603787899017334,
+      "eval_runtime": 23.7147,
+      "eval_samples_per_second": 106.348,
+      "eval_steps_per_second": 6.663,
+      "step": 79500
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 6.81e-06,
+      "loss": 3.418,
+      "step": 79600
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 6.7766666666666664e-06,
+      "loss": 3.4228,
+      "step": 79700
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 6.743333333333334e-06,
+      "loss": 3.4012,
+      "step": 79800
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 6.710000000000001e-06,
+      "loss": 3.4078,
+      "step": 79900
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 6.676666666666667e-06,
+      "loss": 3.4134,
+      "step": 80000
+    },
+    {
+      "epoch": 7.42,
+      "eval_loss": 3.5610873699188232,
+      "eval_runtime": 23.7819,
+      "eval_samples_per_second": 106.047,
+      "eval_steps_per_second": 6.644,
+      "step": 80000
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 6.643333333333334e-06,
+      "loss": 3.4111,
+      "step": 80100
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 6.61e-06,
+      "loss": 3.4028,
+      "step": 80200
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 6.5766666666666675e-06,
+      "loss": 3.4303,
+      "step": 80300
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 6.543333333333334e-06,
+      "loss": 3.3938,
+      "step": 80400
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 6.51e-06,
+      "loss": 3.4403,
+      "step": 80500
+    },
+    {
+      "epoch": 7.46,
+      "eval_loss": 3.563021659851074,
+      "eval_runtime": 23.7781,
+      "eval_samples_per_second": 106.064,
+      "eval_steps_per_second": 6.645,
+      "step": 80500
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 6.476666666666667e-06,
+      "loss": 3.409,
+      "step": 80600
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 6.443333333333334e-06,
+      "loss": 3.4077,
+      "step": 80700
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 6.4100000000000005e-06,
+      "loss": 3.4048,
+      "step": 80800
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 6.376666666666667e-06,
+      "loss": 3.3906,
+      "step": 80900
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 6.343666666666667e-06,
+      "loss": 3.4002,
+      "step": 81000
+    },
+    {
+      "epoch": 7.51,
+      "eval_loss": 3.5600671768188477,
+      "eval_runtime": 23.7876,
+      "eval_samples_per_second": 106.022,
+      "eval_steps_per_second": 6.642,
+      "step": 81000
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 6.310333333333334e-06,
+      "loss": 3.4011,
+      "step": 81100
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 6.277e-06,
+      "loss": 3.4164,
+      "step": 81200
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 6.243666666666667e-06,
+      "loss": 3.4116,
+      "step": 81300
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 6.210333333333334e-06,
+      "loss": 3.3992,
+      "step": 81400
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 6.177e-06,
+      "loss": 3.4147,
+      "step": 81500
+    },
+    {
+      "epoch": 7.55,
+      "eval_loss": 3.557727098464966,
+      "eval_runtime": 23.7591,
+      "eval_samples_per_second": 106.149,
+      "eval_steps_per_second": 6.65,
+      "step": 81500
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 6.143666666666667e-06,
+      "loss": 3.4273,
+      "step": 81600
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 6.110333333333334e-06,
+      "loss": 3.4086,
+      "step": 81700
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 6.0770000000000004e-06,
+      "loss": 3.4102,
+      "step": 81800
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 6.043666666666667e-06,
+      "loss": 3.4214,
+      "step": 81900
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 6.010333333333333e-06,
+      "loss": 3.4068,
+      "step": 82000
+    },
+    {
+      "epoch": 7.6,
+      "eval_loss": 3.5587644577026367,
+      "eval_runtime": 23.8126,
+      "eval_samples_per_second": 105.91,
+      "eval_steps_per_second": 6.635,
+      "step": 82000
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 5.9770000000000005e-06,
+      "loss": 3.422,
+      "step": 82100
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 5.943666666666667e-06,
+      "loss": 3.4128,
+      "step": 82200
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 5.910333333333333e-06,
+      "loss": 3.3978,
+      "step": 82300
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 5.877e-06,
+      "loss": 3.4095,
+      "step": 82400
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 5.843666666666667e-06,
+      "loss": 3.4165,
+      "step": 82500
+    },
+    {
+      "epoch": 7.65,
+      "eval_loss": 3.5612709522247314,
+      "eval_runtime": 23.7739,
+      "eval_samples_per_second": 106.083,
+      "eval_steps_per_second": 6.646,
+      "step": 82500
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 5.8103333333333335e-06,
+      "loss": 3.3894,
+      "step": 82600
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 5.777e-06,
+      "loss": 3.3967,
+      "step": 82700
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 5.743666666666666e-06,
+      "loss": 3.4344,
+      "step": 82800
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 5.710333333333334e-06,
+      "loss": 3.4359,
+      "step": 82900
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 5.677e-06,
+      "loss": 3.409,
+      "step": 83000
+    },
+    {
+      "epoch": 7.69,
+      "eval_loss": 3.55955171585083,
+      "eval_runtime": 23.7881,
+      "eval_samples_per_second": 106.019,
+      "eval_steps_per_second": 6.642,
+      "step": 83000
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 5.6436666666666664e-06,
+      "loss": 3.403,
+      "step": 83100
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 5.610333333333333e-06,
+      "loss": 3.3973,
+      "step": 83200
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 5.577e-06,
+      "loss": 3.4116,
+      "step": 83300
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 5.5436666666666666e-06,
+      "loss": 3.4164,
+      "step": 83400
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 5.510333333333333e-06,
+      "loss": 3.4213,
+      "step": 83500
+    },
+    {
+      "epoch": 7.74,
+      "eval_loss": 3.558309555053711,
+      "eval_runtime": 23.7902,
+      "eval_samples_per_second": 106.01,
+      "eval_steps_per_second": 6.641,
+      "step": 83500
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 5.476999999999999e-06,
+      "loss": 3.4144,
+      "step": 83600
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 5.443666666666667e-06,
+      "loss": 3.4265,
+      "step": 83700
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 5.410333333333334e-06,
+      "loss": 3.4128,
+      "step": 83800
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 5.377e-06,
+      "loss": 3.4057,
+      "step": 83900
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 5.343666666666667e-06,
+      "loss": 3.403,
+      "step": 84000
+    },
+    {
+      "epoch": 7.79,
+      "eval_loss": 3.5601158142089844,
+      "eval_runtime": 23.7948,
+      "eval_samples_per_second": 105.989,
+      "eval_steps_per_second": 6.64,
+      "step": 84000
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 5.310333333333334e-06,
+      "loss": 3.4139,
+      "step": 84100
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 5.2770000000000005e-06,
+      "loss": 3.4101,
+      "step": 84200
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 5.243666666666667e-06,
+      "loss": 3.3735,
+      "step": 84300
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 5.210333333333333e-06,
+      "loss": 3.4077,
+      "step": 84400
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 5.177000000000001e-06,
+      "loss": 3.3819,
+      "step": 84500
+    },
+    {
+      "epoch": 7.83,
+      "eval_loss": 3.5579638481140137,
+      "eval_runtime": 23.8048,
+      "eval_samples_per_second": 105.945,
+      "eval_steps_per_second": 6.637,
+      "step": 84500
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 5.143666666666667e-06,
+      "loss": 3.4202,
+      "step": 84600
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 5.110333333333333e-06,
+      "loss": 3.4047,
+      "step": 84700
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 5.077000000000001e-06,
+      "loss": 3.4063,
+      "step": 84800
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 5.043666666666667e-06,
+      "loss": 3.4038,
+      "step": 84900
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 5.0103333333333335e-06,
+      "loss": 3.4182,
+      "step": 85000
+    },
+    {
+      "epoch": 7.88,
+      "eval_loss": 3.5569992065429688,
+      "eval_runtime": 23.7728,
+      "eval_samples_per_second": 106.088,
+      "eval_steps_per_second": 6.646,
+      "step": 85000
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 4.977e-06,
+      "loss": 3.3997,
+      "step": 85100
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 4.943666666666667e-06,
+      "loss": 3.381,
+      "step": 85200
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 4.910333333333334e-06,
+      "loss": 3.4129,
+      "step": 85300
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 4.877e-06,
+      "loss": 3.4063,
+      "step": 85400
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 4.8436666666666665e-06,
+      "loss": 3.4099,
+      "step": 85500
+    },
+    {
+      "epoch": 7.93,
+      "eval_loss": 3.556957483291626,
+      "eval_runtime": 23.7791,
+      "eval_samples_per_second": 106.06,
+      "eval_steps_per_second": 6.644,
+      "step": 85500
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 4.810333333333334e-06,
+      "loss": 3.4177,
+      "step": 85600
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 4.777e-06,
+      "loss": 3.3911,
+      "step": 85700
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 4.743666666666667e-06,
+      "loss": 3.4014,
+      "step": 85800
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 4.710333333333333e-06,
+      "loss": 3.4199,
+      "step": 85900
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 4.677e-06,
+      "loss": 3.3845,
+      "step": 86000
+    },
+    {
+      "epoch": 7.97,
+      "eval_loss": 3.558166980743408,
+      "eval_runtime": 23.7635,
+      "eval_samples_per_second": 106.129,
+      "eval_steps_per_second": 6.649,
+      "step": 86000
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 4.643666666666667e-06,
+      "loss": 3.3794,
+      "step": 86100
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 4.610333333333333e-06,
+      "loss": 3.4015,
+      "step": 86200
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4.5769999999999995e-06,
+      "loss": 3.3849,
+      "step": 86300
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 4.543666666666667e-06,
+      "loss": 3.4057,
+      "step": 86400
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 4.510333333333333e-06,
+      "loss": 3.411,
+      "step": 86500
+    },
+    {
+      "epoch": 8.02,
+      "eval_loss": 3.5609631538391113,
+      "eval_runtime": 23.7818,
+      "eval_samples_per_second": 106.047,
+      "eval_steps_per_second": 6.644,
+      "step": 86500
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 4.477e-06,
+      "loss": 3.3687,
+      "step": 86600
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 4.443666666666666e-06,
+      "loss": 3.4203,
+      "step": 86700
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 4.410333333333333e-06,
+      "loss": 3.3975,
+      "step": 86800
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 4.377e-06,
+      "loss": 3.3786,
+      "step": 86900
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 4.343666666666667e-06,
+      "loss": 3.3952,
+      "step": 87000
+    },
+    {
+      "epoch": 8.06,
+      "eval_loss": 3.558793544769287,
+      "eval_runtime": 23.7833,
+      "eval_samples_per_second": 106.041,
+      "eval_steps_per_second": 6.643,
+      "step": 87000
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 4.3103333333333335e-06,
+      "loss": 3.4223,
+      "step": 87100
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 4.277000000000001e-06,
+      "loss": 3.3969,
+      "step": 87200
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 4.243666666666667e-06,
+      "loss": 3.4057,
+      "step": 87300
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 4.2103333333333336e-06,
+      "loss": 3.3814,
+      "step": 87400
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 4.177e-06,
+      "loss": 3.4211,
+      "step": 87500
+    },
+    {
+      "epoch": 8.11,
+      "eval_loss": 3.558820962905884,
+      "eval_runtime": 23.7795,
+      "eval_samples_per_second": 106.058,
+      "eval_steps_per_second": 6.644,
+      "step": 87500
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 4.143666666666667e-06,
+      "loss": 3.3888,
+      "step": 87600
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 4.110333333333334e-06,
+      "loss": 3.3956,
+      "step": 87700
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 4.077e-06,
+      "loss": 3.3754,
+      "step": 87800
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 4.043666666666667e-06,
+      "loss": 3.3996,
+      "step": 87900
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 4.010333333333334e-06,
+      "loss": 3.4171,
+      "step": 88000
+    },
+    {
+      "epoch": 8.16,
+      "eval_loss": 3.5569565296173096,
+      "eval_runtime": 23.7758,
+      "eval_samples_per_second": 106.074,
+      "eval_steps_per_second": 6.645,
+      "step": 88000
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 3.977e-06,
+      "loss": 3.4081,
+      "step": 88100
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 3.943666666666667e-06,
+      "loss": 3.3594,
+      "step": 88200
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 3.910333333333334e-06,
+      "loss": 3.4095,
+      "step": 88300
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 3.877e-06,
+      "loss": 3.4008,
+      "step": 88400
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 3.843666666666667e-06,
+      "loss": 3.3825,
+      "step": 88500
+    },
+    {
+      "epoch": 8.2,
+      "eval_loss": 3.56074595451355,
+      "eval_runtime": 23.7071,
+      "eval_samples_per_second": 106.382,
+      "eval_steps_per_second": 6.665,
+      "step": 88500
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 3.810333333333333e-06,
+      "loss": 3.3902,
+      "step": 88600
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 3.7770000000000004e-06,
+      "loss": 3.3889,
+      "step": 88700
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 3.743666666666667e-06,
+      "loss": 3.3736,
+      "step": 88800
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 3.7103333333333333e-06,
+      "loss": 3.3951,
+      "step": 88900
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 3.677e-06,
+      "loss": 3.3807,
+      "step": 89000
+    },
+    {
+      "epoch": 8.25,
+      "eval_loss": 3.5578765869140625,
+      "eval_runtime": 23.792,
+      "eval_samples_per_second": 106.002,
+      "eval_steps_per_second": 6.641,
+      "step": 89000
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 3.644e-06,
+      "loss": 3.3969,
+      "step": 89100
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 3.6106666666666666e-06,
+      "loss": 3.3845,
+      "step": 89200
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 3.5773333333333334e-06,
+      "loss": 3.3756,
+      "step": 89300
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 3.5440000000000003e-06,
+      "loss": 3.3856,
+      "step": 89400
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 3.5106666666666667e-06,
+      "loss": 3.3842,
+      "step": 89500
+    },
+    {
+      "epoch": 8.3,
+      "eval_loss": 3.558271646499634,
+      "eval_runtime": 23.788,
+      "eval_samples_per_second": 106.02,
+      "eval_steps_per_second": 6.642,
+      "step": 89500
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 3.4773333333333336e-06,
+      "loss": 3.3895,
+      "step": 89600
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 3.444e-06,
+      "loss": 3.4004,
+      "step": 89700
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 3.410666666666667e-06,
+      "loss": 3.3814,
+      "step": 89800
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 3.3773333333333332e-06,
+      "loss": 3.3887,
+      "step": 89900
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 3.344e-06,
+      "loss": 3.3809,
+      "step": 90000
+    },
+    {
+      "epoch": 8.34,
+      "eval_loss": 3.5595808029174805,
+      "eval_runtime": 23.7777,
+      "eval_samples_per_second": 106.066,
+      "eval_steps_per_second": 6.645,
+      "step": 90000
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 3.3106666666666665e-06,
+      "loss": 3.3975,
+      "step": 90100
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 3.2773333333333338e-06,
+      "loss": 3.4037,
+      "step": 90200
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 3.244e-06,
+      "loss": 3.4035,
+      "step": 90300
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 3.210666666666667e-06,
+      "loss": 3.4036,
+      "step": 90400
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 3.1773333333333335e-06,
+      "loss": 3.4033,
+      "step": 90500
+    },
+    {
+      "epoch": 8.39,
+      "eval_loss": 3.55897855758667,
+      "eval_runtime": 23.7836,
+      "eval_samples_per_second": 106.039,
+      "eval_steps_per_second": 6.643,
+      "step": 90500
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 3.1440000000000003e-06,
+      "loss": 3.4154,
+      "step": 90600
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 3.1106666666666667e-06,
+      "loss": 3.3953,
+      "step": 90700
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 3.0773333333333336e-06,
+      "loss": 3.3951,
+      "step": 90800
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 3.044e-06,
+      "loss": 3.3892,
+      "step": 90900
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 3.010666666666667e-06,
+      "loss": 3.4156,
+      "step": 91000
+    },
+    {
+      "epoch": 8.44,
+      "eval_loss": 3.5576674938201904,
+      "eval_runtime": 23.7762,
+      "eval_samples_per_second": 106.072,
+      "eval_steps_per_second": 6.645,
+      "step": 91000
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 2.9773333333333333e-06,
+      "loss": 3.4082,
+      "step": 91100
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 2.9443333333333337e-06,
+      "loss": 3.3943,
+      "step": 91200
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 2.911e-06,
+      "loss": 3.3896,
+      "step": 91300
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 2.877666666666667e-06,
+      "loss": 3.415,
+      "step": 91400
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 2.8443333333333334e-06,
+      "loss": 3.3927,
+      "step": 91500
+    },
+    {
+      "epoch": 8.48,
+      "eval_loss": 3.558485269546509,
+      "eval_runtime": 23.7779,
+      "eval_samples_per_second": 106.065,
+      "eval_steps_per_second": 6.645,
+      "step": 91500
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 2.8110000000000003e-06,
+      "loss": 3.3887,
+      "step": 91600
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 2.7776666666666667e-06,
+      "loss": 3.3867,
+      "step": 91700
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 2.7443333333333335e-06,
+      "loss": 3.3735,
+      "step": 91800
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 2.711e-06,
+      "loss": 3.3878,
+      "step": 91900
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 2.677666666666667e-06,
+      "loss": 3.4041,
+      "step": 92000
+    },
+    {
+      "epoch": 8.53,
+      "eval_loss": 3.5595757961273193,
+      "eval_runtime": 23.7777,
+      "eval_samples_per_second": 106.066,
+      "eval_steps_per_second": 6.645,
+      "step": 92000
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 2.6443333333333332e-06,
+      "loss": 3.4155,
+      "step": 92100
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 2.611e-06,
+      "loss": 3.379,
+      "step": 92200
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 2.5776666666666665e-06,
+      "loss": 3.3913,
+      "step": 92300
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 2.5443333333333333e-06,
+      "loss": 3.3749,
+      "step": 92400
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 2.5109999999999998e-06,
+      "loss": 3.4006,
+      "step": 92500
+    },
+    {
+      "epoch": 8.57,
+      "eval_loss": 3.5600292682647705,
+      "eval_runtime": 23.781,
+      "eval_samples_per_second": 106.051,
+      "eval_steps_per_second": 6.644,
+      "step": 92500
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 2.477666666666667e-06,
+      "loss": 3.4026,
+      "step": 92600
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 2.4443333333333334e-06,
+      "loss": 3.4121,
+      "step": 92700
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 2.4110000000000003e-06,
+      "loss": 3.3956,
+      "step": 92800
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 2.3776666666666667e-06,
+      "loss": 3.4069,
+      "step": 92900
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 2.3446666666666668e-06,
+      "loss": 3.4007,
+      "step": 93000
+    },
+    {
+      "epoch": 8.62,
+      "eval_loss": 3.5577518939971924,
+      "eval_runtime": 23.7878,
+      "eval_samples_per_second": 106.021,
+      "eval_steps_per_second": 6.642,
+      "step": 93000
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 2.311333333333333e-06,
+      "loss": 3.3887,
+      "step": 93100
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 2.278e-06,
+      "loss": 3.3724,
+      "step": 93200
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 2.2446666666666665e-06,
+      "loss": 3.3821,
+      "step": 93300
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 2.2113333333333337e-06,
+      "loss": 3.4197,
+      "step": 93400
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 2.178e-06,
+      "loss": 3.4047,
+      "step": 93500
+    },
+    {
+      "epoch": 8.67,
+      "eval_loss": 3.557192087173462,
+      "eval_runtime": 23.7805,
+      "eval_samples_per_second": 106.053,
+      "eval_steps_per_second": 6.644,
+      "step": 93500
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 2.144666666666667e-06,
+      "loss": 3.4009,
+      "step": 93600
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 2.1113333333333334e-06,
+      "loss": 3.3817,
+      "step": 93700
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 2.0780000000000003e-06,
+      "loss": 3.371,
+      "step": 93800
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 2.0446666666666667e-06,
+      "loss": 3.3711,
+      "step": 93900
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 2.0113333333333335e-06,
+      "loss": 3.3904,
+      "step": 94000
+    },
+    {
+      "epoch": 8.71,
+      "eval_loss": 3.557133913040161,
+      "eval_runtime": 23.7814,
+      "eval_samples_per_second": 106.049,
+      "eval_steps_per_second": 6.644,
+      "step": 94000
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 1.978e-06,
+      "loss": 3.4108,
+      "step": 94100
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 1.944666666666667e-06,
+      "loss": 3.4221,
+      "step": 94200
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 1.9113333333333332e-06,
+      "loss": 3.3815,
+      "step": 94300
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 1.878e-06,
+      "loss": 3.3923,
+      "step": 94400
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 1.8446666666666667e-06,
+      "loss": 3.3888,
+      "step": 94500
+    },
+    {
+      "epoch": 8.76,
+      "eval_loss": 3.558104991912842,
+      "eval_runtime": 23.7899,
+      "eval_samples_per_second": 106.011,
+      "eval_steps_per_second": 6.641,
+      "step": 94500
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 1.8113333333333335e-06,
+      "loss": 3.4093,
+      "step": 94600
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 1.7780000000000002e-06,
+      "loss": 3.4228,
+      "step": 94700
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 1.7446666666666668e-06,
+      "loss": 3.4168,
+      "step": 94800
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 1.7113333333333334e-06,
+      "loss": 3.3995,
+      "step": 94900
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 1.678e-06,
+      "loss": 3.3876,
+      "step": 95000
+    },
+    {
+      "epoch": 8.81,
+      "eval_loss": 3.55720591545105,
+      "eval_runtime": 23.7866,
+      "eval_samples_per_second": 106.026,
+      "eval_steps_per_second": 6.642,
+      "step": 95000
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 1.6446666666666667e-06,
+      "loss": 3.3805,
+      "step": 95100
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 1.6113333333333333e-06,
+      "loss": 3.3955,
+      "step": 95200
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 1.578e-06,
+      "loss": 3.391,
+      "step": 95300
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 1.5446666666666668e-06,
+      "loss": 3.401,
+      "step": 95400
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 1.5113333333333334e-06,
+      "loss": 3.3872,
+      "step": 95500
+    },
+    {
+      "epoch": 8.85,
+      "eval_loss": 3.5574591159820557,
+      "eval_runtime": 23.7708,
+      "eval_samples_per_second": 106.096,
+      "eval_steps_per_second": 6.647,
+      "step": 95500
+    },
+    {
+      "epoch": 8.86,
+      "learning_rate": 1.478e-06,
+      "loss": 3.4022,
+      "step": 95600
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 1.4446666666666667e-06,
+      "loss": 3.407,
+      "step": 95700
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 1.4113333333333333e-06,
+      "loss": 3.3884,
+      "step": 95800
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 1.378e-06,
+      "loss": 3.3934,
+      "step": 95900
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 1.3446666666666666e-06,
+      "loss": 3.3753,
+      "step": 96000
+    },
+    {
+      "epoch": 8.9,
+      "eval_loss": 3.5576891899108887,
+      "eval_runtime": 23.8007,
+      "eval_samples_per_second": 105.963,
+      "eval_steps_per_second": 6.638,
+      "step": 96000
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 1.3113333333333332e-06,
+      "loss": 3.39,
+      "step": 96100
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 1.278e-06,
+      "loss": 3.418,
+      "step": 96200
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 1.2446666666666667e-06,
+      "loss": 3.3898,
+      "step": 96300
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 1.2113333333333334e-06,
+      "loss": 3.3955,
+      "step": 96400
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 1.178e-06,
+      "loss": 3.3961,
+      "step": 96500
+    },
+    {
+      "epoch": 8.95,
+      "eval_loss": 3.5567543506622314,
+      "eval_runtime": 23.775,
+      "eval_samples_per_second": 106.078,
+      "eval_steps_per_second": 6.646,
+      "step": 96500
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 1.1446666666666666e-06,
+      "loss": 3.3927,
+      "step": 96600
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 1.1113333333333333e-06,
+      "loss": 3.3778,
+      "step": 96700
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 1.0779999999999999e-06,
+      "loss": 3.4238,
+      "step": 96800
+    },
+    {
+      "epoch": 8.98,
+      "learning_rate": 1.0446666666666665e-06,
+      "loss": 3.3893,
+      "step": 96900
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 1.0113333333333334e-06,
+      "loss": 3.4131,
+      "step": 97000
+    },
+    {
+      "epoch": 8.99,
+      "eval_loss": 3.55790376663208,
+      "eval_runtime": 23.774,
+      "eval_samples_per_second": 106.082,
+      "eval_steps_per_second": 6.646,
+      "step": 97000
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 9.783333333333334e-07,
+      "loss": 3.4256,
+      "step": 97100
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 9.450000000000001e-07,
+      "loss": 3.3537,
+      "step": 97200
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 9.116666666666667e-07,
+      "loss": 3.3696,
+      "step": 97300
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 8.783333333333333e-07,
+      "loss": 3.3778,
+      "step": 97400
+    },
+    {
+      "epoch": 9.04,
+      "learning_rate": 8.453333333333334e-07,
+      "loss": 3.3647,
+      "step": 97500
+    },
+    {
+      "epoch": 9.04,
+      "eval_loss": 3.557291269302368,
+      "eval_runtime": 23.7854,
+      "eval_samples_per_second": 106.032,
+      "eval_steps_per_second": 6.643,
+      "step": 97500
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 8.12e-07,
+      "loss": 3.3852,
+      "step": 97600
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 7.786666666666667e-07,
+      "loss": 3.4056,
+      "step": 97700
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 7.453333333333333e-07,
+      "loss": 3.3955,
+      "step": 97800
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 7.12e-07,
+      "loss": 3.4157,
+      "step": 97900
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 6.786666666666667e-07,
+      "loss": 3.3792,
+      "step": 98000
+    },
+    {
+      "epoch": 9.08,
+      "eval_loss": 3.557647466659546,
+      "eval_runtime": 23.7799,
+      "eval_samples_per_second": 106.056,
+      "eval_steps_per_second": 6.644,
+      "step": 98000
+    },
+    {
+      "epoch": 9.09,
+      "learning_rate": 6.453333333333334e-07,
+      "loss": 3.3977,
+      "step": 98100
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 6.12e-07,
+      "loss": 3.3818,
+      "step": 98200
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 5.786666666666668e-07,
+      "loss": 3.3923,
+      "step": 98300
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 5.453333333333334e-07,
+      "loss": 3.3855,
+      "step": 98400
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 5.123333333333334e-07,
+      "loss": 3.3755,
+      "step": 98500
+    },
+    {
+      "epoch": 9.13,
+      "eval_loss": 3.55747389793396,
+      "eval_runtime": 23.7811,
+      "eval_samples_per_second": 106.051,
+      "eval_steps_per_second": 6.644,
+      "step": 98500
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 4.79e-07,
+      "loss": 3.378,
+      "step": 98600
+    },
+    {
+      "epoch": 9.15,
+      "learning_rate": 4.456666666666667e-07,
+      "loss": 3.3893,
+      "step": 98700
+    },
+    {
+      "epoch": 9.16,
+      "learning_rate": 4.1233333333333336e-07,
+      "loss": 3.387,
+      "step": 98800
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 3.79e-07,
+      "loss": 3.3539,
+      "step": 98900
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 3.456666666666667e-07,
+      "loss": 3.3981,
+      "step": 99000
+    },
+    {
+      "epoch": 9.18,
+      "eval_loss": 3.557342290878296,
+      "eval_runtime": 23.7655,
+      "eval_samples_per_second": 106.12,
+      "eval_steps_per_second": 6.648,
+      "step": 99000
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 3.1266666666666663e-07,
+      "loss": 3.3796,
+      "step": 99100
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 2.793333333333333e-07,
+      "loss": 3.3579,
+      "step": 99200
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 2.46e-07,
+      "loss": 3.3742,
+      "step": 99300
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 2.1266666666666667e-07,
+      "loss": 3.3831,
+      "step": 99400
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 1.7933333333333335e-07,
+      "loss": 3.3914,
+      "step": 99500
+    },
+    {
+      "epoch": 9.22,
+      "eval_loss": 3.5572941303253174,
+      "eval_runtime": 23.7886,
+      "eval_samples_per_second": 106.017,
+      "eval_steps_per_second": 6.642,
+      "step": 99500
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 1.46e-07,
+      "loss": 3.4,
+      "step": 99600
+    },
+    {
+      "epoch": 9.24,
+      "learning_rate": 1.1266666666666667e-07,
+      "loss": 3.3892,
+      "step": 99700
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 7.933333333333334e-08,
+      "loss": 3.3746,
+      "step": 99800
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 4.6e-08,
+      "loss": 3.3924,
+      "step": 99900
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 1.2666666666666668e-08,
+      "loss": 3.4136,
+      "step": 100000
+    },
+    {
+      "epoch": 9.27,
+      "eval_loss": 3.557504177093506,
+      "eval_runtime": 23.7953,
+      "eval_samples_per_second": 105.987,
+      "eval_steps_per_second": 6.64,
+      "step": 100000
+    },
+    {
+      "epoch": 9.27,
+      "step": 100000,
+      "total_flos": 2.2851507373589135e+18,
+      "train_loss": 3.524176220359802,
+      "train_runtime": 67760.9425,
+      "train_samples_per_second": 23.612,
+      "train_steps_per_second": 1.476
+    }
+  ],
+  "max_steps": 100000,
+  "num_train_epochs": 10,
+  "total_flos": 2.2851507373589135e+18,
+  "trial_name": null,
+  "trial_params": null
+}