diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,30930 @@
+{
+  "best_metric": 4.741575241088867,
+  "best_model_checkpoint": "runs/checkpoint-12000",
+  "epoch": 603.6217303822938,
+  "eval_steps": 1000,
+  "global_step": 100000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.12,
+      "learning_rate": 6.666666666666667e-08,
+      "loss": 9.3209,
+      "step": 20
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.3333333333333334e-07,
+      "loss": 9.3182,
+      "step": 40
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 9.3155,
+      "step": 60
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6666666666666667e-07,
+      "loss": 9.311,
+      "step": 80
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 9.3029,
+      "step": 100
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 9.293,
+      "step": 120
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.666666666666667e-07,
+      "loss": 9.2809,
+      "step": 140
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.333333333333333e-07,
+      "loss": 9.2666,
+      "step": 160
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 9.2481,
+      "step": 180
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 9.2245,
+      "step": 200
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 7.333333333333333e-07,
+      "loss": 9.1971,
+      "step": 220
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 9.1603,
+      "step": 240
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 8.666666666666667e-07,
+      "loss": 9.1233,
+      "step": 260
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 9.333333333333334e-07,
+      "loss": 9.0849,
+      "step": 280
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 9.045,
+      "step": 300
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.0666666666666667e-06,
+      "loss": 9.0066,
+      "step": 320
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.1333333333333334e-06,
+      "loss": 8.9707,
+      "step": 340
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 8.9421,
+      "step": 360
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.2666666666666667e-06,
+      "loss": 8.9107,
+      "step": 380
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 8.8775,
+      "step": 400
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 8.8519,
+      "step": 420
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.4666666666666667e-06,
+      "loss": 8.831,
+      "step": 440
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.5333333333333334e-06,
+      "loss": 8.8038,
+      "step": 460
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 8.7751,
+      "step": 480
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 8.7584,
+      "step": 500
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.7333333333333334e-06,
+      "loss": 8.7354,
+      "step": 520
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.8e-06,
+      "loss": 8.7144,
+      "step": 540
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.8666666666666669e-06,
+      "loss": 8.694,
+      "step": 560
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 1.9333333333333336e-06,
+      "loss": 8.6699,
+      "step": 580
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 8.6457,
+      "step": 600
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 2.0666666666666666e-06,
+      "loss": 8.6262,
+      "step": 620
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 2.1333333333333334e-06,
+      "loss": 8.6001,
+      "step": 640
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 2.2e-06,
+      "loss": 8.5778,
+      "step": 660
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 2.266666666666667e-06,
+      "loss": 8.5677,
+      "step": 680
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": 8.541,
+      "step": 700
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 8.5137,
+      "step": 720
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 2.4666666666666666e-06,
+      "loss": 8.4983,
+      "step": 740
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 2.5333333333333334e-06,
+      "loss": 8.4784,
+      "step": 760
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.6e-06,
+      "loss": 8.4527,
+      "step": 780
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 8.4373,
+      "step": 800
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 2.7333333333333336e-06,
+      "loss": 8.4061,
+      "step": 820
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 8.3854,
+      "step": 840
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 2.8666666666666666e-06,
+      "loss": 8.364,
+      "step": 860
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 2.9333333333333333e-06,
+      "loss": 8.3524,
+      "step": 880
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 3e-06,
+      "loss": 8.3238,
+      "step": 900
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 3.066666666666667e-06,
+      "loss": 8.3033,
+      "step": 920
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 3.133333333333333e-06,
+      "loss": 8.2918,
+      "step": 940
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 8.2772,
+      "step": 960
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 3.2666666666666666e-06,
+      "loss": 8.2527,
+      "step": 980
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 8.2359,
+      "step": 1000
+    },
+    {
+      "epoch": 6.04,
+      "eval_accuracy": 0.007023540219893116,
+      "eval_loss": 8.217020988464355,
+      "eval_runtime": 18.6004,
+      "eval_samples_per_second": 109.675,
+      "eval_steps_per_second": 2.312,
+      "step": 1000
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 8.2118,
+      "step": 1020
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 3.466666666666667e-06,
+      "loss": 8.1921,
+      "step": 1040
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 3.5333333333333335e-06,
+      "loss": 8.1792,
+      "step": 1060
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 3.6e-06,
+      "loss": 8.1525,
+      "step": 1080
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 3.666666666666667e-06,
+      "loss": 8.1462,
+      "step": 1100
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 3.7333333333333337e-06,
+      "loss": 8.1221,
+      "step": 1120
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 3.8e-06,
+      "loss": 8.1082,
+      "step": 1140
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 3.866666666666667e-06,
+      "loss": 8.0877,
+      "step": 1160
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 3.9333333333333335e-06,
+      "loss": 8.0805,
+      "step": 1180
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 8.0655,
+      "step": 1200
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 4.066666666666666e-06,
+      "loss": 8.0427,
+      "step": 1220
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 4.133333333333333e-06,
+      "loss": 8.0315,
+      "step": 1240
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 8.0165,
+      "step": 1260
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 4.266666666666667e-06,
+      "loss": 8.009,
+      "step": 1280
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 7.9895,
+      "step": 1300
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 4.4e-06,
+      "loss": 7.9791,
+      "step": 1320
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 4.4666666666666665e-06,
+      "loss": 7.9651,
+      "step": 1340
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 4.533333333333334e-06,
+      "loss": 7.954,
+      "step": 1360
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 4.6e-06,
+      "loss": 7.9467,
+      "step": 1380
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 7.9417,
+      "step": 1400
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 4.7333333333333335e-06,
+      "loss": 7.9304,
+      "step": 1420
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 7.9198,
+      "step": 1440
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 4.866666666666667e-06,
+      "loss": 7.9181,
+      "step": 1460
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 4.933333333333333e-06,
+      "loss": 7.8935,
+      "step": 1480
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 5e-06,
+      "loss": 7.8909,
+      "step": 1500
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 5.066666666666667e-06,
+      "loss": 7.8876,
+      "step": 1520
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 5.133333333333334e-06,
+      "loss": 7.8792,
+      "step": 1540
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 5.2e-06,
+      "loss": 7.8645,
+      "step": 1560
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 5.266666666666667e-06,
+      "loss": 7.8694,
+      "step": 1580
+    },
+    {
+      "epoch": 9.66,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 7.857,
+      "step": 1600
+    },
+    {
+      "epoch": 9.78,
+      "learning_rate": 5.4e-06,
+      "loss": 7.8533,
+      "step": 1620
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 5.466666666666667e-06,
+      "loss": 7.8439,
+      "step": 1640
+    },
+    {
+      "epoch": 10.02,
+      "learning_rate": 5.5333333333333334e-06,
+      "loss": 7.841,
+      "step": 1660
+    },
+    {
+      "epoch": 10.14,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 7.8247,
+      "step": 1680
+    },
+    {
+      "epoch": 10.26,
+      "learning_rate": 5.666666666666667e-06,
+      "loss": 7.8361,
+      "step": 1700
+    },
+    {
+      "epoch": 10.38,
+      "learning_rate": 5.733333333333333e-06,
+      "loss": 7.8296,
+      "step": 1720
+    },
+    {
+      "epoch": 10.5,
+      "learning_rate": 5.8e-06,
+      "loss": 7.8217,
+      "step": 1740
+    },
+    {
+      "epoch": 10.62,
+      "learning_rate": 5.866666666666667e-06,
+      "loss": 7.8181,
+      "step": 1760
+    },
+    {
+      "epoch": 10.74,
+      "learning_rate": 5.933333333333334e-06,
+      "loss": 7.8007,
+      "step": 1780
+    },
+    {
+      "epoch": 10.87,
+      "learning_rate": 6e-06,
+      "loss": 7.8151,
+      "step": 1800
+    },
+    {
+      "epoch": 10.99,
+      "learning_rate": 6.066666666666667e-06,
+      "loss": 7.8132,
+      "step": 1820
+    },
+    {
+      "epoch": 11.11,
+      "learning_rate": 6.133333333333334e-06,
+      "loss": 7.7844,
+      "step": 1840
+    },
+    {
+      "epoch": 11.23,
+      "learning_rate": 6.2e-06,
+      "loss": 7.7854,
+      "step": 1860
+    },
+    {
+      "epoch": 11.35,
+      "learning_rate": 6.266666666666666e-06,
+      "loss": 7.7878,
+      "step": 1880
+    },
+    {
+      "epoch": 11.47,
+      "learning_rate": 6.333333333333334e-06,
+      "loss": 7.7815,
+      "step": 1900
+    },
+    {
+      "epoch": 11.59,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 7.7711,
+      "step": 1920
+    },
+    {
+      "epoch": 11.71,
+      "learning_rate": 6.466666666666667e-06,
+      "loss": 7.7599,
+      "step": 1940
+    },
+    {
+      "epoch": 11.83,
+      "learning_rate": 6.533333333333333e-06,
+      "loss": 7.7511,
+      "step": 1960
+    },
+    {
+      "epoch": 11.95,
+      "learning_rate": 6.6e-06,
+      "loss": 7.7292,
+      "step": 1980
+    },
+    {
+      "epoch": 12.07,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 7.7137,
+      "step": 2000
+    },
+    {
+      "epoch": 12.07,
+      "eval_accuracy": 0.006356254507018601,
+      "eval_loss": 7.700709342956543,
+      "eval_runtime": 18.2793,
+      "eval_samples_per_second": 111.602,
+      "eval_steps_per_second": 2.352,
+      "step": 2000
+    },
+    {
+      "epoch": 12.19,
+      "learning_rate": 6.733333333333333e-06,
+      "loss": 7.6936,
+      "step": 2020
+    },
+    {
+      "epoch": 12.31,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 7.6831,
+      "step": 2040
+    },
+    {
+      "epoch": 12.43,
+      "learning_rate": 6.866666666666667e-06,
+      "loss": 7.6583,
+      "step": 2060
+    },
+    {
+      "epoch": 12.56,
+      "learning_rate": 6.933333333333334e-06,
+      "loss": 7.6526,
+      "step": 2080
+    },
+    {
+      "epoch": 12.68,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 7.6229,
+      "step": 2100
+    },
+    {
+      "epoch": 12.8,
+      "learning_rate": 7.066666666666667e-06,
+      "loss": 7.6069,
+      "step": 2120
+    },
+    {
+      "epoch": 12.92,
+      "learning_rate": 7.133333333333333e-06,
+      "loss": 7.567,
+      "step": 2140
+    },
+    {
+      "epoch": 13.04,
+      "learning_rate": 7.2e-06,
+      "loss": 7.4886,
+      "step": 2160
+    },
+    {
+      "epoch": 13.16,
+      "learning_rate": 7.266666666666668e-06,
+      "loss": 7.4052,
+      "step": 2180
+    },
+    {
+      "epoch": 13.28,
+      "learning_rate": 7.333333333333334e-06,
+      "loss": 7.3529,
+      "step": 2200
+    },
+    {
+      "epoch": 13.4,
+      "learning_rate": 7.4e-06,
+      "loss": 7.2899,
+      "step": 2220
+    },
+    {
+      "epoch": 13.52,
+      "learning_rate": 7.4666666666666675e-06,
+      "loss": 7.2501,
+      "step": 2240
+    },
+    {
+      "epoch": 13.64,
+      "learning_rate": 7.533333333333334e-06,
+      "loss": 7.2155,
+      "step": 2260
+    },
+    {
+      "epoch": 13.76,
+      "learning_rate": 7.6e-06,
+      "loss": 7.1748,
+      "step": 2280
+    },
+    {
+      "epoch": 13.88,
+      "learning_rate": 7.666666666666667e-06,
+      "loss": 7.147,
+      "step": 2300
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 7.733333333333334e-06,
+      "loss": 7.0968,
+      "step": 2320
+    },
+    {
+      "epoch": 14.12,
+      "learning_rate": 7.8e-06,
+      "loss": 7.083,
+      "step": 2340
+    },
+    {
+      "epoch": 14.25,
+      "learning_rate": 7.866666666666667e-06,
+      "loss": 7.0658,
+      "step": 2360
+    },
+    {
+      "epoch": 14.37,
+      "learning_rate": 7.933333333333334e-06,
+      "loss": 7.0429,
+      "step": 2380
+    },
+    {
+      "epoch": 14.49,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 7.0212,
+      "step": 2400
+    },
+    {
+      "epoch": 14.61,
+      "learning_rate": 8.066666666666667e-06,
+      "loss": 7.0132,
+      "step": 2420
+    },
+    {
+      "epoch": 14.73,
+      "learning_rate": 8.133333333333332e-06,
+      "loss": 6.9619,
+      "step": 2440
+    },
+    {
+      "epoch": 14.85,
+      "learning_rate": 8.200000000000001e-06,
+      "loss": 6.9558,
+      "step": 2460
+    },
+    {
+      "epoch": 14.97,
+      "learning_rate": 8.266666666666667e-06,
+      "loss": 6.9292,
+      "step": 2480
+    },
+    {
+      "epoch": 15.09,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 6.8963,
+      "step": 2500
+    },
+    {
+      "epoch": 15.21,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 6.9002,
+      "step": 2520
+    },
+    {
+      "epoch": 15.33,
+      "learning_rate": 8.466666666666666e-06,
+      "loss": 6.8821,
+      "step": 2540
+    },
+    {
+      "epoch": 15.45,
+      "learning_rate": 8.533333333333334e-06,
+      "loss": 6.8669,
+      "step": 2560
+    },
+    {
+      "epoch": 15.57,
+      "learning_rate": 8.599999999999999e-06,
+      "loss": 6.8442,
+      "step": 2580
+    },
+    {
+      "epoch": 15.69,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 6.8243,
+      "step": 2600
+    },
+    {
+      "epoch": 15.81,
+      "learning_rate": 8.733333333333333e-06,
+      "loss": 6.7897,
+      "step": 2620
+    },
+    {
+      "epoch": 15.94,
+      "learning_rate": 8.8e-06,
+      "loss": 6.7863,
+      "step": 2640
+    },
+    {
+      "epoch": 16.06,
+      "learning_rate": 8.866666666666668e-06,
+      "loss": 6.7918,
+      "step": 2660
+    },
+    {
+      "epoch": 16.18,
+      "learning_rate": 8.933333333333333e-06,
+      "loss": 6.7465,
+      "step": 2680
+    },
+    {
+      "epoch": 16.3,
+      "learning_rate": 9e-06,
+      "loss": 6.7454,
+      "step": 2700
+    },
+    {
+      "epoch": 16.42,
+      "learning_rate": 9.066666666666667e-06,
+      "loss": 6.7315,
+      "step": 2720
+    },
+    {
+      "epoch": 16.54,
+      "learning_rate": 9.133333333333335e-06,
+      "loss": 6.7298,
+      "step": 2740
+    },
+    {
+      "epoch": 16.66,
+      "learning_rate": 9.2e-06,
+      "loss": 6.7084,
+      "step": 2760
+    },
+    {
+      "epoch": 16.78,
+      "learning_rate": 9.266666666666667e-06,
+      "loss": 6.6964,
+      "step": 2780
+    },
+    {
+      "epoch": 16.9,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 6.6611,
+      "step": 2800
+    },
+    {
+      "epoch": 17.02,
+      "learning_rate": 9.4e-06,
+      "loss": 6.6561,
+      "step": 2820
+    },
+    {
+      "epoch": 17.14,
+      "learning_rate": 9.466666666666667e-06,
+      "loss": 6.6364,
+      "step": 2840
+    },
+    {
+      "epoch": 17.26,
+      "learning_rate": 9.533333333333334e-06,
+      "loss": 6.6261,
+      "step": 2860
+    },
+    {
+      "epoch": 17.38,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 6.6371,
+      "step": 2880
+    },
+    {
+      "epoch": 17.51,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 6.5983,
+      "step": 2900
+    },
+    {
+      "epoch": 17.63,
+      "learning_rate": 9.733333333333334e-06,
+      "loss": 6.5912,
+      "step": 2920
+    },
+    {
+      "epoch": 17.75,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 6.5684,
+      "step": 2940
+    },
+    {
+      "epoch": 17.87,
+      "learning_rate": 9.866666666666667e-06,
+      "loss": 6.5782,
+      "step": 2960
+    },
+    {
+      "epoch": 17.99,
+      "learning_rate": 9.933333333333334e-06,
+      "loss": 6.5523,
+      "step": 2980
+    },
+    {
+      "epoch": 18.11,
+      "learning_rate": 1e-05,
+      "loss": 6.5277,
+      "step": 3000
+    },
+    {
+      "epoch": 18.11,
+      "eval_accuracy": 4.099534727504421e-05,
+      "eval_loss": 6.525410175323486,
+      "eval_runtime": 18.3845,
+      "eval_samples_per_second": 110.963,
+      "eval_steps_per_second": 2.339,
+      "step": 3000
+    },
+    {
+      "epoch": 18.23,
+      "learning_rate": 1.0066666666666668e-05,
+      "loss": 6.5291,
+      "step": 3020
+    },
+    {
+      "epoch": 18.35,
+      "learning_rate": 1.0133333333333333e-05,
+      "loss": 6.5222,
+      "step": 3040
+    },
+    {
+      "epoch": 18.47,
+      "learning_rate": 1.02e-05,
+      "loss": 6.5028,
+      "step": 3060
+    },
+    {
+      "epoch": 18.59,
+      "learning_rate": 1.0266666666666668e-05,
+      "loss": 6.4887,
+      "step": 3080
+    },
+    {
+      "epoch": 18.71,
+      "learning_rate": 1.0333333333333333e-05,
+      "loss": 6.4745,
+      "step": 3100
+    },
+    {
+      "epoch": 18.83,
+      "learning_rate": 1.04e-05,
+      "loss": 6.4663,
+      "step": 3120
+    },
+    {
+      "epoch": 18.95,
+      "learning_rate": 1.0466666666666668e-05,
+      "loss": 6.4686,
+      "step": 3140
+    },
+    {
+      "epoch": 19.07,
+      "learning_rate": 1.0533333333333335e-05,
+      "loss": 6.4507,
+      "step": 3160
+    },
+    {
+      "epoch": 19.2,
+      "learning_rate": 1.06e-05,
+      "loss": 6.4451,
+      "step": 3180
+    },
+    {
+      "epoch": 19.32,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 6.4153,
+      "step": 3200
+    },
+    {
+      "epoch": 19.44,
+      "learning_rate": 1.0733333333333334e-05,
+      "loss": 6.427,
+      "step": 3220
+    },
+    {
+      "epoch": 19.56,
+      "learning_rate": 1.08e-05,
+      "loss": 6.4189,
+      "step": 3240
+    },
+    {
+      "epoch": 19.68,
+      "learning_rate": 1.0866666666666667e-05,
+      "loss": 6.4027,
+      "step": 3260
+    },
+    {
+      "epoch": 19.8,
+      "learning_rate": 1.0933333333333334e-05,
+      "loss": 6.3801,
+      "step": 3280
+    },
+    {
+      "epoch": 19.92,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 6.3769,
+      "step": 3300
+    },
+    {
+      "epoch": 20.04,
+      "learning_rate": 1.1066666666666667e-05,
+      "loss": 6.358,
+      "step": 3320
+    },
+    {
+      "epoch": 20.16,
+      "learning_rate": 1.1133333333333334e-05,
+      "loss": 6.3591,
+      "step": 3340
+    },
+    {
+      "epoch": 20.28,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 6.3585,
+      "step": 3360
+    },
+    {
+      "epoch": 20.4,
+      "learning_rate": 1.1266666666666667e-05,
+      "loss": 6.3219,
+      "step": 3380
+    },
+    {
+      "epoch": 20.52,
+      "learning_rate": 1.1333333333333334e-05,
+      "loss": 6.3147,
+      "step": 3400
+    },
+    {
+      "epoch": 20.64,
+      "learning_rate": 1.1400000000000001e-05,
+      "loss": 6.3324,
+      "step": 3420
+    },
+    {
+      "epoch": 20.76,
+      "learning_rate": 1.1466666666666666e-05,
+      "loss": 6.316,
+      "step": 3440
+    },
+    {
+      "epoch": 20.89,
+      "learning_rate": 1.1533333333333334e-05,
+      "loss": 6.2978,
+      "step": 3460
+    },
+    {
+      "epoch": 21.01,
+      "learning_rate": 1.16e-05,
+      "loss": 6.2895,
+      "step": 3480
+    },
+    {
+      "epoch": 21.13,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 6.2864,
+      "step": 3500
+    },
+    {
+      "epoch": 21.25,
+      "learning_rate": 1.1733333333333333e-05,
+      "loss": 6.2824,
+      "step": 3520
+    },
+    {
+      "epoch": 21.37,
+      "learning_rate": 1.18e-05,
+      "loss": 6.2451,
+      "step": 3540
+    },
+    {
+      "epoch": 21.49,
+      "learning_rate": 1.1866666666666668e-05,
+      "loss": 6.2508,
+      "step": 3560
+    },
+    {
+      "epoch": 21.61,
+      "learning_rate": 1.1933333333333333e-05,
+      "loss": 6.2341,
+      "step": 3580
+    },
+    {
+      "epoch": 21.73,
+      "learning_rate": 1.2e-05,
+      "loss": 6.2244,
+      "step": 3600
+    },
+    {
+      "epoch": 21.85,
+      "learning_rate": 1.2066666666666667e-05,
+      "loss": 6.2077,
+      "step": 3620
+    },
+    {
+      "epoch": 21.97,
+      "learning_rate": 1.2133333333333335e-05,
+      "loss": 6.2116,
+      "step": 3640
+    },
+    {
+      "epoch": 22.09,
+      "learning_rate": 1.22e-05,
+      "loss": 6.2047,
+      "step": 3660
+    },
+    {
+      "epoch": 22.21,
+      "learning_rate": 1.2266666666666667e-05,
+      "loss": 6.2012,
+      "step": 3680
+    },
+    {
+      "epoch": 22.33,
+      "learning_rate": 1.2333333333333334e-05,
+      "loss": 6.1741,
+      "step": 3700
+    },
+    {
+      "epoch": 22.45,
+      "learning_rate": 1.24e-05,
+      "loss": 6.1616,
+      "step": 3720
+    },
+    {
+      "epoch": 22.58,
+      "learning_rate": 1.2466666666666667e-05,
+      "loss": 6.1643,
+      "step": 3740
+    },
+    {
+      "epoch": 22.7,
+      "learning_rate": 1.2533333333333332e-05,
+      "loss": 6.1585,
+      "step": 3760
+    },
+    {
+      "epoch": 22.82,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 6.1467,
+      "step": 3780
+    },
+    {
+      "epoch": 22.94,
+      "learning_rate": 1.2666666666666668e-05,
+      "loss": 6.1542,
+      "step": 3800
+    },
+    {
+      "epoch": 23.06,
+      "learning_rate": 1.2733333333333334e-05,
+      "loss": 6.1341,
+      "step": 3820
+    },
+    {
+      "epoch": 23.18,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 6.1156,
+      "step": 3840
+    },
+    {
+      "epoch": 23.3,
+      "learning_rate": 1.2866666666666668e-05,
+      "loss": 6.1146,
+      "step": 3860
+    },
+    {
+      "epoch": 23.42,
+      "learning_rate": 1.2933333333333334e-05,
+      "loss": 6.109,
+      "step": 3880
+    },
+    {
+      "epoch": 23.54,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 6.0964,
+      "step": 3900
+    },
+    {
+      "epoch": 23.66,
+      "learning_rate": 1.3066666666666666e-05,
+      "loss": 6.0964,
+      "step": 3920
+    },
+    {
+      "epoch": 23.78,
+      "learning_rate": 1.3133333333333334e-05,
+      "loss": 6.0941,
+      "step": 3940
+    },
+    {
+      "epoch": 23.9,
+      "learning_rate": 1.32e-05,
+      "loss": 6.0569,
+      "step": 3960
+    },
+    {
+      "epoch": 24.02,
+      "learning_rate": 1.3266666666666666e-05,
+      "loss": 6.0643,
+      "step": 3980
+    },
+    {
+      "epoch": 24.14,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 6.0375,
+      "step": 4000
+    },
+    {
+      "epoch": 24.14,
+      "eval_accuracy": 4.494670604854244e-05,
+      "eval_loss": 6.053168773651123,
+      "eval_runtime": 18.3968,
+      "eval_samples_per_second": 110.889,
+      "eval_steps_per_second": 2.337,
+      "step": 4000
+    },
+    {
+      "epoch": 24.27,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 6.04,
+      "step": 4020
+    },
+    {
+      "epoch": 24.39,
+      "learning_rate": 1.3466666666666666e-05,
+      "loss": 6.0531,
+      "step": 4040
+    },
+    {
+      "epoch": 24.51,
+      "learning_rate": 1.3533333333333335e-05,
+      "loss": 6.0362,
+      "step": 4060
+    },
+    {
+      "epoch": 24.63,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 6.0139,
+      "step": 4080
+    },
+    {
+      "epoch": 24.75,
+      "learning_rate": 1.3666666666666666e-05,
+      "loss": 5.994,
+      "step": 4100
+    },
+    {
+      "epoch": 24.87,
+      "learning_rate": 1.3733333333333335e-05,
+      "loss": 6.009,
+      "step": 4120
+    },
+    {
+      "epoch": 24.99,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 5.9937,
+      "step": 4140
+    },
+    {
+      "epoch": 25.11,
+      "learning_rate": 1.3866666666666667e-05,
+      "loss": 5.9837,
+      "step": 4160
+    },
+    {
+      "epoch": 25.23,
+      "learning_rate": 1.3933333333333334e-05,
+      "loss": 5.9619,
+      "step": 4180
+    },
+    {
+      "epoch": 25.35,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 5.9603,
+      "step": 4200
+    },
+    {
+      "epoch": 25.47,
+      "learning_rate": 1.4066666666666667e-05,
+      "loss": 5.9447,
+      "step": 4220
+    },
+    {
+      "epoch": 25.59,
+      "learning_rate": 1.4133333333333334e-05,
+      "loss": 5.9396,
+      "step": 4240
+    },
+    {
+      "epoch": 25.71,
+      "learning_rate": 1.42e-05,
+      "loss": 5.9694,
+      "step": 4260
+    },
+    {
+      "epoch": 25.84,
+      "learning_rate": 1.4266666666666667e-05,
+      "loss": 5.9305,
+      "step": 4280
+    },
+    {
+      "epoch": 25.96,
+      "learning_rate": 1.4333333333333334e-05,
+      "loss": 5.9412,
+      "step": 4300
+    },
+    {
+      "epoch": 26.08,
+      "learning_rate": 1.44e-05,
+      "loss": 5.9241,
+      "step": 4320
+    },
+    {
+      "epoch": 26.2,
+      "learning_rate": 1.4466666666666667e-05,
+      "loss": 5.9065,
+      "step": 4340
+    },
+    {
+      "epoch": 26.32,
+      "learning_rate": 1.4533333333333335e-05,
+      "loss": 5.912,
+      "step": 4360
+    },
+    {
+      "epoch": 26.44,
+      "learning_rate": 1.4599999999999999e-05,
+      "loss": 5.8877,
+      "step": 4380
+    },
+    {
+      "epoch": 26.56,
+      "learning_rate": 1.4666666666666668e-05,
+      "loss": 5.8726,
+      "step": 4400
+    },
+    {
+      "epoch": 26.68,
+      "learning_rate": 1.4733333333333335e-05,
+      "loss": 5.8834,
+      "step": 4420
+    },
+    {
+      "epoch": 26.8,
+      "learning_rate": 1.48e-05,
+      "loss": 5.88,
+      "step": 4440
+    },
+    {
+      "epoch": 26.92,
+      "learning_rate": 1.4866666666666668e-05,
+      "loss": 5.8724,
+      "step": 4460
+    },
+    {
+      "epoch": 27.04,
+      "learning_rate": 1.4933333333333335e-05,
+      "loss": 5.864,
+      "step": 4480
+    },
+    {
+      "epoch": 27.16,
+      "learning_rate": 1.5e-05,
+      "loss": 5.8542,
+      "step": 4500
+    },
+    {
+      "epoch": 27.28,
+      "learning_rate": 1.5066666666666668e-05,
+      "loss": 5.858,
+      "step": 4520
+    },
+    {
+      "epoch": 27.4,
+      "learning_rate": 1.5133333333333333e-05,
+      "loss": 5.8406,
+      "step": 4540
+    },
+    {
+      "epoch": 27.53,
+      "learning_rate": 1.52e-05,
+      "loss": 5.8369,
+      "step": 4560
+    },
+    {
+      "epoch": 27.65,
+      "learning_rate": 1.5266666666666667e-05,
+      "loss": 5.8308,
+      "step": 4580
+    },
+    {
+      "epoch": 27.77,
+      "learning_rate": 1.5333333333333334e-05,
+      "loss": 5.8009,
+      "step": 4600
+    },
+    {
+      "epoch": 27.89,
+      "learning_rate": 1.54e-05,
+      "loss": 5.8264,
+      "step": 4620
+    },
+    {
+      "epoch": 28.01,
+      "learning_rate": 1.546666666666667e-05,
+      "loss": 5.7941,
+      "step": 4640
+    },
+    {
+      "epoch": 28.13,
+      "learning_rate": 1.5533333333333333e-05,
+      "loss": 5.8114,
+      "step": 4660
+    },
+    {
+      "epoch": 28.25,
+      "learning_rate": 1.56e-05,
+      "loss": 5.7903,
+      "step": 4680
+    },
+    {
+      "epoch": 28.37,
+      "learning_rate": 1.5666666666666667e-05,
+      "loss": 5.7821,
+      "step": 4700
+    },
+    {
+      "epoch": 28.49,
+      "learning_rate": 1.5733333333333334e-05,
+      "loss": 5.7837,
+      "step": 4720
+    },
+    {
+      "epoch": 28.61,
+      "learning_rate": 1.58e-05,
+      "loss": 5.774,
+      "step": 4740
+    },
+    {
+      "epoch": 28.73,
+      "learning_rate": 1.586666666666667e-05,
+      "loss": 5.7587,
+      "step": 4760
+    },
+    {
+      "epoch": 28.85,
+      "learning_rate": 1.5933333333333332e-05,
+      "loss": 5.7502,
+      "step": 4780
+    },
+    {
+      "epoch": 28.97,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 5.7666,
+      "step": 4800
+    },
+    {
+      "epoch": 29.09,
+      "learning_rate": 1.606666666666667e-05,
+      "loss": 5.7418,
+      "step": 4820
+    },
+    {
+      "epoch": 29.22,
+      "learning_rate": 1.6133333333333334e-05,
+      "loss": 5.7404,
+      "step": 4840
+    },
+    {
+      "epoch": 29.34,
+      "learning_rate": 1.62e-05,
+      "loss": 5.7338,
+      "step": 4860
+    },
+    {
+      "epoch": 29.46,
+      "learning_rate": 1.6266666666666665e-05,
+      "loss": 5.7212,
+      "step": 4880
+    },
+    {
+      "epoch": 29.58,
+      "learning_rate": 1.6333333333333335e-05,
+      "loss": 5.7091,
+      "step": 4900
+    },
+    {
+      "epoch": 29.7,
+      "learning_rate": 1.6400000000000002e-05,
+      "loss": 5.7103,
+      "step": 4920
+    },
+    {
+      "epoch": 29.82,
+      "learning_rate": 1.6466666666666666e-05,
+      "loss": 5.725,
+      "step": 4940
+    },
+    {
+      "epoch": 29.94,
+      "learning_rate": 1.6533333333333333e-05,
+      "loss": 5.7061,
+      "step": 4960
+    },
+    {
+      "epoch": 30.06,
+      "learning_rate": 1.66e-05,
+      "loss": 5.6879,
+      "step": 4980
+    },
+    {
+      "epoch": 30.18,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 5.6908,
+      "step": 5000
+    },
+    {
+      "epoch": 30.18,
+      "eval_accuracy": 6.618525945609546e-05,
+      "eval_loss": 5.7099995613098145,
+      "eval_runtime": 18.2382,
+      "eval_samples_per_second": 111.853,
+      "eval_steps_per_second": 2.358,
+      "step": 5000
+    },
+    {
+      "epoch": 30.3,
+      "learning_rate": 1.6733333333333335e-05,
+      "loss": 5.6782,
+      "step": 5020
+    },
+    {
+      "epoch": 30.42,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 5.6745,
+      "step": 5040
+    },
+    {
+      "epoch": 30.54,
+      "learning_rate": 1.6866666666666666e-05,
+      "loss": 5.6698,
+      "step": 5060
+    },
+    {
+      "epoch": 30.66,
+      "learning_rate": 1.6933333333333333e-05,
+      "loss": 5.68,
+      "step": 5080
+    },
+    {
+      "epoch": 30.78,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 5.6718,
+      "step": 5100
+    },
+    {
+      "epoch": 30.91,
+      "learning_rate": 1.7066666666666667e-05,
+      "loss": 5.6621,
+      "step": 5120
+    },
+    {
+      "epoch": 31.03,
+      "learning_rate": 1.7133333333333334e-05,
+      "loss": 5.6677,
+      "step": 5140
+    },
+    {
+      "epoch": 31.15,
+      "learning_rate": 1.7199999999999998e-05,
+      "loss": 5.6582,
+      "step": 5160
+    },
+    {
+      "epoch": 31.27,
+      "learning_rate": 1.726666666666667e-05,
+      "loss": 5.6369,
+      "step": 5180
+    },
+    {
+      "epoch": 31.39,
+      "learning_rate": 1.7333333333333336e-05,
+      "loss": 5.631,
+      "step": 5200
+    },
+    {
+      "epoch": 31.51,
+      "learning_rate": 1.74e-05,
+      "loss": 5.6239,
+      "step": 5220
+    },
+    {
+      "epoch": 31.63,
+      "learning_rate": 1.7466666666666667e-05,
+      "loss": 5.6244,
+      "step": 5240
+    },
+    {
+      "epoch": 31.75,
+      "learning_rate": 1.7533333333333334e-05,
+      "loss": 5.629,
+      "step": 5260
+    },
+    {
+      "epoch": 31.87,
+      "learning_rate": 1.76e-05,
+      "loss": 5.6257,
+      "step": 5280
+    },
+    {
+      "epoch": 31.99,
+      "learning_rate": 1.7666666666666668e-05,
+      "loss": 5.6067,
+      "step": 5300
+    },
+    {
+      "epoch": 32.11,
+      "learning_rate": 1.7733333333333335e-05,
+      "loss": 5.5915,
+      "step": 5320
+    },
+    {
+      "epoch": 32.23,
+      "learning_rate": 1.78e-05,
+      "loss": 5.6116,
+      "step": 5340
+    },
+    {
+      "epoch": 32.35,
+      "learning_rate": 1.7866666666666666e-05,
+      "loss": 5.6139,
+      "step": 5360
+    },
+    {
+      "epoch": 32.47,
+      "learning_rate": 1.7933333333333337e-05,
+      "loss": 5.5716,
+      "step": 5380
+    },
+    {
+      "epoch": 32.6,
+      "learning_rate": 1.8e-05,
+      "loss": 5.5983,
+      "step": 5400
+    },
+    {
+      "epoch": 32.72,
+      "learning_rate": 1.8066666666666668e-05,
+      "loss": 5.5821,
+      "step": 5420
+    },
+    {
+      "epoch": 32.84,
+      "learning_rate": 1.8133333333333335e-05,
+      "loss": 5.5799,
+      "step": 5440
+    },
+    {
+      "epoch": 32.96,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 5.558,
+      "step": 5460
+    },
+    {
+      "epoch": 33.08,
+      "learning_rate": 1.826666666666667e-05,
+      "loss": 5.5665,
+      "step": 5480
+    },
+    {
+      "epoch": 33.2,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 5.5493,
+      "step": 5500
+    },
+    {
+      "epoch": 33.32,
+      "learning_rate": 1.84e-05,
+      "loss": 5.5597,
+      "step": 5520
+    },
+    {
+      "epoch": 33.44,
+      "learning_rate": 1.8466666666666667e-05,
+      "loss": 5.5507,
+      "step": 5540
+    },
+    {
+      "epoch": 33.56,
+      "learning_rate": 1.8533333333333334e-05,
+      "loss": 5.5527,
+      "step": 5560
+    },
+    {
+      "epoch": 33.68,
+      "learning_rate": 1.86e-05,
+      "loss": 5.5438,
+      "step": 5580
+    },
+    {
+      "epoch": 33.8,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 5.5427,
+      "step": 5600
+    },
+    {
+      "epoch": 33.92,
+      "learning_rate": 1.8733333333333332e-05,
+      "loss": 5.5477,
+      "step": 5620
+    },
+    {
+      "epoch": 34.04,
+      "learning_rate": 1.88e-05,
+      "loss": 5.5087,
+      "step": 5640
+    },
+    {
+      "epoch": 34.16,
+      "learning_rate": 1.886666666666667e-05,
+      "loss": 5.5358,
+      "step": 5660
+    },
+    {
+      "epoch": 34.29,
+      "learning_rate": 1.8933333333333334e-05,
+      "loss": 5.5069,
+      "step": 5680
+    },
+    {
+      "epoch": 34.41,
+      "learning_rate": 1.9e-05,
+      "loss": 5.5231,
+      "step": 5700
+    },
+    {
+      "epoch": 34.53,
+      "learning_rate": 1.9066666666666668e-05,
+      "loss": 5.5083,
+      "step": 5720
+    },
+    {
+      "epoch": 34.65,
+      "learning_rate": 1.9133333333333332e-05,
+      "loss": 5.5009,
+      "step": 5740
+    },
+    {
+      "epoch": 34.77,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 5.5038,
+      "step": 5760
+    },
+    {
+      "epoch": 34.89,
+      "learning_rate": 1.926666666666667e-05,
+      "loss": 5.5063,
+      "step": 5780
+    },
+    {
+      "epoch": 35.01,
+      "learning_rate": 1.9333333333333333e-05,
+      "loss": 5.4864,
+      "step": 5800
+    },
+    {
+      "epoch": 35.13,
+      "learning_rate": 1.94e-05,
+      "loss": 5.4737,
+      "step": 5820
+    },
+    {
+      "epoch": 35.25,
+      "learning_rate": 1.9466666666666668e-05,
+      "loss": 5.4905,
+      "step": 5840
+    },
+    {
+      "epoch": 35.37,
+      "learning_rate": 1.9533333333333335e-05,
+      "loss": 5.501,
+      "step": 5860
+    },
+    {
+      "epoch": 35.49,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 5.4765,
+      "step": 5880
+    },
+    {
+      "epoch": 35.61,
+      "learning_rate": 1.9666666666666666e-05,
+      "loss": 5.4575,
+      "step": 5900
+    },
+    {
+      "epoch": 35.73,
+      "learning_rate": 1.9733333333333333e-05,
+      "loss": 5.4521,
+      "step": 5920
+    },
+    {
+      "epoch": 35.86,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 5.4681,
+      "step": 5940
+    },
+    {
+      "epoch": 35.98,
+      "learning_rate": 1.9866666666666667e-05,
+      "loss": 5.4621,
+      "step": 5960
+    },
+    {
+      "epoch": 36.1,
+      "learning_rate": 1.9933333333333334e-05,
+      "loss": 5.4473,
+      "step": 5980
+    },
+    {
+      "epoch": 36.22,
+      "learning_rate": 2e-05,
+      "loss": 5.4294,
+      "step": 6000
+    },
+    {
+      "epoch": 36.22,
+      "eval_accuracy": 0.0002054706562219083,
+      "eval_loss": 5.475820541381836,
+      "eval_runtime": 18.2393,
+      "eval_samples_per_second": 111.846,
+      "eval_steps_per_second": 2.358,
+      "step": 6000
+    },
+    {
+      "epoch": 36.34,
+      "learning_rate": 2.0066666666666665e-05,
+      "loss": 5.4475,
+      "step": 6020
+    },
+    {
+      "epoch": 36.46,
+      "learning_rate": 2.0133333333333336e-05,
+      "loss": 5.4346,
+      "step": 6040
+    },
+    {
+      "epoch": 36.58,
+      "learning_rate": 2.0200000000000003e-05,
+      "loss": 5.4358,
+      "step": 6060
+    },
+    {
+      "epoch": 36.7,
+      "learning_rate": 2.0266666666666667e-05,
+      "loss": 5.4414,
+      "step": 6080
+    },
+    {
+      "epoch": 36.82,
+      "learning_rate": 2.0333333333333334e-05,
+      "loss": 5.431,
+      "step": 6100
+    },
+    {
+      "epoch": 36.94,
+      "learning_rate": 2.04e-05,
+      "loss": 5.4276,
+      "step": 6120
+    },
+    {
+      "epoch": 37.06,
+      "learning_rate": 2.046666666666667e-05,
+      "loss": 5.4126,
+      "step": 6140
+    },
+    {
+      "epoch": 37.18,
+      "learning_rate": 2.0533333333333336e-05,
+      "loss": 5.4055,
+      "step": 6160
+    },
+    {
+      "epoch": 37.3,
+      "learning_rate": 2.06e-05,
+      "loss": 5.4142,
+      "step": 6180
+    },
+    {
+      "epoch": 37.42,
+      "learning_rate": 2.0666666666666666e-05,
+      "loss": 5.4029,
+      "step": 6200
+    },
+    {
+      "epoch": 37.55,
+      "learning_rate": 2.0733333333333334e-05,
+      "loss": 5.3825,
+      "step": 6220
+    },
+    {
+      "epoch": 37.67,
+      "learning_rate": 2.08e-05,
+      "loss": 5.3948,
+      "step": 6240
+    },
+    {
+      "epoch": 37.79,
+      "learning_rate": 2.0866666666666668e-05,
+      "loss": 5.4173,
+      "step": 6260
+    },
+    {
+      "epoch": 37.91,
+      "learning_rate": 2.0933333333333335e-05,
+      "loss": 5.3965,
+      "step": 6280
+    },
+    {
+      "epoch": 38.03,
+      "learning_rate": 2.1e-05,
+      "loss": 5.3943,
+      "step": 6300
+    },
+    {
+      "epoch": 38.15,
+      "learning_rate": 2.106666666666667e-05,
+      "loss": 5.372,
+      "step": 6320
+    },
+    {
+      "epoch": 38.27,
+      "learning_rate": 2.1133333333333337e-05,
+      "loss": 5.3875,
+      "step": 6340
+    },
+    {
+      "epoch": 38.39,
+      "learning_rate": 2.12e-05,
+      "loss": 5.3339,
+      "step": 6360
+    },
+    {
+      "epoch": 38.51,
+      "learning_rate": 2.1266666666666667e-05,
+      "loss": 5.3707,
+      "step": 6380
+    },
+    {
+      "epoch": 38.63,
+      "learning_rate": 2.1333333333333335e-05,
+      "loss": 5.3882,
+      "step": 6400
+    },
+    {
+      "epoch": 38.75,
+      "learning_rate": 2.1400000000000002e-05,
+      "loss": 5.3582,
+      "step": 6420
+    },
+    {
+      "epoch": 38.87,
+      "learning_rate": 2.146666666666667e-05,
+      "loss": 5.3476,
+      "step": 6440
+    },
+    {
+      "epoch": 38.99,
+      "learning_rate": 2.1533333333333333e-05,
+      "loss": 5.3561,
+      "step": 6460
+    },
+    {
+      "epoch": 39.11,
+      "learning_rate": 2.16e-05,
+      "loss": 5.3582,
+      "step": 6480
+    },
+    {
+      "epoch": 39.24,
+      "learning_rate": 2.1666666666666667e-05,
+      "loss": 5.3456,
+      "step": 6500
+    },
+    {
+      "epoch": 39.36,
+      "learning_rate": 2.1733333333333334e-05,
+      "loss": 5.3251,
+      "step": 6520
+    },
+    {
+      "epoch": 39.48,
+      "learning_rate": 2.18e-05,
+      "loss": 5.3242,
+      "step": 6540
+    },
+    {
+      "epoch": 39.6,
+      "learning_rate": 2.186666666666667e-05,
+      "loss": 5.3291,
+      "step": 6560
+    },
+    {
+      "epoch": 39.72,
+      "learning_rate": 2.1933333333333332e-05,
+      "loss": 5.3198,
+      "step": 6580
+    },
+    {
+      "epoch": 39.84,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 5.3222,
+      "step": 6600
+    },
+    {
+      "epoch": 39.96,
+      "learning_rate": 2.206666666666667e-05,
+      "loss": 5.3368,
+      "step": 6620
+    },
+    {
+      "epoch": 40.08,
+      "learning_rate": 2.2133333333333334e-05,
+      "loss": 5.2988,
+      "step": 6640
+    },
+    {
+      "epoch": 40.2,
+      "learning_rate": 2.22e-05,
+      "loss": 5.2908,
+      "step": 6660
+    },
+    {
+      "epoch": 40.32,
+      "learning_rate": 2.2266666666666668e-05,
+      "loss": 5.2934,
+      "step": 6680
+    },
+    {
+      "epoch": 40.44,
+      "learning_rate": 2.2333333333333335e-05,
+      "loss": 5.3153,
+      "step": 6700
+    },
+    {
+      "epoch": 40.56,
+      "learning_rate": 2.2400000000000002e-05,
+      "loss": 5.295,
+      "step": 6720
+    },
+    {
+      "epoch": 40.68,
+      "learning_rate": 2.2466666666666666e-05,
+      "loss": 5.2895,
+      "step": 6740
+    },
+    {
+      "epoch": 40.8,
+      "learning_rate": 2.2533333333333333e-05,
+      "loss": 5.3141,
+      "step": 6760
+    },
+    {
+      "epoch": 40.93,
+      "learning_rate": 2.26e-05,
+      "loss": 5.2998,
+      "step": 6780
+    },
+    {
+      "epoch": 41.05,
+      "learning_rate": 2.2666666666666668e-05,
+      "loss": 5.2623,
+      "step": 6800
+    },
+    {
+      "epoch": 41.17,
+      "learning_rate": 2.2733333333333335e-05,
+      "loss": 5.2713,
+      "step": 6820
+    },
+    {
+      "epoch": 41.29,
+      "learning_rate": 2.2800000000000002e-05,
+      "loss": 5.269,
+      "step": 6840
+    },
+    {
+      "epoch": 41.41,
+      "learning_rate": 2.2866666666666666e-05,
+      "loss": 5.2645,
+      "step": 6860
+    },
+    {
+      "epoch": 41.53,
+      "learning_rate": 2.2933333333333333e-05,
+      "loss": 5.2724,
+      "step": 6880
+    },
+    {
+      "epoch": 41.65,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 5.2596,
+      "step": 6900
+    },
+    {
+      "epoch": 41.77,
+      "learning_rate": 2.3066666666666667e-05,
+      "loss": 5.256,
+      "step": 6920
+    },
+    {
+      "epoch": 41.89,
+      "learning_rate": 2.3133333333333334e-05,
+      "loss": 5.2449,
+      "step": 6940
+    },
+    {
+      "epoch": 42.01,
+      "learning_rate": 2.32e-05,
+      "loss": 5.2506,
+      "step": 6960
+    },
+    {
+      "epoch": 42.13,
+      "learning_rate": 2.326666666666667e-05,
+      "loss": 5.2261,
+      "step": 6980
+    },
+    {
+      "epoch": 42.25,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 5.2161,
+      "step": 7000
+    },
+    {
+      "epoch": 42.25,
+      "eval_accuracy": 0.000646541079313649,
+      "eval_loss": 5.289076805114746,
+      "eval_runtime": 18.2424,
+      "eval_samples_per_second": 111.828,
+      "eval_steps_per_second": 2.357,
+      "step": 7000
+    },
+    {
+      "epoch": 42.37,
+      "learning_rate": 2.3400000000000003e-05,
+      "loss": 5.2372,
+      "step": 7020
+    },
+    {
+      "epoch": 42.49,
+      "learning_rate": 2.3466666666666667e-05,
+      "loss": 5.2254,
+      "step": 7040
+    },
+    {
+      "epoch": 42.62,
+      "learning_rate": 2.3533333333333334e-05,
+      "loss": 5.2269,
+      "step": 7060
+    },
+    {
+      "epoch": 42.74,
+      "learning_rate": 2.36e-05,
+      "loss": 5.2126,
+      "step": 7080
+    },
+    {
+      "epoch": 42.86,
+      "learning_rate": 2.3666666666666668e-05,
+      "loss": 5.2207,
+      "step": 7100
+    },
+    {
+      "epoch": 42.98,
+      "learning_rate": 2.3733333333333335e-05,
+      "loss": 5.2211,
+      "step": 7120
+    },
+    {
+      "epoch": 43.1,
+      "learning_rate": 2.38e-05,
+      "loss": 5.1833,
+      "step": 7140
+    },
+    {
+      "epoch": 43.22,
+      "learning_rate": 2.3866666666666666e-05,
+      "loss": 5.1911,
+      "step": 7160
+    },
+    {
+      "epoch": 43.34,
+      "learning_rate": 2.3933333333333337e-05,
+      "loss": 5.1874,
+      "step": 7180
+    },
+    {
+      "epoch": 43.46,
+      "learning_rate": 2.4e-05,
+      "loss": 5.2008,
+      "step": 7200
+    },
+    {
+      "epoch": 43.58,
+      "learning_rate": 2.4066666666666668e-05,
+      "loss": 5.195,
+      "step": 7220
+    },
+    {
+      "epoch": 43.7,
+      "learning_rate": 2.4133333333333335e-05,
+      "loss": 5.183,
+      "step": 7240
+    },
+    {
+      "epoch": 43.82,
+      "learning_rate": 2.4200000000000002e-05,
+      "loss": 5.1729,
+      "step": 7260
+    },
+    {
+      "epoch": 43.94,
+      "learning_rate": 2.426666666666667e-05,
+      "loss": 5.1818,
+      "step": 7280
+    },
+    {
+      "epoch": 44.06,
+      "learning_rate": 2.4333333333333336e-05,
+      "loss": 5.1615,
+      "step": 7300
+    },
+    {
+      "epoch": 44.19,
+      "learning_rate": 2.44e-05,
+      "loss": 5.1769,
+      "step": 7320
+    },
+    {
+      "epoch": 44.31,
+      "learning_rate": 2.4466666666666667e-05,
+      "loss": 5.1479,
+      "step": 7340
+    },
+    {
+      "epoch": 44.43,
+      "learning_rate": 2.4533333333333334e-05,
+      "loss": 5.1789,
+      "step": 7360
+    },
+    {
+      "epoch": 44.55,
+      "learning_rate": 2.46e-05,
+      "loss": 5.1305,
+      "step": 7380
+    },
+    {
+      "epoch": 44.67,
+      "learning_rate": 2.466666666666667e-05,
+      "loss": 5.1425,
+      "step": 7400
+    },
+    {
+      "epoch": 44.79,
+      "learning_rate": 2.4733333333333333e-05,
+      "loss": 5.1452,
+      "step": 7420
+    },
+    {
+      "epoch": 44.91,
+      "learning_rate": 2.48e-05,
+      "loss": 5.1387,
+      "step": 7440
+    },
+    {
+      "epoch": 45.03,
+      "learning_rate": 2.486666666666667e-05,
+      "loss": 5.1278,
+      "step": 7460
+    },
+    {
+      "epoch": 45.15,
+      "learning_rate": 2.4933333333333334e-05,
+      "loss": 5.1048,
+      "step": 7480
+    },
+    {
+      "epoch": 45.27,
+      "learning_rate": 2.5e-05,
+      "loss": 5.1143,
+      "step": 7500
+    },
+    {
+      "epoch": 45.39,
+      "learning_rate": 2.5066666666666665e-05,
+      "loss": 5.1258,
+      "step": 7520
+    },
+    {
+      "epoch": 45.51,
+      "learning_rate": 2.5133333333333336e-05,
+      "loss": 5.1237,
+      "step": 7540
+    },
+    {
+      "epoch": 45.63,
+      "learning_rate": 2.5200000000000003e-05,
+      "loss": 5.1253,
+      "step": 7560
+    },
+    {
+      "epoch": 45.75,
+      "learning_rate": 2.5266666666666666e-05,
+      "loss": 5.1031,
+      "step": 7580
+    },
+    {
+      "epoch": 45.88,
+      "learning_rate": 2.5333333333333337e-05,
+      "loss": 5.0993,
+      "step": 7600
+    },
+    {
+      "epoch": 46.0,
+      "learning_rate": 2.54e-05,
+      "loss": 5.098,
+      "step": 7620
+    },
+    {
+      "epoch": 46.12,
+      "learning_rate": 2.5466666666666668e-05,
+      "loss": 5.0804,
+      "step": 7640
+    },
+    {
+      "epoch": 46.24,
+      "learning_rate": 2.553333333333334e-05,
+      "loss": 5.0841,
+      "step": 7660
+    },
+    {
+      "epoch": 46.36,
+      "learning_rate": 2.5600000000000002e-05,
+      "loss": 5.085,
+      "step": 7680
+    },
+    {
+      "epoch": 46.48,
+      "learning_rate": 2.5666666666666666e-05,
+      "loss": 5.0754,
+      "step": 7700
+    },
+    {
+      "epoch": 46.6,
+      "learning_rate": 2.5733333333333337e-05,
+      "loss": 5.0788,
+      "step": 7720
+    },
+    {
+      "epoch": 46.72,
+      "learning_rate": 2.58e-05,
+      "loss": 5.0837,
+      "step": 7740
+    },
+    {
+      "epoch": 46.84,
+      "learning_rate": 2.5866666666666667e-05,
+      "loss": 5.0867,
+      "step": 7760
+    },
+    {
+      "epoch": 46.96,
+      "learning_rate": 2.5933333333333338e-05,
+      "loss": 5.0632,
+      "step": 7780
+    },
+    {
+      "epoch": 47.08,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 5.0334,
+      "step": 7800
+    },
+    {
+      "epoch": 47.2,
+      "learning_rate": 2.6066666666666666e-05,
+      "loss": 5.051,
+      "step": 7820
+    },
+    {
+      "epoch": 47.32,
+      "learning_rate": 2.6133333333333333e-05,
+      "loss": 5.0631,
+      "step": 7840
+    },
+    {
+      "epoch": 47.44,
+      "learning_rate": 2.6200000000000003e-05,
+      "loss": 5.0541,
+      "step": 7860
+    },
+    {
+      "epoch": 47.57,
+      "learning_rate": 2.6266666666666667e-05,
+      "loss": 5.0481,
+      "step": 7880
+    },
+    {
+      "epoch": 47.69,
+      "learning_rate": 2.633333333333333e-05,
+      "loss": 5.03,
+      "step": 7900
+    },
+    {
+      "epoch": 47.81,
+      "learning_rate": 2.64e-05,
+      "loss": 5.0424,
+      "step": 7920
+    },
+    {
+      "epoch": 47.93,
+      "learning_rate": 2.646666666666667e-05,
+      "loss": 5.0613,
+      "step": 7940
+    },
+    {
+      "epoch": 48.05,
+      "learning_rate": 2.6533333333333332e-05,
+      "loss": 5.0139,
+      "step": 7960
+    },
+    {
+      "epoch": 48.17,
+      "learning_rate": 2.6600000000000003e-05,
+      "loss": 5.0153,
+      "step": 7980
+    },
+    {
+      "epoch": 48.29,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 5.0151,
+      "step": 8000
+    },
+    {
+      "epoch": 48.29,
+      "eval_accuracy": 0.0020700180774663886,
+      "eval_loss": 5.1151652336120605,
+      "eval_runtime": 18.1374,
+      "eval_samples_per_second": 112.475,
+      "eval_steps_per_second": 2.371,
+      "step": 8000
+    },
+    {
+      "epoch": 48.41,
+      "learning_rate": 2.6733333333333334e-05,
+      "loss": 5.0138,
+      "step": 8020
+    },
+    {
+      "epoch": 48.53,
+      "learning_rate": 2.6800000000000004e-05,
+      "loss": 5.0247,
+      "step": 8040
+    },
+    {
+      "epoch": 48.65,
+      "learning_rate": 2.6866666666666668e-05,
+      "loss": 5.0113,
+      "step": 8060
+    },
+    {
+      "epoch": 48.77,
+      "learning_rate": 2.6933333333333332e-05,
+      "loss": 5.0084,
+      "step": 8080
+    },
+    {
+      "epoch": 48.89,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 5.018,
+      "step": 8100
+    },
+    {
+      "epoch": 49.01,
+      "learning_rate": 2.706666666666667e-05,
+      "loss": 4.9896,
+      "step": 8120
+    },
+    {
+      "epoch": 49.13,
+      "learning_rate": 2.7133333333333333e-05,
+      "loss": 4.9871,
+      "step": 8140
+    },
+    {
+      "epoch": 49.26,
+      "learning_rate": 2.7200000000000004e-05,
+      "loss": 4.979,
+      "step": 8160
+    },
+    {
+      "epoch": 49.38,
+      "learning_rate": 2.7266666666666668e-05,
+      "loss": 4.9857,
+      "step": 8180
+    },
+    {
+      "epoch": 49.5,
+      "learning_rate": 2.733333333333333e-05,
+      "loss": 4.9877,
+      "step": 8200
+    },
+    {
+      "epoch": 49.62,
+      "learning_rate": 2.7400000000000002e-05,
+      "loss": 5.0132,
+      "step": 8220
+    },
+    {
+      "epoch": 49.74,
+      "learning_rate": 2.746666666666667e-05,
+      "loss": 4.9614,
+      "step": 8240
+    },
+    {
+      "epoch": 49.86,
+      "learning_rate": 2.7533333333333333e-05,
+      "loss": 4.9745,
+      "step": 8260
+    },
+    {
+      "epoch": 49.98,
+      "learning_rate": 2.7600000000000003e-05,
+      "loss": 4.9696,
+      "step": 8280
+    },
+    {
+      "epoch": 50.1,
+      "learning_rate": 2.7666666666666667e-05,
+      "loss": 4.957,
+      "step": 8300
+    },
+    {
+      "epoch": 50.22,
+      "learning_rate": 2.7733333333333334e-05,
+      "loss": 4.9363,
+      "step": 8320
+    },
+    {
+      "epoch": 50.34,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 4.9432,
+      "step": 8340
+    },
+    {
+      "epoch": 50.46,
+      "learning_rate": 2.786666666666667e-05,
+      "loss": 4.9699,
+      "step": 8360
+    },
+    {
+      "epoch": 50.58,
+      "learning_rate": 2.7933333333333332e-05,
+      "loss": 4.9538,
+      "step": 8380
+    },
+    {
+      "epoch": 50.7,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 4.9687,
+      "step": 8400
+    },
+    {
+      "epoch": 50.82,
+      "learning_rate": 2.806666666666667e-05,
+      "loss": 4.9383,
+      "step": 8420
+    },
+    {
+      "epoch": 50.95,
+      "learning_rate": 2.8133333333333334e-05,
+      "loss": 4.9384,
+      "step": 8440
+    },
+    {
+      "epoch": 51.07,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 4.9388,
+      "step": 8460
+    },
+    {
+      "epoch": 51.19,
+      "learning_rate": 2.8266666666666668e-05,
+      "loss": 4.9146,
+      "step": 8480
+    },
+    {
+      "epoch": 51.31,
+      "learning_rate": 2.8333333333333335e-05,
+      "loss": 4.9131,
+      "step": 8500
+    },
+    {
+      "epoch": 51.43,
+      "learning_rate": 2.84e-05,
+      "loss": 4.9165,
+      "step": 8520
+    },
+    {
+      "epoch": 51.55,
+      "learning_rate": 2.846666666666667e-05,
+      "loss": 4.9345,
+      "step": 8540
+    },
+    {
+      "epoch": 51.67,
+      "learning_rate": 2.8533333333333333e-05,
+      "loss": 4.9299,
+      "step": 8560
+    },
+    {
+      "epoch": 51.79,
+      "learning_rate": 2.86e-05,
+      "loss": 4.9118,
+      "step": 8580
+    },
+    {
+      "epoch": 51.91,
+      "learning_rate": 2.8666666666666668e-05,
+      "loss": 4.9335,
+      "step": 8600
+    },
+    {
+      "epoch": 52.03,
+      "learning_rate": 2.8733333333333335e-05,
+      "loss": 4.9034,
+      "step": 8620
+    },
+    {
+      "epoch": 52.15,
+      "learning_rate": 2.88e-05,
+      "loss": 4.8897,
+      "step": 8640
+    },
+    {
+      "epoch": 52.27,
+      "learning_rate": 2.886666666666667e-05,
+      "loss": 4.8783,
+      "step": 8660
+    },
+    {
+      "epoch": 52.39,
+      "learning_rate": 2.8933333333333333e-05,
+      "loss": 4.9061,
+      "step": 8680
+    },
+    {
+      "epoch": 52.52,
+      "learning_rate": 2.9e-05,
+      "loss": 4.8954,
+      "step": 8700
+    },
+    {
+      "epoch": 52.64,
+      "learning_rate": 2.906666666666667e-05,
+      "loss": 4.8877,
+      "step": 8720
+    },
+    {
+      "epoch": 52.76,
+      "learning_rate": 2.9133333333333334e-05,
+      "loss": 4.893,
+      "step": 8740
+    },
+    {
+      "epoch": 52.88,
+      "learning_rate": 2.9199999999999998e-05,
+      "loss": 4.868,
+      "step": 8760
+    },
+    {
+      "epoch": 53.0,
+      "learning_rate": 2.926666666666667e-05,
+      "loss": 4.9046,
+      "step": 8780
+    },
+    {
+      "epoch": 53.12,
+      "learning_rate": 2.9333333333333336e-05,
+      "loss": 4.8657,
+      "step": 8800
+    },
+    {
+      "epoch": 53.24,
+      "learning_rate": 2.94e-05,
+      "loss": 4.8584,
+      "step": 8820
+    },
+    {
+      "epoch": 53.36,
+      "learning_rate": 2.946666666666667e-05,
+      "loss": 4.8433,
+      "step": 8840
+    },
+    {
+      "epoch": 53.48,
+      "learning_rate": 2.9533333333333334e-05,
+      "loss": 4.8586,
+      "step": 8860
+    },
+    {
+      "epoch": 53.6,
+      "learning_rate": 2.96e-05,
+      "loss": 4.875,
+      "step": 8880
+    },
+    {
+      "epoch": 53.72,
+      "learning_rate": 2.9666666666666672e-05,
+      "loss": 4.8812,
+      "step": 8900
+    },
+    {
+      "epoch": 53.84,
+      "learning_rate": 2.9733333333333336e-05,
+      "loss": 4.8584,
+      "step": 8920
+    },
+    {
+      "epoch": 53.96,
+      "learning_rate": 2.98e-05,
+      "loss": 4.8654,
+      "step": 8940
+    },
+    {
+      "epoch": 54.08,
+      "learning_rate": 2.986666666666667e-05,
+      "loss": 4.8279,
+      "step": 8960
+    },
+    {
+      "epoch": 54.21,
+      "learning_rate": 2.9933333333333337e-05,
+      "loss": 4.8162,
+      "step": 8980
+    },
+    {
+      "epoch": 54.33,
+      "learning_rate": 3e-05,
+      "loss": 4.8349,
+      "step": 9000
+    },
+    {
+      "epoch": 54.33,
+      "eval_accuracy": 0.0019850638638361767,
+      "eval_loss": 4.9847259521484375,
+      "eval_runtime": 18.0947,
+      "eval_samples_per_second": 112.74,
+      "eval_steps_per_second": 2.376,
+      "step": 9000
+    },
+    {
+      "epoch": 54.45,
+      "learning_rate": 3.006666666666667e-05,
+      "loss": 4.8558,
+      "step": 9020
+    },
+    {
+      "epoch": 54.57,
+      "learning_rate": 3.0133333333333335e-05,
+      "loss": 4.8273,
+      "step": 9040
+    },
+    {
+      "epoch": 54.69,
+      "learning_rate": 3.02e-05,
+      "loss": 4.8337,
+      "step": 9060
+    },
+    {
+      "epoch": 54.81,
+      "learning_rate": 3.0266666666666666e-05,
+      "loss": 4.8381,
+      "step": 9080
+    },
+    {
+      "epoch": 54.93,
+      "learning_rate": 3.0333333333333337e-05,
+      "loss": 4.8132,
+      "step": 9100
+    },
+    {
+      "epoch": 55.05,
+      "learning_rate": 3.04e-05,
+      "loss": 4.8152,
+      "step": 9120
+    },
+    {
+      "epoch": 55.17,
+      "learning_rate": 3.0466666666666664e-05,
+      "loss": 4.7843,
+      "step": 9140
+    },
+    {
+      "epoch": 55.29,
+      "learning_rate": 3.0533333333333335e-05,
+      "loss": 4.7997,
+      "step": 9160
+    },
+    {
+      "epoch": 55.41,
+      "learning_rate": 3.06e-05,
+      "loss": 4.8173,
+      "step": 9180
+    },
+    {
+      "epoch": 55.53,
+      "learning_rate": 3.066666666666667e-05,
+      "loss": 4.8259,
+      "step": 9200
+    },
+    {
+      "epoch": 55.65,
+      "learning_rate": 3.073333333333334e-05,
+      "loss": 4.8002,
+      "step": 9220
+    },
+    {
+      "epoch": 55.77,
+      "learning_rate": 3.08e-05,
+      "loss": 4.803,
+      "step": 9240
+    },
+    {
+      "epoch": 55.9,
+      "learning_rate": 3.086666666666667e-05,
+      "loss": 4.7885,
+      "step": 9260
+    },
+    {
+      "epoch": 56.02,
+      "learning_rate": 3.093333333333334e-05,
+      "loss": 4.7864,
+      "step": 9280
+    },
+    {
+      "epoch": 56.14,
+      "learning_rate": 3.1e-05,
+      "loss": 4.7604,
+      "step": 9300
+    },
+    {
+      "epoch": 56.26,
+      "learning_rate": 3.1066666666666665e-05,
+      "loss": 4.7582,
+      "step": 9320
+    },
+    {
+      "epoch": 56.38,
+      "learning_rate": 3.1133333333333336e-05,
+      "loss": 4.8099,
+      "step": 9340
+    },
+    {
+      "epoch": 56.5,
+      "learning_rate": 3.12e-05,
+      "loss": 4.7534,
+      "step": 9360
+    },
+    {
+      "epoch": 56.62,
+      "learning_rate": 3.126666666666666e-05,
+      "loss": 4.7918,
+      "step": 9380
+    },
+    {
+      "epoch": 56.74,
+      "learning_rate": 3.1333333333333334e-05,
+      "loss": 4.7662,
+      "step": 9400
+    },
+    {
+      "epoch": 56.86,
+      "learning_rate": 3.1400000000000004e-05,
+      "loss": 4.783,
+      "step": 9420
+    },
+    {
+      "epoch": 56.98,
+      "learning_rate": 3.146666666666667e-05,
+      "loss": 4.7519,
+      "step": 9440
+    },
+    {
+      "epoch": 57.1,
+      "learning_rate": 3.153333333333334e-05,
+      "loss": 4.7292,
+      "step": 9460
+    },
+    {
+      "epoch": 57.22,
+      "learning_rate": 3.16e-05,
+      "loss": 4.7479,
+      "step": 9480
+    },
+    {
+      "epoch": 57.34,
+      "learning_rate": 3.1666666666666666e-05,
+      "loss": 4.7309,
+      "step": 9500
+    },
+    {
+      "epoch": 57.46,
+      "learning_rate": 3.173333333333334e-05,
+      "loss": 4.7625,
+      "step": 9520
+    },
+    {
+      "epoch": 57.59,
+      "learning_rate": 3.18e-05,
+      "loss": 4.7517,
+      "step": 9540
+    },
+    {
+      "epoch": 57.71,
+      "learning_rate": 3.1866666666666664e-05,
+      "loss": 4.7495,
+      "step": 9560
+    },
+    {
+      "epoch": 57.83,
+      "learning_rate": 3.1933333333333335e-05,
+      "loss": 4.7457,
+      "step": 9580
+    },
+    {
+      "epoch": 57.95,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 4.7366,
+      "step": 9600
+    },
+    {
+      "epoch": 58.07,
+      "learning_rate": 3.206666666666667e-05,
+      "loss": 4.7249,
+      "step": 9620
+    },
+    {
+      "epoch": 58.19,
+      "learning_rate": 3.213333333333334e-05,
+      "loss": 4.7089,
+      "step": 9640
+    },
+    {
+      "epoch": 58.31,
+      "learning_rate": 3.2200000000000003e-05,
+      "loss": 4.7289,
+      "step": 9660
+    },
+    {
+      "epoch": 58.43,
+      "learning_rate": 3.226666666666667e-05,
+      "loss": 4.7049,
+      "step": 9680
+    },
+    {
+      "epoch": 58.55,
+      "learning_rate": 3.233333333333333e-05,
+      "loss": 4.6996,
+      "step": 9700
+    },
+    {
+      "epoch": 58.67,
+      "learning_rate": 3.24e-05,
+      "loss": 4.7265,
+      "step": 9720
+    },
+    {
+      "epoch": 58.79,
+      "learning_rate": 3.2466666666666665e-05,
+      "loss": 4.7109,
+      "step": 9740
+    },
+    {
+      "epoch": 58.91,
+      "learning_rate": 3.253333333333333e-05,
+      "loss": 4.7198,
+      "step": 9760
+    },
+    {
+      "epoch": 59.03,
+      "learning_rate": 3.26e-05,
+      "loss": 4.7008,
+      "step": 9780
+    },
+    {
+      "epoch": 59.15,
+      "learning_rate": 3.266666666666667e-05,
+      "loss": 4.6824,
+      "step": 9800
+    },
+    {
+      "epoch": 59.28,
+      "learning_rate": 3.2733333333333334e-05,
+      "loss": 4.6835,
+      "step": 9820
+    },
+    {
+      "epoch": 59.4,
+      "learning_rate": 3.2800000000000004e-05,
+      "loss": 4.6769,
+      "step": 9840
+    },
+    {
+      "epoch": 59.52,
+      "learning_rate": 3.286666666666667e-05,
+      "loss": 4.6937,
+      "step": 9860
+    },
+    {
+      "epoch": 59.64,
+      "learning_rate": 3.293333333333333e-05,
+      "loss": 4.6854,
+      "step": 9880
+    },
+    {
+      "epoch": 59.76,
+      "learning_rate": 3.3e-05,
+      "loss": 4.6908,
+      "step": 9900
+    },
+    {
+      "epoch": 59.88,
+      "learning_rate": 3.3066666666666666e-05,
+      "loss": 4.6809,
+      "step": 9920
+    },
+    {
+      "epoch": 60.0,
+      "learning_rate": 3.313333333333333e-05,
+      "loss": 4.6611,
+      "step": 9940
+    },
+    {
+      "epoch": 60.12,
+      "learning_rate": 3.32e-05,
+      "loss": 4.6599,
+      "step": 9960
+    },
+    {
+      "epoch": 60.24,
+      "learning_rate": 3.326666666666667e-05,
+      "loss": 4.6588,
+      "step": 9980
+    },
+    {
+      "epoch": 60.36,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 4.6358,
+      "step": 10000
+    },
+    {
+      "epoch": 60.36,
+      "eval_accuracy": 0.002247335302427122,
+      "eval_loss": 4.875439643859863,
+      "eval_runtime": 18.0739,
+      "eval_samples_per_second": 112.87,
+      "eval_steps_per_second": 2.379,
+      "step": 10000
+    },
+    {
+      "epoch": 60.48,
+      "learning_rate": 3.339666666666667e-05,
+      "loss": 4.6601,
+      "step": 10020
+    },
+    {
+      "epoch": 60.6,
+      "learning_rate": 3.3463333333333335e-05,
+      "loss": 4.6622,
+      "step": 10040
+    },
+    {
+      "epoch": 60.72,
+      "learning_rate": 3.353e-05,
+      "loss": 4.6417,
+      "step": 10060
+    },
+    {
+      "epoch": 60.85,
+      "learning_rate": 3.359666666666667e-05,
+      "loss": 4.659,
+      "step": 10080
+    },
+    {
+      "epoch": 60.97,
+      "learning_rate": 3.3663333333333333e-05,
+      "loss": 4.6455,
+      "step": 10100
+    },
+    {
+      "epoch": 61.09,
+      "learning_rate": 3.373e-05,
+      "loss": 4.6091,
+      "step": 10120
+    },
+    {
+      "epoch": 61.21,
+      "learning_rate": 3.379666666666667e-05,
+      "loss": 4.613,
+      "step": 10140
+    },
+    {
+      "epoch": 61.33,
+      "learning_rate": 3.386333333333334e-05,
+      "loss": 4.6148,
+      "step": 10160
+    },
+    {
+      "epoch": 61.45,
+      "learning_rate": 3.393e-05,
+      "loss": 4.6302,
+      "step": 10180
+    },
+    {
+      "epoch": 61.57,
+      "learning_rate": 3.3996666666666666e-05,
+      "loss": 4.6415,
+      "step": 10200
+    },
+    {
+      "epoch": 61.69,
+      "learning_rate": 3.4063333333333336e-05,
+      "loss": 4.6201,
+      "step": 10220
+    },
+    {
+      "epoch": 61.81,
+      "learning_rate": 3.413e-05,
+      "loss": 4.5999,
+      "step": 10240
+    },
+    {
+      "epoch": 61.93,
+      "learning_rate": 3.4196666666666664e-05,
+      "loss": 4.6226,
+      "step": 10260
+    },
+    {
+      "epoch": 62.05,
+      "learning_rate": 3.4263333333333334e-05,
+      "loss": 4.6262,
+      "step": 10280
+    },
+    {
+      "epoch": 62.17,
+      "learning_rate": 3.433e-05,
+      "loss": 4.5851,
+      "step": 10300
+    },
+    {
+      "epoch": 62.29,
+      "learning_rate": 3.439666666666667e-05,
+      "loss": 4.5964,
+      "step": 10320
+    },
+    {
+      "epoch": 62.41,
+      "learning_rate": 3.446333333333334e-05,
+      "loss": 4.5909,
+      "step": 10340
+    },
+    {
+      "epoch": 62.54,
+      "learning_rate": 3.453e-05,
+      "loss": 4.5874,
+      "step": 10360
+    },
+    {
+      "epoch": 62.66,
+      "learning_rate": 3.459666666666667e-05,
+      "loss": 4.5741,
+      "step": 10380
+    },
+    {
+      "epoch": 62.78,
+      "learning_rate": 3.466333333333334e-05,
+      "loss": 4.5983,
+      "step": 10400
+    },
+    {
+      "epoch": 62.9,
+      "learning_rate": 3.473e-05,
+      "loss": 4.5904,
+      "step": 10420
+    },
+    {
+      "epoch": 63.02,
+      "learning_rate": 3.4796666666666665e-05,
+      "loss": 4.5989,
+      "step": 10440
+    },
+    {
+      "epoch": 63.14,
+      "learning_rate": 3.4863333333333336e-05,
+      "loss": 4.5575,
+      "step": 10460
+    },
+    {
+      "epoch": 63.26,
+      "learning_rate": 3.493e-05,
+      "loss": 4.5537,
+      "step": 10480
+    },
+    {
+      "epoch": 63.38,
+      "learning_rate": 3.499666666666667e-05,
+      "loss": 4.5645,
+      "step": 10500
+    },
+    {
+      "epoch": 63.5,
+      "learning_rate": 3.5063333333333334e-05,
+      "loss": 4.5454,
+      "step": 10520
+    },
+    {
+      "epoch": 63.62,
+      "learning_rate": 3.5130000000000004e-05,
+      "loss": 4.5659,
+      "step": 10540
+    },
+    {
+      "epoch": 63.74,
+      "learning_rate": 3.519666666666667e-05,
+      "loss": 4.5412,
+      "step": 10560
+    },
+    {
+      "epoch": 63.86,
+      "learning_rate": 3.526333333333334e-05,
+      "loss": 4.5668,
+      "step": 10580
+    },
+    {
+      "epoch": 63.98,
+      "learning_rate": 3.533e-05,
+      "loss": 4.5536,
+      "step": 10600
+    },
+    {
+      "epoch": 64.1,
+      "learning_rate": 3.5396666666666666e-05,
+      "loss": 4.5191,
+      "step": 10620
+    },
+    {
+      "epoch": 64.23,
+      "learning_rate": 3.5463333333333337e-05,
+      "loss": 4.5262,
+      "step": 10640
+    },
+    {
+      "epoch": 64.35,
+      "learning_rate": 3.553e-05,
+      "loss": 4.544,
+      "step": 10660
+    },
+    {
+      "epoch": 64.47,
+      "learning_rate": 3.5596666666666664e-05,
+      "loss": 4.5408,
+      "step": 10680
+    },
+    {
+      "epoch": 64.59,
+      "learning_rate": 3.5663333333333335e-05,
+      "loss": 4.5145,
+      "step": 10700
+    },
+    {
+      "epoch": 64.71,
+      "learning_rate": 3.5730000000000005e-05,
+      "loss": 4.5178,
+      "step": 10720
+    },
+    {
+      "epoch": 64.83,
+      "learning_rate": 3.579666666666667e-05,
+      "loss": 4.5363,
+      "step": 10740
+    },
+    {
+      "epoch": 64.95,
+      "learning_rate": 3.586333333333334e-05,
+      "loss": 4.5128,
+      "step": 10760
+    },
+    {
+      "epoch": 65.07,
+      "learning_rate": 3.593e-05,
+      "loss": 4.503,
+      "step": 10780
+    },
+    {
+      "epoch": 65.19,
+      "learning_rate": 3.599666666666667e-05,
+      "loss": 4.491,
+      "step": 10800
+    },
+    {
+      "epoch": 65.31,
+      "learning_rate": 3.606333333333333e-05,
+      "loss": 4.4675,
+      "step": 10820
+    },
+    {
+      "epoch": 65.43,
+      "learning_rate": 3.613e-05,
+      "loss": 4.5172,
+      "step": 10840
+    },
+    {
+      "epoch": 65.55,
+      "learning_rate": 3.6196666666666665e-05,
+      "loss": 4.4865,
+      "step": 10860
+    },
+    {
+      "epoch": 65.67,
+      "learning_rate": 3.6263333333333336e-05,
+      "loss": 4.4962,
+      "step": 10880
+    },
+    {
+      "epoch": 65.79,
+      "learning_rate": 3.6330000000000006e-05,
+      "loss": 4.5097,
+      "step": 10900
+    },
+    {
+      "epoch": 65.92,
+      "learning_rate": 3.639666666666667e-05,
+      "loss": 4.5079,
+      "step": 10920
+    },
+    {
+      "epoch": 66.04,
+      "learning_rate": 3.6463333333333334e-05,
+      "loss": 4.4555,
+      "step": 10940
+    },
+    {
+      "epoch": 66.16,
+      "learning_rate": 3.6530000000000004e-05,
+      "loss": 4.4732,
+      "step": 10960
+    },
+    {
+      "epoch": 66.28,
+      "learning_rate": 3.659666666666667e-05,
+      "loss": 4.4547,
+      "step": 10980
+    },
+    {
+      "epoch": 66.4,
+      "learning_rate": 3.666333333333333e-05,
+      "loss": 4.4326,
+      "step": 11000
+    },
+    {
+      "epoch": 66.4,
+      "eval_accuracy": 0.002107062065967935,
+      "eval_loss": 4.780860424041748,
+      "eval_runtime": 18.0696,
+      "eval_samples_per_second": 112.897,
+      "eval_steps_per_second": 2.38,
+      "step": 11000
+    },
+    {
+      "epoch": 66.52,
+      "learning_rate": 3.673e-05,
+      "loss": 4.4623,
+      "step": 11020
+    },
+    {
+      "epoch": 66.64,
+      "learning_rate": 3.6796666666666666e-05,
+      "loss": 4.4712,
+      "step": 11040
+    },
+    {
+      "epoch": 66.76,
+      "learning_rate": 3.686333333333333e-05,
+      "loss": 4.4571,
+      "step": 11060
+    },
+    {
+      "epoch": 66.88,
+      "learning_rate": 3.693e-05,
+      "loss": 4.4793,
+      "step": 11080
+    },
+    {
+      "epoch": 67.0,
+      "learning_rate": 3.699666666666667e-05,
+      "loss": 4.4662,
+      "step": 11100
+    },
+    {
+      "epoch": 67.12,
+      "learning_rate": 3.7063333333333335e-05,
+      "loss": 4.3932,
+      "step": 11120
+    },
+    {
+      "epoch": 67.24,
+      "learning_rate": 3.7130000000000005e-05,
+      "loss": 4.4387,
+      "step": 11140
+    },
+    {
+      "epoch": 67.36,
+      "learning_rate": 3.719666666666667e-05,
+      "loss": 4.426,
+      "step": 11160
+    },
+    {
+      "epoch": 67.48,
+      "learning_rate": 3.726333333333333e-05,
+      "loss": 4.4222,
+      "step": 11180
+    },
+    {
+      "epoch": 67.61,
+      "learning_rate": 3.7330000000000003e-05,
+      "loss": 4.4323,
+      "step": 11200
+    },
+    {
+      "epoch": 67.73,
+      "learning_rate": 3.739666666666667e-05,
+      "loss": 4.4221,
+      "step": 11220
+    },
+    {
+      "epoch": 67.85,
+      "learning_rate": 3.746333333333333e-05,
+      "loss": 4.4385,
+      "step": 11240
+    },
+    {
+      "epoch": 67.97,
+      "learning_rate": 3.753e-05,
+      "loss": 4.4514,
+      "step": 11260
+    },
+    {
+      "epoch": 68.09,
+      "learning_rate": 3.759666666666667e-05,
+      "loss": 4.3912,
+      "step": 11280
+    },
+    {
+      "epoch": 68.21,
+      "learning_rate": 3.7663333333333336e-05,
+      "loss": 4.3944,
+      "step": 11300
+    },
+    {
+      "epoch": 68.33,
+      "learning_rate": 3.7730000000000006e-05,
+      "loss": 4.3894,
+      "step": 11320
+    },
+    {
+      "epoch": 68.45,
+      "learning_rate": 3.779666666666667e-05,
+      "loss": 4.3806,
+      "step": 11340
+    },
+    {
+      "epoch": 68.57,
+      "learning_rate": 3.7863333333333334e-05,
+      "loss": 4.3978,
+      "step": 11360
+    },
+    {
+      "epoch": 68.69,
+      "learning_rate": 3.7930000000000004e-05,
+      "loss": 4.4171,
+      "step": 11380
+    },
+    {
+      "epoch": 68.81,
+      "learning_rate": 3.799666666666667e-05,
+      "loss": 4.4158,
+      "step": 11400
+    },
+    {
+      "epoch": 68.93,
+      "learning_rate": 3.806333333333333e-05,
+      "loss": 4.4008,
+      "step": 11420
+    },
+    {
+      "epoch": 69.05,
+      "learning_rate": 3.8129999999999996e-05,
+      "loss": 4.3798,
+      "step": 11440
+    },
+    {
+      "epoch": 69.18,
+      "learning_rate": 3.8196666666666666e-05,
+      "loss": 4.3538,
+      "step": 11460
+    },
+    {
+      "epoch": 69.3,
+      "learning_rate": 3.826333333333334e-05,
+      "loss": 4.3783,
+      "step": 11480
+    },
+    {
+      "epoch": 69.42,
+      "learning_rate": 3.833e-05,
+      "loss": 4.3597,
+      "step": 11500
+    },
+    {
+      "epoch": 69.54,
+      "learning_rate": 3.839666666666667e-05,
+      "loss": 4.3653,
+      "step": 11520
+    },
+    {
+      "epoch": 69.66,
+      "learning_rate": 3.8463333333333335e-05,
+      "loss": 4.3719,
+      "step": 11540
+    },
+    {
+      "epoch": 69.78,
+      "learning_rate": 3.853e-05,
+      "loss": 4.3391,
+      "step": 11560
+    },
+    {
+      "epoch": 69.9,
+      "learning_rate": 3.859666666666667e-05,
+      "loss": 4.3558,
+      "step": 11580
+    },
+    {
+      "epoch": 70.02,
+      "learning_rate": 3.866333333333333e-05,
+      "loss": 4.3695,
+      "step": 11600
+    },
+    {
+      "epoch": 70.14,
+      "learning_rate": 3.873e-05,
+      "loss": 4.3293,
+      "step": 11620
+    },
+    {
+      "epoch": 70.26,
+      "learning_rate": 3.879666666666667e-05,
+      "loss": 4.3039,
+      "step": 11640
+    },
+    {
+      "epoch": 70.38,
+      "learning_rate": 3.886333333333334e-05,
+      "loss": 4.3328,
+      "step": 11660
+    },
+    {
+      "epoch": 70.5,
+      "learning_rate": 3.893e-05,
+      "loss": 4.3128,
+      "step": 11680
+    },
+    {
+      "epoch": 70.62,
+      "learning_rate": 3.899666666666667e-05,
+      "loss": 4.3224,
+      "step": 11700
+    },
+    {
+      "epoch": 70.74,
+      "learning_rate": 3.9063333333333336e-05,
+      "loss": 4.3377,
+      "step": 11720
+    },
+    {
+      "epoch": 70.87,
+      "learning_rate": 3.913e-05,
+      "loss": 4.3658,
+      "step": 11740
+    },
+    {
+      "epoch": 70.99,
+      "learning_rate": 3.919666666666667e-05,
+      "loss": 4.3473,
+      "step": 11760
+    },
+    {
+      "epoch": 71.11,
+      "learning_rate": 3.9263333333333334e-05,
+      "loss": 4.2763,
+      "step": 11780
+    },
+    {
+      "epoch": 71.23,
+      "learning_rate": 3.933e-05,
+      "loss": 4.2899,
+      "step": 11800
+    },
+    {
+      "epoch": 71.35,
+      "learning_rate": 3.939666666666667e-05,
+      "loss": 4.2896,
+      "step": 11820
+    },
+    {
+      "epoch": 71.47,
+      "learning_rate": 3.946333333333333e-05,
+      "loss": 4.2895,
+      "step": 11840
+    },
+    {
+      "epoch": 71.59,
+      "learning_rate": 3.953e-05,
+      "loss": 4.3077,
+      "step": 11860
+    },
+    {
+      "epoch": 71.71,
+      "learning_rate": 3.959666666666667e-05,
+      "loss": 4.3121,
+      "step": 11880
+    },
+    {
+      "epoch": 71.83,
+      "learning_rate": 3.966333333333334e-05,
+      "loss": 4.3091,
+      "step": 11900
+    },
+    {
+      "epoch": 71.95,
+      "learning_rate": 3.973e-05,
+      "loss": 4.2995,
+      "step": 11920
+    },
+    {
+      "epoch": 72.07,
+      "learning_rate": 3.979666666666667e-05,
+      "loss": 4.2908,
+      "step": 11940
+    },
+    {
+      "epoch": 72.19,
+      "learning_rate": 3.9863333333333335e-05,
+      "loss": 4.2227,
+      "step": 11960
+    },
+    {
+      "epoch": 72.31,
+      "learning_rate": 3.993e-05,
+      "loss": 4.239,
+      "step": 11980
+    },
+    {
+      "epoch": 72.43,
+      "learning_rate": 3.999666666666667e-05,
+      "loss": 4.2632,
+      "step": 12000
+    },
+    {
+      "epoch": 72.43,
+      "eval_accuracy": 0.0017331647420256641,
+      "eval_loss": 4.741575241088867,
+      "eval_runtime": 18.0838,
+      "eval_samples_per_second": 112.808,
+      "eval_steps_per_second": 2.378,
+      "step": 12000
+    },
+    {
+      "epoch": 72.56,
+      "learning_rate": 4.0060000000000006e-05,
+      "loss": 4.2956,
+      "step": 12020
+    },
+    {
+      "epoch": 72.68,
+      "learning_rate": 4.012666666666667e-05,
+      "loss": 4.2654,
+      "step": 12040
+    },
+    {
+      "epoch": 72.8,
+      "learning_rate": 4.0193333333333334e-05,
+      "loss": 4.2661,
+      "step": 12060
+    },
+    {
+      "epoch": 72.92,
+      "learning_rate": 4.0260000000000004e-05,
+      "loss": 4.2876,
+      "step": 12080
+    },
+    {
+      "epoch": 73.04,
+      "learning_rate": 4.032666666666667e-05,
+      "loss": 4.2446,
+      "step": 12100
+    },
+    {
+      "epoch": 73.16,
+      "learning_rate": 4.039333333333333e-05,
+      "loss": 4.2041,
+      "step": 12120
+    },
+    {
+      "epoch": 73.28,
+      "learning_rate": 4.046e-05,
+      "loss": 4.2398,
+      "step": 12140
+    },
+    {
+      "epoch": 73.4,
+      "learning_rate": 4.0526666666666666e-05,
+      "loss": 4.2222,
+      "step": 12160
+    },
+    {
+      "epoch": 73.52,
+      "learning_rate": 4.0593333333333337e-05,
+      "loss": 4.2484,
+      "step": 12180
+    },
+    {
+      "epoch": 73.64,
+      "learning_rate": 4.066e-05,
+      "loss": 4.2174,
+      "step": 12200
+    },
+    {
+      "epoch": 73.76,
+      "learning_rate": 4.072666666666667e-05,
+      "loss": 4.2504,
+      "step": 12220
+    },
+    {
+      "epoch": 73.88,
+      "learning_rate": 4.0793333333333335e-05,
+      "loss": 4.2291,
+      "step": 12240
+    },
+    {
+      "epoch": 74.0,
+      "learning_rate": 4.0860000000000005e-05,
+      "loss": 4.2397,
+      "step": 12260
+    },
+    {
+      "epoch": 74.12,
+      "learning_rate": 4.092666666666667e-05,
+      "loss": 4.1777,
+      "step": 12280
+    },
+    {
+      "epoch": 74.25,
+      "learning_rate": 4.099333333333333e-05,
+      "loss": 4.1897,
+      "step": 12300
+    },
+    {
+      "epoch": 74.37,
+      "learning_rate": 4.106e-05,
+      "loss": 4.1765,
+      "step": 12320
+    },
+    {
+      "epoch": 74.49,
+      "learning_rate": 4.112666666666667e-05,
+      "loss": 4.1897,
+      "step": 12340
+    },
+    {
+      "epoch": 74.61,
+      "learning_rate": 4.119333333333333e-05,
+      "loss": 4.2046,
+      "step": 12360
+    },
+    {
+      "epoch": 74.73,
+      "learning_rate": 4.126e-05,
+      "loss": 4.2271,
+      "step": 12380
+    },
+    {
+      "epoch": 74.85,
+      "learning_rate": 4.132666666666667e-05,
+      "loss": 4.1944,
+      "step": 12400
+    },
+    {
+      "epoch": 74.97,
+      "learning_rate": 4.1393333333333336e-05,
+      "loss": 4.2009,
+      "step": 12420
+    },
+    {
+      "epoch": 75.09,
+      "learning_rate": 4.1460000000000006e-05,
+      "loss": 4.1821,
+      "step": 12440
+    },
+    {
+      "epoch": 75.21,
+      "learning_rate": 4.152666666666667e-05,
+      "loss": 4.1583,
+      "step": 12460
+    },
+    {
+      "epoch": 75.33,
+      "learning_rate": 4.1593333333333334e-05,
+      "loss": 4.1454,
+      "step": 12480
+    },
+    {
+      "epoch": 75.45,
+      "learning_rate": 4.1660000000000004e-05,
+      "loss": 4.1584,
+      "step": 12500
+    },
+    {
+      "epoch": 75.57,
+      "learning_rate": 4.172666666666667e-05,
+      "loss": 4.1799,
+      "step": 12520
+    },
+    {
+      "epoch": 75.69,
+      "learning_rate": 4.179333333333333e-05,
+      "loss": 4.1769,
+      "step": 12540
+    },
+    {
+      "epoch": 75.81,
+      "learning_rate": 4.186e-05,
+      "loss": 4.1602,
+      "step": 12560
+    },
+    {
+      "epoch": 75.94,
+      "learning_rate": 4.192666666666667e-05,
+      "loss": 4.1414,
+      "step": 12580
+    },
+    {
+      "epoch": 76.06,
+      "learning_rate": 4.199333333333334e-05,
+      "loss": 4.1437,
+      "step": 12600
+    },
+    {
+      "epoch": 76.18,
+      "learning_rate": 4.206e-05,
+      "loss": 4.1252,
+      "step": 12620
+    },
+    {
+      "epoch": 76.3,
+      "learning_rate": 4.212666666666667e-05,
+      "loss": 4.1107,
+      "step": 12640
+    },
+    {
+      "epoch": 76.42,
+      "learning_rate": 4.2193333333333335e-05,
+      "loss": 4.1409,
+      "step": 12660
+    },
+    {
+      "epoch": 76.54,
+      "learning_rate": 4.226e-05,
+      "loss": 4.1118,
+      "step": 12680
+    },
+    {
+      "epoch": 76.66,
+      "learning_rate": 4.232666666666667e-05,
+      "loss": 4.1119,
+      "step": 12700
+    },
+    {
+      "epoch": 76.78,
+      "learning_rate": 4.239333333333333e-05,
+      "loss": 4.1528,
+      "step": 12720
+    },
+    {
+      "epoch": 76.9,
+      "learning_rate": 4.246e-05,
+      "loss": 4.1155,
+      "step": 12740
+    },
+    {
+      "epoch": 77.02,
+      "learning_rate": 4.252666666666667e-05,
+      "loss": 4.1292,
+      "step": 12760
+    },
+    {
+      "epoch": 77.14,
+      "learning_rate": 4.259333333333334e-05,
+      "loss": 4.0907,
+      "step": 12780
+    },
+    {
+      "epoch": 77.26,
+      "learning_rate": 4.266e-05,
+      "loss": 4.0803,
+      "step": 12800
+    },
+    {
+      "epoch": 77.38,
+      "learning_rate": 4.272666666666667e-05,
+      "loss": 4.0886,
+      "step": 12820
+    },
+    {
+      "epoch": 77.51,
+      "learning_rate": 4.2793333333333336e-05,
+      "loss": 4.1005,
+      "step": 12840
+    },
+    {
+      "epoch": 77.63,
+      "learning_rate": 4.286e-05,
+      "loss": 4.0972,
+      "step": 12860
+    },
+    {
+      "epoch": 77.75,
+      "learning_rate": 4.292666666666667e-05,
+      "loss": 4.1042,
+      "step": 12880
+    },
+    {
+      "epoch": 77.87,
+      "learning_rate": 4.2993333333333334e-05,
+      "loss": 4.0769,
+      "step": 12900
+    },
+    {
+      "epoch": 77.99,
+      "learning_rate": 4.306e-05,
+      "loss": 4.1025,
+      "step": 12920
+    },
+    {
+      "epoch": 78.11,
+      "learning_rate": 4.312666666666667e-05,
+      "loss": 4.0391,
+      "step": 12940
+    },
+    {
+      "epoch": 78.23,
+      "learning_rate": 4.319333333333334e-05,
+      "loss": 4.0377,
+      "step": 12960
+    },
+    {
+      "epoch": 78.35,
+      "learning_rate": 4.326e-05,
+      "loss": 4.0672,
+      "step": 12980
+    },
+    {
+      "epoch": 78.47,
+      "learning_rate": 4.332666666666667e-05,
+      "loss": 4.0415,
+      "step": 13000
+    },
+    {
+      "epoch": 78.47,
+      "eval_accuracy": 0.001637838211615019,
+      "eval_loss": 4.750297546386719,
+      "eval_runtime": 18.199,
+      "eval_samples_per_second": 112.094,
+      "eval_steps_per_second": 2.363,
+      "step": 13000
+    },
+    {
+      "epoch": 78.59,
+      "learning_rate": 4.339333333333334e-05,
+      "loss": 4.0433,
+      "step": 13020
+    },
+    {
+      "epoch": 78.71,
+      "learning_rate": 4.346e-05,
+      "loss": 4.0433,
+      "step": 13040
+    },
+    {
+      "epoch": 78.83,
+      "learning_rate": 4.352666666666667e-05,
+      "loss": 4.0692,
+      "step": 13060
+    },
+    {
+      "epoch": 78.95,
+      "learning_rate": 4.3593333333333335e-05,
+      "loss": 4.0831,
+      "step": 13080
+    },
+    {
+      "epoch": 79.07,
+      "learning_rate": 4.366e-05,
+      "loss": 4.0378,
+      "step": 13100
+    },
+    {
+      "epoch": 79.2,
+      "learning_rate": 4.372666666666667e-05,
+      "loss": 4.0092,
+      "step": 13120
+    },
+    {
+      "epoch": 79.32,
+      "learning_rate": 4.379333333333333e-05,
+      "loss": 4.0043,
+      "step": 13140
+    },
+    {
+      "epoch": 79.44,
+      "learning_rate": 4.3860000000000004e-05,
+      "loss": 4.0388,
+      "step": 13160
+    },
+    {
+      "epoch": 79.56,
+      "learning_rate": 4.3926666666666674e-05,
+      "loss": 4.0174,
+      "step": 13180
+    },
+    {
+      "epoch": 79.68,
+      "learning_rate": 4.399333333333334e-05,
+      "loss": 3.9951,
+      "step": 13200
+    },
+    {
+      "epoch": 79.8,
+      "learning_rate": 4.406e-05,
+      "loss": 4.0435,
+      "step": 13220
+    },
+    {
+      "epoch": 79.92,
+      "learning_rate": 4.4126666666666665e-05,
+      "loss": 4.0442,
+      "step": 13240
+    },
+    {
+      "epoch": 80.04,
+      "learning_rate": 4.4193333333333336e-05,
+      "loss": 3.9929,
+      "step": 13260
+    },
+    {
+      "epoch": 80.16,
+      "learning_rate": 4.426e-05,
+      "loss": 3.9753,
+      "step": 13280
+    },
+    {
+      "epoch": 80.28,
+      "learning_rate": 4.4326666666666664e-05,
+      "loss": 3.949,
+      "step": 13300
+    },
+    {
+      "epoch": 80.4,
+      "learning_rate": 4.4393333333333334e-05,
+      "loss": 3.9818,
+      "step": 13320
+    },
+    {
+      "epoch": 80.52,
+      "learning_rate": 4.4460000000000005e-05,
+      "loss": 3.9604,
+      "step": 13340
+    },
+    {
+      "epoch": 80.64,
+      "learning_rate": 4.452666666666667e-05,
+      "loss": 3.9965,
+      "step": 13360
+    },
+    {
+      "epoch": 80.76,
+      "learning_rate": 4.459333333333334e-05,
+      "loss": 3.9871,
+      "step": 13380
+    },
+    {
+      "epoch": 80.89,
+      "learning_rate": 4.466e-05,
+      "loss": 4.0184,
+      "step": 13400
+    },
+    {
+      "epoch": 81.01,
+      "learning_rate": 4.4726666666666666e-05,
+      "loss": 3.9988,
+      "step": 13420
+    },
+    {
+      "epoch": 81.13,
+      "learning_rate": 4.479333333333334e-05,
+      "loss": 3.9294,
+      "step": 13440
+    },
+    {
+      "epoch": 81.25,
+      "learning_rate": 4.486e-05,
+      "loss": 3.9239,
+      "step": 13460
+    },
+    {
+      "epoch": 81.37,
+      "learning_rate": 4.4926666666666665e-05,
+      "loss": 3.935,
+      "step": 13480
+    },
+    {
+      "epoch": 81.49,
+      "learning_rate": 4.4993333333333335e-05,
+      "loss": 3.941,
+      "step": 13500
+    },
+    {
+      "epoch": 81.61,
+      "learning_rate": 4.506e-05,
+      "loss": 3.9797,
+      "step": 13520
+    },
+    {
+      "epoch": 81.73,
+      "learning_rate": 4.512666666666667e-05,
+      "loss": 3.9534,
+      "step": 13540
+    },
+    {
+      "epoch": 81.85,
+      "learning_rate": 4.519333333333334e-05,
+      "loss": 3.9324,
+      "step": 13560
+    },
+    {
+      "epoch": 81.97,
+      "learning_rate": 4.5260000000000004e-05,
+      "loss": 3.9686,
+      "step": 13580
+    },
+    {
+      "epoch": 82.09,
+      "learning_rate": 4.532666666666667e-05,
+      "loss": 3.8894,
+      "step": 13600
+    },
+    {
+      "epoch": 82.21,
+      "learning_rate": 4.539333333333334e-05,
+      "loss": 3.8861,
+      "step": 13620
+    },
+    {
+      "epoch": 82.33,
+      "learning_rate": 4.546e-05,
+      "loss": 3.9124,
+      "step": 13640
+    },
+    {
+      "epoch": 82.45,
+      "learning_rate": 4.5526666666666666e-05,
+      "loss": 3.9025,
+      "step": 13660
+    },
+    {
+      "epoch": 82.58,
+      "learning_rate": 4.5593333333333336e-05,
+      "loss": 3.8856,
+      "step": 13680
+    },
+    {
+      "epoch": 82.7,
+      "learning_rate": 4.566e-05,
+      "loss": 3.9063,
+      "step": 13700
+    },
+    {
+      "epoch": 82.82,
+      "learning_rate": 4.572666666666667e-05,
+      "loss": 3.9387,
+      "step": 13720
+    },
+    {
+      "epoch": 82.94,
+      "learning_rate": 4.579333333333334e-05,
+      "loss": 3.9133,
+      "step": 13740
+    },
+    {
+      "epoch": 83.06,
+      "learning_rate": 4.5860000000000005e-05,
+      "loss": 3.8905,
+      "step": 13760
+    },
+    {
+      "epoch": 83.18,
+      "learning_rate": 4.592666666666667e-05,
+      "loss": 3.8375,
+      "step": 13780
+    },
+    {
+      "epoch": 83.3,
+      "learning_rate": 4.599333333333334e-05,
+      "loss": 3.8402,
+      "step": 13800
+    },
+    {
+      "epoch": 83.42,
+      "learning_rate": 4.606e-05,
+      "loss": 3.8487,
+      "step": 13820
+    },
+    {
+      "epoch": 83.54,
+      "learning_rate": 4.612666666666667e-05,
+      "loss": 3.8902,
+      "step": 13840
+    },
+    {
+      "epoch": 83.66,
+      "learning_rate": 4.619333333333333e-05,
+      "loss": 3.8876,
+      "step": 13860
+    },
+    {
+      "epoch": 83.78,
+      "learning_rate": 4.626e-05,
+      "loss": 3.8639,
+      "step": 13880
+    },
+    {
+      "epoch": 83.9,
+      "learning_rate": 4.632666666666667e-05,
+      "loss": 3.9104,
+      "step": 13900
+    },
+    {
+      "epoch": 84.02,
+      "learning_rate": 4.6393333333333335e-05,
+      "loss": 3.8802,
+      "step": 13920
+    },
+    {
+      "epoch": 84.14,
+      "learning_rate": 4.6460000000000006e-05,
+      "loss": 3.8044,
+      "step": 13940
+    },
+    {
+      "epoch": 84.27,
+      "learning_rate": 4.652666666666667e-05,
+      "loss": 3.7955,
+      "step": 13960
+    },
+    {
+      "epoch": 84.39,
+      "learning_rate": 4.659333333333333e-05,
+      "loss": 3.8249,
+      "step": 13980
+    },
+    {
+      "epoch": 84.51,
+      "learning_rate": 4.6660000000000004e-05,
+      "loss": 3.8196,
+      "step": 14000
+    },
+    {
+      "epoch": 84.51,
+      "eval_accuracy": 0.001441258112633482,
+      "eval_loss": 4.847209453582764,
+      "eval_runtime": 18.1965,
+      "eval_samples_per_second": 112.109,
+      "eval_steps_per_second": 2.363,
+      "step": 14000
+    },
+    {
+      "epoch": 84.63,
+      "learning_rate": 4.672666666666667e-05,
+      "loss": 3.8446,
+      "step": 14020
+    },
+    {
+      "epoch": 84.75,
+      "learning_rate": 4.679e-05,
+      "loss": 3.8489,
+      "step": 14040
+    },
+    {
+      "epoch": 84.87,
+      "learning_rate": 4.685666666666667e-05,
+      "loss": 3.8578,
+      "step": 14060
+    },
+    {
+      "epoch": 84.99,
+      "learning_rate": 4.692333333333334e-05,
+      "loss": 3.8582,
+      "step": 14080
+    },
+    {
+      "epoch": 85.11,
+      "learning_rate": 4.699e-05,
+      "loss": 3.7613,
+      "step": 14100
+    },
+    {
+      "epoch": 85.23,
+      "learning_rate": 4.705666666666667e-05,
+      "loss": 3.752,
+      "step": 14120
+    },
+    {
+      "epoch": 85.35,
+      "learning_rate": 4.712333333333334e-05,
+      "loss": 3.7674,
+      "step": 14140
+    },
+    {
+      "epoch": 85.47,
+      "learning_rate": 4.719e-05,
+      "loss": 3.8142,
+      "step": 14160
+    },
+    {
+      "epoch": 85.59,
+      "learning_rate": 4.725666666666667e-05,
+      "loss": 3.7781,
+      "step": 14180
+    },
+    {
+      "epoch": 85.71,
+      "learning_rate": 4.7323333333333335e-05,
+      "loss": 3.8041,
+      "step": 14200
+    },
+    {
+      "epoch": 85.84,
+      "learning_rate": 4.739e-05,
+      "loss": 3.8062,
+      "step": 14220
+    },
+    {
+      "epoch": 85.96,
+      "learning_rate": 4.745666666666667e-05,
+      "loss": 3.8096,
+      "step": 14240
+    },
+    {
+      "epoch": 86.08,
+      "learning_rate": 4.752333333333334e-05,
+      "loss": 3.7798,
+      "step": 14260
+    },
+    {
+      "epoch": 86.2,
+      "learning_rate": 4.7590000000000003e-05,
+      "loss": 3.7221,
+      "step": 14280
+    },
+    {
+      "epoch": 86.32,
+      "learning_rate": 4.7656666666666674e-05,
+      "loss": 3.7424,
+      "step": 14300
+    },
+    {
+      "epoch": 86.44,
+      "learning_rate": 4.772333333333334e-05,
+      "loss": 3.7457,
+      "step": 14320
+    },
+    {
+      "epoch": 86.56,
+      "learning_rate": 4.779e-05,
+      "loss": 3.755,
+      "step": 14340
+    },
+    {
+      "epoch": 86.68,
+      "learning_rate": 4.7856666666666665e-05,
+      "loss": 3.7468,
+      "step": 14360
+    },
+    {
+      "epoch": 86.8,
+      "learning_rate": 4.7923333333333336e-05,
+      "loss": 3.7541,
+      "step": 14380
+    },
+    {
+      "epoch": 86.92,
+      "learning_rate": 4.799e-05,
+      "loss": 3.7615,
+      "step": 14400
+    },
+    {
+      "epoch": 87.04,
+      "learning_rate": 4.805666666666666e-05,
+      "loss": 3.7446,
+      "step": 14420
+    },
+    {
+      "epoch": 87.16,
+      "learning_rate": 4.8123333333333334e-05,
+      "loss": 3.6729,
+      "step": 14440
+    },
+    {
+      "epoch": 87.28,
+      "learning_rate": 4.8190000000000004e-05,
+      "loss": 3.7036,
+      "step": 14460
+    },
+    {
+      "epoch": 87.4,
+      "learning_rate": 4.825666666666667e-05,
+      "loss": 3.702,
+      "step": 14480
+    },
+    {
+      "epoch": 87.53,
+      "learning_rate": 4.832333333333334e-05,
+      "loss": 3.7092,
+      "step": 14500
+    },
+    {
+      "epoch": 87.65,
+      "learning_rate": 4.839e-05,
+      "loss": 3.7291,
+      "step": 14520
+    },
+    {
+      "epoch": 87.77,
+      "learning_rate": 4.8456666666666666e-05,
+      "loss": 3.7032,
+      "step": 14540
+    },
+    {
+      "epoch": 87.89,
+      "learning_rate": 4.852333333333334e-05,
+      "loss": 3.7265,
+      "step": 14560
+    },
+    {
+      "epoch": 88.01,
+      "learning_rate": 4.859e-05,
+      "loss": 3.7322,
+      "step": 14580
+    },
+    {
+      "epoch": 88.13,
+      "learning_rate": 4.865333333333334e-05,
+      "loss": 3.631,
+      "step": 14600
+    },
+    {
+      "epoch": 88.25,
+      "learning_rate": 4.872000000000001e-05,
+      "loss": 3.6268,
+      "step": 14620
+    },
+    {
+      "epoch": 88.37,
+      "learning_rate": 4.878666666666667e-05,
+      "loss": 3.6521,
+      "step": 14640
+    },
+    {
+      "epoch": 88.49,
+      "learning_rate": 4.8853333333333335e-05,
+      "loss": 3.6974,
+      "step": 14660
+    },
+    {
+      "epoch": 88.61,
+      "learning_rate": 4.8920000000000006e-05,
+      "loss": 3.6469,
+      "step": 14680
+    },
+    {
+      "epoch": 88.73,
+      "learning_rate": 4.898666666666667e-05,
+      "loss": 3.681,
+      "step": 14700
+    },
+    {
+      "epoch": 88.85,
+      "learning_rate": 4.9053333333333333e-05,
+      "loss": 3.6776,
+      "step": 14720
+    },
+    {
+      "epoch": 88.97,
+      "learning_rate": 4.9120000000000004e-05,
+      "loss": 3.6987,
+      "step": 14740
+    },
+    {
+      "epoch": 89.09,
+      "learning_rate": 4.918666666666667e-05,
+      "loss": 3.6138,
+      "step": 14760
+    },
+    {
+      "epoch": 89.22,
+      "learning_rate": 4.925333333333333e-05,
+      "loss": 3.6124,
+      "step": 14780
+    },
+    {
+      "epoch": 89.34,
+      "learning_rate": 4.932e-05,
+      "loss": 3.6088,
+      "step": 14800
+    },
+    {
+      "epoch": 89.46,
+      "learning_rate": 4.938666666666667e-05,
+      "loss": 3.6353,
+      "step": 14820
+    },
+    {
+      "epoch": 89.58,
+      "learning_rate": 4.9453333333333336e-05,
+      "loss": 3.6475,
+      "step": 14840
+    },
+    {
+      "epoch": 89.7,
+      "learning_rate": 4.952e-05,
+      "loss": 3.6232,
+      "step": 14860
+    },
+    {
+      "epoch": 89.82,
+      "learning_rate": 4.958666666666667e-05,
+      "loss": 3.6363,
+      "step": 14880
+    },
+    {
+      "epoch": 89.94,
+      "learning_rate": 4.9653333333333335e-05,
+      "loss": 3.6583,
+      "step": 14900
+    },
+    {
+      "epoch": 90.06,
+      "learning_rate": 4.972e-05,
+      "loss": 3.5976,
+      "step": 14920
+    },
+    {
+      "epoch": 90.18,
+      "learning_rate": 4.978666666666667e-05,
+      "loss": 3.5279,
+      "step": 14940
+    },
+    {
+      "epoch": 90.3,
+      "learning_rate": 4.985333333333333e-05,
+      "loss": 3.562,
+      "step": 14960
+    },
+    {
+      "epoch": 90.42,
+      "learning_rate": 4.992e-05,
+      "loss": 3.5838,
+      "step": 14980
+    },
+    {
+      "epoch": 90.54,
+      "learning_rate": 4.9986666666666674e-05,
+      "loss": 3.6207,
+      "step": 15000
+    },
+    {
+      "epoch": 90.54,
+      "eval_accuracy": 0.0014145864409123687,
+      "eval_loss": 5.021492958068848,
+      "eval_runtime": 18.2836,
+      "eval_samples_per_second": 111.575,
+      "eval_steps_per_second": 2.352,
+      "step": 15000
+    },
+    {
+      "epoch": 90.66,
+      "learning_rate": 5.0053333333333344e-05,
+      "loss": 3.6005,
+      "step": 15020
+    },
+    {
+      "epoch": 90.78,
+      "learning_rate": 5.012e-05,
+      "loss": 3.6005,
+      "step": 15040
+    },
+    {
+      "epoch": 90.91,
+      "learning_rate": 5.018666666666667e-05,
+      "loss": 3.5765,
+      "step": 15060
+    },
+    {
+      "epoch": 91.03,
+      "learning_rate": 5.025333333333334e-05,
+      "loss": 3.5912,
+      "step": 15080
+    },
+    {
+      "epoch": 91.15,
+      "learning_rate": 5.032e-05,
+      "loss": 3.4993,
+      "step": 15100
+    },
+    {
+      "epoch": 91.27,
+      "learning_rate": 5.038666666666667e-05,
+      "loss": 3.5075,
+      "step": 15120
+    },
+    {
+      "epoch": 91.39,
+      "learning_rate": 5.045333333333333e-05,
+      "loss": 3.5238,
+      "step": 15140
+    },
+    {
+      "epoch": 91.51,
+      "learning_rate": 5.052e-05,
+      "loss": 3.5347,
+      "step": 15160
+    },
+    {
+      "epoch": 91.63,
+      "learning_rate": 5.058666666666667e-05,
+      "loss": 3.5646,
+      "step": 15180
+    },
+    {
+      "epoch": 91.75,
+      "learning_rate": 5.065333333333333e-05,
+      "loss": 3.5525,
+      "step": 15200
+    },
+    {
+      "epoch": 91.87,
+      "learning_rate": 5.072e-05,
+      "loss": 3.5477,
+      "step": 15220
+    },
+    {
+      "epoch": 91.99,
+      "learning_rate": 5.078666666666667e-05,
+      "loss": 3.5826,
+      "step": 15240
+    },
+    {
+      "epoch": 92.11,
+      "learning_rate": 5.085333333333333e-05,
+      "loss": 3.4527,
+      "step": 15260
+    },
+    {
+      "epoch": 92.23,
+      "learning_rate": 5.092e-05,
+      "loss": 3.4612,
+      "step": 15280
+    },
+    {
+      "epoch": 92.35,
+      "learning_rate": 5.098666666666667e-05,
+      "loss": 3.4665,
+      "step": 15300
+    },
+    {
+      "epoch": 92.47,
+      "learning_rate": 5.105333333333333e-05,
+      "loss": 3.4885,
+      "step": 15320
+    },
+    {
+      "epoch": 92.6,
+      "learning_rate": 5.112e-05,
+      "loss": 3.5039,
+      "step": 15340
+    },
+    {
+      "epoch": 92.72,
+      "learning_rate": 5.118666666666667e-05,
+      "loss": 3.5134,
+      "step": 15360
+    },
+    {
+      "epoch": 92.84,
+      "learning_rate": 5.125333333333333e-05,
+      "loss": 3.5206,
+      "step": 15380
+    },
+    {
+      "epoch": 92.96,
+      "learning_rate": 5.132e-05,
+      "loss": 3.522,
+      "step": 15400
+    },
+    {
+      "epoch": 93.08,
+      "learning_rate": 5.1386666666666674e-05,
+      "loss": 3.4437,
+      "step": 15420
+    },
+    {
+      "epoch": 93.2,
+      "learning_rate": 5.145333333333333e-05,
+      "loss": 3.3961,
+      "step": 15440
+    },
+    {
+      "epoch": 93.32,
+      "learning_rate": 5.152e-05,
+      "loss": 3.4249,
+      "step": 15460
+    },
+    {
+      "epoch": 93.44,
+      "learning_rate": 5.158666666666667e-05,
+      "loss": 3.4609,
+      "step": 15480
+    },
+    {
+      "epoch": 93.56,
+      "learning_rate": 5.165333333333333e-05,
+      "loss": 3.4521,
+      "step": 15500
+    },
+    {
+      "epoch": 93.68,
+      "learning_rate": 5.172e-05,
+      "loss": 3.5057,
+      "step": 15520
+    },
+    {
+      "epoch": 93.8,
+      "learning_rate": 5.178666666666667e-05,
+      "loss": 3.454,
+      "step": 15540
+    },
+    {
+      "epoch": 93.92,
+      "learning_rate": 5.1853333333333334e-05,
+      "loss": 3.4676,
+      "step": 15560
+    },
+    {
+      "epoch": 94.04,
+      "learning_rate": 5.1920000000000004e-05,
+      "loss": 3.4264,
+      "step": 15580
+    },
+    {
+      "epoch": 94.16,
+      "learning_rate": 5.1986666666666675e-05,
+      "loss": 3.3537,
+      "step": 15600
+    },
+    {
+      "epoch": 94.29,
+      "learning_rate": 5.205333333333333e-05,
+      "loss": 3.3604,
+      "step": 15620
+    },
+    {
+      "epoch": 94.41,
+      "learning_rate": 5.212e-05,
+      "loss": 3.4009,
+      "step": 15640
+    },
+    {
+      "epoch": 94.53,
+      "learning_rate": 5.218666666666667e-05,
+      "loss": 3.3898,
+      "step": 15660
+    },
+    {
+      "epoch": 94.65,
+      "learning_rate": 5.225333333333333e-05,
+      "loss": 3.418,
+      "step": 15680
+    },
+    {
+      "epoch": 94.77,
+      "learning_rate": 5.232e-05,
+      "loss": 3.407,
+      "step": 15700
+    },
+    {
+      "epoch": 94.89,
+      "learning_rate": 5.238666666666667e-05,
+      "loss": 3.4534,
+      "step": 15720
+    },
+    {
+      "epoch": 95.01,
+      "learning_rate": 5.2453333333333335e-05,
+      "loss": 3.444,
+      "step": 15740
+    },
+    {
+      "epoch": 95.13,
+      "learning_rate": 5.2520000000000005e-05,
+      "loss": 3.2952,
+      "step": 15760
+    },
+    {
+      "epoch": 95.25,
+      "learning_rate": 5.2586666666666676e-05,
+      "loss": 3.3335,
+      "step": 15780
+    },
+    {
+      "epoch": 95.37,
+      "learning_rate": 5.265333333333333e-05,
+      "loss": 3.3337,
+      "step": 15800
+    },
+    {
+      "epoch": 95.49,
+      "learning_rate": 5.2720000000000003e-05,
+      "loss": 3.3459,
+      "step": 15820
+    },
+    {
+      "epoch": 95.61,
+      "learning_rate": 5.2786666666666674e-05,
+      "loss": 3.3494,
+      "step": 15840
+    },
+    {
+      "epoch": 95.73,
+      "learning_rate": 5.285333333333333e-05,
+      "loss": 3.375,
+      "step": 15860
+    },
+    {
+      "epoch": 95.86,
+      "learning_rate": 5.292e-05,
+      "loss": 3.3768,
+      "step": 15880
+    },
+    {
+      "epoch": 95.98,
+      "learning_rate": 5.298666666666667e-05,
+      "loss": 3.4199,
+      "step": 15900
+    },
+    {
+      "epoch": 96.1,
+      "learning_rate": 5.3053333333333336e-05,
+      "loss": 3.2968,
+      "step": 15920
+    },
+    {
+      "epoch": 96.22,
+      "learning_rate": 5.3120000000000006e-05,
+      "loss": 3.2651,
+      "step": 15940
+    },
+    {
+      "epoch": 96.34,
+      "learning_rate": 5.318666666666667e-05,
+      "loss": 3.2762,
+      "step": 15960
+    },
+    {
+      "epoch": 96.46,
+      "learning_rate": 5.3253333333333334e-05,
+      "loss": 3.3002,
+      "step": 15980
+    },
+    {
+      "epoch": 96.58,
+      "learning_rate": 5.3320000000000004e-05,
+      "loss": 3.3163,
+      "step": 16000
+    },
+    {
+      "epoch": 96.58,
+      "eval_accuracy": 0.0013558099791565826,
+      "eval_loss": 5.293873310089111,
+      "eval_runtime": 18.1779,
+      "eval_samples_per_second": 112.224,
+      "eval_steps_per_second": 2.366,
+      "step": 16000
+    },
+    {
+      "epoch": 96.7,
+      "learning_rate": 5.3386666666666675e-05,
+      "loss": 3.3285,
+      "step": 16020
+    },
+    {
+      "epoch": 96.82,
+      "learning_rate": 5.345333333333333e-05,
+      "loss": 3.3154,
+      "step": 16040
+    },
+    {
+      "epoch": 96.94,
+      "learning_rate": 5.352e-05,
+      "loss": 3.3525,
+      "step": 16060
+    },
+    {
+      "epoch": 97.06,
+      "learning_rate": 5.358666666666667e-05,
+      "loss": 3.2891,
+      "step": 16080
+    },
+    {
+      "epoch": 97.18,
+      "learning_rate": 5.365333333333333e-05,
+      "loss": 3.2113,
+      "step": 16100
+    },
+    {
+      "epoch": 97.3,
+      "learning_rate": 5.372e-05,
+      "loss": 3.2459,
+      "step": 16120
+    },
+    {
+      "epoch": 97.42,
+      "learning_rate": 5.378666666666667e-05,
+      "loss": 3.2479,
+      "step": 16140
+    },
+    {
+      "epoch": 97.55,
+      "learning_rate": 5.3853333333333335e-05,
+      "loss": 3.2572,
+      "step": 16160
+    },
+    {
+      "epoch": 97.67,
+      "learning_rate": 5.3920000000000006e-05,
+      "loss": 3.2825,
+      "step": 16180
+    },
+    {
+      "epoch": 97.79,
+      "learning_rate": 5.3986666666666676e-05,
+      "loss": 3.2782,
+      "step": 16200
+    },
+    {
+      "epoch": 97.91,
+      "learning_rate": 5.405333333333333e-05,
+      "loss": 3.2773,
+      "step": 16220
+    },
+    {
+      "epoch": 98.03,
+      "learning_rate": 5.4120000000000004e-05,
+      "loss": 3.2568,
+      "step": 16240
+    },
+    {
+      "epoch": 98.15,
+      "learning_rate": 5.4186666666666674e-05,
+      "loss": 3.1807,
+      "step": 16260
+    },
+    {
+      "epoch": 98.27,
+      "learning_rate": 5.425333333333333e-05,
+      "loss": 3.1694,
+      "step": 16280
+    },
+    {
+      "epoch": 98.39,
+      "learning_rate": 5.432e-05,
+      "loss": 3.1982,
+      "step": 16300
+    },
+    {
+      "epoch": 98.51,
+      "learning_rate": 5.438666666666667e-05,
+      "loss": 3.2195,
+      "step": 16320
+    },
+    {
+      "epoch": 98.63,
+      "learning_rate": 5.4453333333333336e-05,
+      "loss": 3.2266,
+      "step": 16340
+    },
+    {
+      "epoch": 98.75,
+      "learning_rate": 5.4520000000000007e-05,
+      "loss": 3.2312,
+      "step": 16360
+    },
+    {
+      "epoch": 98.87,
+      "learning_rate": 5.4586666666666664e-05,
+      "loss": 3.2291,
+      "step": 16380
+    },
+    {
+      "epoch": 98.99,
+      "learning_rate": 5.4653333333333334e-05,
+      "loss": 3.2518,
+      "step": 16400
+    },
+    {
+      "epoch": 99.11,
+      "learning_rate": 5.4720000000000005e-05,
+      "loss": 3.1003,
+      "step": 16420
+    },
+    {
+      "epoch": 99.24,
+      "learning_rate": 5.478666666666666e-05,
+      "loss": 3.1198,
+      "step": 16440
+    },
+    {
+      "epoch": 99.36,
+      "learning_rate": 5.485333333333333e-05,
+      "loss": 3.1373,
+      "step": 16460
+    },
+    {
+      "epoch": 99.48,
+      "learning_rate": 5.492e-05,
+      "loss": 3.164,
+      "step": 16480
+    },
+    {
+      "epoch": 99.6,
+      "learning_rate": 5.4986666666666666e-05,
+      "loss": 3.1803,
+      "step": 16500
+    },
+    {
+      "epoch": 99.72,
+      "learning_rate": 5.505333333333334e-05,
+      "loss": 3.1856,
+      "step": 16520
+    },
+    {
+      "epoch": 99.84,
+      "learning_rate": 5.512000000000001e-05,
+      "loss": 3.2084,
+      "step": 16540
+    },
+    {
+      "epoch": 99.96,
+      "learning_rate": 5.5186666666666665e-05,
+      "loss": 3.2043,
+      "step": 16560
+    },
+    {
+      "epoch": 100.08,
+      "learning_rate": 5.5253333333333335e-05,
+      "loss": 3.1133,
+      "step": 16580
+    },
+    {
+      "epoch": 100.2,
+      "learning_rate": 5.531666666666667e-05,
+      "loss": 3.0598,
+      "step": 16600
+    },
+    {
+      "epoch": 100.32,
+      "learning_rate": 5.538333333333333e-05,
+      "loss": 3.0833,
+      "step": 16620
+    },
+    {
+      "epoch": 100.44,
+      "learning_rate": 5.545e-05,
+      "loss": 3.0921,
+      "step": 16640
+    },
+    {
+      "epoch": 100.56,
+      "learning_rate": 5.551666666666667e-05,
+      "loss": 3.1288,
+      "step": 16660
+    },
+    {
+      "epoch": 100.68,
+      "learning_rate": 5.5583333333333334e-05,
+      "loss": 3.1343,
+      "step": 16680
+    },
+    {
+      "epoch": 100.8,
+      "learning_rate": 5.5650000000000004e-05,
+      "loss": 3.1412,
+      "step": 16700
+    },
+    {
+      "epoch": 100.93,
+      "learning_rate": 5.5716666666666675e-05,
+      "loss": 3.1505,
+      "step": 16720
+    },
+    {
+      "epoch": 101.05,
+      "learning_rate": 5.578333333333333e-05,
+      "loss": 3.109,
+      "step": 16740
+    },
+    {
+      "epoch": 101.17,
+      "learning_rate": 5.585e-05,
+      "loss": 3.0122,
+      "step": 16760
+    },
+    {
+      "epoch": 101.29,
+      "learning_rate": 5.591666666666667e-05,
+      "loss": 3.0209,
+      "step": 16780
+    },
+    {
+      "epoch": 101.41,
+      "learning_rate": 5.598333333333333e-05,
+      "loss": 3.0625,
+      "step": 16800
+    },
+    {
+      "epoch": 101.53,
+      "learning_rate": 5.605e-05,
+      "loss": 3.0545,
+      "step": 16820
+    },
+    {
+      "epoch": 101.65,
+      "learning_rate": 5.611666666666667e-05,
+      "loss": 3.0823,
+      "step": 16840
+    },
+    {
+      "epoch": 101.77,
+      "learning_rate": 5.6183333333333335e-05,
+      "loss": 3.0862,
+      "step": 16860
+    },
+    {
+      "epoch": 101.89,
+      "learning_rate": 5.6250000000000005e-05,
+      "loss": 3.0877,
+      "step": 16880
+    },
+    {
+      "epoch": 102.01,
+      "learning_rate": 5.6316666666666676e-05,
+      "loss": 3.0987,
+      "step": 16900
+    },
+    {
+      "epoch": 102.13,
+      "learning_rate": 5.638333333333333e-05,
+      "loss": 2.9435,
+      "step": 16920
+    },
+    {
+      "epoch": 102.25,
+      "learning_rate": 5.645e-05,
+      "loss": 2.9804,
+      "step": 16940
+    },
+    {
+      "epoch": 102.37,
+      "learning_rate": 5.6516666666666674e-05,
+      "loss": 3.0011,
+      "step": 16960
+    },
+    {
+      "epoch": 102.49,
+      "learning_rate": 5.658333333333333e-05,
+      "loss": 3.0206,
+      "step": 16980
+    },
+    {
+      "epoch": 102.62,
+      "learning_rate": 5.665e-05,
+      "loss": 3.0377,
+      "step": 17000
+    },
+    {
+      "epoch": 102.62,
+      "eval_accuracy": 0.001351364700536397,
+      "eval_loss": 5.668473720550537,
+      "eval_runtime": 18.2311,
+      "eval_samples_per_second": 111.897,
+      "eval_steps_per_second": 2.359,
+      "step": 17000
+    },
+    {
+      "epoch": 102.74,
+      "learning_rate": 5.671666666666667e-05,
+      "loss": 3.0418,
+      "step": 17020
+    },
+    {
+      "epoch": 102.86,
+      "learning_rate": 5.6783333333333336e-05,
+      "loss": 3.043,
+      "step": 17040
+    },
+    {
+      "epoch": 102.98,
+      "learning_rate": 5.6850000000000006e-05,
+      "loss": 3.0512,
+      "step": 17060
+    },
+    {
+      "epoch": 103.1,
+      "learning_rate": 5.691666666666668e-05,
+      "loss": 2.9331,
+      "step": 17080
+    },
+    {
+      "epoch": 103.22,
+      "learning_rate": 5.6983333333333334e-05,
+      "loss": 2.9127,
+      "step": 17100
+    },
+    {
+      "epoch": 103.34,
+      "learning_rate": 5.7050000000000004e-05,
+      "loss": 2.9382,
+      "step": 17120
+    },
+    {
+      "epoch": 103.46,
+      "learning_rate": 5.7116666666666675e-05,
+      "loss": 2.9569,
+      "step": 17140
+    },
+    {
+      "epoch": 103.58,
+      "learning_rate": 5.718333333333333e-05,
+      "loss": 2.9815,
+      "step": 17160
+    },
+    {
+      "epoch": 103.7,
+      "learning_rate": 5.725e-05,
+      "loss": 2.9712,
+      "step": 17180
+    },
+    {
+      "epoch": 103.82,
+      "learning_rate": 5.731666666666667e-05,
+      "loss": 3.0054,
+      "step": 17200
+    },
+    {
+      "epoch": 103.94,
+      "learning_rate": 5.738333333333334e-05,
+      "loss": 2.9977,
+      "step": 17220
+    },
+    {
+      "epoch": 104.06,
+      "learning_rate": 5.745e-05,
+      "loss": 2.942,
+      "step": 17240
+    },
+    {
+      "epoch": 104.19,
+      "learning_rate": 5.751666666666667e-05,
+      "loss": 2.8658,
+      "step": 17260
+    },
+    {
+      "epoch": 104.31,
+      "learning_rate": 5.7583333333333335e-05,
+      "loss": 2.8741,
+      "step": 17280
+    },
+    {
+      "epoch": 104.43,
+      "learning_rate": 5.7650000000000005e-05,
+      "loss": 2.8884,
+      "step": 17300
+    },
+    {
+      "epoch": 104.55,
+      "learning_rate": 5.7716666666666676e-05,
+      "loss": 2.9327,
+      "step": 17320
+    },
+    {
+      "epoch": 104.67,
+      "learning_rate": 5.778333333333333e-05,
+      "loss": 2.9238,
+      "step": 17340
+    },
+    {
+      "epoch": 104.79,
+      "learning_rate": 5.7850000000000003e-05,
+      "loss": 2.9496,
+      "step": 17360
+    },
+    {
+      "epoch": 104.91,
+      "learning_rate": 5.7916666666666674e-05,
+      "loss": 2.9621,
+      "step": 17380
+    },
+    {
+      "epoch": 105.03,
+      "learning_rate": 5.798333333333333e-05,
+      "loss": 2.9189,
+      "step": 17400
+    },
+    {
+      "epoch": 105.15,
+      "learning_rate": 5.805e-05,
+      "loss": 2.7945,
+      "step": 17420
+    },
+    {
+      "epoch": 105.27,
+      "learning_rate": 5.811666666666667e-05,
+      "loss": 2.8316,
+      "step": 17440
+    },
+    {
+      "epoch": 105.39,
+      "learning_rate": 5.8183333333333336e-05,
+      "loss": 2.8449,
+      "step": 17460
+    },
+    {
+      "epoch": 105.51,
+      "learning_rate": 5.8250000000000006e-05,
+      "loss": 2.8638,
+      "step": 17480
+    },
+    {
+      "epoch": 105.63,
+      "learning_rate": 5.831666666666668e-05,
+      "loss": 2.8714,
+      "step": 17500
+    },
+    {
+      "epoch": 105.75,
+      "learning_rate": 5.8383333333333334e-05,
+      "loss": 2.9079,
+      "step": 17520
+    },
+    {
+      "epoch": 105.88,
+      "learning_rate": 5.8450000000000005e-05,
+      "loss": 2.9015,
+      "step": 17540
+    },
+    {
+      "epoch": 106.0,
+      "learning_rate": 5.851666666666666e-05,
+      "loss": 2.9151,
+      "step": 17560
+    },
+    {
+      "epoch": 106.12,
+      "learning_rate": 5.858333333333333e-05,
+      "loss": 2.7488,
+      "step": 17580
+    },
+    {
+      "epoch": 106.24,
+      "learning_rate": 5.865e-05,
+      "loss": 2.7721,
+      "step": 17600
+    },
+    {
+      "epoch": 106.36,
+      "learning_rate": 5.8716666666666666e-05,
+      "loss": 2.7912,
+      "step": 17620
+    },
+    {
+      "epoch": 106.48,
+      "learning_rate": 5.878333333333334e-05,
+      "loss": 2.8061,
+      "step": 17640
+    },
+    {
+      "epoch": 106.6,
+      "learning_rate": 5.885000000000001e-05,
+      "loss": 2.819,
+      "step": 17660
+    },
+    {
+      "epoch": 106.72,
+      "learning_rate": 5.8916666666666664e-05,
+      "loss": 2.8298,
+      "step": 17680
+    },
+    {
+      "epoch": 106.84,
+      "learning_rate": 5.8983333333333335e-05,
+      "loss": 2.8507,
+      "step": 17700
+    },
+    {
+      "epoch": 106.96,
+      "learning_rate": 5.9050000000000006e-05,
+      "loss": 2.8622,
+      "step": 17720
+    },
+    {
+      "epoch": 107.08,
+      "learning_rate": 5.911666666666666e-05,
+      "loss": 2.7413,
+      "step": 17740
+    },
+    {
+      "epoch": 107.2,
+      "learning_rate": 5.918333333333333e-05,
+      "loss": 2.7069,
+      "step": 17760
+    },
+    {
+      "epoch": 107.32,
+      "learning_rate": 5.9250000000000004e-05,
+      "loss": 2.7273,
+      "step": 17780
+    },
+    {
+      "epoch": 107.44,
+      "learning_rate": 5.931666666666667e-05,
+      "loss": 2.7585,
+      "step": 17800
+    },
+    {
+      "epoch": 107.57,
+      "learning_rate": 5.938333333333334e-05,
+      "loss": 2.7499,
+      "step": 17820
+    },
+    {
+      "epoch": 107.69,
+      "learning_rate": 5.945000000000001e-05,
+      "loss": 2.778,
+      "step": 17840
+    },
+    {
+      "epoch": 107.81,
+      "learning_rate": 5.9516666666666665e-05,
+      "loss": 2.7969,
+      "step": 17860
+    },
+    {
+      "epoch": 107.93,
+      "learning_rate": 5.9583333333333336e-05,
+      "loss": 2.8039,
+      "step": 17880
+    },
+    {
+      "epoch": 108.05,
+      "learning_rate": 5.9650000000000007e-05,
+      "loss": 2.7593,
+      "step": 17900
+    },
+    {
+      "epoch": 108.17,
+      "learning_rate": 5.9716666666666664e-05,
+      "loss": 2.6348,
+      "step": 17920
+    },
+    {
+      "epoch": 108.29,
+      "learning_rate": 5.9783333333333334e-05,
+      "loss": 2.6874,
+      "step": 17940
+    },
+    {
+      "epoch": 108.41,
+      "learning_rate": 5.9850000000000005e-05,
+      "loss": 2.6992,
+      "step": 17960
+    },
+    {
+      "epoch": 108.53,
+      "learning_rate": 5.991666666666667e-05,
+      "loss": 2.7062,
+      "step": 17980
+    },
+    {
+      "epoch": 108.65,
+      "learning_rate": 5.998333333333334e-05,
+      "loss": 2.7272,
+      "step": 18000
+    },
+    {
+      "epoch": 108.65,
+      "eval_accuracy": 0.0013049362349477926,
+      "eval_loss": 6.164906024932861,
+      "eval_runtime": 18.2042,
+      "eval_samples_per_second": 112.062,
+      "eval_steps_per_second": 2.362,
+      "step": 18000
+    },
+    {
+      "epoch": 108.77,
+      "learning_rate": 6.005000000000001e-05,
+      "loss": 2.7555,
+      "step": 18020
+    },
+    {
+      "epoch": 108.89,
+      "learning_rate": 6.0116666666666667e-05,
+      "loss": 2.7582,
+      "step": 18040
+    },
+    {
+      "epoch": 109.01,
+      "learning_rate": 6.018333333333334e-05,
+      "loss": 2.7407,
+      "step": 18060
+    },
+    {
+      "epoch": 109.13,
+      "learning_rate": 6.025000000000001e-05,
+      "loss": 2.5885,
+      "step": 18080
+    },
+    {
+      "epoch": 109.26,
+      "learning_rate": 6.0316666666666665e-05,
+      "loss": 2.6007,
+      "step": 18100
+    },
+    {
+      "epoch": 109.38,
+      "learning_rate": 6.0383333333333335e-05,
+      "loss": 2.6265,
+      "step": 18120
+    },
+    {
+      "epoch": 109.5,
+      "learning_rate": 6.0450000000000006e-05,
+      "loss": 2.6611,
+      "step": 18140
+    },
+    {
+      "epoch": 109.62,
+      "learning_rate": 6.051666666666666e-05,
+      "loss": 2.6709,
+      "step": 18160
+    },
+    {
+      "epoch": 109.74,
+      "learning_rate": 6.058333333333333e-05,
+      "loss": 2.694,
+      "step": 18180
+    },
+    {
+      "epoch": 109.86,
+      "learning_rate": 6.0650000000000004e-05,
+      "loss": 2.7067,
+      "step": 18200
+    },
+    {
+      "epoch": 109.98,
+      "learning_rate": 6.071666666666667e-05,
+      "loss": 2.7346,
+      "step": 18220
+    },
+    {
+      "epoch": 110.1,
+      "learning_rate": 6.078333333333334e-05,
+      "loss": 2.5661,
+      "step": 18240
+    },
+    {
+      "epoch": 110.22,
+      "learning_rate": 6.085000000000001e-05,
+      "loss": 2.5621,
+      "step": 18260
+    },
+    {
+      "epoch": 110.34,
+      "learning_rate": 6.0916666666666666e-05,
+      "loss": 2.5762,
+      "step": 18280
+    },
+    {
+      "epoch": 110.46,
+      "learning_rate": 6.0983333333333336e-05,
+      "loss": 2.5957,
+      "step": 18300
+    },
+    {
+      "epoch": 110.58,
+      "learning_rate": 6.105e-05,
+      "loss": 2.6202,
+      "step": 18320
+    },
+    {
+      "epoch": 110.7,
+      "learning_rate": 6.111666666666667e-05,
+      "loss": 2.6272,
+      "step": 18340
+    },
+    {
+      "epoch": 110.82,
+      "learning_rate": 6.118000000000001e-05,
+      "loss": 2.6594,
+      "step": 18360
+    },
+    {
+      "epoch": 110.95,
+      "learning_rate": 6.124666666666667e-05,
+      "loss": 2.6702,
+      "step": 18380
+    },
+    {
+      "epoch": 111.07,
+      "learning_rate": 6.131333333333333e-05,
+      "loss": 2.5795,
+      "step": 18400
+    },
+    {
+      "epoch": 111.19,
+      "learning_rate": 6.138e-05,
+      "loss": 2.4937,
+      "step": 18420
+    },
+    {
+      "epoch": 111.31,
+      "learning_rate": 6.144666666666668e-05,
+      "loss": 2.5201,
+      "step": 18440
+    },
+    {
+      "epoch": 111.43,
+      "learning_rate": 6.151333333333334e-05,
+      "loss": 2.5366,
+      "step": 18460
+    },
+    {
+      "epoch": 111.55,
+      "learning_rate": 6.158e-05,
+      "loss": 2.5571,
+      "step": 18480
+    },
+    {
+      "epoch": 111.67,
+      "learning_rate": 6.164666666666668e-05,
+      "loss": 2.5815,
+      "step": 18500
+    },
+    {
+      "epoch": 111.79,
+      "learning_rate": 6.171333333333333e-05,
+      "loss": 2.5902,
+      "step": 18520
+    },
+    {
+      "epoch": 111.91,
+      "learning_rate": 6.178000000000001e-05,
+      "loss": 2.6134,
+      "step": 18540
+    },
+    {
+      "epoch": 112.03,
+      "learning_rate": 6.184666666666667e-05,
+      "loss": 2.5729,
+      "step": 18560
+    },
+    {
+      "epoch": 112.15,
+      "learning_rate": 6.191333333333334e-05,
+      "loss": 2.4462,
+      "step": 18580
+    },
+    {
+      "epoch": 112.27,
+      "learning_rate": 6.198e-05,
+      "loss": 2.4578,
+      "step": 18600
+    },
+    {
+      "epoch": 112.39,
+      "learning_rate": 6.204666666666668e-05,
+      "loss": 2.4986,
+      "step": 18620
+    },
+    {
+      "epoch": 112.52,
+      "learning_rate": 6.211333333333334e-05,
+      "loss": 2.5043,
+      "step": 18640
+    },
+    {
+      "epoch": 112.64,
+      "learning_rate": 6.218e-05,
+      "loss": 2.5323,
+      "step": 18660
+    },
+    {
+      "epoch": 112.76,
+      "learning_rate": 6.224666666666667e-05,
+      "loss": 2.5318,
+      "step": 18680
+    },
+    {
+      "epoch": 112.88,
+      "learning_rate": 6.231333333333333e-05,
+      "loss": 2.5624,
+      "step": 18700
+    },
+    {
+      "epoch": 113.0,
+      "learning_rate": 6.238000000000001e-05,
+      "loss": 2.563,
+      "step": 18720
+    },
+    {
+      "epoch": 113.12,
+      "learning_rate": 6.244666666666666e-05,
+      "loss": 2.3918,
+      "step": 18740
+    },
+    {
+      "epoch": 113.24,
+      "learning_rate": 6.251333333333334e-05,
+      "loss": 2.4037,
+      "step": 18760
+    },
+    {
+      "epoch": 113.36,
+      "learning_rate": 6.258e-05,
+      "loss": 2.4364,
+      "step": 18780
+    },
+    {
+      "epoch": 113.48,
+      "learning_rate": 6.264666666666666e-05,
+      "loss": 2.4501,
+      "step": 18800
+    },
+    {
+      "epoch": 113.6,
+      "learning_rate": 6.271333333333334e-05,
+      "loss": 2.4688,
+      "step": 18820
+    },
+    {
+      "epoch": 113.72,
+      "learning_rate": 6.278e-05,
+      "loss": 2.4888,
+      "step": 18840
+    },
+    {
+      "epoch": 113.84,
+      "learning_rate": 6.284666666666667e-05,
+      "loss": 2.4992,
+      "step": 18860
+    },
+    {
+      "epoch": 113.96,
+      "learning_rate": 6.291333333333333e-05,
+      "loss": 2.5299,
+      "step": 18880
+    },
+    {
+      "epoch": 114.08,
+      "learning_rate": 6.298000000000001e-05,
+      "loss": 2.396,
+      "step": 18900
+    },
+    {
+      "epoch": 114.21,
+      "learning_rate": 6.304666666666666e-05,
+      "loss": 2.348,
+      "step": 18920
+    },
+    {
+      "epoch": 114.33,
+      "learning_rate": 6.311333333333334e-05,
+      "loss": 2.3742,
+      "step": 18940
+    },
+    {
+      "epoch": 114.45,
+      "learning_rate": 6.318e-05,
+      "loss": 2.3905,
+      "step": 18960
+    },
+    {
+      "epoch": 114.57,
+      "learning_rate": 6.324666666666667e-05,
+      "loss": 2.4142,
+      "step": 18980
+    },
+    {
+      "epoch": 114.69,
+      "learning_rate": 6.331333333333333e-05,
+      "loss": 2.4319,
+      "step": 19000
+    },
+    {
+      "epoch": 114.69,
+      "eval_accuracy": 0.0012856733609269888,
+      "eval_loss": 6.755620956420898,
+      "eval_runtime": 18.1595,
+      "eval_samples_per_second": 112.338,
+      "eval_steps_per_second": 2.368,
+      "step": 19000
+    },
+    {
+      "epoch": 114.81,
+      "learning_rate": 6.338e-05,
+      "loss": 2.4442,
+      "step": 19020
+    },
+    {
+      "epoch": 114.93,
+      "learning_rate": 6.344666666666667e-05,
+      "loss": 2.4764,
+      "step": 19040
+    },
+    {
+      "epoch": 115.05,
+      "learning_rate": 6.351333333333333e-05,
+      "loss": 2.4088,
+      "step": 19060
+    },
+    {
+      "epoch": 115.17,
+      "learning_rate": 6.358000000000001e-05,
+      "loss": 2.2868,
+      "step": 19080
+    },
+    {
+      "epoch": 115.29,
+      "learning_rate": 6.364666666666666e-05,
+      "loss": 2.3098,
+      "step": 19100
+    },
+    {
+      "epoch": 115.41,
+      "learning_rate": 6.371333333333334e-05,
+      "loss": 2.3324,
+      "step": 19120
+    },
+    {
+      "epoch": 115.53,
+      "learning_rate": 6.378e-05,
+      "loss": 2.3646,
+      "step": 19140
+    },
+    {
+      "epoch": 115.65,
+      "learning_rate": 6.384666666666667e-05,
+      "loss": 2.3865,
+      "step": 19160
+    },
+    {
+      "epoch": 115.77,
+      "learning_rate": 6.391333333333333e-05,
+      "loss": 2.3991,
+      "step": 19180
+    },
+    {
+      "epoch": 115.9,
+      "learning_rate": 6.398000000000001e-05,
+      "loss": 2.4262,
+      "step": 19200
+    },
+    {
+      "epoch": 116.02,
+      "learning_rate": 6.404666666666667e-05,
+      "loss": 2.3915,
+      "step": 19220
+    },
+    {
+      "epoch": 116.14,
+      "learning_rate": 6.411333333333333e-05,
+      "loss": 2.2261,
+      "step": 19240
+    },
+    {
+      "epoch": 116.26,
+      "learning_rate": 6.418000000000001e-05,
+      "loss": 2.2716,
+      "step": 19260
+    },
+    {
+      "epoch": 116.38,
+      "learning_rate": 6.424666666666666e-05,
+      "loss": 2.2805,
+      "step": 19280
+    },
+    {
+      "epoch": 116.5,
+      "learning_rate": 6.431333333333334e-05,
+      "loss": 2.3009,
+      "step": 19300
+    },
+    {
+      "epoch": 116.62,
+      "learning_rate": 6.438e-05,
+      "loss": 2.3275,
+      "step": 19320
+    },
+    {
+      "epoch": 116.74,
+      "learning_rate": 6.444666666666667e-05,
+      "loss": 2.3381,
+      "step": 19340
+    },
+    {
+      "epoch": 116.86,
+      "learning_rate": 6.451333333333333e-05,
+      "loss": 2.366,
+      "step": 19360
+    },
+    {
+      "epoch": 116.98,
+      "learning_rate": 6.458000000000001e-05,
+      "loss": 2.381,
+      "step": 19380
+    },
+    {
+      "epoch": 117.1,
+      "learning_rate": 6.464666666666667e-05,
+      "loss": 2.2173,
+      "step": 19400
+    },
+    {
+      "epoch": 117.22,
+      "learning_rate": 6.471333333333334e-05,
+      "loss": 2.1985,
+      "step": 19420
+    },
+    {
+      "epoch": 117.34,
+      "learning_rate": 6.478000000000001e-05,
+      "loss": 2.2065,
+      "step": 19440
+    },
+    {
+      "epoch": 117.46,
+      "learning_rate": 6.484666666666666e-05,
+      "loss": 2.2554,
+      "step": 19460
+    },
+    {
+      "epoch": 117.59,
+      "learning_rate": 6.491333333333334e-05,
+      "loss": 2.2683,
+      "step": 19480
+    },
+    {
+      "epoch": 117.71,
+      "learning_rate": 6.498e-05,
+      "loss": 2.2806,
+      "step": 19500
+    },
+    {
+      "epoch": 117.83,
+      "learning_rate": 6.504666666666667e-05,
+      "loss": 2.3162,
+      "step": 19520
+    },
+    {
+      "epoch": 117.95,
+      "learning_rate": 6.511333333333333e-05,
+      "loss": 2.3214,
+      "step": 19540
+    },
+    {
+      "epoch": 118.07,
+      "learning_rate": 6.518000000000001e-05,
+      "loss": 2.2072,
+      "step": 19560
+    },
+    {
+      "epoch": 118.19,
+      "learning_rate": 6.524666666666667e-05,
+      "loss": 2.1393,
+      "step": 19580
+    },
+    {
+      "epoch": 118.31,
+      "learning_rate": 6.531333333333334e-05,
+      "loss": 2.1773,
+      "step": 19600
+    },
+    {
+      "epoch": 118.43,
+      "learning_rate": 6.538000000000001e-05,
+      "loss": 2.192,
+      "step": 19620
+    },
+    {
+      "epoch": 118.55,
+      "learning_rate": 6.544666666666666e-05,
+      "loss": 2.2329,
+      "step": 19640
+    },
+    {
+      "epoch": 118.67,
+      "learning_rate": 6.551333333333334e-05,
+      "loss": 2.2384,
+      "step": 19660
+    },
+    {
+      "epoch": 118.79,
+      "learning_rate": 6.558e-05,
+      "loss": 2.2479,
+      "step": 19680
+    },
+    {
+      "epoch": 118.91,
+      "learning_rate": 6.564666666666667e-05,
+      "loss": 2.2883,
+      "step": 19700
+    },
+    {
+      "epoch": 119.03,
+      "learning_rate": 6.571333333333333e-05,
+      "loss": 2.2437,
+      "step": 19720
+    },
+    {
+      "epoch": 119.15,
+      "learning_rate": 6.578000000000001e-05,
+      "loss": 2.0835,
+      "step": 19740
+    },
+    {
+      "epoch": 119.28,
+      "learning_rate": 6.584666666666667e-05,
+      "loss": 2.1107,
+      "step": 19760
+    },
+    {
+      "epoch": 119.4,
+      "learning_rate": 6.591333333333334e-05,
+      "loss": 2.1388,
+      "step": 19780
+    },
+    {
+      "epoch": 119.52,
+      "learning_rate": 6.598e-05,
+      "loss": 2.1651,
+      "step": 19800
+    },
+    {
+      "epoch": 119.64,
+      "learning_rate": 6.604666666666667e-05,
+      "loss": 2.1889,
+      "step": 19820
+    },
+    {
+      "epoch": 119.76,
+      "learning_rate": 6.611333333333334e-05,
+      "loss": 2.1927,
+      "step": 19840
+    },
+    {
+      "epoch": 119.88,
+      "learning_rate": 6.618e-05,
+      "loss": 2.2085,
+      "step": 19860
+    },
+    {
+      "epoch": 120.0,
+      "learning_rate": 6.624666666666667e-05,
+      "loss": 2.2624,
+      "step": 19880
+    },
+    {
+      "epoch": 120.12,
+      "learning_rate": 6.631333333333333e-05,
+      "loss": 2.0292,
+      "step": 19900
+    },
+    {
+      "epoch": 120.24,
+      "learning_rate": 6.638e-05,
+      "loss": 2.0654,
+      "step": 19920
+    },
+    {
+      "epoch": 120.36,
+      "learning_rate": 6.644666666666666e-05,
+      "loss": 2.0827,
+      "step": 19940
+    },
+    {
+      "epoch": 120.48,
+      "learning_rate": 6.651333333333334e-05,
+      "loss": 2.1191,
+      "step": 19960
+    },
+    {
+      "epoch": 120.6,
+      "learning_rate": 6.658e-05,
+      "loss": 2.1327,
+      "step": 19980
+    },
+    {
+      "epoch": 120.72,
+      "learning_rate": 6.664666666666667e-05,
+      "loss": 2.1647,
+      "step": 20000
+    },
+    {
+      "epoch": 120.72,
+      "eval_accuracy": 0.0013138267921881637,
+      "eval_loss": 7.395051002502441,
+      "eval_runtime": 18.1458,
+      "eval_samples_per_second": 112.423,
+      "eval_steps_per_second": 2.37,
+      "step": 20000
+    },
+    {
+      "epoch": 120.85,
+      "learning_rate": 6.671333333333334e-05,
+      "loss": 2.1718,
+      "step": 20020
+    },
+    {
+      "epoch": 120.97,
+      "learning_rate": 6.678e-05,
+      "loss": 2.179,
+      "step": 20040
+    },
+    {
+      "epoch": 121.09,
+      "learning_rate": 6.684666666666667e-05,
+      "loss": 2.0547,
+      "step": 20060
+    },
+    {
+      "epoch": 121.21,
+      "learning_rate": 6.691333333333334e-05,
+      "loss": 2.0058,
+      "step": 20080
+    },
+    {
+      "epoch": 121.33,
+      "learning_rate": 6.698e-05,
+      "loss": 2.0423,
+      "step": 20100
+    },
+    {
+      "epoch": 121.45,
+      "learning_rate": 6.704666666666666e-05,
+      "loss": 2.0595,
+      "step": 20120
+    },
+    {
+      "epoch": 121.57,
+      "learning_rate": 6.711333333333334e-05,
+      "loss": 2.0777,
+      "step": 20140
+    },
+    {
+      "epoch": 121.69,
+      "learning_rate": 6.718e-05,
+      "loss": 2.0979,
+      "step": 20160
+    },
+    {
+      "epoch": 121.81,
+      "learning_rate": 6.724666666666667e-05,
+      "loss": 2.132,
+      "step": 20180
+    },
+    {
+      "epoch": 121.93,
+      "learning_rate": 6.731333333333335e-05,
+      "loss": 2.1595,
+      "step": 20200
+    },
+    {
+      "epoch": 122.05,
+      "learning_rate": 6.738e-05,
+      "loss": 2.0667,
+      "step": 20220
+    },
+    {
+      "epoch": 122.17,
+      "learning_rate": 6.744666666666667e-05,
+      "loss": 1.9628,
+      "step": 20240
+    },
+    {
+      "epoch": 122.29,
+      "learning_rate": 6.751333333333334e-05,
+      "loss": 1.9853,
+      "step": 20260
+    },
+    {
+      "epoch": 122.41,
+      "learning_rate": 6.758e-05,
+      "loss": 2.0203,
+      "step": 20280
+    },
+    {
+      "epoch": 122.54,
+      "learning_rate": 6.764666666666666e-05,
+      "loss": 2.0288,
+      "step": 20300
+    },
+    {
+      "epoch": 122.66,
+      "learning_rate": 6.771333333333334e-05,
+      "loss": 2.0528,
+      "step": 20320
+    },
+    {
+      "epoch": 122.78,
+      "learning_rate": 6.778e-05,
+      "loss": 2.0722,
+      "step": 20340
+    },
+    {
+      "epoch": 122.9,
+      "learning_rate": 6.784666666666667e-05,
+      "loss": 2.0802,
+      "step": 20360
+    },
+    {
+      "epoch": 123.02,
+      "learning_rate": 6.791e-05,
+      "loss": 2.0712,
+      "step": 20380
+    },
+    {
+      "epoch": 123.14,
+      "learning_rate": 6.797666666666667e-05,
+      "loss": 1.9094,
+      "step": 20400
+    },
+    {
+      "epoch": 123.26,
+      "learning_rate": 6.804333333333333e-05,
+      "loss": 1.9334,
+      "step": 20420
+    },
+    {
+      "epoch": 123.38,
+      "learning_rate": 6.811000000000001e-05,
+      "loss": 1.9429,
+      "step": 20440
+    },
+    {
+      "epoch": 123.5,
+      "learning_rate": 6.817666666666666e-05,
+      "loss": 1.9631,
+      "step": 20460
+    },
+    {
+      "epoch": 123.62,
+      "learning_rate": 6.824333333333334e-05,
+      "loss": 2.0011,
+      "step": 20480
+    },
+    {
+      "epoch": 123.74,
+      "learning_rate": 6.831e-05,
+      "loss": 2.0226,
+      "step": 20500
+    },
+    {
+      "epoch": 123.86,
+      "learning_rate": 6.837666666666667e-05,
+      "loss": 2.0459,
+      "step": 20520
+    },
+    {
+      "epoch": 123.98,
+      "learning_rate": 6.844333333333334e-05,
+      "loss": 2.072,
+      "step": 20540
+    },
+    {
+      "epoch": 124.1,
+      "learning_rate": 6.851e-05,
+      "loss": 1.8929,
+      "step": 20560
+    },
+    {
+      "epoch": 124.23,
+      "learning_rate": 6.857666666666667e-05,
+      "loss": 1.8904,
+      "step": 20580
+    },
+    {
+      "epoch": 124.35,
+      "learning_rate": 6.864333333333333e-05,
+      "loss": 1.9023,
+      "step": 20600
+    },
+    {
+      "epoch": 124.47,
+      "learning_rate": 6.871000000000001e-05,
+      "loss": 1.9269,
+      "step": 20620
+    },
+    {
+      "epoch": 124.59,
+      "learning_rate": 6.877666666666666e-05,
+      "loss": 1.9467,
+      "step": 20640
+    },
+    {
+      "epoch": 124.71,
+      "learning_rate": 6.884333333333334e-05,
+      "loss": 1.9685,
+      "step": 20660
+    },
+    {
+      "epoch": 124.83,
+      "learning_rate": 6.891e-05,
+      "loss": 1.9962,
+      "step": 20680
+    },
+    {
+      "epoch": 124.95,
+      "learning_rate": 6.897666666666667e-05,
+      "loss": 2.0065,
+      "step": 20700
+    },
+    {
+      "epoch": 125.07,
+      "learning_rate": 6.904333333333334e-05,
+      "loss": 1.9111,
+      "step": 20720
+    },
+    {
+      "epoch": 125.19,
+      "learning_rate": 6.911000000000001e-05,
+      "loss": 1.8209,
+      "step": 20740
+    },
+    {
+      "epoch": 125.31,
+      "learning_rate": 6.917666666666667e-05,
+      "loss": 1.8422,
+      "step": 20760
+    },
+    {
+      "epoch": 125.43,
+      "learning_rate": 6.924333333333334e-05,
+      "loss": 1.8881,
+      "step": 20780
+    },
+    {
+      "epoch": 125.55,
+      "learning_rate": 6.931000000000001e-05,
+      "loss": 1.895,
+      "step": 20800
+    },
+    {
+      "epoch": 125.67,
+      "learning_rate": 6.937666666666666e-05,
+      "loss": 1.9153,
+      "step": 20820
+    },
+    {
+      "epoch": 125.79,
+      "learning_rate": 6.944333333333334e-05,
+      "loss": 1.9444,
+      "step": 20840
+    },
+    {
+      "epoch": 125.92,
+      "learning_rate": 6.951e-05,
+      "loss": 1.9533,
+      "step": 20860
+    },
+    {
+      "epoch": 126.04,
+      "learning_rate": 6.957666666666667e-05,
+      "loss": 1.9184,
+      "step": 20880
+    },
+    {
+      "epoch": 126.16,
+      "learning_rate": 6.964333333333334e-05,
+      "loss": 1.7681,
+      "step": 20900
+    },
+    {
+      "epoch": 126.28,
+      "learning_rate": 6.971000000000001e-05,
+      "loss": 1.7999,
+      "step": 20920
+    },
+    {
+      "epoch": 126.4,
+      "learning_rate": 6.977666666666667e-05,
+      "loss": 1.8303,
+      "step": 20940
+    },
+    {
+      "epoch": 126.52,
+      "learning_rate": 6.984333333333334e-05,
+      "loss": 1.8537,
+      "step": 20960
+    },
+    {
+      "epoch": 126.64,
+      "learning_rate": 6.991000000000001e-05,
+      "loss": 1.8724,
+      "step": 20980
+    },
+    {
+      "epoch": 126.76,
+      "learning_rate": 6.997666666666666e-05,
+      "loss": 1.9001,
+      "step": 21000
+    },
+    {
+      "epoch": 126.76,
+      "eval_accuracy": 0.001323211269275222,
+      "eval_loss": 8.08230972290039,
+      "eval_runtime": 18.3096,
+      "eval_samples_per_second": 111.417,
+      "eval_steps_per_second": 2.348,
+      "step": 21000
+    },
+    {
+      "epoch": 126.88,
+      "learning_rate": 7.004333333333334e-05,
+      "loss": 1.9246,
+      "step": 21020
+    },
+    {
+      "epoch": 127.0,
+      "learning_rate": 7.011e-05,
+      "loss": 1.9363,
+      "step": 21040
+    },
+    {
+      "epoch": 127.12,
+      "learning_rate": 7.017666666666667e-05,
+      "loss": 1.7457,
+      "step": 21060
+    },
+    {
+      "epoch": 127.24,
+      "learning_rate": 7.024333333333333e-05,
+      "loss": 1.7626,
+      "step": 21080
+    },
+    {
+      "epoch": 127.36,
+      "learning_rate": 7.031e-05,
+      "loss": 1.785,
+      "step": 21100
+    },
+    {
+      "epoch": 127.48,
+      "learning_rate": 7.037666666666667e-05,
+      "loss": 1.7905,
+      "step": 21120
+    },
+    {
+      "epoch": 127.61,
+      "learning_rate": 7.044333333333334e-05,
+      "loss": 1.8258,
+      "step": 21140
+    },
+    {
+      "epoch": 127.73,
+      "learning_rate": 7.051e-05,
+      "loss": 1.8491,
+      "step": 21160
+    },
+    {
+      "epoch": 127.85,
+      "learning_rate": 7.057666666666666e-05,
+      "loss": 1.8689,
+      "step": 21180
+    },
+    {
+      "epoch": 127.97,
+      "learning_rate": 7.064333333333334e-05,
+      "loss": 1.8854,
+      "step": 21200
+    },
+    {
+      "epoch": 128.09,
+      "learning_rate": 7.070999999999999e-05,
+      "loss": 1.7481,
+      "step": 21220
+    },
+    {
+      "epoch": 128.21,
+      "learning_rate": 7.077666666666667e-05,
+      "loss": 1.7011,
+      "step": 21240
+    },
+    {
+      "epoch": 128.33,
+      "learning_rate": 7.084333333333333e-05,
+      "loss": 1.7239,
+      "step": 21260
+    },
+    {
+      "epoch": 128.45,
+      "learning_rate": 7.091e-05,
+      "loss": 1.7696,
+      "step": 21280
+    },
+    {
+      "epoch": 128.57,
+      "learning_rate": 7.097666666666667e-05,
+      "loss": 1.7884,
+      "step": 21300
+    },
+    {
+      "epoch": 128.69,
+      "learning_rate": 7.104333333333334e-05,
+      "loss": 1.8025,
+      "step": 21320
+    },
+    {
+      "epoch": 128.81,
+      "learning_rate": 7.111e-05,
+      "loss": 1.8398,
+      "step": 21340
+    },
+    {
+      "epoch": 128.93,
+      "learning_rate": 7.117666666666667e-05,
+      "loss": 1.8538,
+      "step": 21360
+    },
+    {
+      "epoch": 129.05,
+      "learning_rate": 7.124333333333334e-05,
+      "loss": 1.7678,
+      "step": 21380
+    },
+    {
+      "epoch": 129.18,
+      "learning_rate": 7.130999999999999e-05,
+      "loss": 1.6649,
+      "step": 21400
+    },
+    {
+      "epoch": 129.3,
+      "learning_rate": 7.137666666666667e-05,
+      "loss": 1.695,
+      "step": 21420
+    },
+    {
+      "epoch": 129.42,
+      "learning_rate": 7.144333333333333e-05,
+      "loss": 1.7073,
+      "step": 21440
+    },
+    {
+      "epoch": 129.54,
+      "learning_rate": 7.151e-05,
+      "loss": 1.7488,
+      "step": 21460
+    },
+    {
+      "epoch": 129.66,
+      "learning_rate": 7.157666666666668e-05,
+      "loss": 1.7523,
+      "step": 21480
+    },
+    {
+      "epoch": 129.78,
+      "learning_rate": 7.164333333333334e-05,
+      "loss": 1.792,
+      "step": 21500
+    },
+    {
+      "epoch": 129.9,
+      "learning_rate": 7.171e-05,
+      "loss": 1.8004,
+      "step": 21520
+    },
+    {
+      "epoch": 130.02,
+      "learning_rate": 7.177666666666667e-05,
+      "loss": 1.7849,
+      "step": 21540
+    },
+    {
+      "epoch": 130.14,
+      "learning_rate": 7.184333333333334e-05,
+      "loss": 1.629,
+      "step": 21560
+    },
+    {
+      "epoch": 130.26,
+      "learning_rate": 7.191e-05,
+      "loss": 1.6424,
+      "step": 21580
+    },
+    {
+      "epoch": 130.38,
+      "learning_rate": 7.197666666666667e-05,
+      "loss": 1.6661,
+      "step": 21600
+    },
+    {
+      "epoch": 130.5,
+      "learning_rate": 7.204333333333334e-05,
+      "loss": 1.7055,
+      "step": 21620
+    },
+    {
+      "epoch": 130.62,
+      "learning_rate": 7.211e-05,
+      "loss": 1.7159,
+      "step": 21640
+    },
+    {
+      "epoch": 130.74,
+      "learning_rate": 7.217666666666668e-05,
+      "loss": 1.7365,
+      "step": 21660
+    },
+    {
+      "epoch": 130.87,
+      "learning_rate": 7.224333333333334e-05,
+      "loss": 1.7562,
+      "step": 21680
+    },
+    {
+      "epoch": 130.99,
+      "learning_rate": 7.231e-05,
+      "loss": 1.7823,
+      "step": 21700
+    },
+    {
+      "epoch": 131.11,
+      "learning_rate": 7.237666666666667e-05,
+      "loss": 1.6083,
+      "step": 21720
+    },
+    {
+      "epoch": 131.23,
+      "learning_rate": 7.244333333333335e-05,
+      "loss": 1.6106,
+      "step": 21740
+    },
+    {
+      "epoch": 131.35,
+      "learning_rate": 7.251e-05,
+      "loss": 1.6318,
+      "step": 21760
+    },
+    {
+      "epoch": 131.47,
+      "learning_rate": 7.257666666666667e-05,
+      "loss": 1.6561,
+      "step": 21780
+    },
+    {
+      "epoch": 131.59,
+      "learning_rate": 7.264333333333334e-05,
+      "loss": 1.6767,
+      "step": 21800
+    },
+    {
+      "epoch": 131.71,
+      "learning_rate": 7.271e-05,
+      "loss": 1.681,
+      "step": 21820
+    },
+    {
+      "epoch": 131.83,
+      "learning_rate": 7.277666666666668e-05,
+      "loss": 1.7072,
+      "step": 21840
+    },
+    {
+      "epoch": 131.95,
+      "learning_rate": 7.284333333333334e-05,
+      "loss": 1.7437,
+      "step": 21860
+    },
+    {
+      "epoch": 132.07,
+      "learning_rate": 7.291e-05,
+      "loss": 1.6319,
+      "step": 21880
+    },
+    {
+      "epoch": 132.19,
+      "learning_rate": 7.297666666666667e-05,
+      "loss": 1.547,
+      "step": 21900
+    },
+    {
+      "epoch": 132.31,
+      "learning_rate": 7.304333333333335e-05,
+      "loss": 1.5772,
+      "step": 21920
+    },
+    {
+      "epoch": 132.43,
+      "learning_rate": 7.311e-05,
+      "loss": 1.5979,
+      "step": 21940
+    },
+    {
+      "epoch": 132.56,
+      "learning_rate": 7.317666666666667e-05,
+      "loss": 1.6208,
+      "step": 21960
+    },
+    {
+      "epoch": 132.68,
+      "learning_rate": 7.324333333333334e-05,
+      "loss": 1.6501,
+      "step": 21980
+    },
+    {
+      "epoch": 132.8,
+      "learning_rate": 7.331e-05,
+      "loss": 1.6708,
+      "step": 22000
+    },
+    {
+      "epoch": 132.8,
+      "eval_accuracy": 0.001301972715867669,
+      "eval_loss": 8.823031425476074,
+      "eval_runtime": 18.1374,
+      "eval_samples_per_second": 112.475,
+      "eval_steps_per_second": 2.371,
+      "step": 22000
+    },
+    {
+      "epoch": 132.92,
+      "learning_rate": 7.337666666666666e-05,
+      "loss": 1.6895,
+      "step": 22020
+    },
+    {
+      "epoch": 133.04,
+      "learning_rate": 7.344333333333334e-05,
+      "loss": 1.6339,
+      "step": 22040
+    },
+    {
+      "epoch": 133.16,
+      "learning_rate": 7.351e-05,
+      "loss": 1.4999,
+      "step": 22060
+    },
+    {
+      "epoch": 133.28,
+      "learning_rate": 7.357666666666667e-05,
+      "loss": 1.5272,
+      "step": 22080
+    },
+    {
+      "epoch": 133.4,
+      "learning_rate": 7.364333333333335e-05,
+      "loss": 1.5672,
+      "step": 22100
+    },
+    {
+      "epoch": 133.52,
+      "learning_rate": 7.371e-05,
+      "loss": 1.5869,
+      "step": 22120
+    },
+    {
+      "epoch": 133.64,
+      "learning_rate": 7.377666666666667e-05,
+      "loss": 1.6036,
+      "step": 22140
+    },
+    {
+      "epoch": 133.76,
+      "learning_rate": 7.384333333333334e-05,
+      "loss": 1.6253,
+      "step": 22160
+    },
+    {
+      "epoch": 133.88,
+      "learning_rate": 7.391e-05,
+      "loss": 1.6508,
+      "step": 22180
+    },
+    {
+      "epoch": 134.0,
+      "learning_rate": 7.397666666666667e-05,
+      "loss": 1.6691,
+      "step": 22200
+    },
+    {
+      "epoch": 134.12,
+      "learning_rate": 7.404333333333334e-05,
+      "loss": 1.4677,
+      "step": 22220
+    },
+    {
+      "epoch": 134.25,
+      "learning_rate": 7.411000000000001e-05,
+      "loss": 1.4871,
+      "step": 22240
+    },
+    {
+      "epoch": 134.37,
+      "learning_rate": 7.417666666666667e-05,
+      "loss": 1.5172,
+      "step": 22260
+    },
+    {
+      "epoch": 134.49,
+      "learning_rate": 7.424333333333333e-05,
+      "loss": 1.5336,
+      "step": 22280
+    },
+    {
+      "epoch": 134.61,
+      "learning_rate": 7.431e-05,
+      "loss": 1.5616,
+      "step": 22300
+    },
+    {
+      "epoch": 134.73,
+      "learning_rate": 7.437666666666668e-05,
+      "loss": 1.5933,
+      "step": 22320
+    },
+    {
+      "epoch": 134.85,
+      "learning_rate": 7.444333333333333e-05,
+      "loss": 1.6153,
+      "step": 22340
+    },
+    {
+      "epoch": 134.97,
+      "learning_rate": 7.451e-05,
+      "loss": 1.6363,
+      "step": 22360
+    },
+    {
+      "epoch": 135.09,
+      "learning_rate": 7.457333333333334e-05,
+      "loss": 1.4673,
+      "step": 22380
+    },
+    {
+      "epoch": 135.21,
+      "learning_rate": 7.464e-05,
+      "loss": 1.4428,
+      "step": 22400
+    },
+    {
+      "epoch": 135.33,
+      "learning_rate": 7.470666666666667e-05,
+      "loss": 1.4685,
+      "step": 22420
+    },
+    {
+      "epoch": 135.45,
+      "learning_rate": 7.477333333333334e-05,
+      "loss": 1.5001,
+      "step": 22440
+    },
+    {
+      "epoch": 135.57,
+      "learning_rate": 7.484e-05,
+      "loss": 1.5205,
+      "step": 22460
+    },
+    {
+      "epoch": 135.69,
+      "learning_rate": 7.490666666666667e-05,
+      "loss": 1.5495,
+      "step": 22480
+    },
+    {
+      "epoch": 135.81,
+      "learning_rate": 7.497333333333334e-05,
+      "loss": 1.5563,
+      "step": 22500
+    },
+    {
+      "epoch": 135.94,
+      "learning_rate": 7.504e-05,
+      "loss": 1.5845,
+      "step": 22520
+    },
+    {
+      "epoch": 136.06,
+      "learning_rate": 7.510666666666666e-05,
+      "loss": 1.498,
+      "step": 22540
+    },
+    {
+      "epoch": 136.18,
+      "learning_rate": 7.517333333333334e-05,
+      "loss": 1.4094,
+      "step": 22560
+    },
+    {
+      "epoch": 136.3,
+      "learning_rate": 7.524e-05,
+      "loss": 1.4375,
+      "step": 22580
+    },
+    {
+      "epoch": 136.42,
+      "learning_rate": 7.530666666666667e-05,
+      "loss": 1.4701,
+      "step": 22600
+    },
+    {
+      "epoch": 136.54,
+      "learning_rate": 7.537333333333335e-05,
+      "loss": 1.5002,
+      "step": 22620
+    },
+    {
+      "epoch": 136.66,
+      "learning_rate": 7.544e-05,
+      "loss": 1.512,
+      "step": 22640
+    },
+    {
+      "epoch": 136.78,
+      "learning_rate": 7.550666666666667e-05,
+      "loss": 1.5179,
+      "step": 22660
+    },
+    {
+      "epoch": 136.9,
+      "learning_rate": 7.557333333333334e-05,
+      "loss": 1.5384,
+      "step": 22680
+    },
+    {
+      "epoch": 137.02,
+      "learning_rate": 7.564e-05,
+      "loss": 1.5182,
+      "step": 22700
+    },
+    {
+      "epoch": 137.14,
+      "learning_rate": 7.570666666666666e-05,
+      "loss": 1.3657,
+      "step": 22720
+    },
+    {
+      "epoch": 137.26,
+      "learning_rate": 7.577333333333334e-05,
+      "loss": 1.3809,
+      "step": 22740
+    },
+    {
+      "epoch": 137.38,
+      "learning_rate": 7.584e-05,
+      "loss": 1.4188,
+      "step": 22760
+    },
+    {
+      "epoch": 137.51,
+      "learning_rate": 7.590666666666667e-05,
+      "loss": 1.4384,
+      "step": 22780
+    },
+    {
+      "epoch": 137.63,
+      "learning_rate": 7.597333333333335e-05,
+      "loss": 1.4652,
+      "step": 22800
+    },
+    {
+      "epoch": 137.75,
+      "learning_rate": 7.604e-05,
+      "loss": 1.4989,
+      "step": 22820
+    },
+    {
+      "epoch": 137.87,
+      "learning_rate": 7.610666666666667e-05,
+      "loss": 1.5065,
+      "step": 22840
+    },
+    {
+      "epoch": 137.99,
+      "learning_rate": 7.617333333333334e-05,
+      "loss": 1.5319,
+      "step": 22860
+    },
+    {
+      "epoch": 138.11,
+      "learning_rate": 7.624e-05,
+      "loss": 1.3482,
+      "step": 22880
+    },
+    {
+      "epoch": 138.23,
+      "learning_rate": 7.630666666666667e-05,
+      "loss": 1.3541,
+      "step": 22900
+    },
+    {
+      "epoch": 138.35,
+      "learning_rate": 7.637333333333334e-05,
+      "loss": 1.3673,
+      "step": 22920
+    },
+    {
+      "epoch": 138.47,
+      "learning_rate": 7.644e-05,
+      "loss": 1.4019,
+      "step": 22940
+    },
+    {
+      "epoch": 138.59,
+      "learning_rate": 7.650666666666667e-05,
+      "loss": 1.4269,
+      "step": 22960
+    },
+    {
+      "epoch": 138.71,
+      "learning_rate": 7.657333333333335e-05,
+      "loss": 1.4475,
+      "step": 22980
+    },
+    {
+      "epoch": 138.83,
+      "learning_rate": 7.664e-05,
+      "loss": 1.4762,
+      "step": 23000
+    },
+    {
+      "epoch": 138.83,
+      "eval_accuracy": 0.001289624719700487,
+      "eval_loss": 9.533530235290527,
+      "eval_runtime": 18.0601,
+      "eval_samples_per_second": 112.956,
+      "eval_steps_per_second": 2.381,
+      "step": 23000
+    },
+    {
+      "epoch": 138.95,
+      "learning_rate": 7.670666666666668e-05,
+      "loss": 1.5101,
+      "step": 23020
+    },
+    {
+      "epoch": 139.07,
+      "learning_rate": 7.677333333333334e-05,
+      "loss": 1.3644,
+      "step": 23040
+    },
+    {
+      "epoch": 139.2,
+      "learning_rate": 7.684e-05,
+      "loss": 1.3088,
+      "step": 23060
+    },
+    {
+      "epoch": 139.32,
+      "learning_rate": 7.690666666666667e-05,
+      "loss": 1.3408,
+      "step": 23080
+    },
+    {
+      "epoch": 139.44,
+      "learning_rate": 7.697333333333334e-05,
+      "loss": 1.3495,
+      "step": 23100
+    },
+    {
+      "epoch": 139.56,
+      "learning_rate": 7.704000000000001e-05,
+      "loss": 1.3906,
+      "step": 23120
+    },
+    {
+      "epoch": 139.68,
+      "learning_rate": 7.710666666666667e-05,
+      "loss": 1.408,
+      "step": 23140
+    },
+    {
+      "epoch": 139.8,
+      "learning_rate": 7.717333333333334e-05,
+      "loss": 1.4335,
+      "step": 23160
+    },
+    {
+      "epoch": 139.92,
+      "learning_rate": 7.724e-05,
+      "loss": 1.4561,
+      "step": 23180
+    },
+    {
+      "epoch": 140.04,
+      "learning_rate": 7.730666666666668e-05,
+      "loss": 1.3944,
+      "step": 23200
+    },
+    {
+      "epoch": 140.16,
+      "learning_rate": 7.737333333333334e-05,
+      "loss": 1.2728,
+      "step": 23220
+    },
+    {
+      "epoch": 140.28,
+      "learning_rate": 7.744e-05,
+      "loss": 1.2974,
+      "step": 23240
+    },
+    {
+      "epoch": 140.4,
+      "learning_rate": 7.750666666666667e-05,
+      "loss": 1.321,
+      "step": 23260
+    },
+    {
+      "epoch": 140.52,
+      "learning_rate": 7.757333333333335e-05,
+      "loss": 1.3482,
+      "step": 23280
+    },
+    {
+      "epoch": 140.64,
+      "learning_rate": 7.764e-05,
+      "loss": 1.3878,
+      "step": 23300
+    },
+    {
+      "epoch": 140.76,
+      "learning_rate": 7.770666666666667e-05,
+      "loss": 1.3908,
+      "step": 23320
+    },
+    {
+      "epoch": 140.89,
+      "learning_rate": 7.777333333333334e-05,
+      "loss": 1.4063,
+      "step": 23340
+    },
+    {
+      "epoch": 141.01,
+      "learning_rate": 7.784e-05,
+      "loss": 1.4223,
+      "step": 23360
+    },
+    {
+      "epoch": 141.13,
+      "learning_rate": 7.790666666666668e-05,
+      "loss": 1.2325,
+      "step": 23380
+    },
+    {
+      "epoch": 141.25,
+      "learning_rate": 7.797333333333333e-05,
+      "loss": 1.2659,
+      "step": 23400
+    },
+    {
+      "epoch": 141.37,
+      "learning_rate": 7.804e-05,
+      "loss": 1.2973,
+      "step": 23420
+    },
+    {
+      "epoch": 141.49,
+      "learning_rate": 7.810666666666667e-05,
+      "loss": 1.3071,
+      "step": 23440
+    },
+    {
+      "epoch": 141.61,
+      "learning_rate": 7.817333333333333e-05,
+      "loss": 1.3359,
+      "step": 23460
+    },
+    {
+      "epoch": 141.73,
+      "learning_rate": 7.824e-05,
+      "loss": 1.3585,
+      "step": 23480
+    },
+    {
+      "epoch": 141.85,
+      "learning_rate": 7.830666666666667e-05,
+      "loss": 1.3678,
+      "step": 23500
+    },
+    {
+      "epoch": 141.97,
+      "learning_rate": 7.837333333333334e-05,
+      "loss": 1.379,
+      "step": 23520
+    },
+    {
+      "epoch": 142.09,
+      "learning_rate": 7.844e-05,
+      "loss": 1.2398,
+      "step": 23540
+    },
+    {
+      "epoch": 142.21,
+      "learning_rate": 7.850666666666668e-05,
+      "loss": 1.2069,
+      "step": 23560
+    },
+    {
+      "epoch": 142.33,
+      "learning_rate": 7.857333333333333e-05,
+      "loss": 1.2478,
+      "step": 23580
+    },
+    {
+      "epoch": 142.45,
+      "learning_rate": 7.864e-05,
+      "loss": 1.2768,
+      "step": 23600
+    },
+    {
+      "epoch": 142.58,
+      "learning_rate": 7.870666666666667e-05,
+      "loss": 1.2963,
+      "step": 23620
+    },
+    {
+      "epoch": 142.7,
+      "learning_rate": 7.876999999999999e-05,
+      "loss": 1.3087,
+      "step": 23640
+    },
+    {
+      "epoch": 142.82,
+      "learning_rate": 7.883666666666667e-05,
+      "loss": 1.3438,
+      "step": 23660
+    },
+    {
+      "epoch": 142.94,
+      "learning_rate": 7.890333333333333e-05,
+      "loss": 1.3481,
+      "step": 23680
+    },
+    {
+      "epoch": 143.06,
+      "learning_rate": 7.897e-05,
+      "loss": 1.2644,
+      "step": 23700
+    },
+    {
+      "epoch": 143.18,
+      "learning_rate": 7.903666666666668e-05,
+      "loss": 1.176,
+      "step": 23720
+    },
+    {
+      "epoch": 143.3,
+      "learning_rate": 7.910333333333334e-05,
+      "loss": 1.206,
+      "step": 23740
+    },
+    {
+      "epoch": 143.42,
+      "learning_rate": 7.917e-05,
+      "loss": 1.23,
+      "step": 23760
+    },
+    {
+      "epoch": 143.54,
+      "learning_rate": 7.923666666666667e-05,
+      "loss": 1.2664,
+      "step": 23780
+    },
+    {
+      "epoch": 143.66,
+      "learning_rate": 7.930333333333334e-05,
+      "loss": 1.2763,
+      "step": 23800
+    },
+    {
+      "epoch": 143.78,
+      "learning_rate": 7.937e-05,
+      "loss": 1.3145,
+      "step": 23820
+    },
+    {
+      "epoch": 143.9,
+      "learning_rate": 7.943666666666667e-05,
+      "loss": 1.3303,
+      "step": 23840
+    },
+    {
+      "epoch": 144.02,
+      "learning_rate": 7.950333333333334e-05,
+      "loss": 1.305,
+      "step": 23860
+    },
+    {
+      "epoch": 144.14,
+      "learning_rate": 7.957e-05,
+      "loss": 1.152,
+      "step": 23880
+    },
+    {
+      "epoch": 144.27,
+      "learning_rate": 7.963666666666668e-05,
+      "loss": 1.1715,
+      "step": 23900
+    },
+    {
+      "epoch": 144.39,
+      "learning_rate": 7.970333333333334e-05,
+      "loss": 1.2116,
+      "step": 23920
+    },
+    {
+      "epoch": 144.51,
+      "learning_rate": 7.977e-05,
+      "loss": 1.2291,
+      "step": 23940
+    },
+    {
+      "epoch": 144.63,
+      "learning_rate": 7.983666666666667e-05,
+      "loss": 1.2752,
+      "step": 23960
+    },
+    {
+      "epoch": 144.75,
+      "learning_rate": 7.990333333333334e-05,
+      "loss": 1.267,
+      "step": 23980
+    },
+    {
+      "epoch": 144.87,
+      "learning_rate": 7.997e-05,
+      "loss": 1.2833,
+      "step": 24000
+    },
+    {
+      "epoch": 144.87,
+      "eval_accuracy": 0.001314320712034851,
+      "eval_loss": 10.197283744812012,
+      "eval_runtime": 18.1169,
+      "eval_samples_per_second": 112.602,
+      "eval_steps_per_second": 2.373,
+      "step": 24000
+    },
+    {
+      "epoch": 144.99,
+      "learning_rate": 8.003666666666667e-05,
+      "loss": 1.319,
+      "step": 24020
+    },
+    {
+      "epoch": 145.11,
+      "learning_rate": 8.010333333333334e-05,
+      "loss": 1.1368,
+      "step": 24040
+    },
+    {
+      "epoch": 145.23,
+      "learning_rate": 8.017e-05,
+      "loss": 1.1447,
+      "step": 24060
+    },
+    {
+      "epoch": 145.35,
+      "learning_rate": 8.023666666666668e-05,
+      "loss": 1.1814,
+      "step": 24080
+    },
+    {
+      "epoch": 145.47,
+      "learning_rate": 8.030333333333334e-05,
+      "loss": 1.2116,
+      "step": 24100
+    },
+    {
+      "epoch": 145.59,
+      "learning_rate": 8.037e-05,
+      "loss": 1.221,
+      "step": 24120
+    },
+    {
+      "epoch": 145.71,
+      "learning_rate": 8.043666666666667e-05,
+      "loss": 1.2411,
+      "step": 24140
+    },
+    {
+      "epoch": 145.84,
+      "learning_rate": 8.050333333333335e-05,
+      "loss": 1.2624,
+      "step": 24160
+    },
+    {
+      "epoch": 145.96,
+      "learning_rate": 8.057e-05,
+      "loss": 1.2827,
+      "step": 24180
+    },
+    {
+      "epoch": 146.08,
+      "learning_rate": 8.063666666666667e-05,
+      "loss": 1.1684,
+      "step": 24200
+    },
+    {
+      "epoch": 146.2,
+      "learning_rate": 8.070333333333334e-05,
+      "loss": 1.1256,
+      "step": 24220
+    },
+    {
+      "epoch": 146.32,
+      "learning_rate": 8.077e-05,
+      "loss": 1.1634,
+      "step": 24240
+    },
+    {
+      "epoch": 146.44,
+      "learning_rate": 8.083666666666668e-05,
+      "loss": 1.1625,
+      "step": 24260
+    },
+    {
+      "epoch": 146.56,
+      "learning_rate": 8.090333333333334e-05,
+      "loss": 1.1606,
+      "step": 24280
+    },
+    {
+      "epoch": 146.68,
+      "learning_rate": 8.097e-05,
+      "loss": 1.1903,
+      "step": 24300
+    },
+    {
+      "epoch": 146.8,
+      "learning_rate": 8.103666666666667e-05,
+      "loss": 1.218,
+      "step": 24320
+    },
+    {
+      "epoch": 146.92,
+      "learning_rate": 8.110333333333335e-05,
+      "loss": 1.2402,
+      "step": 24340
+    },
+    {
+      "epoch": 147.04,
+      "learning_rate": 8.117e-05,
+      "loss": 1.1858,
+      "step": 24360
+    },
+    {
+      "epoch": 147.16,
+      "learning_rate": 8.123666666666667e-05,
+      "loss": 1.0605,
+      "step": 24380
+    },
+    {
+      "epoch": 147.28,
+      "learning_rate": 8.130333333333334e-05,
+      "loss": 1.1045,
+      "step": 24400
+    },
+    {
+      "epoch": 147.4,
+      "learning_rate": 8.137e-05,
+      "loss": 1.1225,
+      "step": 24420
+    },
+    {
+      "epoch": 147.53,
+      "learning_rate": 8.143666666666667e-05,
+      "loss": 1.1464,
+      "step": 24440
+    },
+    {
+      "epoch": 147.65,
+      "learning_rate": 8.150333333333334e-05,
+      "loss": 1.1694,
+      "step": 24460
+    },
+    {
+      "epoch": 147.77,
+      "learning_rate": 8.157e-05,
+      "loss": 1.1967,
+      "step": 24480
+    },
+    {
+      "epoch": 147.89,
+      "learning_rate": 8.163666666666667e-05,
+      "loss": 1.2042,
+      "step": 24500
+    },
+    {
+      "epoch": 148.01,
+      "learning_rate": 8.170333333333333e-05,
+      "loss": 1.2241,
+      "step": 24520
+    },
+    {
+      "epoch": 148.13,
+      "learning_rate": 8.177e-05,
+      "loss": 1.0313,
+      "step": 24540
+    },
+    {
+      "epoch": 148.25,
+      "learning_rate": 8.183666666666668e-05,
+      "loss": 1.0624,
+      "step": 24560
+    },
+    {
+      "epoch": 148.37,
+      "learning_rate": 8.190333333333333e-05,
+      "loss": 1.0905,
+      "step": 24580
+    },
+    {
+      "epoch": 148.49,
+      "learning_rate": 8.197e-05,
+      "loss": 1.1182,
+      "step": 24600
+    },
+    {
+      "epoch": 148.61,
+      "learning_rate": 8.203666666666667e-05,
+      "loss": 1.1554,
+      "step": 24620
+    },
+    {
+      "epoch": 148.73,
+      "learning_rate": 8.210333333333333e-05,
+      "loss": 1.1825,
+      "step": 24640
+    },
+    {
+      "epoch": 148.85,
+      "learning_rate": 8.217000000000001e-05,
+      "loss": 1.1987,
+      "step": 24660
+    },
+    {
+      "epoch": 148.97,
+      "learning_rate": 8.223666666666667e-05,
+      "loss": 1.2207,
+      "step": 24680
+    },
+    {
+      "epoch": 149.09,
+      "learning_rate": 8.230333333333334e-05,
+      "loss": 1.0685,
+      "step": 24700
+    },
+    {
+      "epoch": 149.22,
+      "learning_rate": 8.237e-05,
+      "loss": 1.0407,
+      "step": 24720
+    },
+    {
+      "epoch": 149.34,
+      "learning_rate": 8.243666666666668e-05,
+      "loss": 1.0582,
+      "step": 24740
+    },
+    {
+      "epoch": 149.46,
+      "learning_rate": 8.250333333333333e-05,
+      "loss": 1.0789,
+      "step": 24760
+    },
+    {
+      "epoch": 149.58,
+      "learning_rate": 8.257e-05,
+      "loss": 1.1121,
+      "step": 24780
+    },
+    {
+      "epoch": 149.7,
+      "learning_rate": 8.263666666666667e-05,
+      "loss": 1.1414,
+      "step": 24800
+    },
+    {
+      "epoch": 149.82,
+      "learning_rate": 8.270333333333333e-05,
+      "loss": 1.1653,
+      "step": 24820
+    },
+    {
+      "epoch": 149.94,
+      "learning_rate": 8.277000000000001e-05,
+      "loss": 1.1782,
+      "step": 24840
+    },
+    {
+      "epoch": 150.06,
+      "learning_rate": 8.283666666666667e-05,
+      "loss": 1.0871,
+      "step": 24860
+    },
+    {
+      "epoch": 150.18,
+      "learning_rate": 8.290333333333334e-05,
+      "loss": 1.0067,
+      "step": 24880
+    },
+    {
+      "epoch": 150.3,
+      "learning_rate": 8.297e-05,
+      "loss": 1.0524,
+      "step": 24900
+    },
+    {
+      "epoch": 150.42,
+      "learning_rate": 8.303666666666668e-05,
+      "loss": 1.0559,
+      "step": 24920
+    },
+    {
+      "epoch": 150.54,
+      "learning_rate": 8.310333333333333e-05,
+      "loss": 1.0838,
+      "step": 24940
+    },
+    {
+      "epoch": 150.66,
+      "learning_rate": 8.317e-05,
+      "loss": 1.1119,
+      "step": 24960
+    },
+    {
+      "epoch": 150.78,
+      "learning_rate": 8.323666666666667e-05,
+      "loss": 1.1226,
+      "step": 24980
+    },
+    {
+      "epoch": 150.91,
+      "learning_rate": 8.330333333333333e-05,
+      "loss": 1.1451,
+      "step": 25000
+    },
+    {
+      "epoch": 150.91,
+      "eval_accuracy": 0.0012935760784739852,
+      "eval_loss": 10.821310043334961,
+      "eval_runtime": 18.1019,
+      "eval_samples_per_second": 112.695,
+      "eval_steps_per_second": 2.375,
+      "step": 25000
+    },
+    {
+      "epoch": 151.03,
+      "learning_rate": 8.337000000000001e-05,
+      "loss": 1.12,
+      "step": 25020
+    },
+    {
+      "epoch": 151.15,
+      "learning_rate": 8.343666666666667e-05,
+      "loss": 0.9728,
+      "step": 25040
+    },
+    {
+      "epoch": 151.27,
+      "learning_rate": 8.350333333333334e-05,
+      "loss": 1.0058,
+      "step": 25060
+    },
+    {
+      "epoch": 151.39,
+      "learning_rate": 8.357e-05,
+      "loss": 1.0314,
+      "step": 25080
+    },
+    {
+      "epoch": 151.51,
+      "learning_rate": 8.363666666666668e-05,
+      "loss": 1.0595,
+      "step": 25100
+    },
+    {
+      "epoch": 151.63,
+      "learning_rate": 8.370333333333333e-05,
+      "loss": 1.0719,
+      "step": 25120
+    },
+    {
+      "epoch": 151.75,
+      "learning_rate": 8.377e-05,
+      "loss": 1.1014,
+      "step": 25140
+    },
+    {
+      "epoch": 151.87,
+      "learning_rate": 8.383666666666667e-05,
+      "loss": 1.1169,
+      "step": 25160
+    },
+    {
+      "epoch": 151.99,
+      "learning_rate": 8.390333333333333e-05,
+      "loss": 1.1255,
+      "step": 25180
+    },
+    {
+      "epoch": 152.11,
+      "learning_rate": 8.397000000000001e-05,
+      "loss": 0.9556,
+      "step": 25200
+    },
+    {
+      "epoch": 152.23,
+      "learning_rate": 8.403666666666667e-05,
+      "loss": 0.9681,
+      "step": 25220
+    },
+    {
+      "epoch": 152.35,
+      "learning_rate": 8.410333333333334e-05,
+      "loss": 1.0043,
+      "step": 25240
+    },
+    {
+      "epoch": 152.47,
+      "learning_rate": 8.417e-05,
+      "loss": 1.0042,
+      "step": 25260
+    },
+    {
+      "epoch": 152.6,
+      "learning_rate": 8.423666666666668e-05,
+      "loss": 1.0228,
+      "step": 25280
+    },
+    {
+      "epoch": 152.72,
+      "learning_rate": 8.430333333333333e-05,
+      "loss": 1.0566,
+      "step": 25300
+    },
+    {
+      "epoch": 152.84,
+      "learning_rate": 8.437000000000001e-05,
+      "loss": 1.0815,
+      "step": 25320
+    },
+    {
+      "epoch": 152.96,
+      "learning_rate": 8.443666666666667e-05,
+      "loss": 1.1157,
+      "step": 25340
+    },
+    {
+      "epoch": 153.08,
+      "learning_rate": 8.450333333333333e-05,
+      "loss": 0.9935,
+      "step": 25360
+    },
+    {
+      "epoch": 153.2,
+      "learning_rate": 8.457e-05,
+      "loss": 0.9415,
+      "step": 25380
+    },
+    {
+      "epoch": 153.32,
+      "learning_rate": 8.463666666666668e-05,
+      "loss": 0.9601,
+      "step": 25400
+    },
+    {
+      "epoch": 153.44,
+      "learning_rate": 8.470333333333334e-05,
+      "loss": 0.9884,
+      "step": 25420
+    },
+    {
+      "epoch": 153.56,
+      "learning_rate": 8.477e-05,
+      "loss": 1.0064,
+      "step": 25440
+    },
+    {
+      "epoch": 153.68,
+      "learning_rate": 8.483666666666668e-05,
+      "loss": 1.0422,
+      "step": 25460
+    },
+    {
+      "epoch": 153.8,
+      "learning_rate": 8.490333333333333e-05,
+      "loss": 1.0665,
+      "step": 25480
+    },
+    {
+      "epoch": 153.92,
+      "learning_rate": 8.497000000000001e-05,
+      "loss": 1.0851,
+      "step": 25500
+    },
+    {
+      "epoch": 154.04,
+      "learning_rate": 8.503666666666667e-05,
+      "loss": 1.0234,
+      "step": 25520
+    },
+    {
+      "epoch": 154.16,
+      "learning_rate": 8.510333333333334e-05,
+      "loss": 0.9137,
+      "step": 25540
+    },
+    {
+      "epoch": 154.29,
+      "learning_rate": 8.517e-05,
+      "loss": 0.9448,
+      "step": 25560
+    },
+    {
+      "epoch": 154.41,
+      "learning_rate": 8.523666666666668e-05,
+      "loss": 0.9683,
+      "step": 25580
+    },
+    {
+      "epoch": 154.53,
+      "learning_rate": 8.530333333333334e-05,
+      "loss": 1.0072,
+      "step": 25600
+    },
+    {
+      "epoch": 154.65,
+      "learning_rate": 8.537e-05,
+      "loss": 0.9969,
+      "step": 25620
+    },
+    {
+      "epoch": 154.77,
+      "learning_rate": 8.543666666666668e-05,
+      "loss": 1.022,
+      "step": 25640
+    },
+    {
+      "epoch": 154.89,
+      "learning_rate": 8.55e-05,
+      "loss": 1.0452,
+      "step": 25660
+    },
+    {
+      "epoch": 155.01,
+      "learning_rate": 8.556666666666667e-05,
+      "loss": 1.0589,
+      "step": 25680
+    },
+    {
+      "epoch": 155.13,
+      "learning_rate": 8.563333333333333e-05,
+      "loss": 0.8943,
+      "step": 25700
+    },
+    {
+      "epoch": 155.25,
+      "learning_rate": 8.57e-05,
+      "loss": 0.9057,
+      "step": 25720
+    },
+    {
+      "epoch": 155.37,
+      "learning_rate": 8.576666666666667e-05,
+      "loss": 0.943,
+      "step": 25740
+    },
+    {
+      "epoch": 155.49,
+      "learning_rate": 8.583333333333334e-05,
+      "loss": 0.9658,
+      "step": 25760
+    },
+    {
+      "epoch": 155.61,
+      "learning_rate": 8.59e-05,
+      "loss": 0.9853,
+      "step": 25780
+    },
+    {
+      "epoch": 155.73,
+      "learning_rate": 8.596666666666668e-05,
+      "loss": 1.0105,
+      "step": 25800
+    },
+    {
+      "epoch": 155.86,
+      "learning_rate": 8.603333333333333e-05,
+      "loss": 1.0318,
+      "step": 25820
+    },
+    {
+      "epoch": 155.98,
+      "learning_rate": 8.61e-05,
+      "loss": 1.0523,
+      "step": 25840
+    },
+    {
+      "epoch": 156.1,
+      "learning_rate": 8.616666666666667e-05,
+      "loss": 0.8952,
+      "step": 25860
+    },
+    {
+      "epoch": 156.22,
+      "learning_rate": 8.623333333333333e-05,
+      "loss": 0.899,
+      "step": 25880
+    },
+    {
+      "epoch": 156.34,
+      "learning_rate": 8.63e-05,
+      "loss": 0.9027,
+      "step": 25900
+    },
+    {
+      "epoch": 156.46,
+      "learning_rate": 8.636666666666667e-05,
+      "loss": 0.9333,
+      "step": 25920
+    },
+    {
+      "epoch": 156.58,
+      "learning_rate": 8.643333333333334e-05,
+      "loss": 0.962,
+      "step": 25940
+    },
+    {
+      "epoch": 156.7,
+      "learning_rate": 8.65e-05,
+      "loss": 0.9726,
+      "step": 25960
+    },
+    {
+      "epoch": 156.82,
+      "learning_rate": 8.656666666666668e-05,
+      "loss": 0.9879,
+      "step": 25980
+    },
+    {
+      "epoch": 156.94,
+      "learning_rate": 8.663333333333333e-05,
+      "loss": 1.0251,
+      "step": 26000
+    },
+    {
+      "epoch": 156.94,
+      "eval_accuracy": 0.0012832037616935523,
+      "eval_loss": 11.440205574035645,
+      "eval_runtime": 18.1944,
+      "eval_samples_per_second": 112.123,
+      "eval_steps_per_second": 2.363,
+      "step": 26000
+    },
+    {
+      "epoch": 157.06,
+      "learning_rate": 8.67e-05,
+      "loss": 0.9266,
+      "step": 26020
+    },
+    {
+      "epoch": 157.18,
+      "learning_rate": 8.676666666666667e-05,
+      "loss": 0.8623,
+      "step": 26040
+    },
+    {
+      "epoch": 157.3,
+      "learning_rate": 8.683333333333333e-05,
+      "loss": 0.887,
+      "step": 26060
+    },
+    {
+      "epoch": 157.42,
+      "learning_rate": 8.69e-05,
+      "loss": 0.9145,
+      "step": 26080
+    },
+    {
+      "epoch": 157.55,
+      "learning_rate": 8.696666666666668e-05,
+      "loss": 0.9028,
+      "step": 26100
+    },
+    {
+      "epoch": 157.67,
+      "learning_rate": 8.703333333333334e-05,
+      "loss": 0.9591,
+      "step": 26120
+    },
+    {
+      "epoch": 157.79,
+      "learning_rate": 8.71e-05,
+      "loss": 0.98,
+      "step": 26140
+    },
+    {
+      "epoch": 157.91,
+      "learning_rate": 8.716666666666668e-05,
+      "loss": 1.0075,
+      "step": 26160
+    },
+    {
+      "epoch": 158.03,
+      "learning_rate": 8.723333333333333e-05,
+      "loss": 0.9595,
+      "step": 26180
+    },
+    {
+      "epoch": 158.15,
+      "learning_rate": 8.730000000000001e-05,
+      "loss": 0.8239,
+      "step": 26200
+    },
+    {
+      "epoch": 158.27,
+      "learning_rate": 8.736666666666667e-05,
+      "loss": 0.8542,
+      "step": 26220
+    },
+    {
+      "epoch": 158.39,
+      "learning_rate": 8.743333333333334e-05,
+      "loss": 0.8868,
+      "step": 26240
+    },
+    {
+      "epoch": 158.51,
+      "learning_rate": 8.75e-05,
+      "loss": 0.8996,
+      "step": 26260
+    },
+    {
+      "epoch": 158.63,
+      "learning_rate": 8.756666666666668e-05,
+      "loss": 0.9325,
+      "step": 26280
+    },
+    {
+      "epoch": 158.75,
+      "learning_rate": 8.763333333333334e-05,
+      "loss": 0.9553,
+      "step": 26300
+    },
+    {
+      "epoch": 158.87,
+      "learning_rate": 8.77e-05,
+      "loss": 0.9839,
+      "step": 26320
+    },
+    {
+      "epoch": 158.99,
+      "learning_rate": 8.776666666666668e-05,
+      "loss": 1.0011,
+      "step": 26340
+    },
+    {
+      "epoch": 159.11,
+      "learning_rate": 8.783333333333333e-05,
+      "loss": 0.8096,
+      "step": 26360
+    },
+    {
+      "epoch": 159.24,
+      "learning_rate": 8.790000000000001e-05,
+      "loss": 0.826,
+      "step": 26380
+    },
+    {
+      "epoch": 159.36,
+      "learning_rate": 8.796666666666667e-05,
+      "loss": 0.8569,
+      "step": 26400
+    },
+    {
+      "epoch": 159.48,
+      "learning_rate": 8.803333333333334e-05,
+      "loss": 0.8801,
+      "step": 26420
+    },
+    {
+      "epoch": 159.6,
+      "learning_rate": 8.81e-05,
+      "loss": 0.8991,
+      "step": 26440
+    },
+    {
+      "epoch": 159.72,
+      "learning_rate": 8.816666666666668e-05,
+      "loss": 0.9162,
+      "step": 26460
+    },
+    {
+      "epoch": 159.84,
+      "learning_rate": 8.823333333333334e-05,
+      "loss": 0.9429,
+      "step": 26480
+    },
+    {
+      "epoch": 159.96,
+      "learning_rate": 8.83e-05,
+      "loss": 0.9653,
+      "step": 26500
+    },
+    {
+      "epoch": 160.08,
+      "learning_rate": 8.836666666666667e-05,
+      "loss": 0.8525,
+      "step": 26520
+    },
+    {
+      "epoch": 160.2,
+      "learning_rate": 8.843333333333333e-05,
+      "loss": 0.8147,
+      "step": 26540
+    },
+    {
+      "epoch": 160.32,
+      "learning_rate": 8.850000000000001e-05,
+      "loss": 0.829,
+      "step": 26560
+    },
+    {
+      "epoch": 160.44,
+      "learning_rate": 8.856666666666667e-05,
+      "loss": 0.8521,
+      "step": 26580
+    },
+    {
+      "epoch": 160.56,
+      "learning_rate": 8.863333333333334e-05,
+      "loss": 0.8666,
+      "step": 26600
+    },
+    {
+      "epoch": 160.68,
+      "learning_rate": 8.87e-05,
+      "loss": 0.9125,
+      "step": 26620
+    },
+    {
+      "epoch": 160.8,
+      "learning_rate": 8.876666666666668e-05,
+      "loss": 0.9217,
+      "step": 26640
+    },
+    {
+      "epoch": 160.93,
+      "learning_rate": 8.883333333333333e-05,
+      "loss": 0.9398,
+      "step": 26660
+    },
+    {
+      "epoch": 161.05,
+      "learning_rate": 8.89e-05,
+      "loss": 0.8752,
+      "step": 26680
+    },
+    {
+      "epoch": 161.17,
+      "learning_rate": 8.896666666666667e-05,
+      "loss": 0.7788,
+      "step": 26700
+    },
+    {
+      "epoch": 161.29,
+      "learning_rate": 8.903333333333333e-05,
+      "loss": 0.8097,
+      "step": 26720
+    },
+    {
+      "epoch": 161.41,
+      "learning_rate": 8.910000000000001e-05,
+      "loss": 0.8351,
+      "step": 26740
+    },
+    {
+      "epoch": 161.53,
+      "learning_rate": 8.916666666666667e-05,
+      "loss": 0.8531,
+      "step": 26760
+    },
+    {
+      "epoch": 161.65,
+      "learning_rate": 8.923333333333334e-05,
+      "loss": 0.8707,
+      "step": 26780
+    },
+    {
+      "epoch": 161.77,
+      "learning_rate": 8.93e-05,
+      "loss": 0.8952,
+      "step": 26800
+    },
+    {
+      "epoch": 161.89,
+      "learning_rate": 8.936666666666668e-05,
+      "loss": 0.9155,
+      "step": 26820
+    },
+    {
+      "epoch": 162.01,
+      "learning_rate": 8.943333333333333e-05,
+      "loss": 0.9079,
+      "step": 26840
+    },
+    {
+      "epoch": 162.13,
+      "learning_rate": 8.950000000000001e-05,
+      "loss": 0.7534,
+      "step": 26860
+    },
+    {
+      "epoch": 162.25,
+      "learning_rate": 8.956666666666667e-05,
+      "loss": 0.7953,
+      "step": 26880
+    },
+    {
+      "epoch": 162.37,
+      "learning_rate": 8.963333333333333e-05,
+      "loss": 0.8226,
+      "step": 26900
+    },
+    {
+      "epoch": 162.49,
+      "learning_rate": 8.970000000000001e-05,
+      "loss": 0.8424,
+      "step": 26920
+    },
+    {
+      "epoch": 162.62,
+      "learning_rate": 8.976666666666666e-05,
+      "loss": 0.8411,
+      "step": 26940
+    },
+    {
+      "epoch": 162.74,
+      "learning_rate": 8.983333333333334e-05,
+      "loss": 0.877,
+      "step": 26960
+    },
+    {
+      "epoch": 162.86,
+      "learning_rate": 8.99e-05,
+      "loss": 0.8957,
+      "step": 26980
+    },
+    {
+      "epoch": 162.98,
+      "learning_rate": 8.996666666666667e-05,
+      "loss": 0.9164,
+      "step": 27000
+    },
+    {
+      "epoch": 162.98,
+      "eval_accuracy": 0.0012822159220001778,
+      "eval_loss": 11.99949836730957,
+      "eval_runtime": 18.2112,
+      "eval_samples_per_second": 112.019,
+      "eval_steps_per_second": 2.361,
+      "step": 27000
+    },
+    {
+      "epoch": 163.1,
+      "learning_rate": 9.003333333333333e-05,
+      "loss": 0.7606,
+      "step": 27020
+    },
+    {
+      "epoch": 163.22,
+      "learning_rate": 9.010000000000001e-05,
+      "loss": 0.7527,
+      "step": 27040
+    },
+    {
+      "epoch": 163.34,
+      "learning_rate": 9.016666666666667e-05,
+      "loss": 0.7913,
+      "step": 27060
+    },
+    {
+      "epoch": 163.46,
+      "learning_rate": 9.023333333333334e-05,
+      "loss": 0.8033,
+      "step": 27080
+    },
+    {
+      "epoch": 163.58,
+      "learning_rate": 9.030000000000001e-05,
+      "loss": 0.8187,
+      "step": 27100
+    },
+    {
+      "epoch": 163.7,
+      "learning_rate": 9.036666666666666e-05,
+      "loss": 0.8593,
+      "step": 27120
+    },
+    {
+      "epoch": 163.82,
+      "learning_rate": 9.043333333333334e-05,
+      "loss": 0.8624,
+      "step": 27140
+    },
+    {
+      "epoch": 163.94,
+      "learning_rate": 9.05e-05,
+      "loss": 0.884,
+      "step": 27160
+    },
+    {
+      "epoch": 164.06,
+      "learning_rate": 9.056666666666667e-05,
+      "loss": 0.8033,
+      "step": 27180
+    },
+    {
+      "epoch": 164.19,
+      "learning_rate": 9.063333333333333e-05,
+      "loss": 0.7372,
+      "step": 27200
+    },
+    {
+      "epoch": 164.31,
+      "learning_rate": 9.070000000000001e-05,
+      "loss": 0.7514,
+      "step": 27220
+    },
+    {
+      "epoch": 164.43,
+      "learning_rate": 9.076666666666667e-05,
+      "loss": 0.7856,
+      "step": 27240
+    },
+    {
+      "epoch": 164.55,
+      "learning_rate": 9.083333333333334e-05,
+      "loss": 0.8032,
+      "step": 27260
+    },
+    {
+      "epoch": 164.67,
+      "learning_rate": 9.090000000000001e-05,
+      "loss": 0.8261,
+      "step": 27280
+    },
+    {
+      "epoch": 164.79,
+      "learning_rate": 9.096666666666666e-05,
+      "loss": 0.8518,
+      "step": 27300
+    },
+    {
+      "epoch": 164.91,
+      "learning_rate": 9.103333333333334e-05,
+      "loss": 0.8898,
+      "step": 27320
+    },
+    {
+      "epoch": 165.03,
+      "learning_rate": 9.11e-05,
+      "loss": 0.8442,
+      "step": 27340
+    },
+    {
+      "epoch": 165.15,
+      "learning_rate": 9.116666666666667e-05,
+      "loss": 0.7088,
+      "step": 27360
+    },
+    {
+      "epoch": 165.27,
+      "learning_rate": 9.123333333333333e-05,
+      "loss": 0.7288,
+      "step": 27380
+    },
+    {
+      "epoch": 165.39,
+      "learning_rate": 9.130000000000001e-05,
+      "loss": 0.7667,
+      "step": 27400
+    },
+    {
+      "epoch": 165.51,
+      "learning_rate": 9.136666666666666e-05,
+      "loss": 0.7884,
+      "step": 27420
+    },
+    {
+      "epoch": 165.63,
+      "learning_rate": 9.143333333333334e-05,
+      "loss": 0.815,
+      "step": 27440
+    },
+    {
+      "epoch": 165.75,
+      "learning_rate": 9.15e-05,
+      "loss": 0.8417,
+      "step": 27460
+    },
+    {
+      "epoch": 165.88,
+      "learning_rate": 9.156666666666667e-05,
+      "loss": 0.8541,
+      "step": 27480
+    },
+    {
+      "epoch": 166.0,
+      "learning_rate": 9.163333333333334e-05,
+      "loss": 0.8833,
+      "step": 27500
+    },
+    {
+      "epoch": 166.12,
+      "learning_rate": 9.17e-05,
+      "loss": 0.7026,
+      "step": 27520
+    },
+    {
+      "epoch": 166.24,
+      "learning_rate": 9.176666666666667e-05,
+      "loss": 0.7185,
+      "step": 27540
+    },
+    {
+      "epoch": 166.36,
+      "learning_rate": 9.183333333333333e-05,
+      "loss": 0.7563,
+      "step": 27560
+    },
+    {
+      "epoch": 166.48,
+      "learning_rate": 9.190000000000001e-05,
+      "loss": 0.7845,
+      "step": 27580
+    },
+    {
+      "epoch": 166.6,
+      "learning_rate": 9.196666666666666e-05,
+      "loss": 0.7935,
+      "step": 27600
+    },
+    {
+      "epoch": 166.72,
+      "learning_rate": 9.203333333333334e-05,
+      "loss": 0.8099,
+      "step": 27620
+    },
+    {
+      "epoch": 166.84,
+      "learning_rate": 9.21e-05,
+      "loss": 0.8281,
+      "step": 27640
+    },
+    {
+      "epoch": 166.96,
+      "learning_rate": 9.216666666666667e-05,
+      "loss": 0.8446,
+      "step": 27660
+    },
+    {
+      "epoch": 167.08,
+      "learning_rate": 9.223e-05,
+      "loss": 0.7318,
+      "step": 27680
+    },
+    {
+      "epoch": 167.2,
+      "learning_rate": 9.229666666666668e-05,
+      "loss": 0.7045,
+      "step": 27700
+    },
+    {
+      "epoch": 167.32,
+      "learning_rate": 9.236333333333333e-05,
+      "loss": 0.7265,
+      "step": 27720
+    },
+    {
+      "epoch": 167.44,
+      "learning_rate": 9.243000000000001e-05,
+      "loss": 0.7518,
+      "step": 27740
+    },
+    {
+      "epoch": 167.57,
+      "learning_rate": 9.249666666666667e-05,
+      "loss": 0.7368,
+      "step": 27760
+    },
+    {
+      "epoch": 167.69,
+      "learning_rate": 9.256333333333334e-05,
+      "loss": 0.774,
+      "step": 27780
+    },
+    {
+      "epoch": 167.81,
+      "learning_rate": 9.263e-05,
+      "loss": 0.8074,
+      "step": 27800
+    },
+    {
+      "epoch": 167.93,
+      "learning_rate": 9.269666666666668e-05,
+      "loss": 0.8461,
+      "step": 27820
+    },
+    {
+      "epoch": 168.05,
+      "learning_rate": 9.276333333333334e-05,
+      "loss": 0.7798,
+      "step": 27840
+    },
+    {
+      "epoch": 168.17,
+      "learning_rate": 9.283e-05,
+      "loss": 0.6831,
+      "step": 27860
+    },
+    {
+      "epoch": 168.29,
+      "learning_rate": 9.289666666666668e-05,
+      "loss": 0.7007,
+      "step": 27880
+    },
+    {
+      "epoch": 168.41,
+      "learning_rate": 9.296333333333333e-05,
+      "loss": 0.7333,
+      "step": 27900
+    },
+    {
+      "epoch": 168.53,
+      "learning_rate": 9.303000000000001e-05,
+      "loss": 0.7524,
+      "step": 27920
+    },
+    {
+      "epoch": 168.65,
+      "learning_rate": 9.309666666666667e-05,
+      "loss": 0.779,
+      "step": 27940
+    },
+    {
+      "epoch": 168.77,
+      "learning_rate": 9.316333333333334e-05,
+      "loss": 0.7919,
+      "step": 27960
+    },
+    {
+      "epoch": 168.89,
+      "learning_rate": 9.323e-05,
+      "loss": 0.8171,
+      "step": 27980
+    },
+    {
+      "epoch": 169.01,
+      "learning_rate": 9.329666666666668e-05,
+      "loss": 0.8174,
+      "step": 28000
+    },
+    {
+      "epoch": 169.01,
+      "eval_accuracy": 0.0012678922464462467,
+      "eval_loss": 12.56796646118164,
+      "eval_runtime": 18.1166,
+      "eval_samples_per_second": 112.604,
+      "eval_steps_per_second": 2.374,
+      "step": 28000
+    },
+    {
+      "epoch": 169.13,
+      "learning_rate": 9.336333333333334e-05,
+      "loss": 0.6442,
+      "step": 28020
+    },
+    {
+      "epoch": 169.26,
+      "learning_rate": 9.343e-05,
+      "loss": 0.6902,
+      "step": 28040
+    },
+    {
+      "epoch": 169.38,
+      "learning_rate": 9.349666666666667e-05,
+      "loss": 0.7125,
+      "step": 28060
+    },
+    {
+      "epoch": 169.5,
+      "learning_rate": 9.356333333333333e-05,
+      "loss": 0.735,
+      "step": 28080
+    },
+    {
+      "epoch": 169.62,
+      "learning_rate": 9.363000000000001e-05,
+      "loss": 0.7575,
+      "step": 28100
+    },
+    {
+      "epoch": 169.74,
+      "learning_rate": 9.369666666666666e-05,
+      "loss": 0.7694,
+      "step": 28120
+    },
+    {
+      "epoch": 169.86,
+      "learning_rate": 9.376333333333334e-05,
+      "loss": 0.799,
+      "step": 28140
+    },
+    {
+      "epoch": 169.98,
+      "learning_rate": 9.383e-05,
+      "loss": 0.8383,
+      "step": 28160
+    },
+    {
+      "epoch": 170.1,
+      "learning_rate": 9.389666666666666e-05,
+      "loss": 0.7015,
+      "step": 28180
+    },
+    {
+      "epoch": 170.22,
+      "learning_rate": 9.396333333333334e-05,
+      "loss": 0.6759,
+      "step": 28200
+    },
+    {
+      "epoch": 170.34,
+      "learning_rate": 9.403e-05,
+      "loss": 0.6986,
+      "step": 28220
+    },
+    {
+      "epoch": 170.46,
+      "learning_rate": 9.409666666666667e-05,
+      "loss": 0.7233,
+      "step": 28240
+    },
+    {
+      "epoch": 170.58,
+      "learning_rate": 9.416333333333333e-05,
+      "loss": 0.747,
+      "step": 28260
+    },
+    {
+      "epoch": 170.7,
+      "learning_rate": 9.423000000000001e-05,
+      "loss": 0.7691,
+      "step": 28280
+    },
+    {
+      "epoch": 170.82,
+      "learning_rate": 9.429666666666666e-05,
+      "loss": 0.7904,
+      "step": 28300
+    },
+    {
+      "epoch": 170.95,
+      "learning_rate": 9.436333333333334e-05,
+      "loss": 0.8325,
+      "step": 28320
+    },
+    {
+      "epoch": 171.07,
+      "learning_rate": 9.443e-05,
+      "loss": 0.7161,
+      "step": 28340
+    },
+    {
+      "epoch": 171.19,
+      "learning_rate": 9.449666666666667e-05,
+      "loss": 0.6488,
+      "step": 28360
+    },
+    {
+      "epoch": 171.31,
+      "learning_rate": 9.456333333333334e-05,
+      "loss": 0.6754,
+      "step": 28380
+    },
+    {
+      "epoch": 171.43,
+      "learning_rate": 9.463000000000001e-05,
+      "loss": 0.6955,
+      "step": 28400
+    },
+    {
+      "epoch": 171.55,
+      "learning_rate": 9.469666666666667e-05,
+      "loss": 0.7172,
+      "step": 28420
+    },
+    {
+      "epoch": 171.67,
+      "learning_rate": 9.476333333333333e-05,
+      "loss": 0.7461,
+      "step": 28440
+    },
+    {
+      "epoch": 171.79,
+      "learning_rate": 9.483000000000001e-05,
+      "loss": 0.7593,
+      "step": 28460
+    },
+    {
+      "epoch": 171.91,
+      "learning_rate": 9.489666666666666e-05,
+      "loss": 0.7793,
+      "step": 28480
+    },
+    {
+      "epoch": 172.03,
+      "learning_rate": 9.496333333333334e-05,
+      "loss": 0.75,
+      "step": 28500
+    },
+    {
+      "epoch": 172.15,
+      "learning_rate": 9.503e-05,
+      "loss": 0.6312,
+      "step": 28520
+    },
+    {
+      "epoch": 172.27,
+      "learning_rate": 9.509666666666667e-05,
+      "loss": 0.645,
+      "step": 28540
+    },
+    {
+      "epoch": 172.39,
+      "learning_rate": 9.516333333333334e-05,
+      "loss": 0.6736,
+      "step": 28560
+    },
+    {
+      "epoch": 172.52,
+      "learning_rate": 9.523000000000001e-05,
+      "loss": 0.6944,
+      "step": 28580
+    },
+    {
+      "epoch": 172.64,
+      "learning_rate": 9.529666666666667e-05,
+      "loss": 0.7285,
+      "step": 28600
+    },
+    {
+      "epoch": 172.76,
+      "learning_rate": 9.536333333333334e-05,
+      "loss": 0.7417,
+      "step": 28620
+    },
+    {
+      "epoch": 172.88,
+      "learning_rate": 9.543000000000001e-05,
+      "loss": 0.7498,
+      "step": 28640
+    },
+    {
+      "epoch": 173.0,
+      "learning_rate": 9.549666666666666e-05,
+      "loss": 0.7812,
+      "step": 28660
+    },
+    {
+      "epoch": 173.12,
+      "learning_rate": 9.556333333333334e-05,
+      "loss": 0.5998,
+      "step": 28680
+    },
+    {
+      "epoch": 173.24,
+      "learning_rate": 9.563e-05,
+      "loss": 0.6117,
+      "step": 28700
+    },
+    {
+      "epoch": 173.36,
+      "learning_rate": 9.569666666666667e-05,
+      "loss": 0.6423,
+      "step": 28720
+    },
+    {
+      "epoch": 173.48,
+      "learning_rate": 9.576333333333333e-05,
+      "loss": 0.679,
+      "step": 28740
+    },
+    {
+      "epoch": 173.6,
+      "learning_rate": 9.583000000000001e-05,
+      "loss": 0.7041,
+      "step": 28760
+    },
+    {
+      "epoch": 173.72,
+      "learning_rate": 9.589666666666667e-05,
+      "loss": 0.7316,
+      "step": 28780
+    },
+    {
+      "epoch": 173.84,
+      "learning_rate": 9.596333333333334e-05,
+      "loss": 0.7618,
+      "step": 28800
+    },
+    {
+      "epoch": 173.96,
+      "learning_rate": 9.603000000000001e-05,
+      "loss": 0.7617,
+      "step": 28820
+    },
+    {
+      "epoch": 174.08,
+      "learning_rate": 9.609666666666666e-05,
+      "loss": 0.6526,
+      "step": 28840
+    },
+    {
+      "epoch": 174.21,
+      "learning_rate": 9.616333333333334e-05,
+      "loss": 0.6295,
+      "step": 28860
+    },
+    {
+      "epoch": 174.33,
+      "learning_rate": 9.623e-05,
+      "loss": 0.6592,
+      "step": 28880
+    },
+    {
+      "epoch": 174.45,
+      "learning_rate": 9.629666666666667e-05,
+      "loss": 0.6655,
+      "step": 28900
+    },
+    {
+      "epoch": 174.57,
+      "learning_rate": 9.636333333333333e-05,
+      "loss": 0.6944,
+      "step": 28920
+    },
+    {
+      "epoch": 174.69,
+      "learning_rate": 9.643000000000001e-05,
+      "loss": 0.7155,
+      "step": 28940
+    },
+    {
+      "epoch": 174.81,
+      "learning_rate": 9.649666666666667e-05,
+      "loss": 0.7335,
+      "step": 28960
+    },
+    {
+      "epoch": 174.93,
+      "learning_rate": 9.656333333333334e-05,
+      "loss": 0.7532,
+      "step": 28980
+    },
+    {
+      "epoch": 175.05,
+      "learning_rate": 9.663000000000002e-05,
+      "loss": 0.6862,
+      "step": 29000
+    },
+    {
+      "epoch": 175.05,
+      "eval_accuracy": 0.0012698679258329958,
+      "eval_loss": 13.00501537322998,
+      "eval_runtime": 18.2714,
+      "eval_samples_per_second": 111.65,
+      "eval_steps_per_second": 2.353,
+      "step": 29000
+    },
+    {
+      "epoch": 175.17,
+      "learning_rate": 9.669666666666667e-05,
+      "loss": 0.6244,
+      "step": 29020
+    },
+    {
+      "epoch": 175.29,
+      "learning_rate": 9.676333333333334e-05,
+      "loss": 0.6274,
+      "step": 29040
+    },
+    {
+      "epoch": 175.41,
+      "learning_rate": 9.683e-05,
+      "loss": 0.6504,
+      "step": 29060
+    },
+    {
+      "epoch": 175.53,
+      "learning_rate": 9.689666666666667e-05,
+      "loss": 0.6514,
+      "step": 29080
+    },
+    {
+      "epoch": 175.65,
+      "learning_rate": 9.696333333333333e-05,
+      "loss": 0.6818,
+      "step": 29100
+    },
+    {
+      "epoch": 175.77,
+      "learning_rate": 9.703000000000001e-05,
+      "loss": 0.7134,
+      "step": 29120
+    },
+    {
+      "epoch": 175.9,
+      "learning_rate": 9.709666666666667e-05,
+      "loss": 0.7197,
+      "step": 29140
+    },
+    {
+      "epoch": 176.02,
+      "learning_rate": 9.716333333333334e-05,
+      "loss": 0.7298,
+      "step": 29160
+    },
+    {
+      "epoch": 176.14,
+      "learning_rate": 9.723000000000002e-05,
+      "loss": 0.5845,
+      "step": 29180
+    },
+    {
+      "epoch": 176.26,
+      "learning_rate": 9.729666666666667e-05,
+      "loss": 0.6012,
+      "step": 29200
+    },
+    {
+      "epoch": 176.38,
+      "learning_rate": 9.736333333333334e-05,
+      "loss": 0.6107,
+      "step": 29220
+    },
+    {
+      "epoch": 176.5,
+      "learning_rate": 9.743000000000001e-05,
+      "loss": 0.6382,
+      "step": 29240
+    },
+    {
+      "epoch": 176.62,
+      "learning_rate": 9.749666666666667e-05,
+      "loss": 0.6693,
+      "step": 29260
+    },
+    {
+      "epoch": 176.74,
+      "learning_rate": 9.756333333333333e-05,
+      "loss": 0.6891,
+      "step": 29280
+    },
+    {
+      "epoch": 176.86,
+      "learning_rate": 9.763e-05,
+      "loss": 0.7019,
+      "step": 29300
+    },
+    {
+      "epoch": 176.98,
+      "learning_rate": 9.769666666666668e-05,
+      "loss": 0.7324,
+      "step": 29320
+    },
+    {
+      "epoch": 177.1,
+      "learning_rate": 9.776333333333334e-05,
+      "loss": 0.5831,
+      "step": 29340
+    },
+    {
+      "epoch": 177.22,
+      "learning_rate": 9.783e-05,
+      "loss": 0.5856,
+      "step": 29360
+    },
+    {
+      "epoch": 177.34,
+      "learning_rate": 9.789666666666667e-05,
+      "loss": 0.6299,
+      "step": 29380
+    },
+    {
+      "epoch": 177.46,
+      "learning_rate": 9.796333333333334e-05,
+      "loss": 0.6551,
+      "step": 29400
+    },
+    {
+      "epoch": 177.59,
+      "learning_rate": 9.803e-05,
+      "loss": 0.6866,
+      "step": 29420
+    },
+    {
+      "epoch": 177.71,
+      "learning_rate": 9.809666666666667e-05,
+      "loss": 0.6927,
+      "step": 29440
+    },
+    {
+      "epoch": 177.83,
+      "learning_rate": 9.816333333333334e-05,
+      "loss": 0.7127,
+      "step": 29460
+    },
+    {
+      "epoch": 177.95,
+      "learning_rate": 9.823e-05,
+      "loss": 0.729,
+      "step": 29480
+    },
+    {
+      "epoch": 178.07,
+      "learning_rate": 9.829666666666666e-05,
+      "loss": 0.6424,
+      "step": 29500
+    },
+    {
+      "epoch": 178.19,
+      "learning_rate": 9.836333333333334e-05,
+      "loss": 0.587,
+      "step": 29520
+    },
+    {
+      "epoch": 178.31,
+      "learning_rate": 9.843e-05,
+      "loss": 0.6095,
+      "step": 29540
+    },
+    {
+      "epoch": 178.43,
+      "learning_rate": 9.849666666666667e-05,
+      "loss": 0.6249,
+      "step": 29560
+    },
+    {
+      "epoch": 178.55,
+      "learning_rate": 9.856333333333335e-05,
+      "loss": 0.6561,
+      "step": 29580
+    },
+    {
+      "epoch": 178.67,
+      "learning_rate": 9.863e-05,
+      "loss": 0.6764,
+      "step": 29600
+    },
+    {
+      "epoch": 178.79,
+      "learning_rate": 9.869666666666667e-05,
+      "loss": 0.6963,
+      "step": 29620
+    },
+    {
+      "epoch": 178.91,
+      "learning_rate": 9.876333333333334e-05,
+      "loss": 0.7177,
+      "step": 29640
+    },
+    {
+      "epoch": 179.03,
+      "learning_rate": 9.883e-05,
+      "loss": 0.6766,
+      "step": 29660
+    },
+    {
+      "epoch": 179.15,
+      "learning_rate": 9.889333333333334e-05,
+      "loss": 0.5581,
+      "step": 29680
+    },
+    {
+      "epoch": 179.28,
+      "learning_rate": 9.896000000000001e-05,
+      "loss": 0.5755,
+      "step": 29700
+    },
+    {
+      "epoch": 179.4,
+      "learning_rate": 9.902666666666666e-05,
+      "loss": 0.6193,
+      "step": 29720
+    },
+    {
+      "epoch": 179.52,
+      "learning_rate": 9.909333333333334e-05,
+      "loss": 0.6339,
+      "step": 29740
+    },
+    {
+      "epoch": 179.64,
+      "learning_rate": 9.916e-05,
+      "loss": 0.6576,
+      "step": 29760
+    },
+    {
+      "epoch": 179.76,
+      "learning_rate": 9.922666666666667e-05,
+      "loss": 0.666,
+      "step": 29780
+    },
+    {
+      "epoch": 179.88,
+      "learning_rate": 9.929333333333333e-05,
+      "loss": 0.6945,
+      "step": 29800
+    },
+    {
+      "epoch": 180.0,
+      "learning_rate": 9.936000000000001e-05,
+      "loss": 0.7119,
+      "step": 29820
+    },
+    {
+      "epoch": 180.12,
+      "learning_rate": 9.942666666666667e-05,
+      "loss": 0.5469,
+      "step": 29840
+    },
+    {
+      "epoch": 180.24,
+      "learning_rate": 9.949333333333334e-05,
+      "loss": 0.5791,
+      "step": 29860
+    },
+    {
+      "epoch": 180.36,
+      "learning_rate": 9.956e-05,
+      "loss": 0.6019,
+      "step": 29880
+    },
+    {
+      "epoch": 180.48,
+      "learning_rate": 9.962666666666667e-05,
+      "loss": 0.6169,
+      "step": 29900
+    },
+    {
+      "epoch": 180.6,
+      "learning_rate": 9.969333333333334e-05,
+      "loss": 0.6365,
+      "step": 29920
+    },
+    {
+      "epoch": 180.72,
+      "learning_rate": 9.976000000000001e-05,
+      "loss": 0.6589,
+      "step": 29940
+    },
+    {
+      "epoch": 180.85,
+      "learning_rate": 9.982666666666667e-05,
+      "loss": 0.699,
+      "step": 29960
+    },
+    {
+      "epoch": 180.97,
+      "learning_rate": 9.989333333333333e-05,
+      "loss": 0.7137,
+      "step": 29980
+    },
+    {
+      "epoch": 181.09,
+      "learning_rate": 9.996000000000001e-05,
+      "loss": 0.5738,
+      "step": 30000
+    },
+    {
+      "epoch": 181.09,
+      "eval_accuracy": 0.0012624591281326866,
+      "eval_loss": 13.469202041625977,
+      "eval_runtime": 18.2911,
+      "eval_samples_per_second": 111.529,
+      "eval_steps_per_second": 2.351,
+      "step": 30000
+    },
+    {
+      "epoch": 181.21,
+      "learning_rate": 9.999999677727208e-05,
+      "loss": 0.5442,
+      "step": 30020
+    },
+    {
+      "epoch": 181.33,
+      "learning_rate": 9.99999605215876e-05,
+      "loss": 0.5689,
+      "step": 30040
+    },
+    {
+      "epoch": 181.45,
+      "learning_rate": 9.999988398183804e-05,
+      "loss": 0.6313,
+      "step": 30060
+    },
+    {
+      "epoch": 181.57,
+      "learning_rate": 9.999976715808505e-05,
+      "loss": 0.6337,
+      "step": 30080
+    },
+    {
+      "epoch": 181.69,
+      "learning_rate": 9.999961005042278e-05,
+      "loss": 0.6481,
+      "step": 30100
+    },
+    {
+      "epoch": 181.81,
+      "learning_rate": 9.999941265897779e-05,
+      "loss": 0.6627,
+      "step": 30120
+    },
+    {
+      "epoch": 181.93,
+      "learning_rate": 9.999917498390912e-05,
+      "loss": 0.6641,
+      "step": 30140
+    },
+    {
+      "epoch": 182.05,
+      "learning_rate": 9.999889702540826e-05,
+      "loss": 0.6085,
+      "step": 30160
+    },
+    {
+      "epoch": 182.17,
+      "learning_rate": 9.999857878369916e-05,
+      "loss": 0.5334,
+      "step": 30180
+    },
+    {
+      "epoch": 182.29,
+      "learning_rate": 9.99982202590382e-05,
+      "loss": 0.5593,
+      "step": 30200
+    },
+    {
+      "epoch": 182.41,
+      "learning_rate": 9.999782145171425e-05,
+      "loss": 0.5836,
+      "step": 30220
+    },
+    {
+      "epoch": 182.54,
+      "learning_rate": 9.999738236204865e-05,
+      "loss": 0.5933,
+      "step": 30240
+    },
+    {
+      "epoch": 182.66,
+      "learning_rate": 9.999690299039513e-05,
+      "loss": 0.6157,
+      "step": 30260
+    },
+    {
+      "epoch": 182.78,
+      "learning_rate": 9.999638333713993e-05,
+      "loss": 0.648,
+      "step": 30280
+    },
+    {
+      "epoch": 182.9,
+      "learning_rate": 9.99958234027017e-05,
+      "loss": 0.6773,
+      "step": 30300
+    },
+    {
+      "epoch": 183.02,
+      "learning_rate": 9.99952231875316e-05,
+      "loss": 0.6561,
+      "step": 30320
+    },
+    {
+      "epoch": 183.14,
+      "learning_rate": 9.999458269211321e-05,
+      "loss": 0.5231,
+      "step": 30340
+    },
+    {
+      "epoch": 183.26,
+      "learning_rate": 9.999390191696255e-05,
+      "loss": 0.5543,
+      "step": 30360
+    },
+    {
+      "epoch": 183.38,
+      "learning_rate": 9.99931808626281e-05,
+      "loss": 0.5525,
+      "step": 30380
+    },
+    {
+      "epoch": 183.5,
+      "learning_rate": 9.999241952969083e-05,
+      "loss": 0.5855,
+      "step": 30400
+    },
+    {
+      "epoch": 183.62,
+      "learning_rate": 9.99916179187641e-05,
+      "loss": 0.6019,
+      "step": 30420
+    },
+    {
+      "epoch": 183.74,
+      "learning_rate": 9.999077603049378e-05,
+      "loss": 0.6313,
+      "step": 30440
+    },
+    {
+      "epoch": 183.86,
+      "learning_rate": 9.998989386555814e-05,
+      "loss": 0.6601,
+      "step": 30460
+    },
+    {
+      "epoch": 183.98,
+      "learning_rate": 9.998897142466797e-05,
+      "loss": 0.6642,
+      "step": 30480
+    },
+    {
+      "epoch": 184.1,
+      "learning_rate": 9.99880087085664e-05,
+      "loss": 0.5154,
+      "step": 30500
+    },
+    {
+      "epoch": 184.23,
+      "learning_rate": 9.998700571802912e-05,
+      "loss": 0.5116,
+      "step": 30520
+    },
+    {
+      "epoch": 184.35,
+      "learning_rate": 9.998596245386422e-05,
+      "loss": 0.5457,
+      "step": 30540
+    },
+    {
+      "epoch": 184.47,
+      "learning_rate": 9.998487891691219e-05,
+      "loss": 0.5773,
+      "step": 30560
+    },
+    {
+      "epoch": 184.59,
+      "learning_rate": 9.998375510804609e-05,
+      "loss": 0.5974,
+      "step": 30580
+    },
+    {
+      "epoch": 184.71,
+      "learning_rate": 9.998259102817129e-05,
+      "loss": 0.6244,
+      "step": 30600
+    },
+    {
+      "epoch": 184.83,
+      "learning_rate": 9.99813866782257e-05,
+      "loss": 0.6283,
+      "step": 30620
+    },
+    {
+      "epoch": 184.95,
+      "learning_rate": 9.998014205917963e-05,
+      "loss": 0.6439,
+      "step": 30640
+    },
+    {
+      "epoch": 185.07,
+      "learning_rate": 9.997885717203586e-05,
+      "loss": 0.5588,
+      "step": 30660
+    },
+    {
+      "epoch": 185.19,
+      "learning_rate": 9.997753201782961e-05,
+      "loss": 0.5076,
+      "step": 30680
+    },
+    {
+      "epoch": 185.31,
+      "learning_rate": 9.997616659762848e-05,
+      "loss": 0.5282,
+      "step": 30700
+    },
+    {
+      "epoch": 185.43,
+      "learning_rate": 9.997476091253261e-05,
+      "loss": 0.5626,
+      "step": 30720
+    },
+    {
+      "epoch": 185.55,
+      "learning_rate": 9.997331496367455e-05,
+      "loss": 0.5749,
+      "step": 30740
+    },
+    {
+      "epoch": 185.67,
+      "learning_rate": 9.997182875221922e-05,
+      "loss": 0.5809,
+      "step": 30760
+    },
+    {
+      "epoch": 185.79,
+      "learning_rate": 9.997030227936408e-05,
+      "loss": 0.6128,
+      "step": 30780
+    },
+    {
+      "epoch": 185.92,
+      "learning_rate": 9.996873554633897e-05,
+      "loss": 0.6223,
+      "step": 30800
+    },
+    {
+      "epoch": 186.04,
+      "learning_rate": 9.996712855440616e-05,
+      "loss": 0.5699,
+      "step": 30820
+    },
+    {
+      "epoch": 186.16,
+      "learning_rate": 9.99654813048604e-05,
+      "loss": 0.4744,
+      "step": 30840
+    },
+    {
+      "epoch": 186.28,
+      "learning_rate": 9.996379379902883e-05,
+      "loss": 0.5115,
+      "step": 30860
+    },
+    {
+      "epoch": 186.4,
+      "learning_rate": 9.996206603827105e-05,
+      "loss": 0.5336,
+      "step": 30880
+    },
+    {
+      "epoch": 186.52,
+      "learning_rate": 9.99602980239791e-05,
+      "loss": 0.5699,
+      "step": 30900
+    },
+    {
+      "epoch": 186.64,
+      "learning_rate": 9.99584897575774e-05,
+      "loss": 0.5892,
+      "step": 30920
+    },
+    {
+      "epoch": 186.76,
+      "learning_rate": 9.995664124052288e-05,
+      "loss": 0.5947,
+      "step": 30940
+    },
+    {
+      "epoch": 186.88,
+      "learning_rate": 9.995475247430483e-05,
+      "loss": 0.6197,
+      "step": 30960
+    },
+    {
+      "epoch": 187.0,
+      "learning_rate": 9.995282346044501e-05,
+      "loss": 0.6107,
+      "step": 30980
+    },
+    {
+      "epoch": 187.12,
+      "learning_rate": 9.995085420049757e-05,
+      "loss": 0.4524,
+      "step": 31000
+    },
+    {
+      "epoch": 187.12,
+      "eval_accuracy": 0.0012782645632266796,
+      "eval_loss": 13.922017097473145,
+      "eval_runtime": 18.1472,
+      "eval_samples_per_second": 112.414,
+      "eval_steps_per_second": 2.37,
+      "step": 31000
+    },
+    {
+      "epoch": 187.24,
+      "learning_rate": 9.994884469604912e-05,
+      "loss": 0.488,
+      "step": 31020
+    },
+    {
+      "epoch": 187.36,
+      "learning_rate": 9.99467949487187e-05,
+      "loss": 0.5118,
+      "step": 31040
+    },
+    {
+      "epoch": 187.48,
+      "learning_rate": 9.994470496015773e-05,
+      "loss": 0.527,
+      "step": 31060
+    },
+    {
+      "epoch": 187.61,
+      "learning_rate": 9.994257473205009e-05,
+      "loss": 0.5565,
+      "step": 31080
+    },
+    {
+      "epoch": 187.73,
+      "learning_rate": 9.994040426611204e-05,
+      "loss": 0.5922,
+      "step": 31100
+    },
+    {
+      "epoch": 187.85,
+      "learning_rate": 9.993819356409233e-05,
+      "loss": 0.6074,
+      "step": 31120
+    },
+    {
+      "epoch": 187.97,
+      "learning_rate": 9.993594262777203e-05,
+      "loss": 0.6096,
+      "step": 31140
+    },
+    {
+      "epoch": 188.09,
+      "learning_rate": 9.993365145896473e-05,
+      "loss": 0.4983,
+      "step": 31160
+    },
+    {
+      "epoch": 188.21,
+      "learning_rate": 9.993132005951635e-05,
+      "loss": 0.4859,
+      "step": 31180
+    },
+    {
+      "epoch": 188.33,
+      "learning_rate": 9.992894843130526e-05,
+      "loss": 0.5026,
+      "step": 31200
+    },
+    {
+      "epoch": 188.45,
+      "learning_rate": 9.992653657624225e-05,
+      "loss": 0.5297,
+      "step": 31220
+    },
+    {
+      "epoch": 188.57,
+      "learning_rate": 9.992408449627051e-05,
+      "loss": 0.5384,
+      "step": 31240
+    },
+    {
+      "epoch": 188.69,
+      "learning_rate": 9.992159219336561e-05,
+      "loss": 0.5536,
+      "step": 31260
+    },
+    {
+      "epoch": 188.81,
+      "learning_rate": 9.991905966953559e-05,
+      "loss": 0.5885,
+      "step": 31280
+    },
+    {
+      "epoch": 188.93,
+      "learning_rate": 9.991648692682083e-05,
+      "loss": 0.612,
+      "step": 31300
+    },
+    {
+      "epoch": 189.05,
+      "learning_rate": 9.991387396729415e-05,
+      "loss": 0.5315,
+      "step": 31320
+    },
+    {
+      "epoch": 189.18,
+      "learning_rate": 9.99112207930608e-05,
+      "loss": 0.4821,
+      "step": 31340
+    },
+    {
+      "epoch": 189.3,
+      "learning_rate": 9.990852740625831e-05,
+      "loss": 0.5144,
+      "step": 31360
+    },
+    {
+      "epoch": 189.42,
+      "learning_rate": 9.990579380905678e-05,
+      "loss": 0.5196,
+      "step": 31380
+    },
+    {
+      "epoch": 189.54,
+      "learning_rate": 9.990302000365858e-05,
+      "loss": 0.5276,
+      "step": 31400
+    },
+    {
+      "epoch": 189.66,
+      "learning_rate": 9.99002059922985e-05,
+      "loss": 0.5463,
+      "step": 31420
+    },
+    {
+      "epoch": 189.78,
+      "learning_rate": 9.989735177724378e-05,
+      "loss": 0.5747,
+      "step": 31440
+    },
+    {
+      "epoch": 189.9,
+      "learning_rate": 9.989445736079397e-05,
+      "loss": 0.5867,
+      "step": 31460
+    },
+    {
+      "epoch": 190.02,
+      "learning_rate": 9.989152274528109e-05,
+      "loss": 0.5554,
+      "step": 31480
+    },
+    {
+      "epoch": 190.14,
+      "learning_rate": 9.988854793306947e-05,
+      "loss": 0.4461,
+      "step": 31500
+    },
+    {
+      "epoch": 190.26,
+      "learning_rate": 9.988553292655587e-05,
+      "loss": 0.4767,
+      "step": 31520
+    },
+    {
+      "epoch": 190.38,
+      "learning_rate": 9.988247772816945e-05,
+      "loss": 0.4925,
+      "step": 31540
+    },
+    {
+      "epoch": 190.5,
+      "learning_rate": 9.987938234037168e-05,
+      "loss": 0.5114,
+      "step": 31560
+    },
+    {
+      "epoch": 190.62,
+      "learning_rate": 9.987624676565652e-05,
+      "loss": 0.5341,
+      "step": 31580
+    },
+    {
+      "epoch": 190.74,
+      "learning_rate": 9.987307100655019e-05,
+      "loss": 0.5667,
+      "step": 31600
+    },
+    {
+      "epoch": 190.87,
+      "learning_rate": 9.986985506561139e-05,
+      "loss": 0.5664,
+      "step": 31620
+    },
+    {
+      "epoch": 190.99,
+      "learning_rate": 9.98665989454311e-05,
+      "loss": 0.5933,
+      "step": 31640
+    },
+    {
+      "epoch": 191.11,
+      "learning_rate": 9.986330264863275e-05,
+      "loss": 0.449,
+      "step": 31660
+    },
+    {
+      "epoch": 191.23,
+      "learning_rate": 9.985996617787208e-05,
+      "loss": 0.4608,
+      "step": 31680
+    },
+    {
+      "epoch": 191.35,
+      "learning_rate": 9.985658953583725e-05,
+      "loss": 0.4939,
+      "step": 31700
+    },
+    {
+      "epoch": 191.47,
+      "learning_rate": 9.985334451973896e-05,
+      "loss": 0.5193,
+      "step": 31720
+    },
+    {
+      "epoch": 191.59,
+      "learning_rate": 9.984988955157381e-05,
+      "loss": 0.5285,
+      "step": 31740
+    },
+    {
+      "epoch": 191.71,
+      "learning_rate": 9.984639442025306e-05,
+      "loss": 0.5442,
+      "step": 31760
+    },
+    {
+      "epoch": 191.83,
+      "learning_rate": 9.984285912859263e-05,
+      "loss": 0.5525,
+      "step": 31780
+    },
+    {
+      "epoch": 191.95,
+      "learning_rate": 9.98392836794409e-05,
+      "loss": 0.5815,
+      "step": 31800
+    },
+    {
+      "epoch": 192.07,
+      "learning_rate": 9.98356680756785e-05,
+      "loss": 0.4852,
+      "step": 31820
+    },
+    {
+      "epoch": 192.19,
+      "learning_rate": 9.983201232021847e-05,
+      "loss": 0.4503,
+      "step": 31840
+    },
+    {
+      "epoch": 192.31,
+      "learning_rate": 9.982831641600618e-05,
+      "loss": 0.4759,
+      "step": 31860
+    },
+    {
+      "epoch": 192.43,
+      "learning_rate": 9.982458036601939e-05,
+      "loss": 0.4892,
+      "step": 31880
+    },
+    {
+      "epoch": 192.56,
+      "learning_rate": 9.982080417326811e-05,
+      "loss": 0.5049,
+      "step": 31900
+    },
+    {
+      "epoch": 192.68,
+      "learning_rate": 9.98169878407948e-05,
+      "loss": 0.5061,
+      "step": 31920
+    },
+    {
+      "epoch": 192.8,
+      "learning_rate": 9.981313137167419e-05,
+      "loss": 0.5286,
+      "step": 31940
+    },
+    {
+      "epoch": 192.92,
+      "learning_rate": 9.980923476901333e-05,
+      "loss": 0.5424,
+      "step": 31960
+    },
+    {
+      "epoch": 193.04,
+      "learning_rate": 9.980529803595172e-05,
+      "loss": 0.5141,
+      "step": 31980
+    },
+    {
+      "epoch": 193.16,
+      "learning_rate": 9.980132117566106e-05,
+      "loss": 0.4252,
+      "step": 32000
+    },
+    {
+      "epoch": 193.16,
+      "eval_accuracy": 0.0012767828036866177,
+      "eval_loss": 14.33401107788086,
+      "eval_runtime": 18.147,
+      "eval_samples_per_second": 112.415,
+      "eval_steps_per_second": 2.37,
+      "step": 32000
+    },
+    {
+      "epoch": 193.28,
+      "learning_rate": 9.979730419134546e-05,
+      "loss": 0.4494,
+      "step": 32020
+    },
+    {
+      "epoch": 193.4,
+      "learning_rate": 9.979324708624129e-05,
+      "loss": 0.4821,
+      "step": 32040
+    },
+    {
+      "epoch": 193.52,
+      "learning_rate": 9.978914986361734e-05,
+      "loss": 0.5057,
+      "step": 32060
+    },
+    {
+      "epoch": 193.64,
+      "learning_rate": 9.978501252677464e-05,
+      "loss": 0.5145,
+      "step": 32080
+    },
+    {
+      "epoch": 193.76,
+      "learning_rate": 9.978083507904659e-05,
+      "loss": 0.5297,
+      "step": 32100
+    },
+    {
+      "epoch": 193.88,
+      "learning_rate": 9.977661752379883e-05,
+      "loss": 0.5479,
+      "step": 32120
+    },
+    {
+      "epoch": 194.0,
+      "learning_rate": 9.977235986442943e-05,
+      "loss": 0.5598,
+      "step": 32140
+    },
+    {
+      "epoch": 194.12,
+      "learning_rate": 9.976806210436868e-05,
+      "loss": 0.3987,
+      "step": 32160
+    },
+    {
+      "epoch": 194.25,
+      "learning_rate": 9.97637242470792e-05,
+      "loss": 0.4324,
+      "step": 32180
+    },
+    {
+      "epoch": 194.37,
+      "learning_rate": 9.975934629605595e-05,
+      "loss": 0.4709,
+      "step": 32200
+    },
+    {
+      "epoch": 194.49,
+      "learning_rate": 9.975492825482615e-05,
+      "loss": 0.4809,
+      "step": 32220
+    },
+    {
+      "epoch": 194.61,
+      "learning_rate": 9.975047012694933e-05,
+      "loss": 0.4829,
+      "step": 32240
+    },
+    {
+      "epoch": 194.73,
+      "learning_rate": 9.974597191601733e-05,
+      "loss": 0.5059,
+      "step": 32260
+    },
+    {
+      "epoch": 194.85,
+      "learning_rate": 9.97414336256543e-05,
+      "loss": 0.5159,
+      "step": 32280
+    },
+    {
+      "epoch": 194.97,
+      "learning_rate": 9.97368552595166e-05,
+      "loss": 0.534,
+      "step": 32300
+    },
+    {
+      "epoch": 195.09,
+      "learning_rate": 9.9732236821293e-05,
+      "loss": 0.4287,
+      "step": 32320
+    },
+    {
+      "epoch": 195.21,
+      "learning_rate": 9.972757831470445e-05,
+      "loss": 0.4337,
+      "step": 32340
+    },
+    {
+      "epoch": 195.33,
+      "learning_rate": 9.972287974350425e-05,
+      "loss": 0.4485,
+      "step": 32360
+    },
+    {
+      "epoch": 195.45,
+      "learning_rate": 9.971814111147793e-05,
+      "loss": 0.4649,
+      "step": 32380
+    },
+    {
+      "epoch": 195.57,
+      "learning_rate": 9.971336242244333e-05,
+      "loss": 0.4701,
+      "step": 32400
+    },
+    {
+      "epoch": 195.69,
+      "learning_rate": 9.970854368025057e-05,
+      "loss": 0.4826,
+      "step": 32420
+    },
+    {
+      "epoch": 195.81,
+      "learning_rate": 9.970368488878202e-05,
+      "loss": 0.4994,
+      "step": 32440
+    },
+    {
+      "epoch": 195.94,
+      "learning_rate": 9.969878605195231e-05,
+      "loss": 0.5269,
+      "step": 32460
+    },
+    {
+      "epoch": 196.06,
+      "learning_rate": 9.969384717370834e-05,
+      "loss": 0.4673,
+      "step": 32480
+    },
+    {
+      "epoch": 196.18,
+      "learning_rate": 9.968886825802926e-05,
+      "loss": 0.3981,
+      "step": 32500
+    },
+    {
+      "epoch": 196.3,
+      "learning_rate": 9.968384930892653e-05,
+      "loss": 0.4332,
+      "step": 32520
+    },
+    {
+      "epoch": 196.42,
+      "learning_rate": 9.967879033044382e-05,
+      "loss": 0.4525,
+      "step": 32540
+    },
+    {
+      "epoch": 196.54,
+      "learning_rate": 9.967369132665703e-05,
+      "loss": 0.4606,
+      "step": 32560
+    },
+    {
+      "epoch": 196.66,
+      "learning_rate": 9.966855230167436e-05,
+      "loss": 0.4709,
+      "step": 32580
+    },
+    {
+      "epoch": 196.78,
+      "learning_rate": 9.966337325963624e-05,
+      "loss": 0.4942,
+      "step": 32600
+    },
+    {
+      "epoch": 196.9,
+      "learning_rate": 9.96581542047153e-05,
+      "loss": 0.5128,
+      "step": 32620
+    },
+    {
+      "epoch": 197.02,
+      "learning_rate": 9.965289514111644e-05,
+      "loss": 0.4951,
+      "step": 32640
+    },
+    {
+      "epoch": 197.14,
+      "learning_rate": 9.96475960730768e-05,
+      "loss": 0.4019,
+      "step": 32660
+    },
+    {
+      "epoch": 197.26,
+      "learning_rate": 9.964225700486576e-05,
+      "loss": 0.4134,
+      "step": 32680
+    },
+    {
+      "epoch": 197.38,
+      "learning_rate": 9.96368779407849e-05,
+      "loss": 0.4361,
+      "step": 32700
+    },
+    {
+      "epoch": 197.51,
+      "learning_rate": 9.963145888516803e-05,
+      "loss": 0.4558,
+      "step": 32720
+    },
+    {
+      "epoch": 197.63,
+      "learning_rate": 9.96259998423812e-05,
+      "loss": 0.47,
+      "step": 32740
+    },
+    {
+      "epoch": 197.75,
+      "learning_rate": 9.962050081682261e-05,
+      "loss": 0.4927,
+      "step": 32760
+    },
+    {
+      "epoch": 197.87,
+      "learning_rate": 9.96149618129228e-05,
+      "loss": 0.5016,
+      "step": 32780
+    },
+    {
+      "epoch": 197.99,
+      "learning_rate": 9.96093828351444e-05,
+      "loss": 0.5179,
+      "step": 32800
+    },
+    {
+      "epoch": 198.11,
+      "learning_rate": 9.960376388798232e-05,
+      "loss": 0.4012,
+      "step": 32820
+    },
+    {
+      "epoch": 198.23,
+      "learning_rate": 9.959810497596361e-05,
+      "loss": 0.4098,
+      "step": 32840
+    },
+    {
+      "epoch": 198.35,
+      "learning_rate": 9.959240610364757e-05,
+      "loss": 0.4298,
+      "step": 32860
+    },
+    {
+      "epoch": 198.47,
+      "learning_rate": 9.958666727562568e-05,
+      "loss": 0.4416,
+      "step": 32880
+    },
+    {
+      "epoch": 198.59,
+      "learning_rate": 9.95808884965216e-05,
+      "loss": 0.4643,
+      "step": 32900
+    },
+    {
+      "epoch": 198.71,
+      "learning_rate": 9.957506977099123e-05,
+      "loss": 0.4748,
+      "step": 32920
+    },
+    {
+      "epoch": 198.83,
+      "learning_rate": 9.956921110372255e-05,
+      "loss": 0.4831,
+      "step": 32940
+    },
+    {
+      "epoch": 198.95,
+      "learning_rate": 9.956331249943583e-05,
+      "loss": 0.502,
+      "step": 32960
+    },
+    {
+      "epoch": 199.07,
+      "learning_rate": 9.955737396288343e-05,
+      "loss": 0.417,
+      "step": 32980
+    },
+    {
+      "epoch": 199.2,
+      "learning_rate": 9.955139549884995e-05,
+      "loss": 0.3952,
+      "step": 33000
+    },
+    {
+      "epoch": 199.2,
+      "eval_accuracy": 0.001259495609052563,
+      "eval_loss": 14.796114921569824,
+      "eval_runtime": 18.1968,
+      "eval_samples_per_second": 112.108,
+      "eval_steps_per_second": 2.363,
+      "step": 33000
+    },
+    {
+      "epoch": 199.32,
+      "learning_rate": 9.954537711215212e-05,
+      "loss": 0.4078,
+      "step": 33020
+    },
+    {
+      "epoch": 199.44,
+      "learning_rate": 9.953931880763885e-05,
+      "loss": 0.4446,
+      "step": 33040
+    },
+    {
+      "epoch": 199.56,
+      "learning_rate": 9.953322059019118e-05,
+      "loss": 0.4559,
+      "step": 33060
+    },
+    {
+      "epoch": 199.68,
+      "learning_rate": 9.95270824647224e-05,
+      "loss": 0.4593,
+      "step": 33080
+    },
+    {
+      "epoch": 199.8,
+      "learning_rate": 9.952090443617783e-05,
+      "loss": 0.4724,
+      "step": 33100
+    },
+    {
+      "epoch": 199.92,
+      "learning_rate": 9.951468650953499e-05,
+      "loss": 0.4951,
+      "step": 33120
+    },
+    {
+      "epoch": 200.04,
+      "learning_rate": 9.95084286898036e-05,
+      "loss": 0.4568,
+      "step": 33140
+    },
+    {
+      "epoch": 200.16,
+      "learning_rate": 9.950213098202543e-05,
+      "loss": 0.3755,
+      "step": 33160
+    },
+    {
+      "epoch": 200.28,
+      "learning_rate": 9.949579339127444e-05,
+      "loss": 0.4004,
+      "step": 33180
+    },
+    {
+      "epoch": 200.4,
+      "learning_rate": 9.948941592265673e-05,
+      "loss": 0.4046,
+      "step": 33200
+    },
+    {
+      "epoch": 200.52,
+      "learning_rate": 9.948299858131049e-05,
+      "loss": 0.4194,
+      "step": 33220
+    },
+    {
+      "epoch": 200.64,
+      "learning_rate": 9.947654137240606e-05,
+      "loss": 0.4467,
+      "step": 33240
+    },
+    {
+      "epoch": 200.76,
+      "learning_rate": 9.94700443011459e-05,
+      "loss": 0.4544,
+      "step": 33260
+    },
+    {
+      "epoch": 200.89,
+      "learning_rate": 9.946350737276459e-05,
+      "loss": 0.476,
+      "step": 33280
+    },
+    {
+      "epoch": 201.01,
+      "learning_rate": 9.945693059252881e-05,
+      "loss": 0.4785,
+      "step": 33300
+    },
+    {
+      "epoch": 201.13,
+      "learning_rate": 9.945031396573734e-05,
+      "loss": 0.3558,
+      "step": 33320
+    },
+    {
+      "epoch": 201.25,
+      "learning_rate": 9.94436574977211e-05,
+      "loss": 0.3768,
+      "step": 33340
+    },
+    {
+      "epoch": 201.37,
+      "learning_rate": 9.943696119384305e-05,
+      "loss": 0.4028,
+      "step": 33360
+    },
+    {
+      "epoch": 201.49,
+      "learning_rate": 9.943022505949833e-05,
+      "loss": 0.4206,
+      "step": 33380
+    },
+    {
+      "epoch": 201.61,
+      "learning_rate": 9.942344910011407e-05,
+      "loss": 0.4348,
+      "step": 33400
+    },
+    {
+      "epoch": 201.73,
+      "learning_rate": 9.941663332114959e-05,
+      "loss": 0.456,
+      "step": 33420
+    },
+    {
+      "epoch": 201.85,
+      "learning_rate": 9.940977772809621e-05,
+      "loss": 0.4605,
+      "step": 33440
+    },
+    {
+      "epoch": 201.97,
+      "learning_rate": 9.940288232647736e-05,
+      "loss": 0.4742,
+      "step": 33460
+    },
+    {
+      "epoch": 202.09,
+      "learning_rate": 9.939594712184854e-05,
+      "loss": 0.3742,
+      "step": 33480
+    },
+    {
+      "epoch": 202.21,
+      "learning_rate": 9.938897211979732e-05,
+      "loss": 0.3726,
+      "step": 33500
+    },
+    {
+      "epoch": 202.33,
+      "learning_rate": 9.938195732594335e-05,
+      "loss": 0.4019,
+      "step": 33520
+    },
+    {
+      "epoch": 202.45,
+      "learning_rate": 9.937490274593828e-05,
+      "loss": 0.4086,
+      "step": 33540
+    },
+    {
+      "epoch": 202.58,
+      "learning_rate": 9.93678083854659e-05,
+      "loss": 0.4346,
+      "step": 33560
+    },
+    {
+      "epoch": 202.7,
+      "learning_rate": 9.936067425024201e-05,
+      "loss": 0.4373,
+      "step": 33580
+    },
+    {
+      "epoch": 202.82,
+      "learning_rate": 9.93535003460144e-05,
+      "loss": 0.453,
+      "step": 33600
+    },
+    {
+      "epoch": 202.94,
+      "learning_rate": 9.934628667856303e-05,
+      "loss": 0.4692,
+      "step": 33620
+    },
+    {
+      "epoch": 203.06,
+      "learning_rate": 9.933903325369977e-05,
+      "loss": 0.4171,
+      "step": 33640
+    },
+    {
+      "epoch": 203.18,
+      "learning_rate": 9.933174007726858e-05,
+      "loss": 0.3703,
+      "step": 33660
+    },
+    {
+      "epoch": 203.3,
+      "learning_rate": 9.932440715514545e-05,
+      "loss": 0.3899,
+      "step": 33680
+    },
+    {
+      "epoch": 203.42,
+      "learning_rate": 9.93170344932384e-05,
+      "loss": 0.4087,
+      "step": 33700
+    },
+    {
+      "epoch": 203.54,
+      "learning_rate": 9.930962209748741e-05,
+      "loss": 0.4244,
+      "step": 33720
+    },
+    {
+      "epoch": 203.66,
+      "learning_rate": 9.930254352349008e-05,
+      "loss": 0.4306,
+      "step": 33740
+    },
+    {
+      "epoch": 203.78,
+      "learning_rate": 9.929505366394967e-05,
+      "loss": 0.443,
+      "step": 33760
+    },
+    {
+      "epoch": 203.9,
+      "learning_rate": 9.92875240882749e-05,
+      "loss": 0.4556,
+      "step": 33780
+    },
+    {
+      "epoch": 204.02,
+      "learning_rate": 9.92799548025322e-05,
+      "loss": 0.4405,
+      "step": 33800
+    },
+    {
+      "epoch": 204.14,
+      "learning_rate": 9.927234581282004e-05,
+      "loss": 0.3568,
+      "step": 33820
+    },
+    {
+      "epoch": 204.27,
+      "learning_rate": 9.92646971252688e-05,
+      "loss": 0.3736,
+      "step": 33840
+    },
+    {
+      "epoch": 204.39,
+      "learning_rate": 9.925700874604093e-05,
+      "loss": 0.3884,
+      "step": 33860
+    },
+    {
+      "epoch": 204.51,
+      "learning_rate": 9.92492806813308e-05,
+      "loss": 0.408,
+      "step": 33880
+    },
+    {
+      "epoch": 204.63,
+      "learning_rate": 9.924151293736476e-05,
+      "loss": 0.4256,
+      "step": 33900
+    },
+    {
+      "epoch": 204.75,
+      "learning_rate": 9.923370552040116e-05,
+      "loss": 0.4364,
+      "step": 33920
+    },
+    {
+      "epoch": 204.87,
+      "learning_rate": 9.922585843673031e-05,
+      "loss": 0.4657,
+      "step": 33940
+    },
+    {
+      "epoch": 204.99,
+      "learning_rate": 9.921797169267443e-05,
+      "loss": 0.4655,
+      "step": 33960
+    },
+    {
+      "epoch": 205.11,
+      "learning_rate": 9.921004529458772e-05,
+      "loss": 0.35,
+      "step": 33980
+    },
+    {
+      "epoch": 205.23,
+      "learning_rate": 9.920207924885638e-05,
+      "loss": 0.3684,
+      "step": 34000
+    },
+    {
+      "epoch": 205.23,
+      "eval_accuracy": 0.0012590016892058756,
+      "eval_loss": 15.24209976196289,
+      "eval_runtime": 18.177,
+      "eval_samples_per_second": 112.23,
+      "eval_steps_per_second": 2.366,
+      "step": 34000
+    },
+    {
+      "epoch": 205.35,
+      "learning_rate": 9.919407356189849e-05,
+      "loss": 0.3822,
+      "step": 34020
+    },
+    {
+      "epoch": 205.47,
+      "learning_rate": 9.918602824016408e-05,
+      "loss": 0.3956,
+      "step": 34040
+    },
+    {
+      "epoch": 205.59,
+      "learning_rate": 9.91779432901351e-05,
+      "loss": 0.4029,
+      "step": 34060
+    },
+    {
+      "epoch": 205.71,
+      "learning_rate": 9.916981871832549e-05,
+      "loss": 0.4162,
+      "step": 34080
+    },
+    {
+      "epoch": 205.84,
+      "learning_rate": 9.916165453128106e-05,
+      "loss": 0.4329,
+      "step": 34100
+    },
+    {
+      "epoch": 205.96,
+      "learning_rate": 9.915345073557953e-05,
+      "loss": 0.4465,
+      "step": 34120
+    },
+    {
+      "epoch": 206.08,
+      "learning_rate": 9.914520733783055e-05,
+      "loss": 0.3748,
+      "step": 34140
+    },
+    {
+      "epoch": 206.2,
+      "learning_rate": 9.913692434467569e-05,
+      "loss": 0.3555,
+      "step": 34160
+    },
+    {
+      "epoch": 206.32,
+      "learning_rate": 9.912860176278842e-05,
+      "loss": 0.3747,
+      "step": 34180
+    },
+    {
+      "epoch": 206.44,
+      "learning_rate": 9.912023959887408e-05,
+      "loss": 0.3954,
+      "step": 34200
+    },
+    {
+      "epoch": 206.56,
+      "learning_rate": 9.91118378596699e-05,
+      "loss": 0.403,
+      "step": 34220
+    },
+    {
+      "epoch": 206.68,
+      "learning_rate": 9.910339655194502e-05,
+      "loss": 0.4143,
+      "step": 34240
+    },
+    {
+      "epoch": 206.8,
+      "learning_rate": 9.909491568250047e-05,
+      "loss": 0.4167,
+      "step": 34260
+    },
+    {
+      "epoch": 206.92,
+      "learning_rate": 9.90863952581691e-05,
+      "loss": 0.434,
+      "step": 34280
+    },
+    {
+      "epoch": 207.04,
+      "learning_rate": 9.907783528581568e-05,
+      "loss": 0.4013,
+      "step": 34300
+    },
+    {
+      "epoch": 207.16,
+      "learning_rate": 9.906923577233682e-05,
+      "loss": 0.3298,
+      "step": 34320
+    },
+    {
+      "epoch": 207.28,
+      "learning_rate": 9.906059672466101e-05,
+      "loss": 0.3502,
+      "step": 34340
+    },
+    {
+      "epoch": 207.4,
+      "learning_rate": 9.905191814974854e-05,
+      "loss": 0.3611,
+      "step": 34360
+    },
+    {
+      "epoch": 207.53,
+      "learning_rate": 9.90432000545916e-05,
+      "loss": 0.3774,
+      "step": 34380
+    },
+    {
+      "epoch": 207.65,
+      "learning_rate": 9.903444244621423e-05,
+      "loss": 0.3982,
+      "step": 34400
+    },
+    {
+      "epoch": 207.77,
+      "learning_rate": 9.902564533167222e-05,
+      "loss": 0.4134,
+      "step": 34420
+    },
+    {
+      "epoch": 207.89,
+      "learning_rate": 9.901680871805329e-05,
+      "loss": 0.4233,
+      "step": 34440
+    },
+    {
+      "epoch": 208.01,
+      "learning_rate": 9.900793261247692e-05,
+      "loss": 0.4307,
+      "step": 34460
+    },
+    {
+      "epoch": 208.13,
+      "learning_rate": 9.899901702209445e-05,
+      "loss": 0.3294,
+      "step": 34480
+    },
+    {
+      "epoch": 208.25,
+      "learning_rate": 9.899006195408898e-05,
+      "loss": 0.3552,
+      "step": 34500
+    },
+    {
+      "epoch": 208.37,
+      "learning_rate": 9.898106741567545e-05,
+      "loss": 0.366,
+      "step": 34520
+    },
+    {
+      "epoch": 208.49,
+      "learning_rate": 9.897203341410062e-05,
+      "loss": 0.3707,
+      "step": 34540
+    },
+    {
+      "epoch": 208.61,
+      "learning_rate": 9.8962959956643e-05,
+      "loss": 0.402,
+      "step": 34560
+    },
+    {
+      "epoch": 208.73,
+      "learning_rate": 9.895384705061292e-05,
+      "loss": 0.3991,
+      "step": 34580
+    },
+    {
+      "epoch": 208.85,
+      "learning_rate": 9.894469470335251e-05,
+      "loss": 0.4095,
+      "step": 34600
+    },
+    {
+      "epoch": 208.97,
+      "learning_rate": 9.89355029222356e-05,
+      "loss": 0.4253,
+      "step": 34620
+    },
+    {
+      "epoch": 209.09,
+      "learning_rate": 9.892627171466786e-05,
+      "loss": 0.3422,
+      "step": 34640
+    },
+    {
+      "epoch": 209.22,
+      "learning_rate": 9.891700108808674e-05,
+      "loss": 0.3367,
+      "step": 34660
+    },
+    {
+      "epoch": 209.34,
+      "learning_rate": 9.890769104996137e-05,
+      "loss": 0.3543,
+      "step": 34680
+    },
+    {
+      "epoch": 209.46,
+      "learning_rate": 9.889834160779271e-05,
+      "loss": 0.378,
+      "step": 34700
+    },
+    {
+      "epoch": 209.58,
+      "learning_rate": 9.888895276911342e-05,
+      "loss": 0.384,
+      "step": 34720
+    },
+    {
+      "epoch": 209.7,
+      "learning_rate": 9.887952454148792e-05,
+      "loss": 0.4011,
+      "step": 34740
+    },
+    {
+      "epoch": 209.82,
+      "learning_rate": 9.88700569325124e-05,
+      "loss": 0.4053,
+      "step": 34760
+    },
+    {
+      "epoch": 209.94,
+      "learning_rate": 9.886054994981466e-05,
+      "loss": 0.4203,
+      "step": 34780
+    },
+    {
+      "epoch": 210.06,
+      "learning_rate": 9.88510036010544e-05,
+      "loss": 0.3665,
+      "step": 34800
+    },
+    {
+      "epoch": 210.18,
+      "learning_rate": 9.884141789392285e-05,
+      "loss": 0.3291,
+      "step": 34820
+    },
+    {
+      "epoch": 210.3,
+      "learning_rate": 9.883179283614311e-05,
+      "loss": 0.3487,
+      "step": 34840
+    },
+    {
+      "epoch": 210.42,
+      "learning_rate": 9.882212843546988e-05,
+      "loss": 0.3575,
+      "step": 34860
+    },
+    {
+      "epoch": 210.54,
+      "learning_rate": 9.881242469968962e-05,
+      "loss": 0.3729,
+      "step": 34880
+    },
+    {
+      "epoch": 210.66,
+      "learning_rate": 9.880268163662042e-05,
+      "loss": 0.3939,
+      "step": 34900
+    },
+    {
+      "epoch": 210.78,
+      "learning_rate": 9.879289925411212e-05,
+      "loss": 0.3978,
+      "step": 34920
+    },
+    {
+      "epoch": 210.91,
+      "learning_rate": 9.878307756004619e-05,
+      "loss": 0.4146,
+      "step": 34940
+    },
+    {
+      "epoch": 211.03,
+      "learning_rate": 9.87732165623358e-05,
+      "loss": 0.3925,
+      "step": 34960
+    },
+    {
+      "epoch": 211.15,
+      "learning_rate": 9.876331626892576e-05,
+      "loss": 0.3063,
+      "step": 34980
+    },
+    {
+      "epoch": 211.27,
+      "learning_rate": 9.875337668779259e-05,
+      "loss": 0.3338,
+      "step": 35000
+    },
+    {
+      "epoch": 211.27,
+      "eval_accuracy": 0.001261471288439312,
+      "eval_loss": 15.643324851989746,
+      "eval_runtime": 18.2037,
+      "eval_samples_per_second": 112.065,
+      "eval_steps_per_second": 2.362,
+      "step": 35000
+    },
+    {
+      "epoch": 211.39,
+      "learning_rate": 9.874339782694441e-05,
+      "loss": 0.3398,
+      "step": 35020
+    },
+    {
+      "epoch": 211.51,
+      "learning_rate": 9.873337969442101e-05,
+      "loss": 0.3385,
+      "step": 35040
+    },
+    {
+      "epoch": 211.63,
+      "learning_rate": 9.872332229829383e-05,
+      "loss": 0.3627,
+      "step": 35060
+    },
+    {
+      "epoch": 211.75,
+      "learning_rate": 9.871322564666592e-05,
+      "loss": 0.3795,
+      "step": 35080
+    },
+    {
+      "epoch": 211.87,
+      "learning_rate": 9.870308974767197e-05,
+      "loss": 0.3977,
+      "step": 35100
+    },
+    {
+      "epoch": 211.99,
+      "learning_rate": 9.869291460947829e-05,
+      "loss": 0.4096,
+      "step": 35120
+    },
+    {
+      "epoch": 212.11,
+      "learning_rate": 9.86827002402828e-05,
+      "loss": 0.3138,
+      "step": 35140
+    },
+    {
+      "epoch": 212.23,
+      "learning_rate": 9.867244664831507e-05,
+      "loss": 0.3286,
+      "step": 35160
+    },
+    {
+      "epoch": 212.35,
+      "learning_rate": 9.866215384183619e-05,
+      "loss": 0.3414,
+      "step": 35180
+    },
+    {
+      "epoch": 212.47,
+      "learning_rate": 9.86518218291389e-05,
+      "loss": 0.3603,
+      "step": 35200
+    },
+    {
+      "epoch": 212.6,
+      "learning_rate": 9.864145061854753e-05,
+      "loss": 0.3646,
+      "step": 35220
+    },
+    {
+      "epoch": 212.72,
+      "learning_rate": 9.863104021841796e-05,
+      "loss": 0.3821,
+      "step": 35240
+    },
+    {
+      "epoch": 212.84,
+      "learning_rate": 9.862059063713767e-05,
+      "loss": 0.3876,
+      "step": 35260
+    },
+    {
+      "epoch": 212.96,
+      "learning_rate": 9.861010188312568e-05,
+      "loss": 0.3981,
+      "step": 35280
+    },
+    {
+      "epoch": 213.08,
+      "learning_rate": 9.859957396483262e-05,
+      "loss": 0.3357,
+      "step": 35300
+    },
+    {
+      "epoch": 213.2,
+      "learning_rate": 9.858900689074064e-05,
+      "loss": 0.3274,
+      "step": 35320
+    },
+    {
+      "epoch": 213.32,
+      "learning_rate": 9.857840066936341e-05,
+      "loss": 0.3389,
+      "step": 35340
+    },
+    {
+      "epoch": 213.44,
+      "learning_rate": 9.856775530924621e-05,
+      "loss": 0.3445,
+      "step": 35360
+    },
+    {
+      "epoch": 213.56,
+      "learning_rate": 9.855707081896579e-05,
+      "loss": 0.3609,
+      "step": 35380
+    },
+    {
+      "epoch": 213.68,
+      "learning_rate": 9.854634720713047e-05,
+      "loss": 0.3758,
+      "step": 35400
+    },
+    {
+      "epoch": 213.8,
+      "learning_rate": 9.853558448238003e-05,
+      "loss": 0.3877,
+      "step": 35420
+    },
+    {
+      "epoch": 213.92,
+      "learning_rate": 9.852478265338585e-05,
+      "loss": 0.3981,
+      "step": 35440
+    },
+    {
+      "epoch": 214.04,
+      "learning_rate": 9.851394172885074e-05,
+      "loss": 0.3666,
+      "step": 35460
+    },
+    {
+      "epoch": 214.16,
+      "learning_rate": 9.850306171750905e-05,
+      "loss": 0.2927,
+      "step": 35480
+    },
+    {
+      "epoch": 214.29,
+      "learning_rate": 9.84921426281266e-05,
+      "loss": 0.3201,
+      "step": 35500
+    },
+    {
+      "epoch": 214.41,
+      "learning_rate": 9.848118446950072e-05,
+      "loss": 0.3414,
+      "step": 35520
+    },
+    {
+      "epoch": 214.53,
+      "learning_rate": 9.847018725046018e-05,
+      "loss": 0.3555,
+      "step": 35540
+    },
+    {
+      "epoch": 214.65,
+      "learning_rate": 9.845915097986526e-05,
+      "loss": 0.3696,
+      "step": 35560
+    },
+    {
+      "epoch": 214.77,
+      "learning_rate": 9.844807566660765e-05,
+      "loss": 0.3677,
+      "step": 35580
+    },
+    {
+      "epoch": 214.89,
+      "learning_rate": 9.843696131961058e-05,
+      "loss": 0.3831,
+      "step": 35600
+    },
+    {
+      "epoch": 215.01,
+      "learning_rate": 9.842580794782863e-05,
+      "loss": 0.3879,
+      "step": 35620
+    },
+    {
+      "epoch": 215.13,
+      "learning_rate": 9.841461556024789e-05,
+      "loss": 0.2884,
+      "step": 35640
+    },
+    {
+      "epoch": 215.25,
+      "learning_rate": 9.840338416588588e-05,
+      "loss": 0.3135,
+      "step": 35660
+    },
+    {
+      "epoch": 215.37,
+      "learning_rate": 9.839211377379151e-05,
+      "loss": 0.3296,
+      "step": 35680
+    },
+    {
+      "epoch": 215.49,
+      "learning_rate": 9.838080439304515e-05,
+      "loss": 0.3421,
+      "step": 35700
+    },
+    {
+      "epoch": 215.61,
+      "learning_rate": 9.836945603275853e-05,
+      "loss": 0.3571,
+      "step": 35720
+    },
+    {
+      "epoch": 215.73,
+      "learning_rate": 9.835806870207487e-05,
+      "loss": 0.3734,
+      "step": 35740
+    },
+    {
+      "epoch": 215.86,
+      "learning_rate": 9.834664241016869e-05,
+      "loss": 0.372,
+      "step": 35760
+    },
+    {
+      "epoch": 215.98,
+      "learning_rate": 9.833575135341002e-05,
+      "loss": 0.3917,
+      "step": 35780
+    },
+    {
+      "epoch": 216.1,
+      "learning_rate": 9.832424911362719e-05,
+      "loss": 0.3021,
+      "step": 35800
+    },
+    {
+      "epoch": 216.22,
+      "learning_rate": 9.831270793986968e-05,
+      "loss": 0.3098,
+      "step": 35820
+    },
+    {
+      "epoch": 216.34,
+      "learning_rate": 9.830112784143601e-05,
+      "loss": 0.3254,
+      "step": 35840
+    },
+    {
+      "epoch": 216.46,
+      "learning_rate": 9.828950882765608e-05,
+      "loss": 0.3275,
+      "step": 35860
+    },
+    {
+      "epoch": 216.58,
+      "learning_rate": 9.827785090789109e-05,
+      "loss": 0.3395,
+      "step": 35880
+    },
+    {
+      "epoch": 216.7,
+      "learning_rate": 9.826615409153362e-05,
+      "loss": 0.3466,
+      "step": 35900
+    },
+    {
+      "epoch": 216.82,
+      "learning_rate": 9.82544183880076e-05,
+      "loss": 0.361,
+      "step": 35920
+    },
+    {
+      "epoch": 216.94,
+      "learning_rate": 9.824264380676826e-05,
+      "loss": 0.3701,
+      "step": 35940
+    },
+    {
+      "epoch": 217.06,
+      "learning_rate": 9.823083035730218e-05,
+      "loss": 0.329,
+      "step": 35960
+    },
+    {
+      "epoch": 217.18,
+      "learning_rate": 9.821897804912722e-05,
+      "loss": 0.3044,
+      "step": 35980
+    },
+    {
+      "epoch": 217.3,
+      "learning_rate": 9.820708689179259e-05,
+      "loss": 0.307,
+      "step": 36000
+    },
+    {
+      "epoch": 217.3,
+      "eval_accuracy": 0.0012698679258329958,
+      "eval_loss": 16.018239974975586,
+      "eval_runtime": 18.2272,
+      "eval_samples_per_second": 111.921,
+      "eval_steps_per_second": 2.359,
+      "step": 36000
+    },
+    {
+      "epoch": 217.42,
+      "learning_rate": 9.819515689487877e-05,
+      "loss": 0.3246,
+      "step": 36020
+    },
+    {
+      "epoch": 217.55,
+      "learning_rate": 9.818318806799755e-05,
+      "loss": 0.3441,
+      "step": 36040
+    },
+    {
+      "epoch": 217.67,
+      "learning_rate": 9.817118042079201e-05,
+      "loss": 0.3581,
+      "step": 36060
+    },
+    {
+      "epoch": 217.79,
+      "learning_rate": 9.815913396293648e-05,
+      "loss": 0.3714,
+      "step": 36080
+    },
+    {
+      "epoch": 217.91,
+      "learning_rate": 9.814704870413657e-05,
+      "loss": 0.3816,
+      "step": 36100
+    },
+    {
+      "epoch": 218.03,
+      "learning_rate": 9.813492465412916e-05,
+      "loss": 0.3529,
+      "step": 36120
+    },
+    {
+      "epoch": 218.15,
+      "learning_rate": 9.812276182268236e-05,
+      "loss": 0.2744,
+      "step": 36140
+    },
+    {
+      "epoch": 218.27,
+      "learning_rate": 9.811056021959559e-05,
+      "loss": 0.2969,
+      "step": 36160
+    },
+    {
+      "epoch": 218.39,
+      "learning_rate": 9.809831985469942e-05,
+      "loss": 0.3186,
+      "step": 36180
+    },
+    {
+      "epoch": 218.51,
+      "learning_rate": 9.80860407378557e-05,
+      "loss": 0.3259,
+      "step": 36200
+    },
+    {
+      "epoch": 218.63,
+      "learning_rate": 9.807372287895748e-05,
+      "loss": 0.3433,
+      "step": 36220
+    },
+    {
+      "epoch": 218.75,
+      "learning_rate": 9.806136628792907e-05,
+      "loss": 0.3623,
+      "step": 36240
+    },
+    {
+      "epoch": 218.87,
+      "learning_rate": 9.804897097472594e-05,
+      "loss": 0.3708,
+      "step": 36260
+    },
+    {
+      "epoch": 218.99,
+      "learning_rate": 9.803653694933476e-05,
+      "loss": 0.3843,
+      "step": 36280
+    },
+    {
+      "epoch": 219.11,
+      "learning_rate": 9.802406422177339e-05,
+      "loss": 0.285,
+      "step": 36300
+    },
+    {
+      "epoch": 219.24,
+      "learning_rate": 9.801155280209091e-05,
+      "loss": 0.2975,
+      "step": 36320
+    },
+    {
+      "epoch": 219.36,
+      "learning_rate": 9.799900270036753e-05,
+      "loss": 0.3218,
+      "step": 36340
+    },
+    {
+      "epoch": 219.48,
+      "learning_rate": 9.798641392671463e-05,
+      "loss": 0.3277,
+      "step": 36360
+    },
+    {
+      "epoch": 219.6,
+      "learning_rate": 9.797378649127478e-05,
+      "loss": 0.3469,
+      "step": 36380
+    },
+    {
+      "epoch": 219.72,
+      "learning_rate": 9.796112040422165e-05,
+      "loss": 0.3561,
+      "step": 36400
+    },
+    {
+      "epoch": 219.84,
+      "learning_rate": 9.794841567576011e-05,
+      "loss": 0.3646,
+      "step": 36420
+    },
+    {
+      "epoch": 219.96,
+      "learning_rate": 9.79356723161261e-05,
+      "loss": 0.3528,
+      "step": 36440
+    },
+    {
+      "epoch": 220.08,
+      "learning_rate": 9.792289033558673e-05,
+      "loss": 0.2793,
+      "step": 36460
+    },
+    {
+      "epoch": 220.2,
+      "learning_rate": 9.791006974444022e-05,
+      "loss": 0.2655,
+      "step": 36480
+    },
+    {
+      "epoch": 220.32,
+      "learning_rate": 9.789721055301586e-05,
+      "loss": 0.2856,
+      "step": 36500
+    },
+    {
+      "epoch": 220.44,
+      "learning_rate": 9.788431277167409e-05,
+      "loss": 0.3194,
+      "step": 36520
+    },
+    {
+      "epoch": 220.56,
+      "learning_rate": 9.787137641080641e-05,
+      "loss": 0.3235,
+      "step": 36540
+    },
+    {
+      "epoch": 220.68,
+      "learning_rate": 9.785840148083543e-05,
+      "loss": 0.3404,
+      "step": 36560
+    },
+    {
+      "epoch": 220.8,
+      "learning_rate": 9.784538799221478e-05,
+      "loss": 0.3541,
+      "step": 36580
+    },
+    {
+      "epoch": 220.93,
+      "learning_rate": 9.783233595542923e-05,
+      "loss": 0.3574,
+      "step": 36600
+    },
+    {
+      "epoch": 221.05,
+      "learning_rate": 9.781924538099456e-05,
+      "loss": 0.3257,
+      "step": 36620
+    },
+    {
+      "epoch": 221.17,
+      "learning_rate": 9.78061162794576e-05,
+      "loss": 0.2748,
+      "step": 36640
+    },
+    {
+      "epoch": 221.29,
+      "learning_rate": 9.779294866139624e-05,
+      "loss": 0.293,
+      "step": 36660
+    },
+    {
+      "epoch": 221.41,
+      "learning_rate": 9.777974253741938e-05,
+      "loss": 0.3191,
+      "step": 36680
+    },
+    {
+      "epoch": 221.53,
+      "learning_rate": 9.776649791816698e-05,
+      "loss": 0.3214,
+      "step": 36700
+    },
+    {
+      "epoch": 221.65,
+      "learning_rate": 9.775321481430996e-05,
+      "loss": 0.3387,
+      "step": 36720
+    },
+    {
+      "epoch": 221.77,
+      "learning_rate": 9.773989323655028e-05,
+      "loss": 0.3565,
+      "step": 36740
+    },
+    {
+      "epoch": 221.89,
+      "learning_rate": 9.772653319562093e-05,
+      "loss": 0.3568,
+      "step": 36760
+    },
+    {
+      "epoch": 222.01,
+      "learning_rate": 9.771313470228583e-05,
+      "loss": 0.3475,
+      "step": 36780
+    },
+    {
+      "epoch": 222.13,
+      "learning_rate": 9.769969776733988e-05,
+      "loss": 0.2739,
+      "step": 36800
+    },
+    {
+      "epoch": 222.25,
+      "learning_rate": 9.768622240160901e-05,
+      "loss": 0.2889,
+      "step": 36820
+    },
+    {
+      "epoch": 222.37,
+      "learning_rate": 9.767270861595005e-05,
+      "loss": 0.3142,
+      "step": 36840
+    },
+    {
+      "epoch": 222.49,
+      "learning_rate": 9.765915642125087e-05,
+      "loss": 0.3283,
+      "step": 36860
+    },
+    {
+      "epoch": 222.62,
+      "learning_rate": 9.764556582843016e-05,
+      "loss": 0.3417,
+      "step": 36880
+    },
+    {
+      "epoch": 222.74,
+      "learning_rate": 9.763193684843765e-05,
+      "loss": 0.3398,
+      "step": 36900
+    },
+    {
+      "epoch": 222.86,
+      "learning_rate": 9.761826949225397e-05,
+      "loss": 0.3485,
+      "step": 36920
+    },
+    {
+      "epoch": 222.98,
+      "learning_rate": 9.760456377089062e-05,
+      "loss": 0.3611,
+      "step": 36940
+    },
+    {
+      "epoch": 223.1,
+      "learning_rate": 9.759081969539012e-05,
+      "loss": 0.2758,
+      "step": 36960
+    },
+    {
+      "epoch": 223.22,
+      "learning_rate": 9.757703727682574e-05,
+      "loss": 0.2811,
+      "step": 36980
+    },
+    {
+      "epoch": 223.34,
+      "learning_rate": 9.75632165263018e-05,
+      "loss": 0.2951,
+      "step": 37000
+    },
+    {
+      "epoch": 223.34,
+      "eval_accuracy": 0.0012827098418468652,
+      "eval_loss": 16.308704376220703,
+      "eval_runtime": 18.151,
+      "eval_samples_per_second": 112.391,
+      "eval_steps_per_second": 2.369,
+      "step": 37000
+    },
+    {
+      "epoch": 223.46,
+      "learning_rate": 9.75493574549534e-05,
+      "loss": 0.3059,
+      "step": 37020
+    },
+    {
+      "epoch": 223.58,
+      "learning_rate": 9.753546007394654e-05,
+      "loss": 0.3237,
+      "step": 37040
+    },
+    {
+      "epoch": 223.7,
+      "learning_rate": 9.752152439447808e-05,
+      "loss": 0.3246,
+      "step": 37060
+    },
+    {
+      "epoch": 223.82,
+      "learning_rate": 9.750755042777577e-05,
+      "loss": 0.3269,
+      "step": 37080
+    },
+    {
+      "epoch": 223.94,
+      "learning_rate": 9.749353818509817e-05,
+      "loss": 0.3511,
+      "step": 37100
+    },
+    {
+      "epoch": 224.06,
+      "learning_rate": 9.74794876777347e-05,
+      "loss": 0.3038,
+      "step": 37120
+    },
+    {
+      "epoch": 224.19,
+      "learning_rate": 9.746539891700556e-05,
+      "loss": 0.2703,
+      "step": 37140
+    },
+    {
+      "epoch": 224.31,
+      "learning_rate": 9.745127191426187e-05,
+      "loss": 0.2879,
+      "step": 37160
+    },
+    {
+      "epoch": 224.43,
+      "learning_rate": 9.743710668088545e-05,
+      "loss": 0.2977,
+      "step": 37180
+    },
+    {
+      "epoch": 224.55,
+      "learning_rate": 9.7422903228289e-05,
+      "loss": 0.3238,
+      "step": 37200
+    },
+    {
+      "epoch": 224.67,
+      "learning_rate": 9.7408661567916e-05,
+      "loss": 0.3337,
+      "step": 37220
+    },
+    {
+      "epoch": 224.79,
+      "learning_rate": 9.739438171124064e-05,
+      "loss": 0.338,
+      "step": 37240
+    },
+    {
+      "epoch": 224.91,
+      "learning_rate": 9.7380063669768e-05,
+      "loss": 0.3417,
+      "step": 37260
+    },
+    {
+      "epoch": 225.03,
+      "learning_rate": 9.736570745503386e-05,
+      "loss": 0.3337,
+      "step": 37280
+    },
+    {
+      "epoch": 225.15,
+      "learning_rate": 9.735131307860475e-05,
+      "loss": 0.2796,
+      "step": 37300
+    },
+    {
+      "epoch": 225.27,
+      "learning_rate": 9.733688055207794e-05,
+      "loss": 0.2922,
+      "step": 37320
+    },
+    {
+      "epoch": 225.39,
+      "learning_rate": 9.732240988708148e-05,
+      "loss": 0.3005,
+      "step": 37340
+    },
+    {
+      "epoch": 225.51,
+      "learning_rate": 9.730790109527412e-05,
+      "loss": 0.3055,
+      "step": 37360
+    },
+    {
+      "epoch": 225.63,
+      "learning_rate": 9.729335418834534e-05,
+      "loss": 0.3104,
+      "step": 37380
+    },
+    {
+      "epoch": 225.75,
+      "learning_rate": 9.72787691780153e-05,
+      "loss": 0.3269,
+      "step": 37400
+    },
+    {
+      "epoch": 225.88,
+      "learning_rate": 9.726414607603492e-05,
+      "loss": 0.3465,
+      "step": 37420
+    },
+    {
+      "epoch": 226.0,
+      "learning_rate": 9.724948489418571e-05,
+      "loss": 0.3467,
+      "step": 37440
+    },
+    {
+      "epoch": 226.12,
+      "learning_rate": 9.723478564427995e-05,
+      "loss": 0.2534,
+      "step": 37460
+    },
+    {
+      "epoch": 226.24,
+      "learning_rate": 9.722004833816056e-05,
+      "loss": 0.273,
+      "step": 37480
+    },
+    {
+      "epoch": 226.36,
+      "learning_rate": 9.720527298770112e-05,
+      "loss": 0.2769,
+      "step": 37500
+    },
+    {
+      "epoch": 226.48,
+      "learning_rate": 9.719045960480588e-05,
+      "loss": 0.2951,
+      "step": 37520
+    },
+    {
+      "epoch": 226.6,
+      "learning_rate": 9.717560820140969e-05,
+      "loss": 0.3103,
+      "step": 37540
+    },
+    {
+      "epoch": 226.72,
+      "learning_rate": 9.716071878947805e-05,
+      "loss": 0.315,
+      "step": 37560
+    },
+    {
+      "epoch": 226.84,
+      "learning_rate": 9.714579138100712e-05,
+      "loss": 0.3223,
+      "step": 37580
+    },
+    {
+      "epoch": 226.96,
+      "learning_rate": 9.713082598802364e-05,
+      "loss": 0.3421,
+      "step": 37600
+    },
+    {
+      "epoch": 227.08,
+      "learning_rate": 9.711582262258493e-05,
+      "loss": 0.27,
+      "step": 37620
+    },
+    {
+      "epoch": 227.2,
+      "learning_rate": 9.710078129677896e-05,
+      "loss": 0.2533,
+      "step": 37640
+    },
+    {
+      "epoch": 227.32,
+      "learning_rate": 9.708570202272422e-05,
+      "loss": 0.2818,
+      "step": 37660
+    },
+    {
+      "epoch": 227.44,
+      "learning_rate": 9.707058481256985e-05,
+      "loss": 0.2896,
+      "step": 37680
+    },
+    {
+      "epoch": 227.57,
+      "learning_rate": 9.705542967849547e-05,
+      "loss": 0.3054,
+      "step": 37700
+    },
+    {
+      "epoch": 227.69,
+      "learning_rate": 9.704023663271136e-05,
+      "loss": 0.3231,
+      "step": 37720
+    },
+    {
+      "epoch": 227.81,
+      "learning_rate": 9.702500568745821e-05,
+      "loss": 0.325,
+      "step": 37740
+    },
+    {
+      "epoch": 227.93,
+      "learning_rate": 9.700973685500735e-05,
+      "loss": 0.3357,
+      "step": 37760
+    },
+    {
+      "epoch": 228.05,
+      "learning_rate": 9.69951963823665e-05,
+      "loss": 0.2986,
+      "step": 37780
+    },
+    {
+      "epoch": 228.17,
+      "learning_rate": 9.697985370529101e-05,
+      "loss": 0.2518,
+      "step": 37800
+    },
+    {
+      "epoch": 228.29,
+      "learning_rate": 9.696447317739598e-05,
+      "loss": 0.2729,
+      "step": 37820
+    },
+    {
+      "epoch": 228.41,
+      "learning_rate": 9.694905481107317e-05,
+      "loss": 0.2929,
+      "step": 37840
+    },
+    {
+      "epoch": 228.53,
+      "learning_rate": 9.693359861874491e-05,
+      "loss": 0.3086,
+      "step": 37860
+    },
+    {
+      "epoch": 228.65,
+      "learning_rate": 9.691810461286397e-05,
+      "loss": 0.3131,
+      "step": 37880
+    },
+    {
+      "epoch": 228.77,
+      "learning_rate": 9.690257280591358e-05,
+      "loss": 0.3224,
+      "step": 37900
+    },
+    {
+      "epoch": 228.89,
+      "learning_rate": 9.688700321040744e-05,
+      "loss": 0.3387,
+      "step": 37920
+    },
+    {
+      "epoch": 229.01,
+      "learning_rate": 9.687139583888972e-05,
+      "loss": 0.3309,
+      "step": 37940
+    },
+    {
+      "epoch": 229.13,
+      "learning_rate": 9.685575070393494e-05,
+      "loss": 0.2475,
+      "step": 37960
+    },
+    {
+      "epoch": 229.26,
+      "learning_rate": 9.684006781814814e-05,
+      "loss": 0.2696,
+      "step": 37980
+    },
+    {
+      "epoch": 229.38,
+      "learning_rate": 9.682434719416473e-05,
+      "loss": 0.28,
+      "step": 38000
+    },
+    {
+      "epoch": 229.38,
+      "eval_accuracy": 0.0013024666357143563,
+      "eval_loss": 16.655616760253906,
+      "eval_runtime": 18.1057,
+      "eval_samples_per_second": 112.672,
+      "eval_steps_per_second": 2.375,
+      "step": 38000
+    },
+    {
+      "epoch": 229.5,
+      "learning_rate": 9.680858884465054e-05,
+      "loss": 0.3057,
+      "step": 38020
+    },
+    {
+      "epoch": 229.62,
+      "learning_rate": 9.679279278230179e-05,
+      "loss": 0.3062,
+      "step": 38040
+    },
+    {
+      "epoch": 229.74,
+      "learning_rate": 9.677695901984507e-05,
+      "loss": 0.3179,
+      "step": 38060
+    },
+    {
+      "epoch": 229.86,
+      "learning_rate": 9.676108757003735e-05,
+      "loss": 0.3286,
+      "step": 38080
+    },
+    {
+      "epoch": 229.98,
+      "learning_rate": 9.6745178445666e-05,
+      "loss": 0.3341,
+      "step": 38100
+    },
+    {
+      "epoch": 230.1,
+      "learning_rate": 9.672923165954868e-05,
+      "loss": 0.2633,
+      "step": 38120
+    },
+    {
+      "epoch": 230.22,
+      "learning_rate": 9.671324722453345e-05,
+      "loss": 0.2591,
+      "step": 38140
+    },
+    {
+      "epoch": 230.34,
+      "learning_rate": 9.669722515349868e-05,
+      "loss": 0.267,
+      "step": 38160
+    },
+    {
+      "epoch": 230.46,
+      "learning_rate": 9.668116545935304e-05,
+      "loss": 0.285,
+      "step": 38180
+    },
+    {
+      "epoch": 230.58,
+      "learning_rate": 9.666506815503558e-05,
+      "loss": 0.3014,
+      "step": 38200
+    },
+    {
+      "epoch": 230.7,
+      "learning_rate": 9.664893325351555e-05,
+      "loss": 0.3129,
+      "step": 38220
+    },
+    {
+      "epoch": 230.82,
+      "learning_rate": 9.66327607677926e-05,
+      "loss": 0.3232,
+      "step": 38240
+    },
+    {
+      "epoch": 230.95,
+      "learning_rate": 9.661655071089657e-05,
+      "loss": 0.3232,
+      "step": 38260
+    },
+    {
+      "epoch": 231.07,
+      "learning_rate": 9.660030309588764e-05,
+      "loss": 0.2646,
+      "step": 38280
+    },
+    {
+      "epoch": 231.19,
+      "learning_rate": 9.65840179358562e-05,
+      "loss": 0.2464,
+      "step": 38300
+    },
+    {
+      "epoch": 231.31,
+      "learning_rate": 9.656769524392292e-05,
+      "loss": 0.2627,
+      "step": 38320
+    },
+    {
+      "epoch": 231.43,
+      "learning_rate": 9.65513350332387e-05,
+      "loss": 0.2748,
+      "step": 38340
+    },
+    {
+      "epoch": 231.55,
+      "learning_rate": 9.653493731698467e-05,
+      "loss": 0.2917,
+      "step": 38360
+    },
+    {
+      "epoch": 231.67,
+      "learning_rate": 9.651850210837215e-05,
+      "loss": 0.3009,
+      "step": 38380
+    },
+    {
+      "epoch": 231.79,
+      "learning_rate": 9.65020294206427e-05,
+      "loss": 0.3167,
+      "step": 38400
+    },
+    {
+      "epoch": 231.91,
+      "learning_rate": 9.64855192670681e-05,
+      "loss": 0.325,
+      "step": 38420
+    },
+    {
+      "epoch": 232.03,
+      "learning_rate": 9.646897166095024e-05,
+      "loss": 0.2969,
+      "step": 38440
+    },
+    {
+      "epoch": 232.15,
+      "learning_rate": 9.645238661562124e-05,
+      "loss": 0.2487,
+      "step": 38460
+    },
+    {
+      "epoch": 232.27,
+      "learning_rate": 9.643576414444338e-05,
+      "loss": 0.2682,
+      "step": 38480
+    },
+    {
+      "epoch": 232.39,
+      "learning_rate": 9.641910426080908e-05,
+      "loss": 0.2745,
+      "step": 38500
+    },
+    {
+      "epoch": 232.52,
+      "learning_rate": 9.64024069781409e-05,
+      "loss": 0.2934,
+      "step": 38520
+    },
+    {
+      "epoch": 232.64,
+      "learning_rate": 9.638567230989155e-05,
+      "loss": 0.305,
+      "step": 38540
+    },
+    {
+      "epoch": 232.76,
+      "learning_rate": 9.636890026954385e-05,
+      "loss": 0.306,
+      "step": 38560
+    },
+    {
+      "epoch": 232.88,
+      "learning_rate": 9.635209087061072e-05,
+      "loss": 0.3161,
+      "step": 38580
+    },
+    {
+      "epoch": 233.0,
+      "learning_rate": 9.63352441266352e-05,
+      "loss": 0.3189,
+      "step": 38600
+    },
+    {
+      "epoch": 233.12,
+      "learning_rate": 9.63183600511904e-05,
+      "loss": 0.2318,
+      "step": 38620
+    },
+    {
+      "epoch": 233.24,
+      "learning_rate": 9.630143865787951e-05,
+      "loss": 0.2566,
+      "step": 38640
+    },
+    {
+      "epoch": 233.36,
+      "learning_rate": 9.62844799603358e-05,
+      "loss": 0.2733,
+      "step": 38660
+    },
+    {
+      "epoch": 233.48,
+      "learning_rate": 9.62674839722226e-05,
+      "loss": 0.2854,
+      "step": 38680
+    },
+    {
+      "epoch": 233.6,
+      "learning_rate": 9.625045070723324e-05,
+      "loss": 0.2851,
+      "step": 38700
+    },
+    {
+      "epoch": 233.72,
+      "learning_rate": 9.623338017909113e-05,
+      "loss": 0.2984,
+      "step": 38720
+    },
+    {
+      "epoch": 233.84,
+      "learning_rate": 9.621627240154968e-05,
+      "loss": 0.3112,
+      "step": 38740
+    },
+    {
+      "epoch": 233.96,
+      "learning_rate": 9.619912738839233e-05,
+      "loss": 0.3176,
+      "step": 38760
+    },
+    {
+      "epoch": 234.08,
+      "learning_rate": 9.61819451534325e-05,
+      "loss": 0.257,
+      "step": 38780
+    },
+    {
+      "epoch": 234.21,
+      "learning_rate": 9.61647257105136e-05,
+      "loss": 0.2461,
+      "step": 38800
+    },
+    {
+      "epoch": 234.33,
+      "learning_rate": 9.614746907350904e-05,
+      "loss": 0.2659,
+      "step": 38820
+    },
+    {
+      "epoch": 234.45,
+      "learning_rate": 9.613017525632215e-05,
+      "loss": 0.2738,
+      "step": 38840
+    },
+    {
+      "epoch": 234.57,
+      "learning_rate": 9.611284427288628e-05,
+      "loss": 0.2759,
+      "step": 38860
+    },
+    {
+      "epoch": 234.69,
+      "learning_rate": 9.609547613716468e-05,
+      "loss": 0.288,
+      "step": 38880
+    },
+    {
+      "epoch": 234.81,
+      "learning_rate": 9.607807086315055e-05,
+      "loss": 0.2997,
+      "step": 38900
+    },
+    {
+      "epoch": 234.93,
+      "learning_rate": 9.606062846486698e-05,
+      "loss": 0.3054,
+      "step": 38920
+    },
+    {
+      "epoch": 235.05,
+      "learning_rate": 9.604314895636702e-05,
+      "loss": 0.2719,
+      "step": 38940
+    },
+    {
+      "epoch": 235.17,
+      "learning_rate": 9.602563235173356e-05,
+      "loss": 0.2404,
+      "step": 38960
+    },
+    {
+      "epoch": 235.29,
+      "learning_rate": 9.600807866507946e-05,
+      "loss": 0.2632,
+      "step": 38980
+    },
+    {
+      "epoch": 235.41,
+      "learning_rate": 9.59904879105474e-05,
+      "loss": 0.2688,
+      "step": 39000
+    },
+    {
+      "epoch": 235.41,
+      "eval_accuracy": 0.0012797463227667415,
+      "eval_loss": 16.930341720581055,
+      "eval_runtime": 18.3698,
+      "eval_samples_per_second": 111.052,
+      "eval_steps_per_second": 2.341,
+      "step": 39000
+    },
+    {
+      "epoch": 235.53,
+      "learning_rate": 9.597286010230989e-05,
+      "loss": 0.2756,
+      "step": 39020
+    },
+    {
+      "epoch": 235.65,
+      "learning_rate": 9.595519525456937e-05,
+      "loss": 0.2901,
+      "step": 39040
+    },
+    {
+      "epoch": 235.77,
+      "learning_rate": 9.593749338155809e-05,
+      "loss": 0.3015,
+      "step": 39060
+    },
+    {
+      "epoch": 235.9,
+      "learning_rate": 9.591975449753812e-05,
+      "loss": 0.3011,
+      "step": 39080
+    },
+    {
+      "epoch": 236.02,
+      "learning_rate": 9.590197861680137e-05,
+      "loss": 0.2924,
+      "step": 39100
+    },
+    {
+      "epoch": 236.14,
+      "learning_rate": 9.588416575366953e-05,
+      "loss": 0.2287,
+      "step": 39120
+    },
+    {
+      "epoch": 236.26,
+      "learning_rate": 9.58663159224941e-05,
+      "loss": 0.2561,
+      "step": 39140
+    },
+    {
+      "epoch": 236.38,
+      "learning_rate": 9.58484291376564e-05,
+      "loss": 0.269,
+      "step": 39160
+    },
+    {
+      "epoch": 236.5,
+      "learning_rate": 9.583050541356747e-05,
+      "loss": 0.2768,
+      "step": 39180
+    },
+    {
+      "epoch": 236.62,
+      "learning_rate": 9.58125447646681e-05,
+      "loss": 0.2751,
+      "step": 39200
+    },
+    {
+      "epoch": 236.74,
+      "learning_rate": 9.579454720542889e-05,
+      "loss": 0.2796,
+      "step": 39220
+    },
+    {
+      "epoch": 236.86,
+      "learning_rate": 9.577651275035015e-05,
+      "loss": 0.294,
+      "step": 39240
+    },
+    {
+      "epoch": 236.98,
+      "learning_rate": 9.575844141396191e-05,
+      "loss": 0.3024,
+      "step": 39260
+    },
+    {
+      "epoch": 237.1,
+      "learning_rate": 9.574033321082392e-05,
+      "loss": 0.225,
+      "step": 39280
+    },
+    {
+      "epoch": 237.22,
+      "learning_rate": 9.572218815552565e-05,
+      "loss": 0.2385,
+      "step": 39300
+    },
+    {
+      "epoch": 237.34,
+      "learning_rate": 9.57040062626862e-05,
+      "loss": 0.2572,
+      "step": 39320
+    },
+    {
+      "epoch": 237.46,
+      "learning_rate": 9.568578754695442e-05,
+      "loss": 0.2704,
+      "step": 39340
+    },
+    {
+      "epoch": 237.59,
+      "learning_rate": 9.566753202300881e-05,
+      "loss": 0.2814,
+      "step": 39360
+    },
+    {
+      "epoch": 237.71,
+      "learning_rate": 9.56492397055575e-05,
+      "loss": 0.2763,
+      "step": 39380
+    },
+    {
+      "epoch": 237.83,
+      "learning_rate": 9.563091060933829e-05,
+      "loss": 0.2887,
+      "step": 39400
+    },
+    {
+      "epoch": 237.95,
+      "learning_rate": 9.561254474911858e-05,
+      "loss": 0.3002,
+      "step": 39420
+    },
+    {
+      "epoch": 238.07,
+      "learning_rate": 9.559414213969543e-05,
+      "loss": 0.248,
+      "step": 39440
+    },
+    {
+      "epoch": 238.19,
+      "learning_rate": 9.55757027958955e-05,
+      "loss": 0.2257,
+      "step": 39460
+    },
+    {
+      "epoch": 238.31,
+      "learning_rate": 9.555722673257501e-05,
+      "loss": 0.2484,
+      "step": 39480
+    },
+    {
+      "epoch": 238.43,
+      "learning_rate": 9.55387139646198e-05,
+      "loss": 0.2583,
+      "step": 39500
+    },
+    {
+      "epoch": 238.55,
+      "learning_rate": 9.552016450694529e-05,
+      "loss": 0.2678,
+      "step": 39520
+    },
+    {
+      "epoch": 238.67,
+      "learning_rate": 9.550157837449643e-05,
+      "loss": 0.277,
+      "step": 39540
+    },
+    {
+      "epoch": 238.79,
+      "learning_rate": 9.548295558224771e-05,
+      "loss": 0.289,
+      "step": 39560
+    },
+    {
+      "epoch": 238.91,
+      "learning_rate": 9.546429614520322e-05,
+      "loss": 0.2988,
+      "step": 39580
+    },
+    {
+      "epoch": 239.03,
+      "learning_rate": 9.54456000783965e-05,
+      "loss": 0.2735,
+      "step": 39600
+    },
+    {
+      "epoch": 239.15,
+      "learning_rate": 9.542686739689065e-05,
+      "loss": 0.2326,
+      "step": 39620
+    },
+    {
+      "epoch": 239.28,
+      "learning_rate": 9.540809811577823e-05,
+      "loss": 0.2474,
+      "step": 39640
+    },
+    {
+      "epoch": 239.4,
+      "learning_rate": 9.538929225018133e-05,
+      "loss": 0.2511,
+      "step": 39660
+    },
+    {
+      "epoch": 239.52,
+      "learning_rate": 9.537044981525151e-05,
+      "loss": 0.2598,
+      "step": 39680
+    },
+    {
+      "epoch": 239.64,
+      "learning_rate": 9.535157082616975e-05,
+      "loss": 0.2636,
+      "step": 39700
+    },
+    {
+      "epoch": 239.76,
+      "learning_rate": 9.533265529814653e-05,
+      "loss": 0.2777,
+      "step": 39720
+    },
+    {
+      "epoch": 239.88,
+      "learning_rate": 9.531370324642175e-05,
+      "loss": 0.291,
+      "step": 39740
+    },
+    {
+      "epoch": 240.0,
+      "learning_rate": 9.529471468626472e-05,
+      "loss": 0.2991,
+      "step": 39760
+    },
+    {
+      "epoch": 240.12,
+      "learning_rate": 9.527568963297418e-05,
+      "loss": 0.221,
+      "step": 39780
+    },
+    {
+      "epoch": 240.24,
+      "learning_rate": 9.525662810187827e-05,
+      "loss": 0.2425,
+      "step": 39800
+    },
+    {
+      "epoch": 240.36,
+      "learning_rate": 9.523753010833456e-05,
+      "loss": 0.2525,
+      "step": 39820
+    },
+    {
+      "epoch": 240.48,
+      "learning_rate": 9.521839566772993e-05,
+      "loss": 0.254,
+      "step": 39840
+    },
+    {
+      "epoch": 240.6,
+      "learning_rate": 9.520018420410632e-05,
+      "loss": 0.2664,
+      "step": 39860
+    },
+    {
+      "epoch": 240.72,
+      "learning_rate": 9.518097873610068e-05,
+      "loss": 0.2762,
+      "step": 39880
+    },
+    {
+      "epoch": 240.85,
+      "learning_rate": 9.516173686659653e-05,
+      "loss": 0.2776,
+      "step": 39900
+    },
+    {
+      "epoch": 240.97,
+      "learning_rate": 9.51424586110967e-05,
+      "loss": 0.2882,
+      "step": 39920
+    },
+    {
+      "epoch": 241.09,
+      "learning_rate": 9.512314398513335e-05,
+      "loss": 0.2368,
+      "step": 39940
+    },
+    {
+      "epoch": 241.21,
+      "learning_rate": 9.510379300426791e-05,
+      "loss": 0.234,
+      "step": 39960
+    },
+    {
+      "epoch": 241.33,
+      "learning_rate": 9.508440568409111e-05,
+      "loss": 0.2474,
+      "step": 39980
+    },
+    {
+      "epoch": 241.45,
+      "learning_rate": 9.506498204022297e-05,
+      "loss": 0.2582,
+      "step": 40000
+    },
+    {
+      "epoch": 241.45,
+      "eval_accuracy": 0.001275794963993243,
+      "eval_loss": 17.22085189819336,
+      "eval_runtime": 18.3419,
+      "eval_samples_per_second": 111.221,
+      "eval_steps_per_second": 2.344,
+      "step": 40000
+    },
+    {
+      "epoch": 241.57,
+      "learning_rate": 9.504552208831279e-05,
+      "loss": 0.2672,
+      "step": 40020
+    },
+    {
+      "epoch": 241.69,
+      "learning_rate": 9.502602584403908e-05,
+      "loss": 0.2703,
+      "step": 40040
+    },
+    {
+      "epoch": 241.81,
+      "learning_rate": 9.50064933231096e-05,
+      "loss": 0.2776,
+      "step": 40060
+    },
+    {
+      "epoch": 241.93,
+      "learning_rate": 9.498692454126137e-05,
+      "loss": 0.2848,
+      "step": 40080
+    },
+    {
+      "epoch": 242.05,
+      "learning_rate": 9.496731951426059e-05,
+      "loss": 0.2521,
+      "step": 40100
+    },
+    {
+      "epoch": 242.17,
+      "learning_rate": 9.494767825790269e-05,
+      "loss": 0.2219,
+      "step": 40120
+    },
+    {
+      "epoch": 242.29,
+      "learning_rate": 9.492800078801228e-05,
+      "loss": 0.2415,
+      "step": 40140
+    },
+    {
+      "epoch": 242.41,
+      "learning_rate": 9.490828712044312e-05,
+      "loss": 0.2456,
+      "step": 40160
+    },
+    {
+      "epoch": 242.54,
+      "learning_rate": 9.488853727107818e-05,
+      "loss": 0.254,
+      "step": 40180
+    },
+    {
+      "epoch": 242.66,
+      "learning_rate": 9.486875125582955e-05,
+      "loss": 0.2649,
+      "step": 40200
+    },
+    {
+      "epoch": 242.78,
+      "learning_rate": 9.484892909063845e-05,
+      "loss": 0.2692,
+      "step": 40220
+    },
+    {
+      "epoch": 242.9,
+      "learning_rate": 9.482907079147524e-05,
+      "loss": 0.2733,
+      "step": 40240
+    },
+    {
+      "epoch": 243.02,
+      "learning_rate": 9.480917637433942e-05,
+      "loss": 0.2743,
+      "step": 40260
+    },
+    {
+      "epoch": 243.14,
+      "learning_rate": 9.478924585525955e-05,
+      "loss": 0.2128,
+      "step": 40280
+    },
+    {
+      "epoch": 243.26,
+      "learning_rate": 9.476927925029326e-05,
+      "loss": 0.2237,
+      "step": 40300
+    },
+    {
+      "epoch": 243.38,
+      "learning_rate": 9.474927657552733e-05,
+      "loss": 0.2399,
+      "step": 40320
+    },
+    {
+      "epoch": 243.5,
+      "learning_rate": 9.472923784707752e-05,
+      "loss": 0.2543,
+      "step": 40340
+    },
+    {
+      "epoch": 243.62,
+      "learning_rate": 9.47091630810887e-05,
+      "loss": 0.2638,
+      "step": 40360
+    },
+    {
+      "epoch": 243.74,
+      "learning_rate": 9.468905229373472e-05,
+      "loss": 0.274,
+      "step": 40380
+    },
+    {
+      "epoch": 243.86,
+      "learning_rate": 9.466890550121849e-05,
+      "loss": 0.2794,
+      "step": 40400
+    },
+    {
+      "epoch": 243.98,
+      "learning_rate": 9.464872271977189e-05,
+      "loss": 0.2843,
+      "step": 40420
+    },
+    {
+      "epoch": 244.1,
+      "learning_rate": 9.462850396565588e-05,
+      "loss": 0.2136,
+      "step": 40440
+    },
+    {
+      "epoch": 244.23,
+      "learning_rate": 9.460824925516031e-05,
+      "loss": 0.2276,
+      "step": 40460
+    },
+    {
+      "epoch": 244.35,
+      "learning_rate": 9.458795860460403e-05,
+      "loss": 0.2371,
+      "step": 40480
+    },
+    {
+      "epoch": 244.47,
+      "learning_rate": 9.456763203033485e-05,
+      "loss": 0.2377,
+      "step": 40500
+    },
+    {
+      "epoch": 244.59,
+      "learning_rate": 9.454726954872954e-05,
+      "loss": 0.2525,
+      "step": 40520
+    },
+    {
+      "epoch": 244.71,
+      "learning_rate": 9.452687117619377e-05,
+      "loss": 0.2647,
+      "step": 40540
+    },
+    {
+      "epoch": 244.83,
+      "learning_rate": 9.450643692916216e-05,
+      "loss": 0.2757,
+      "step": 40560
+    },
+    {
+      "epoch": 244.95,
+      "learning_rate": 9.448596682409819e-05,
+      "loss": 0.2834,
+      "step": 40580
+    },
+    {
+      "epoch": 245.07,
+      "learning_rate": 9.446546087749425e-05,
+      "loss": 0.2351,
+      "step": 40600
+    },
+    {
+      "epoch": 245.19,
+      "learning_rate": 9.444491910587163e-05,
+      "loss": 0.214,
+      "step": 40620
+    },
+    {
+      "epoch": 245.31,
+      "learning_rate": 9.442434152578046e-05,
+      "loss": 0.2268,
+      "step": 40640
+    },
+    {
+      "epoch": 245.43,
+      "learning_rate": 9.440372815379973e-05,
+      "loss": 0.2443,
+      "step": 40660
+    },
+    {
+      "epoch": 245.55,
+      "learning_rate": 9.438307900653724e-05,
+      "loss": 0.2506,
+      "step": 40680
+    },
+    {
+      "epoch": 245.67,
+      "learning_rate": 9.436239410062966e-05,
+      "loss": 0.2556,
+      "step": 40700
+    },
+    {
+      "epoch": 245.79,
+      "learning_rate": 9.434167345274244e-05,
+      "loss": 0.2669,
+      "step": 40720
+    },
+    {
+      "epoch": 245.92,
+      "learning_rate": 9.432091707956981e-05,
+      "loss": 0.2705,
+      "step": 40740
+    },
+    {
+      "epoch": 246.04,
+      "learning_rate": 9.430012499783482e-05,
+      "loss": 0.2517,
+      "step": 40760
+    },
+    {
+      "epoch": 246.16,
+      "learning_rate": 9.42792972242893e-05,
+      "loss": 0.2108,
+      "step": 40780
+    },
+    {
+      "epoch": 246.28,
+      "learning_rate": 9.425843377571377e-05,
+      "loss": 0.2259,
+      "step": 40800
+    },
+    {
+      "epoch": 246.4,
+      "learning_rate": 9.423753466891754e-05,
+      "loss": 0.2342,
+      "step": 40820
+    },
+    {
+      "epoch": 246.52,
+      "learning_rate": 9.421659992073866e-05,
+      "loss": 0.2451,
+      "step": 40840
+    },
+    {
+      "epoch": 246.64,
+      "learning_rate": 9.419562954804388e-05,
+      "loss": 0.2526,
+      "step": 40860
+    },
+    {
+      "epoch": 246.76,
+      "learning_rate": 9.417462356772863e-05,
+      "loss": 0.2691,
+      "step": 40880
+    },
+    {
+      "epoch": 246.88,
+      "learning_rate": 9.415358199671707e-05,
+      "loss": 0.2647,
+      "step": 40900
+    },
+    {
+      "epoch": 247.0,
+      "learning_rate": 9.4132504851962e-05,
+      "loss": 0.2669,
+      "step": 40920
+    },
+    {
+      "epoch": 247.12,
+      "learning_rate": 9.411139215044491e-05,
+      "loss": 0.1954,
+      "step": 40940
+    },
+    {
+      "epoch": 247.24,
+      "learning_rate": 9.409024390917591e-05,
+      "loss": 0.2133,
+      "step": 40960
+    },
+    {
+      "epoch": 247.36,
+      "learning_rate": 9.406906014519374e-05,
+      "loss": 0.228,
+      "step": 40980
+    },
+    {
+      "epoch": 247.48,
+      "learning_rate": 9.404784087556582e-05,
+      "loss": 0.238,
+      "step": 41000
+    },
+    {
+      "epoch": 247.48,
+      "eval_accuracy": 0.0012669044067528722,
+      "eval_loss": 17.53106689453125,
+      "eval_runtime": 18.3103,
+      "eval_samples_per_second": 111.413,
+      "eval_steps_per_second": 2.348,
+      "step": 41000
+    },
+    {
+      "epoch": 247.61,
+      "learning_rate": 9.402658611738809e-05,
+      "loss": 0.2404,
+      "step": 41020
+    },
+    {
+      "epoch": 247.73,
+      "learning_rate": 9.400529588778513e-05,
+      "loss": 0.246,
+      "step": 41040
+    },
+    {
+      "epoch": 247.85,
+      "learning_rate": 9.398397020391012e-05,
+      "loss": 0.2531,
+      "step": 41060
+    },
+    {
+      "epoch": 247.97,
+      "learning_rate": 9.396260908294476e-05,
+      "loss": 0.2728,
+      "step": 41080
+    },
+    {
+      "epoch": 248.09,
+      "learning_rate": 9.394121254209932e-05,
+      "loss": 0.2107,
+      "step": 41100
+    },
+    {
+      "epoch": 248.21,
+      "learning_rate": 9.391978059861261e-05,
+      "loss": 0.2131,
+      "step": 41120
+    },
+    {
+      "epoch": 248.33,
+      "learning_rate": 9.389831326975196e-05,
+      "loss": 0.2284,
+      "step": 41140
+    },
+    {
+      "epoch": 248.45,
+      "learning_rate": 9.38768105728132e-05,
+      "loss": 0.2374,
+      "step": 41160
+    },
+    {
+      "epoch": 248.57,
+      "learning_rate": 9.385527252512068e-05,
+      "loss": 0.2467,
+      "step": 41180
+    },
+    {
+      "epoch": 248.69,
+      "learning_rate": 9.38336991440272e-05,
+      "loss": 0.2539,
+      "step": 41200
+    },
+    {
+      "epoch": 248.81,
+      "learning_rate": 9.381209044691405e-05,
+      "loss": 0.2612,
+      "step": 41220
+    },
+    {
+      "epoch": 248.93,
+      "learning_rate": 9.379044645119098e-05,
+      "loss": 0.2592,
+      "step": 41240
+    },
+    {
+      "epoch": 249.05,
+      "learning_rate": 9.376876717429615e-05,
+      "loss": 0.2225,
+      "step": 41260
+    },
+    {
+      "epoch": 249.18,
+      "learning_rate": 9.374705263369617e-05,
+      "loss": 0.2056,
+      "step": 41280
+    },
+    {
+      "epoch": 249.3,
+      "learning_rate": 9.372530284688603e-05,
+      "loss": 0.2156,
+      "step": 41300
+    },
+    {
+      "epoch": 249.42,
+      "learning_rate": 9.370351783138917e-05,
+      "loss": 0.2303,
+      "step": 41320
+    },
+    {
+      "epoch": 249.54,
+      "learning_rate": 9.36816976047574e-05,
+      "loss": 0.2363,
+      "step": 41340
+    },
+    {
+      "epoch": 249.66,
+      "learning_rate": 9.365984218457083e-05,
+      "loss": 0.2437,
+      "step": 41360
+    },
+    {
+      "epoch": 249.78,
+      "learning_rate": 9.3637951588438e-05,
+      "loss": 0.258,
+      "step": 41380
+    },
+    {
+      "epoch": 249.9,
+      "learning_rate": 9.361602583399578e-05,
+      "loss": 0.2636,
+      "step": 41400
+    },
+    {
+      "epoch": 250.02,
+      "learning_rate": 9.359406493890934e-05,
+      "loss": 0.2554,
+      "step": 41420
+    },
+    {
+      "epoch": 250.14,
+      "learning_rate": 9.357206892087219e-05,
+      "loss": 0.2029,
+      "step": 41440
+    },
+    {
+      "epoch": 250.26,
+      "learning_rate": 9.355003779760612e-05,
+      "loss": 0.2184,
+      "step": 41460
+    },
+    {
+      "epoch": 250.38,
+      "learning_rate": 9.352797158686119e-05,
+      "loss": 0.223,
+      "step": 41480
+    },
+    {
+      "epoch": 250.5,
+      "learning_rate": 9.350587030641578e-05,
+      "loss": 0.2358,
+      "step": 41500
+    },
+    {
+      "epoch": 250.62,
+      "learning_rate": 9.348373397407646e-05,
+      "loss": 0.2397,
+      "step": 41520
+    },
+    {
+      "epoch": 250.74,
+      "learning_rate": 9.346156260767809e-05,
+      "loss": 0.2508,
+      "step": 41540
+    },
+    {
+      "epoch": 250.87,
+      "learning_rate": 9.343935622508373e-05,
+      "loss": 0.2566,
+      "step": 41560
+    },
+    {
+      "epoch": 250.99,
+      "learning_rate": 9.341711484418468e-05,
+      "loss": 0.2656,
+      "step": 41580
+    },
+    {
+      "epoch": 251.11,
+      "learning_rate": 9.33948384829004e-05,
+      "loss": 0.1967,
+      "step": 41600
+    },
+    {
+      "epoch": 251.23,
+      "learning_rate": 9.337252715917857e-05,
+      "loss": 0.2034,
+      "step": 41620
+    },
+    {
+      "epoch": 251.35,
+      "learning_rate": 9.3350180890995e-05,
+      "loss": 0.2193,
+      "step": 41640
+    },
+    {
+      "epoch": 251.47,
+      "learning_rate": 9.332779969635369e-05,
+      "loss": 0.2288,
+      "step": 41660
+    },
+    {
+      "epoch": 251.59,
+      "learning_rate": 9.330538359328675e-05,
+      "loss": 0.2342,
+      "step": 41680
+    },
+    {
+      "epoch": 251.71,
+      "learning_rate": 9.328293259985446e-05,
+      "loss": 0.2438,
+      "step": 41700
+    },
+    {
+      "epoch": 251.83,
+      "learning_rate": 9.326044673414513e-05,
+      "loss": 0.2479,
+      "step": 41720
+    },
+    {
+      "epoch": 251.95,
+      "learning_rate": 9.323792601427526e-05,
+      "loss": 0.2524,
+      "step": 41740
+    },
+    {
+      "epoch": 252.07,
+      "learning_rate": 9.321537045838936e-05,
+      "loss": 0.2124,
+      "step": 41760
+    },
+    {
+      "epoch": 252.19,
+      "learning_rate": 9.319278008466006e-05,
+      "loss": 0.204,
+      "step": 41780
+    },
+    {
+      "epoch": 252.31,
+      "learning_rate": 9.317015491128797e-05,
+      "loss": 0.2212,
+      "step": 41800
+    },
+    {
+      "epoch": 252.43,
+      "learning_rate": 9.314749495650186e-05,
+      "loss": 0.2324,
+      "step": 41820
+    },
+    {
+      "epoch": 252.56,
+      "learning_rate": 9.312480023855836e-05,
+      "loss": 0.2432,
+      "step": 41840
+    },
+    {
+      "epoch": 252.68,
+      "learning_rate": 9.310207077574227e-05,
+      "loss": 0.2483,
+      "step": 41860
+    },
+    {
+      "epoch": 252.8,
+      "learning_rate": 9.307930658636626e-05,
+      "loss": 0.2507,
+      "step": 41880
+    },
+    {
+      "epoch": 252.92,
+      "learning_rate": 9.305650768877104e-05,
+      "loss": 0.2496,
+      "step": 41900
+    },
+    {
+      "epoch": 253.04,
+      "learning_rate": 9.303367410132526e-05,
+      "loss": 0.2302,
+      "step": 41920
+    },
+    {
+      "epoch": 253.16,
+      "learning_rate": 9.301080584242554e-05,
+      "loss": 0.1952,
+      "step": 41940
+    },
+    {
+      "epoch": 253.28,
+      "learning_rate": 9.298790293049641e-05,
+      "loss": 0.2053,
+      "step": 41960
+    },
+    {
+      "epoch": 253.4,
+      "learning_rate": 9.296496538399036e-05,
+      "loss": 0.2208,
+      "step": 41980
+    },
+    {
+      "epoch": 253.52,
+      "learning_rate": 9.294199322138773e-05,
+      "loss": 0.2261,
+      "step": 42000
+    },
+    {
+      "epoch": 253.52,
+      "eval_accuracy": 0.0012703618456796832,
+      "eval_loss": 17.773109436035156,
+      "eval_runtime": 18.2679,
+      "eval_samples_per_second": 111.672,
+      "eval_steps_per_second": 2.354,
+      "step": 42000
+    },
+    {
+      "epoch": 253.64,
+      "learning_rate": 9.29189864611968e-05,
+      "loss": 0.2384,
+      "step": 42020
+    },
+    {
+      "epoch": 253.76,
+      "learning_rate": 9.289594512195368e-05,
+      "loss": 0.2474,
+      "step": 42040
+    },
+    {
+      "epoch": 253.88,
+      "learning_rate": 9.287286922222239e-05,
+      "loss": 0.2504,
+      "step": 42060
+    },
+    {
+      "epoch": 254.0,
+      "learning_rate": 9.28497587805947e-05,
+      "loss": 0.2465,
+      "step": 42080
+    },
+    {
+      "epoch": 254.12,
+      "learning_rate": 9.282661381569036e-05,
+      "loss": 0.1798,
+      "step": 42100
+    },
+    {
+      "epoch": 254.25,
+      "learning_rate": 9.280343434615679e-05,
+      "loss": 0.2024,
+      "step": 42120
+    },
+    {
+      "epoch": 254.37,
+      "learning_rate": 9.27802203906693e-05,
+      "loss": 0.2118,
+      "step": 42140
+    },
+    {
+      "epoch": 254.49,
+      "learning_rate": 9.275697196793094e-05,
+      "loss": 0.2175,
+      "step": 42160
+    },
+    {
+      "epoch": 254.61,
+      "learning_rate": 9.273368909667255e-05,
+      "loss": 0.2289,
+      "step": 42180
+    },
+    {
+      "epoch": 254.73,
+      "learning_rate": 9.271037179565272e-05,
+      "loss": 0.2415,
+      "step": 42200
+    },
+    {
+      "epoch": 254.85,
+      "learning_rate": 9.268702008365777e-05,
+      "loss": 0.2444,
+      "step": 42220
+    },
+    {
+      "epoch": 254.97,
+      "learning_rate": 9.266363397950178e-05,
+      "loss": 0.2511,
+      "step": 42240
+    },
+    {
+      "epoch": 255.09,
+      "learning_rate": 9.264021350202646e-05,
+      "loss": 0.1973,
+      "step": 42260
+    },
+    {
+      "epoch": 255.21,
+      "learning_rate": 9.26167586701013e-05,
+      "loss": 0.203,
+      "step": 42280
+    },
+    {
+      "epoch": 255.33,
+      "learning_rate": 9.259326950262345e-05,
+      "loss": 0.2129,
+      "step": 42300
+    },
+    {
+      "epoch": 255.45,
+      "learning_rate": 9.256974601851767e-05,
+      "loss": 0.2219,
+      "step": 42320
+    },
+    {
+      "epoch": 255.57,
+      "learning_rate": 9.254618823673644e-05,
+      "loss": 0.2311,
+      "step": 42340
+    },
+    {
+      "epoch": 255.69,
+      "learning_rate": 9.252259617625983e-05,
+      "loss": 0.2318,
+      "step": 42360
+    },
+    {
+      "epoch": 255.81,
+      "learning_rate": 9.249896985609551e-05,
+      "loss": 0.2426,
+      "step": 42380
+    },
+    {
+      "epoch": 255.94,
+      "learning_rate": 9.247530929527881e-05,
+      "loss": 0.2493,
+      "step": 42400
+    },
+    {
+      "epoch": 256.06,
+      "learning_rate": 9.245161451287261e-05,
+      "loss": 0.213,
+      "step": 42420
+    },
+    {
+      "epoch": 256.18,
+      "learning_rate": 9.242788552796735e-05,
+      "loss": 0.1958,
+      "step": 42440
+    },
+    {
+      "epoch": 256.3,
+      "learning_rate": 9.240412235968108e-05,
+      "loss": 0.2034,
+      "step": 42460
+    },
+    {
+      "epoch": 256.42,
+      "learning_rate": 9.238032502715931e-05,
+      "loss": 0.2114,
+      "step": 42480
+    },
+    {
+      "epoch": 256.54,
+      "learning_rate": 9.235649354957517e-05,
+      "loss": 0.215,
+      "step": 42500
+    },
+    {
+      "epoch": 256.66,
+      "learning_rate": 9.233262794612923e-05,
+      "loss": 0.2266,
+      "step": 42520
+    },
+    {
+      "epoch": 256.78,
+      "learning_rate": 9.230872823604956e-05,
+      "loss": 0.2381,
+      "step": 42540
+    },
+    {
+      "epoch": 256.9,
+      "learning_rate": 9.228479443859173e-05,
+      "loss": 0.2428,
+      "step": 42560
+    },
+    {
+      "epoch": 257.02,
+      "learning_rate": 9.226082657303879e-05,
+      "loss": 0.2346,
+      "step": 42580
+    },
+    {
+      "epoch": 257.14,
+      "learning_rate": 9.22368246587012e-05,
+      "loss": 0.1906,
+      "step": 42600
+    },
+    {
+      "epoch": 257.26,
+      "learning_rate": 9.221278871491688e-05,
+      "loss": 0.1992,
+      "step": 42620
+    },
+    {
+      "epoch": 257.38,
+      "learning_rate": 9.218871876105113e-05,
+      "loss": 0.2092,
+      "step": 42640
+    },
+    {
+      "epoch": 257.51,
+      "learning_rate": 9.21646148164967e-05,
+      "loss": 0.215,
+      "step": 42660
+    },
+    {
+      "epoch": 257.63,
+      "learning_rate": 9.214047690067369e-05,
+      "loss": 0.2278,
+      "step": 42680
+    },
+    {
+      "epoch": 257.75,
+      "learning_rate": 9.21163050330296e-05,
+      "loss": 0.2331,
+      "step": 42700
+    },
+    {
+      "epoch": 257.87,
+      "learning_rate": 9.209209923303925e-05,
+      "loss": 0.2367,
+      "step": 42720
+    },
+    {
+      "epoch": 257.99,
+      "learning_rate": 9.206785952020482e-05,
+      "loss": 0.2468,
+      "step": 42740
+    },
+    {
+      "epoch": 258.11,
+      "learning_rate": 9.204358591405582e-05,
+      "loss": 0.1846,
+      "step": 42760
+    },
+    {
+      "epoch": 258.23,
+      "learning_rate": 9.201927843414904e-05,
+      "loss": 0.1935,
+      "step": 42780
+    },
+    {
+      "epoch": 258.35,
+      "learning_rate": 9.199493710006859e-05,
+      "loss": 0.2078,
+      "step": 42800
+    },
+    {
+      "epoch": 258.47,
+      "learning_rate": 9.197056193142583e-05,
+      "loss": 0.215,
+      "step": 42820
+    },
+    {
+      "epoch": 258.59,
+      "learning_rate": 9.194615294785942e-05,
+      "loss": 0.2207,
+      "step": 42840
+    },
+    {
+      "epoch": 258.71,
+      "learning_rate": 9.192171016903521e-05,
+      "loss": 0.2314,
+      "step": 42860
+    },
+    {
+      "epoch": 258.83,
+      "learning_rate": 9.189723361464631e-05,
+      "loss": 0.2314,
+      "step": 42880
+    },
+    {
+      "epoch": 258.95,
+      "learning_rate": 9.187272330441303e-05,
+      "loss": 0.2301,
+      "step": 42900
+    },
+    {
+      "epoch": 259.07,
+      "learning_rate": 9.184817925808291e-05,
+      "loss": 0.196,
+      "step": 42920
+    },
+    {
+      "epoch": 259.2,
+      "learning_rate": 9.182360149543063e-05,
+      "loss": 0.1875,
+      "step": 42940
+    },
+    {
+      "epoch": 259.32,
+      "learning_rate": 9.179899003625804e-05,
+      "loss": 0.1949,
+      "step": 42960
+    },
+    {
+      "epoch": 259.44,
+      "learning_rate": 9.177434490039418e-05,
+      "loss": 0.2029,
+      "step": 42980
+    },
+    {
+      "epoch": 259.56,
+      "learning_rate": 9.174966610769516e-05,
+      "loss": 0.21,
+      "step": 43000
+    },
+    {
+      "epoch": 259.56,
+      "eval_accuracy": 0.0012723375250664323,
+      "eval_loss": 18.020519256591797,
+      "eval_runtime": 18.2997,
+      "eval_samples_per_second": 111.477,
+      "eval_steps_per_second": 2.35,
+      "step": 43000
+    },
+    {
+      "epoch": 259.68,
+      "learning_rate": 9.172495367804423e-05,
+      "loss": 0.22,
+      "step": 43020
+    },
+    {
+      "epoch": 259.8,
+      "learning_rate": 9.170020763135179e-05,
+      "loss": 0.2269,
+      "step": 43040
+    },
+    {
+      "epoch": 259.92,
+      "learning_rate": 9.16766677673682e-05,
+      "loss": 0.2345,
+      "step": 43060
+    },
+    {
+      "epoch": 260.04,
+      "learning_rate": 9.165185622481447e-05,
+      "loss": 0.211,
+      "step": 43080
+    },
+    {
+      "epoch": 260.16,
+      "learning_rate": 9.162701112411252e-05,
+      "loss": 0.1793,
+      "step": 43100
+    },
+    {
+      "epoch": 260.28,
+      "learning_rate": 9.160213248527955e-05,
+      "loss": 0.1949,
+      "step": 43120
+    },
+    {
+      "epoch": 260.4,
+      "learning_rate": 9.15772203283599e-05,
+      "loss": 0.2091,
+      "step": 43140
+    },
+    {
+      "epoch": 260.52,
+      "learning_rate": 9.155227467342477e-05,
+      "loss": 0.2117,
+      "step": 43160
+    },
+    {
+      "epoch": 260.64,
+      "learning_rate": 9.152729554057249e-05,
+      "loss": 0.2188,
+      "step": 43180
+    },
+    {
+      "epoch": 260.76,
+      "learning_rate": 9.150228294992826e-05,
+      "loss": 0.2277,
+      "step": 43200
+    },
+    {
+      "epoch": 260.89,
+      "learning_rate": 9.147723692164427e-05,
+      "loss": 0.2348,
+      "step": 43220
+    },
+    {
+      "epoch": 261.01,
+      "learning_rate": 9.145215747589966e-05,
+      "loss": 0.2308,
+      "step": 43240
+    },
+    {
+      "epoch": 261.13,
+      "learning_rate": 9.142704463290049e-05,
+      "loss": 0.1738,
+      "step": 43260
+    },
+    {
+      "epoch": 261.25,
+      "learning_rate": 9.140189841287972e-05,
+      "loss": 0.1887,
+      "step": 43280
+    },
+    {
+      "epoch": 261.37,
+      "learning_rate": 9.137671883609721e-05,
+      "loss": 0.198,
+      "step": 43300
+    },
+    {
+      "epoch": 261.49,
+      "learning_rate": 9.13515059228397e-05,
+      "loss": 0.2048,
+      "step": 43320
+    },
+    {
+      "epoch": 261.61,
+      "learning_rate": 9.132625969342075e-05,
+      "loss": 0.2194,
+      "step": 43340
+    },
+    {
+      "epoch": 261.73,
+      "learning_rate": 9.13009801681808e-05,
+      "loss": 0.2212,
+      "step": 43360
+    },
+    {
+      "epoch": 261.85,
+      "learning_rate": 9.127566736748714e-05,
+      "loss": 0.2255,
+      "step": 43380
+    },
+    {
+      "epoch": 261.97,
+      "learning_rate": 9.12503213117338e-05,
+      "loss": 0.2328,
+      "step": 43400
+    },
+    {
+      "epoch": 262.09,
+      "learning_rate": 9.122494202134162e-05,
+      "loss": 0.1815,
+      "step": 43420
+    },
+    {
+      "epoch": 262.21,
+      "learning_rate": 9.11995295167583e-05,
+      "loss": 0.1861,
+      "step": 43440
+    },
+    {
+      "epoch": 262.33,
+      "learning_rate": 9.117408381845817e-05,
+      "loss": 0.19,
+      "step": 43460
+    },
+    {
+      "epoch": 262.45,
+      "learning_rate": 9.114860494694244e-05,
+      "loss": 0.1978,
+      "step": 43480
+    },
+    {
+      "epoch": 262.58,
+      "learning_rate": 9.112309292273891e-05,
+      "loss": 0.2094,
+      "step": 43500
+    },
+    {
+      "epoch": 262.7,
+      "learning_rate": 9.10975477664022e-05,
+      "loss": 0.2218,
+      "step": 43520
+    },
+    {
+      "epoch": 262.82,
+      "learning_rate": 9.107196949851356e-05,
+      "loss": 0.2242,
+      "step": 43540
+    },
+    {
+      "epoch": 262.94,
+      "learning_rate": 9.104635813968093e-05,
+      "loss": 0.2231,
+      "step": 43560
+    },
+    {
+      "epoch": 263.06,
+      "learning_rate": 9.102071371053896e-05,
+      "loss": 0.1983,
+      "step": 43580
+    },
+    {
+      "epoch": 263.18,
+      "learning_rate": 9.099503623174885e-05,
+      "loss": 0.1759,
+      "step": 43600
+    },
+    {
+      "epoch": 263.3,
+      "learning_rate": 9.096932572399852e-05,
+      "loss": 0.1882,
+      "step": 43620
+    },
+    {
+      "epoch": 263.42,
+      "learning_rate": 9.094358220800243e-05,
+      "loss": 0.2008,
+      "step": 43640
+    },
+    {
+      "epoch": 263.54,
+      "learning_rate": 9.09178057045017e-05,
+      "loss": 0.2054,
+      "step": 43660
+    },
+    {
+      "epoch": 263.66,
+      "learning_rate": 9.089199623426398e-05,
+      "loss": 0.2105,
+      "step": 43680
+    },
+    {
+      "epoch": 263.78,
+      "learning_rate": 9.08661538180835e-05,
+      "loss": 0.2176,
+      "step": 43700
+    },
+    {
+      "epoch": 263.9,
+      "learning_rate": 9.0840278476781e-05,
+      "loss": 0.2194,
+      "step": 43720
+    },
+    {
+      "epoch": 264.02,
+      "learning_rate": 9.081437023120381e-05,
+      "loss": 0.2127,
+      "step": 43740
+    },
+    {
+      "epoch": 264.14,
+      "learning_rate": 9.078842910222573e-05,
+      "loss": 0.1643,
+      "step": 43760
+    },
+    {
+      "epoch": 264.27,
+      "learning_rate": 9.076245511074703e-05,
+      "loss": 0.1765,
+      "step": 43780
+    },
+    {
+      "epoch": 264.39,
+      "learning_rate": 9.073644827769451e-05,
+      "loss": 0.1886,
+      "step": 43800
+    },
+    {
+      "epoch": 264.51,
+      "learning_rate": 9.071040862402141e-05,
+      "loss": 0.1927,
+      "step": 43820
+    },
+    {
+      "epoch": 264.63,
+      "learning_rate": 9.06843361707074e-05,
+      "loss": 0.2045,
+      "step": 43840
+    },
+    {
+      "epoch": 264.75,
+      "learning_rate": 9.065823093875858e-05,
+      "loss": 0.2113,
+      "step": 43860
+    },
+    {
+      "epoch": 264.87,
+      "learning_rate": 9.063209294920746e-05,
+      "loss": 0.2232,
+      "step": 43880
+    },
+    {
+      "epoch": 264.99,
+      "learning_rate": 9.060592222311296e-05,
+      "loss": 0.2263,
+      "step": 43900
+    },
+    {
+      "epoch": 265.11,
+      "learning_rate": 9.057971878156036e-05,
+      "loss": 0.1684,
+      "step": 43920
+    },
+    {
+      "epoch": 265.23,
+      "learning_rate": 9.05534826456613e-05,
+      "loss": 0.1869,
+      "step": 43940
+    },
+    {
+      "epoch": 265.35,
+      "learning_rate": 9.052721383655374e-05,
+      "loss": 0.1947,
+      "step": 43960
+    },
+    {
+      "epoch": 265.47,
+      "learning_rate": 9.050091237540201e-05,
+      "loss": 0.1987,
+      "step": 43980
+    },
+    {
+      "epoch": 265.59,
+      "learning_rate": 9.047457828339672e-05,
+      "loss": 0.2073,
+      "step": 44000
+    },
+    {
+      "epoch": 265.59,
+      "eval_accuracy": 0.001287649040313738,
+      "eval_loss": 18.269285202026367,
+      "eval_runtime": 18.1707,
+      "eval_samples_per_second": 112.268,
+      "eval_steps_per_second": 2.366,
+      "step": 44000
+    },
+    {
+      "epoch": 265.71,
+      "learning_rate": 9.044821158175476e-05,
+      "loss": 0.2091,
+      "step": 44020
+    },
+    {
+      "epoch": 265.84,
+      "learning_rate": 9.042181229171931e-05,
+      "loss": 0.2144,
+      "step": 44040
+    },
+    {
+      "epoch": 265.96,
+      "learning_rate": 9.03953804345598e-05,
+      "loss": 0.2186,
+      "step": 44060
+    },
+    {
+      "epoch": 266.08,
+      "learning_rate": 9.03689160315719e-05,
+      "loss": 0.1812,
+      "step": 44080
+    },
+    {
+      "epoch": 266.2,
+      "learning_rate": 9.03424191040775e-05,
+      "loss": 0.1765,
+      "step": 44100
+    },
+    {
+      "epoch": 266.32,
+      "learning_rate": 9.031588967342473e-05,
+      "loss": 0.1837,
+      "step": 44120
+    },
+    {
+      "epoch": 266.44,
+      "learning_rate": 9.028932776098781e-05,
+      "loss": 0.1924,
+      "step": 44140
+    },
+    {
+      "epoch": 266.56,
+      "learning_rate": 9.026273338816724e-05,
+      "loss": 0.2053,
+      "step": 44160
+    },
+    {
+      "epoch": 266.68,
+      "learning_rate": 9.02361065763896e-05,
+      "loss": 0.2062,
+      "step": 44180
+    },
+    {
+      "epoch": 266.8,
+      "learning_rate": 9.020944734710766e-05,
+      "loss": 0.2131,
+      "step": 44200
+    },
+    {
+      "epoch": 266.92,
+      "learning_rate": 9.018275572180024e-05,
+      "loss": 0.2233,
+      "step": 44220
+    },
+    {
+      "epoch": 267.04,
+      "learning_rate": 9.015603172197233e-05,
+      "loss": 0.1966,
+      "step": 44240
+    },
+    {
+      "epoch": 267.16,
+      "learning_rate": 9.012927536915497e-05,
+      "loss": 0.1641,
+      "step": 44260
+    },
+    {
+      "epoch": 267.28,
+      "learning_rate": 9.010248668490528e-05,
+      "loss": 0.1751,
+      "step": 44280
+    },
+    {
+      "epoch": 267.4,
+      "learning_rate": 9.007566569080639e-05,
+      "loss": 0.1921,
+      "step": 44300
+    },
+    {
+      "epoch": 267.53,
+      "learning_rate": 9.004881240846751e-05,
+      "loss": 0.1983,
+      "step": 44320
+    },
+    {
+      "epoch": 267.65,
+      "learning_rate": 9.002192685952385e-05,
+      "loss": 0.2062,
+      "step": 44340
+    },
+    {
+      "epoch": 267.77,
+      "learning_rate": 8.999500906563658e-05,
+      "loss": 0.2132,
+      "step": 44360
+    },
+    {
+      "epoch": 267.89,
+      "learning_rate": 8.996805904849292e-05,
+      "loss": 0.2181,
+      "step": 44380
+    },
+    {
+      "epoch": 268.01,
+      "learning_rate": 8.9941076829806e-05,
+      "loss": 0.224,
+      "step": 44400
+    },
+    {
+      "epoch": 268.13,
+      "learning_rate": 8.991406243131489e-05,
+      "loss": 0.1613,
+      "step": 44420
+    },
+    {
+      "epoch": 268.25,
+      "learning_rate": 8.98870158747846e-05,
+      "loss": 0.1759,
+      "step": 44440
+    },
+    {
+      "epoch": 268.37,
+      "learning_rate": 8.985993718200608e-05,
+      "loss": 0.1872,
+      "step": 44460
+    },
+    {
+      "epoch": 268.49,
+      "learning_rate": 8.983282637479614e-05,
+      "loss": 0.1948,
+      "step": 44480
+    },
+    {
+      "epoch": 268.61,
+      "learning_rate": 8.980568347499744e-05,
+      "loss": 0.2063,
+      "step": 44500
+    },
+    {
+      "epoch": 268.73,
+      "learning_rate": 8.977850850447855e-05,
+      "loss": 0.2108,
+      "step": 44520
+    },
+    {
+      "epoch": 268.85,
+      "learning_rate": 8.975130148513384e-05,
+      "loss": 0.2153,
+      "step": 44540
+    },
+    {
+      "epoch": 268.97,
+      "learning_rate": 8.972406243888351e-05,
+      "loss": 0.2124,
+      "step": 44560
+    },
+    {
+      "epoch": 269.09,
+      "learning_rate": 8.969679138767359e-05,
+      "loss": 0.1699,
+      "step": 44580
+    },
+    {
+      "epoch": 269.22,
+      "learning_rate": 8.966948835347586e-05,
+      "loss": 0.1748,
+      "step": 44600
+    },
+    {
+      "epoch": 269.34,
+      "learning_rate": 8.964215335828787e-05,
+      "loss": 0.1807,
+      "step": 44620
+    },
+    {
+      "epoch": 269.46,
+      "learning_rate": 8.961478642413294e-05,
+      "loss": 0.1875,
+      "step": 44640
+    },
+    {
+      "epoch": 269.58,
+      "learning_rate": 8.958738757306013e-05,
+      "loss": 0.2025,
+      "step": 44660
+    },
+    {
+      "epoch": 269.7,
+      "learning_rate": 8.955995682714417e-05,
+      "loss": 0.2071,
+      "step": 44680
+    },
+    {
+      "epoch": 269.82,
+      "learning_rate": 8.953249420848555e-05,
+      "loss": 0.2106,
+      "step": 44700
+    },
+    {
+      "epoch": 269.94,
+      "learning_rate": 8.950499973921038e-05,
+      "loss": 0.2192,
+      "step": 44720
+    },
+    {
+      "epoch": 270.06,
+      "learning_rate": 8.947747344147049e-05,
+      "loss": 0.183,
+      "step": 44740
+    },
+    {
+      "epoch": 270.18,
+      "learning_rate": 8.944991533744327e-05,
+      "loss": 0.163,
+      "step": 44760
+    },
+    {
+      "epoch": 270.3,
+      "learning_rate": 8.942232544933182e-05,
+      "loss": 0.1736,
+      "step": 44780
+    },
+    {
+      "epoch": 270.42,
+      "learning_rate": 8.93947037993648e-05,
+      "loss": 0.1891,
+      "step": 44800
+    },
+    {
+      "epoch": 270.54,
+      "learning_rate": 8.93670504097965e-05,
+      "loss": 0.195,
+      "step": 44820
+    },
+    {
+      "epoch": 270.66,
+      "learning_rate": 8.933936530290672e-05,
+      "loss": 0.2006,
+      "step": 44840
+    },
+    {
+      "epoch": 270.78,
+      "learning_rate": 8.931164850100086e-05,
+      "loss": 0.21,
+      "step": 44860
+    },
+    {
+      "epoch": 270.91,
+      "learning_rate": 8.928390002640989e-05,
+      "loss": 0.2132,
+      "step": 44880
+    },
+    {
+      "epoch": 271.03,
+      "learning_rate": 8.925611990149021e-05,
+      "loss": 0.1937,
+      "step": 44900
+    },
+    {
+      "epoch": 271.15,
+      "learning_rate": 8.922830814862377e-05,
+      "loss": 0.1565,
+      "step": 44920
+    },
+    {
+      "epoch": 271.27,
+      "learning_rate": 8.920046479021799e-05,
+      "loss": 0.1704,
+      "step": 44940
+    },
+    {
+      "epoch": 271.39,
+      "learning_rate": 8.91725898487058e-05,
+      "loss": 0.1823,
+      "step": 44960
+    },
+    {
+      "epoch": 271.51,
+      "learning_rate": 8.91446833465455e-05,
+      "loss": 0.1915,
+      "step": 44980
+    },
+    {
+      "epoch": 271.63,
+      "learning_rate": 8.911674530622088e-05,
+      "loss": 0.1976,
+      "step": 45000
+    },
+    {
+      "epoch": 271.63,
+      "eval_accuracy": 0.0012748071242998686,
+      "eval_loss": 18.4633846282959,
+      "eval_runtime": 18.1499,
+      "eval_samples_per_second": 112.397,
+      "eval_steps_per_second": 2.369,
+      "step": 45000
+    },
+    {
+      "epoch": 271.75,
+      "learning_rate": 8.908877575024107e-05,
+      "loss": 0.2002,
+      "step": 45020
+    },
+    {
+      "epoch": 271.87,
+      "learning_rate": 8.906077470114069e-05,
+      "loss": 0.2083,
+      "step": 45040
+    },
+    {
+      "epoch": 271.99,
+      "learning_rate": 8.903274218147963e-05,
+      "loss": 0.2086,
+      "step": 45060
+    },
+    {
+      "epoch": 272.11,
+      "learning_rate": 8.900467821384324e-05,
+      "loss": 0.1585,
+      "step": 45080
+    },
+    {
+      "epoch": 272.23,
+      "learning_rate": 8.897658282084211e-05,
+      "loss": 0.1718,
+      "step": 45100
+    },
+    {
+      "epoch": 272.35,
+      "learning_rate": 8.894845602511219e-05,
+      "loss": 0.1782,
+      "step": 45120
+    },
+    {
+      "epoch": 272.47,
+      "learning_rate": 8.892029784931476e-05,
+      "loss": 0.1915,
+      "step": 45140
+    },
+    {
+      "epoch": 272.6,
+      "learning_rate": 8.889210831613634e-05,
+      "loss": 0.1949,
+      "step": 45160
+    },
+    {
+      "epoch": 272.72,
+      "learning_rate": 8.886388744828872e-05,
+      "loss": 0.1967,
+      "step": 45180
+    },
+    {
+      "epoch": 272.84,
+      "learning_rate": 8.883563526850894e-05,
+      "loss": 0.1999,
+      "step": 45200
+    },
+    {
+      "epoch": 272.96,
+      "learning_rate": 8.88073517995593e-05,
+      "loss": 0.2079,
+      "step": 45220
+    },
+    {
+      "epoch": 273.08,
+      "learning_rate": 8.877903706422724e-05,
+      "loss": 0.1664,
+      "step": 45240
+    },
+    {
+      "epoch": 273.2,
+      "learning_rate": 8.875069108532546e-05,
+      "loss": 0.1624,
+      "step": 45260
+    },
+    {
+      "epoch": 273.32,
+      "learning_rate": 8.87223138856918e-05,
+      "loss": 0.1712,
+      "step": 45280
+    },
+    {
+      "epoch": 273.44,
+      "learning_rate": 8.869390548818923e-05,
+      "loss": 0.185,
+      "step": 45300
+    },
+    {
+      "epoch": 273.56,
+      "learning_rate": 8.866546591570592e-05,
+      "loss": 0.1908,
+      "step": 45320
+    },
+    {
+      "epoch": 273.68,
+      "learning_rate": 8.863699519115511e-05,
+      "loss": 0.199,
+      "step": 45340
+    },
+    {
+      "epoch": 273.8,
+      "learning_rate": 8.860849333747514e-05,
+      "loss": 0.199,
+      "step": 45360
+    },
+    {
+      "epoch": 273.92,
+      "learning_rate": 8.857996037762944e-05,
+      "loss": 0.2075,
+      "step": 45380
+    },
+    {
+      "epoch": 274.04,
+      "learning_rate": 8.855139633460649e-05,
+      "loss": 0.1818,
+      "step": 45400
+    },
+    {
+      "epoch": 274.16,
+      "learning_rate": 8.852280123141984e-05,
+      "loss": 0.1547,
+      "step": 45420
+    },
+    {
+      "epoch": 274.29,
+      "learning_rate": 8.849417509110805e-05,
+      "loss": 0.1717,
+      "step": 45440
+    },
+    {
+      "epoch": 274.41,
+      "learning_rate": 8.846551793673467e-05,
+      "loss": 0.1772,
+      "step": 45460
+    },
+    {
+      "epoch": 274.53,
+      "learning_rate": 8.843682979138825e-05,
+      "loss": 0.1837,
+      "step": 45480
+    },
+    {
+      "epoch": 274.65,
+      "learning_rate": 8.840811067818233e-05,
+      "loss": 0.1893,
+      "step": 45500
+    },
+    {
+      "epoch": 274.77,
+      "learning_rate": 8.837936062025538e-05,
+      "loss": 0.1939,
+      "step": 45520
+    },
+    {
+      "epoch": 274.89,
+      "learning_rate": 8.835057964077079e-05,
+      "loss": 0.1976,
+      "step": 45540
+    },
+    {
+      "epoch": 275.01,
+      "learning_rate": 8.832176776291688e-05,
+      "loss": 0.1941,
+      "step": 45560
+    },
+    {
+      "epoch": 275.13,
+      "learning_rate": 8.829292500990683e-05,
+      "loss": 0.1437,
+      "step": 45580
+    },
+    {
+      "epoch": 275.25,
+      "learning_rate": 8.826405140497878e-05,
+      "loss": 0.1559,
+      "step": 45600
+    },
+    {
+      "epoch": 275.37,
+      "learning_rate": 8.823514697139564e-05,
+      "loss": 0.166,
+      "step": 45620
+    },
+    {
+      "epoch": 275.49,
+      "learning_rate": 8.820621173244519e-05,
+      "loss": 0.1785,
+      "step": 45640
+    },
+    {
+      "epoch": 275.61,
+      "learning_rate": 8.817724571144004e-05,
+      "loss": 0.1846,
+      "step": 45660
+    },
+    {
+      "epoch": 275.73,
+      "learning_rate": 8.814824893171758e-05,
+      "loss": 0.1921,
+      "step": 45680
+    },
+    {
+      "epoch": 275.86,
+      "learning_rate": 8.811922141664e-05,
+      "loss": 0.1994,
+      "step": 45700
+    },
+    {
+      "epoch": 275.98,
+      "learning_rate": 8.809016318959424e-05,
+      "loss": 0.2001,
+      "step": 45720
+    },
+    {
+      "epoch": 276.1,
+      "learning_rate": 8.806107427399197e-05,
+      "loss": 0.1521,
+      "step": 45740
+    },
+    {
+      "epoch": 276.22,
+      "learning_rate": 8.803195469326964e-05,
+      "loss": 0.1565,
+      "step": 45760
+    },
+    {
+      "epoch": 276.34,
+      "learning_rate": 8.800280447088836e-05,
+      "loss": 0.1699,
+      "step": 45780
+    },
+    {
+      "epoch": 276.46,
+      "learning_rate": 8.797362363033392e-05,
+      "loss": 0.1753,
+      "step": 45800
+    },
+    {
+      "epoch": 276.58,
+      "learning_rate": 8.794441219511681e-05,
+      "loss": 0.1801,
+      "step": 45820
+    },
+    {
+      "epoch": 276.7,
+      "learning_rate": 8.791517018877216e-05,
+      "loss": 0.1878,
+      "step": 45840
+    },
+    {
+      "epoch": 276.82,
+      "learning_rate": 8.78858976348597e-05,
+      "loss": 0.191,
+      "step": 45860
+    },
+    {
+      "epoch": 276.94,
+      "learning_rate": 8.785659455696384e-05,
+      "loss": 0.1946,
+      "step": 45880
+    },
+    {
+      "epoch": 277.06,
+      "learning_rate": 8.782726097869349e-05,
+      "loss": 0.1662,
+      "step": 45900
+    },
+    {
+      "epoch": 277.18,
+      "learning_rate": 8.779789692368223e-05,
+      "loss": 0.1548,
+      "step": 45920
+    },
+    {
+      "epoch": 277.3,
+      "learning_rate": 8.776850241558814e-05,
+      "loss": 0.1647,
+      "step": 45940
+    },
+    {
+      "epoch": 277.42,
+      "learning_rate": 8.773907747809383e-05,
+      "loss": 0.1762,
+      "step": 45960
+    },
+    {
+      "epoch": 277.55,
+      "learning_rate": 8.770962213490643e-05,
+      "loss": 0.1846,
+      "step": 45980
+    },
+    {
+      "epoch": 277.67,
+      "learning_rate": 8.768013640975761e-05,
+      "loss": 0.1865,
+      "step": 46000
+    },
+    {
+      "epoch": 277.67,
+      "eval_accuracy": 0.0012427023342651954,
+      "eval_loss": 18.721508026123047,
+      "eval_runtime": 18.1526,
+      "eval_samples_per_second": 112.381,
+      "eval_steps_per_second": 2.369,
+      "step": 46000
+    },
+    {
+      "epoch": 277.79,
+      "learning_rate": 8.765062032640346e-05,
+      "loss": 0.1917,
+      "step": 46020
+    },
+    {
+      "epoch": 277.91,
+      "learning_rate": 8.762107390862455e-05,
+      "loss": 0.1932,
+      "step": 46040
+    },
+    {
+      "epoch": 278.03,
+      "learning_rate": 8.759149718022594e-05,
+      "loss": 0.1826,
+      "step": 46060
+    },
+    {
+      "epoch": 278.15,
+      "learning_rate": 8.756189016503702e-05,
+      "loss": 0.149,
+      "step": 46080
+    },
+    {
+      "epoch": 278.27,
+      "learning_rate": 8.753225288691165e-05,
+      "loss": 0.1608,
+      "step": 46100
+    },
+    {
+      "epoch": 278.39,
+      "learning_rate": 8.750258536972804e-05,
+      "loss": 0.1662,
+      "step": 46120
+    },
+    {
+      "epoch": 278.51,
+      "learning_rate": 8.747288763738877e-05,
+      "loss": 0.175,
+      "step": 46140
+    },
+    {
+      "epoch": 278.63,
+      "learning_rate": 8.744315971382078e-05,
+      "loss": 0.1816,
+      "step": 46160
+    },
+    {
+      "epoch": 278.75,
+      "learning_rate": 8.741340162297531e-05,
+      "loss": 0.1871,
+      "step": 46180
+    },
+    {
+      "epoch": 278.87,
+      "learning_rate": 8.738361338882792e-05,
+      "loss": 0.1913,
+      "step": 46200
+    },
+    {
+      "epoch": 278.99,
+      "learning_rate": 8.735379503537844e-05,
+      "loss": 0.1944,
+      "step": 46220
+    },
+    {
+      "epoch": 279.11,
+      "learning_rate": 8.732394658665101e-05,
+      "loss": 0.1461,
+      "step": 46240
+    },
+    {
+      "epoch": 279.24,
+      "learning_rate": 8.729406806669396e-05,
+      "loss": 0.1549,
+      "step": 46260
+    },
+    {
+      "epoch": 279.36,
+      "learning_rate": 8.726415949957987e-05,
+      "loss": 0.1621,
+      "step": 46280
+    },
+    {
+      "epoch": 279.48,
+      "learning_rate": 8.723422090940555e-05,
+      "loss": 0.1694,
+      "step": 46300
+    },
+    {
+      "epoch": 279.6,
+      "learning_rate": 8.720425232029198e-05,
+      "loss": 0.1767,
+      "step": 46320
+    },
+    {
+      "epoch": 279.72,
+      "learning_rate": 8.717425375638429e-05,
+      "loss": 0.1835,
+      "step": 46340
+    },
+    {
+      "epoch": 279.84,
+      "learning_rate": 8.714422524185181e-05,
+      "loss": 0.1895,
+      "step": 46360
+    },
+    {
+      "epoch": 279.96,
+      "learning_rate": 8.711416680088795e-05,
+      "loss": 0.1945,
+      "step": 46380
+    },
+    {
+      "epoch": 280.08,
+      "learning_rate": 8.708407845771027e-05,
+      "loss": 0.1564,
+      "step": 46400
+    },
+    {
+      "epoch": 280.2,
+      "learning_rate": 8.70539602365604e-05,
+      "loss": 0.1489,
+      "step": 46420
+    },
+    {
+      "epoch": 280.32,
+      "learning_rate": 8.702381216170403e-05,
+      "loss": 0.1611,
+      "step": 46440
+    },
+    {
+      "epoch": 280.44,
+      "learning_rate": 8.699363425743093e-05,
+      "loss": 0.1717,
+      "step": 46460
+    },
+    {
+      "epoch": 280.56,
+      "learning_rate": 8.69634265480549e-05,
+      "loss": 0.1738,
+      "step": 46480
+    },
+    {
+      "epoch": 280.68,
+      "learning_rate": 8.693318905791375e-05,
+      "loss": 0.1804,
+      "step": 46500
+    },
+    {
+      "epoch": 280.8,
+      "learning_rate": 8.690292181136924e-05,
+      "loss": 0.1828,
+      "step": 46520
+    },
+    {
+      "epoch": 280.93,
+      "learning_rate": 8.687262483280719e-05,
+      "loss": 0.193,
+      "step": 46540
+    },
+    {
+      "epoch": 281.05,
+      "learning_rate": 8.684229814663731e-05,
+      "loss": 0.173,
+      "step": 46560
+    },
+    {
+      "epoch": 281.17,
+      "learning_rate": 8.681194177729328e-05,
+      "loss": 0.1497,
+      "step": 46580
+    },
+    {
+      "epoch": 281.29,
+      "learning_rate": 8.678155574923265e-05,
+      "loss": 0.1582,
+      "step": 46600
+    },
+    {
+      "epoch": 281.41,
+      "learning_rate": 8.675114008693689e-05,
+      "loss": 0.1657,
+      "step": 46620
+    },
+    {
+      "epoch": 281.53,
+      "learning_rate": 8.672069481491141e-05,
+      "loss": 0.1739,
+      "step": 46640
+    },
+    {
+      "epoch": 281.65,
+      "learning_rate": 8.669021995768534e-05,
+      "loss": 0.1777,
+      "step": 46660
+    },
+    {
+      "epoch": 281.77,
+      "learning_rate": 8.665971553981175e-05,
+      "loss": 0.1774,
+      "step": 46680
+    },
+    {
+      "epoch": 281.89,
+      "learning_rate": 8.662918158586753e-05,
+      "loss": 0.1802,
+      "step": 46700
+    },
+    {
+      "epoch": 282.01,
+      "learning_rate": 8.65986181204533e-05,
+      "loss": 0.1809,
+      "step": 46720
+    },
+    {
+      "epoch": 282.13,
+      "learning_rate": 8.656802516819349e-05,
+      "loss": 0.14,
+      "step": 46740
+    },
+    {
+      "epoch": 282.25,
+      "learning_rate": 8.653740275373631e-05,
+      "loss": 0.153,
+      "step": 46760
+    },
+    {
+      "epoch": 282.37,
+      "learning_rate": 8.650675090175366e-05,
+      "loss": 0.1604,
+      "step": 46780
+    },
+    {
+      "epoch": 282.49,
+      "learning_rate": 8.647606963694122e-05,
+      "loss": 0.1675,
+      "step": 46800
+    },
+    {
+      "epoch": 282.62,
+      "learning_rate": 8.644535898401831e-05,
+      "loss": 0.168,
+      "step": 46820
+    },
+    {
+      "epoch": 282.74,
+      "learning_rate": 8.641461896772793e-05,
+      "loss": 0.1735,
+      "step": 46840
+    },
+    {
+      "epoch": 282.86,
+      "learning_rate": 8.638384961283679e-05,
+      "loss": 0.1833,
+      "step": 46860
+    },
+    {
+      "epoch": 282.98,
+      "learning_rate": 8.63530509441352e-05,
+      "loss": 0.1921,
+      "step": 46880
+    },
+    {
+      "epoch": 283.1,
+      "learning_rate": 8.632222298643706e-05,
+      "loss": 0.1446,
+      "step": 46900
+    },
+    {
+      "epoch": 283.22,
+      "learning_rate": 8.629136576457991e-05,
+      "loss": 0.1552,
+      "step": 46920
+    },
+    {
+      "epoch": 283.34,
+      "learning_rate": 8.626047930342488e-05,
+      "loss": 0.1601,
+      "step": 46940
+    },
+    {
+      "epoch": 283.46,
+      "learning_rate": 8.622956362785662e-05,
+      "loss": 0.1659,
+      "step": 46960
+    },
+    {
+      "epoch": 283.58,
+      "learning_rate": 8.619861876278332e-05,
+      "loss": 0.1737,
+      "step": 46980
+    },
+    {
+      "epoch": 283.7,
+      "learning_rate": 8.616764473313671e-05,
+      "loss": 0.1769,
+      "step": 47000
+    },
+    {
+      "epoch": 283.7,
+      "eval_accuracy": 0.0012644348075194357,
+      "eval_loss": 18.946748733520508,
+      "eval_runtime": 18.1402,
+      "eval_samples_per_second": 112.457,
+      "eval_steps_per_second": 2.37,
+      "step": 47000
+    },
+    {
+      "epoch": 283.82,
+      "learning_rate": 8.6136641563872e-05,
+      "loss": 0.1784,
+      "step": 47020
+    },
+    {
+      "epoch": 283.94,
+      "learning_rate": 8.610560927996788e-05,
+      "loss": 0.1857,
+      "step": 47040
+    },
+    {
+      "epoch": 284.06,
+      "learning_rate": 8.607610166559625e-05,
+      "loss": 0.1554,
+      "step": 47060
+    },
+    {
+      "epoch": 284.19,
+      "learning_rate": 8.60450126800791e-05,
+      "loss": 0.1436,
+      "step": 47080
+    },
+    {
+      "epoch": 284.31,
+      "learning_rate": 8.601389465374627e-05,
+      "loss": 0.1576,
+      "step": 47100
+    },
+    {
+      "epoch": 284.43,
+      "learning_rate": 8.598274761166903e-05,
+      "loss": 0.1629,
+      "step": 47120
+    },
+    {
+      "epoch": 284.55,
+      "learning_rate": 8.595157157894194e-05,
+      "loss": 0.1713,
+      "step": 47140
+    },
+    {
+      "epoch": 284.67,
+      "learning_rate": 8.592036658068301e-05,
+      "loss": 0.1786,
+      "step": 47160
+    },
+    {
+      "epoch": 284.79,
+      "learning_rate": 8.588913264203351e-05,
+      "loss": 0.1788,
+      "step": 47180
+    },
+    {
+      "epoch": 284.91,
+      "learning_rate": 8.585786978815808e-05,
+      "loss": 0.1832,
+      "step": 47200
+    },
+    {
+      "epoch": 285.03,
+      "learning_rate": 8.582657804424465e-05,
+      "loss": 0.1693,
+      "step": 47220
+    },
+    {
+      "epoch": 285.15,
+      "learning_rate": 8.579525743550438e-05,
+      "loss": 0.1376,
+      "step": 47240
+    },
+    {
+      "epoch": 285.27,
+      "learning_rate": 8.576390798717174e-05,
+      "loss": 0.1543,
+      "step": 47260
+    },
+    {
+      "epoch": 285.39,
+      "learning_rate": 8.57325297245044e-05,
+      "loss": 0.1552,
+      "step": 47280
+    },
+    {
+      "epoch": 285.51,
+      "learning_rate": 8.570112267278328e-05,
+      "loss": 0.1635,
+      "step": 47300
+    },
+    {
+      "epoch": 285.63,
+      "learning_rate": 8.566968685731245e-05,
+      "loss": 0.1733,
+      "step": 47320
+    },
+    {
+      "epoch": 285.75,
+      "learning_rate": 8.563822230341919e-05,
+      "loss": 0.1781,
+      "step": 47340
+    },
+    {
+      "epoch": 285.88,
+      "learning_rate": 8.560672903645392e-05,
+      "loss": 0.1807,
+      "step": 47360
+    },
+    {
+      "epoch": 286.0,
+      "learning_rate": 8.557520708179021e-05,
+      "loss": 0.1837,
+      "step": 47380
+    },
+    {
+      "epoch": 286.12,
+      "learning_rate": 8.554365646482472e-05,
+      "loss": 0.1383,
+      "step": 47400
+    },
+    {
+      "epoch": 286.24,
+      "learning_rate": 8.551207721097721e-05,
+      "loss": 0.1507,
+      "step": 47420
+    },
+    {
+      "epoch": 286.36,
+      "learning_rate": 8.548046934569051e-05,
+      "loss": 0.1589,
+      "step": 47440
+    },
+    {
+      "epoch": 286.48,
+      "learning_rate": 8.544883289443053e-05,
+      "loss": 0.1627,
+      "step": 47460
+    },
+    {
+      "epoch": 286.6,
+      "learning_rate": 8.541716788268617e-05,
+      "loss": 0.1676,
+      "step": 47480
+    },
+    {
+      "epoch": 286.72,
+      "learning_rate": 8.538547433596933e-05,
+      "loss": 0.1768,
+      "step": 47500
+    },
+    {
+      "epoch": 286.84,
+      "learning_rate": 8.535375227981497e-05,
+      "loss": 0.1743,
+      "step": 47520
+    },
+    {
+      "epoch": 286.96,
+      "learning_rate": 8.532200173978097e-05,
+      "loss": 0.178,
+      "step": 47540
+    },
+    {
+      "epoch": 287.08,
+      "learning_rate": 8.529022274144816e-05,
+      "loss": 0.1468,
+      "step": 47560
+    },
+    {
+      "epoch": 287.2,
+      "learning_rate": 8.525841531042031e-05,
+      "loss": 0.1407,
+      "step": 47580
+    },
+    {
+      "epoch": 287.32,
+      "learning_rate": 8.522657947232407e-05,
+      "loss": 0.1505,
+      "step": 47600
+    },
+    {
+      "epoch": 287.44,
+      "learning_rate": 8.519471525280903e-05,
+      "loss": 0.1569,
+      "step": 47620
+    },
+    {
+      "epoch": 287.57,
+      "learning_rate": 8.516282267754761e-05,
+      "loss": 0.1657,
+      "step": 47640
+    },
+    {
+      "epoch": 287.69,
+      "learning_rate": 8.513090177223506e-05,
+      "loss": 0.1651,
+      "step": 47660
+    },
+    {
+      "epoch": 287.81,
+      "learning_rate": 8.509895256258948e-05,
+      "loss": 0.169,
+      "step": 47680
+    },
+    {
+      "epoch": 287.93,
+      "learning_rate": 8.506697507435182e-05,
+      "loss": 0.1762,
+      "step": 47700
+    },
+    {
+      "epoch": 288.05,
+      "learning_rate": 8.50349693332857e-05,
+      "loss": 0.1584,
+      "step": 47720
+    },
+    {
+      "epoch": 288.17,
+      "learning_rate": 8.50029353651776e-05,
+      "loss": 0.1408,
+      "step": 47740
+    },
+    {
+      "epoch": 288.29,
+      "learning_rate": 8.497087319583672e-05,
+      "loss": 0.1439,
+      "step": 47760
+    },
+    {
+      "epoch": 288.41,
+      "learning_rate": 8.493878285109495e-05,
+      "loss": 0.1497,
+      "step": 47780
+    },
+    {
+      "epoch": 288.53,
+      "learning_rate": 8.49066643568069e-05,
+      "loss": 0.1571,
+      "step": 47800
+    },
+    {
+      "epoch": 288.65,
+      "learning_rate": 8.487451773884987e-05,
+      "loss": 0.1613,
+      "step": 47820
+    },
+    {
+      "epoch": 288.77,
+      "learning_rate": 8.484234302312382e-05,
+      "loss": 0.1688,
+      "step": 47840
+    },
+    {
+      "epoch": 288.89,
+      "learning_rate": 8.48101402355513e-05,
+      "loss": 0.174,
+      "step": 47860
+    },
+    {
+      "epoch": 289.01,
+      "learning_rate": 8.477790940207756e-05,
+      "loss": 0.1721,
+      "step": 47880
+    },
+    {
+      "epoch": 289.13,
+      "learning_rate": 8.474565054867037e-05,
+      "loss": 0.134,
+      "step": 47900
+    },
+    {
+      "epoch": 289.26,
+      "learning_rate": 8.471336370132012e-05,
+      "loss": 0.1472,
+      "step": 47920
+    },
+    {
+      "epoch": 289.38,
+      "learning_rate": 8.468104888603973e-05,
+      "loss": 0.1517,
+      "step": 47940
+    },
+    {
+      "epoch": 289.5,
+      "learning_rate": 8.464870612886467e-05,
+      "loss": 0.1545,
+      "step": 47960
+    },
+    {
+      "epoch": 289.62,
+      "learning_rate": 8.46163354558529e-05,
+      "loss": 0.1612,
+      "step": 47980
+    },
+    {
+      "epoch": 289.74,
+      "learning_rate": 8.458393689308491e-05,
+      "loss": 0.1649,
+      "step": 48000
+    },
+    {
+      "epoch": 289.74,
+      "eval_accuracy": 0.0012802402426134287,
+      "eval_loss": 19.142297744750977,
+      "eval_runtime": 18.2303,
+      "eval_samples_per_second": 111.902,
+      "eval_steps_per_second": 2.359,
+      "step": 48000
+    },
+    {
+      "epoch": 289.86,
+      "learning_rate": 8.455151046666363e-05,
+      "loss": 0.1753,
+      "step": 48020
+    },
+    {
+      "epoch": 289.98,
+      "learning_rate": 8.451905620271443e-05,
+      "loss": 0.1782,
+      "step": 48040
+    },
+    {
+      "epoch": 290.1,
+      "learning_rate": 8.448657412738515e-05,
+      "loss": 0.1353,
+      "step": 48060
+    },
+    {
+      "epoch": 290.22,
+      "learning_rate": 8.445406426684598e-05,
+      "loss": 0.1392,
+      "step": 48080
+    },
+    {
+      "epoch": 290.34,
+      "learning_rate": 8.442152664728958e-05,
+      "loss": 0.1525,
+      "step": 48100
+    },
+    {
+      "epoch": 290.46,
+      "learning_rate": 8.438896129493086e-05,
+      "loss": 0.1546,
+      "step": 48120
+    },
+    {
+      "epoch": 290.58,
+      "learning_rate": 8.435636823600716e-05,
+      "loss": 0.1624,
+      "step": 48140
+    },
+    {
+      "epoch": 290.7,
+      "learning_rate": 8.432374749677814e-05,
+      "loss": 0.1696,
+      "step": 48160
+    },
+    {
+      "epoch": 290.82,
+      "learning_rate": 8.42910991035257e-05,
+      "loss": 0.1724,
+      "step": 48180
+    },
+    {
+      "epoch": 290.95,
+      "learning_rate": 8.425842308255412e-05,
+      "loss": 0.1678,
+      "step": 48200
+    },
+    {
+      "epoch": 291.07,
+      "learning_rate": 8.422571946018982e-05,
+      "loss": 0.1391,
+      "step": 48220
+    },
+    {
+      "epoch": 291.19,
+      "learning_rate": 8.419298826278154e-05,
+      "loss": 0.1272,
+      "step": 48240
+    },
+    {
+      "epoch": 291.31,
+      "learning_rate": 8.416022951670022e-05,
+      "loss": 0.1432,
+      "step": 48260
+    },
+    {
+      "epoch": 291.43,
+      "learning_rate": 8.412744324833898e-05,
+      "loss": 0.1465,
+      "step": 48280
+    },
+    {
+      "epoch": 291.55,
+      "learning_rate": 8.409462948411315e-05,
+      "loss": 0.1572,
+      "step": 48300
+    },
+    {
+      "epoch": 291.67,
+      "learning_rate": 8.406178825046015e-05,
+      "loss": 0.1593,
+      "step": 48320
+    },
+    {
+      "epoch": 291.79,
+      "learning_rate": 8.402891957383959e-05,
+      "loss": 0.1619,
+      "step": 48340
+    },
+    {
+      "epoch": 291.91,
+      "learning_rate": 8.399602348073316e-05,
+      "loss": 0.1699,
+      "step": 48360
+    },
+    {
+      "epoch": 292.03,
+      "learning_rate": 8.396309999764467e-05,
+      "loss": 0.1575,
+      "step": 48380
+    },
+    {
+      "epoch": 292.15,
+      "learning_rate": 8.393014915109995e-05,
+      "loss": 0.1269,
+      "step": 48400
+    },
+    {
+      "epoch": 292.27,
+      "learning_rate": 8.389717096764691e-05,
+      "loss": 0.1392,
+      "step": 48420
+    },
+    {
+      "epoch": 292.39,
+      "learning_rate": 8.386416547385547e-05,
+      "loss": 0.1476,
+      "step": 48440
+    },
+    {
+      "epoch": 292.52,
+      "learning_rate": 8.383113269631757e-05,
+      "loss": 0.1546,
+      "step": 48460
+    },
+    {
+      "epoch": 292.64,
+      "learning_rate": 8.379807266164714e-05,
+      "loss": 0.1575,
+      "step": 48480
+    },
+    {
+      "epoch": 292.76,
+      "learning_rate": 8.376498539648001e-05,
+      "loss": 0.1614,
+      "step": 48500
+    },
+    {
+      "epoch": 292.88,
+      "learning_rate": 8.373187092747403e-05,
+      "loss": 0.1648,
+      "step": 48520
+    },
+    {
+      "epoch": 293.0,
+      "learning_rate": 8.369872928130891e-05,
+      "loss": 0.1679,
+      "step": 48540
+    },
+    {
+      "epoch": 293.12,
+      "learning_rate": 8.366556048468628e-05,
+      "loss": 0.1246,
+      "step": 48560
+    },
+    {
+      "epoch": 293.24,
+      "learning_rate": 8.363236456432964e-05,
+      "loss": 0.1359,
+      "step": 48580
+    },
+    {
+      "epoch": 293.36,
+      "learning_rate": 8.359914154698434e-05,
+      "loss": 0.1425,
+      "step": 48600
+    },
+    {
+      "epoch": 293.48,
+      "learning_rate": 8.356589145941757e-05,
+      "loss": 0.1506,
+      "step": 48620
+    },
+    {
+      "epoch": 293.6,
+      "learning_rate": 8.353261432841832e-05,
+      "loss": 0.1561,
+      "step": 48640
+    },
+    {
+      "epoch": 293.72,
+      "learning_rate": 8.34993101807974e-05,
+      "loss": 0.1618,
+      "step": 48660
+    },
+    {
+      "epoch": 293.84,
+      "learning_rate": 8.346597904338731e-05,
+      "loss": 0.1633,
+      "step": 48680
+    },
+    {
+      "epoch": 293.96,
+      "learning_rate": 8.343262094304238e-05,
+      "loss": 0.1679,
+      "step": 48700
+    },
+    {
+      "epoch": 294.08,
+      "learning_rate": 8.339923590663863e-05,
+      "loss": 0.1352,
+      "step": 48720
+    },
+    {
+      "epoch": 294.21,
+      "learning_rate": 8.336582396107378e-05,
+      "loss": 0.1327,
+      "step": 48740
+    },
+    {
+      "epoch": 294.33,
+      "learning_rate": 8.33323851332672e-05,
+      "loss": 0.1416,
+      "step": 48760
+    },
+    {
+      "epoch": 294.45,
+      "learning_rate": 8.329891945015998e-05,
+      "loss": 0.1473,
+      "step": 48780
+    },
+    {
+      "epoch": 294.57,
+      "learning_rate": 8.326542693871482e-05,
+      "loss": 0.1527,
+      "step": 48800
+    },
+    {
+      "epoch": 294.69,
+      "learning_rate": 8.323190762591601e-05,
+      "loss": 0.156,
+      "step": 48820
+    },
+    {
+      "epoch": 294.81,
+      "learning_rate": 8.319836153876947e-05,
+      "loss": 0.1582,
+      "step": 48840
+    },
+    {
+      "epoch": 294.93,
+      "learning_rate": 8.316478870430269e-05,
+      "loss": 0.1618,
+      "step": 48860
+    },
+    {
+      "epoch": 295.05,
+      "learning_rate": 8.313118914956466e-05,
+      "loss": 0.145,
+      "step": 48880
+    },
+    {
+      "epoch": 295.17,
+      "learning_rate": 8.309756290162595e-05,
+      "loss": 0.1271,
+      "step": 48900
+    },
+    {
+      "epoch": 295.29,
+      "learning_rate": 8.306390998757863e-05,
+      "loss": 0.1355,
+      "step": 48920
+    },
+    {
+      "epoch": 295.41,
+      "learning_rate": 8.303023043453624e-05,
+      "loss": 0.1441,
+      "step": 48940
+    },
+    {
+      "epoch": 295.53,
+      "learning_rate": 8.299652426963379e-05,
+      "loss": 0.1485,
+      "step": 48960
+    },
+    {
+      "epoch": 295.65,
+      "learning_rate": 8.296279152002771e-05,
+      "loss": 0.1496,
+      "step": 48980
+    },
+    {
+      "epoch": 295.77,
+      "learning_rate": 8.29290322128959e-05,
+      "loss": 0.1517,
+      "step": 49000
+    },
+    {
+      "epoch": 295.77,
+      "eval_accuracy": 0.0012590016892058756,
+      "eval_loss": 19.363754272460938,
+      "eval_runtime": 18.183,
+      "eval_samples_per_second": 112.193,
+      "eval_steps_per_second": 2.365,
+      "step": 49000
+    },
+    {
+      "epoch": 295.9,
+      "learning_rate": 8.28952463754376e-05,
+      "loss": 0.1588,
+      "step": 49020
+    },
+    {
+      "epoch": 296.02,
+      "learning_rate": 8.286143403487345e-05,
+      "loss": 0.1548,
+      "step": 49040
+    },
+    {
+      "epoch": 296.14,
+      "learning_rate": 8.282759521844545e-05,
+      "loss": 0.1199,
+      "step": 49060
+    },
+    {
+      "epoch": 296.26,
+      "learning_rate": 8.27954238444015e-05,
+      "loss": 0.1325,
+      "step": 49080
+    },
+    {
+      "epoch": 296.38,
+      "learning_rate": 8.276153347847452e-05,
+      "loss": 0.1418,
+      "step": 49100
+    },
+    {
+      "epoch": 296.5,
+      "learning_rate": 8.272761671717178e-05,
+      "loss": 0.1464,
+      "step": 49120
+    },
+    {
+      "epoch": 296.62,
+      "learning_rate": 8.269367358781937e-05,
+      "loss": 0.1504,
+      "step": 49140
+    },
+    {
+      "epoch": 296.74,
+      "learning_rate": 8.265970411776467e-05,
+      "loss": 0.1542,
+      "step": 49160
+    },
+    {
+      "epoch": 296.86,
+      "learning_rate": 8.262570833437626e-05,
+      "loss": 0.1546,
+      "step": 49180
+    },
+    {
+      "epoch": 296.98,
+      "learning_rate": 8.259168626504395e-05,
+      "loss": 0.1581,
+      "step": 49200
+    },
+    {
+      "epoch": 297.1,
+      "learning_rate": 8.255763793717868e-05,
+      "loss": 0.1237,
+      "step": 49220
+    },
+    {
+      "epoch": 297.22,
+      "learning_rate": 8.25235633782126e-05,
+      "loss": 0.131,
+      "step": 49240
+    },
+    {
+      "epoch": 297.34,
+      "learning_rate": 8.248946261559893e-05,
+      "loss": 0.1359,
+      "step": 49260
+    },
+    {
+      "epoch": 297.46,
+      "learning_rate": 8.245533567681208e-05,
+      "loss": 0.1455,
+      "step": 49280
+    },
+    {
+      "epoch": 297.59,
+      "learning_rate": 8.24211825893475e-05,
+      "loss": 0.1492,
+      "step": 49300
+    },
+    {
+      "epoch": 297.71,
+      "learning_rate": 8.238700338072167e-05,
+      "loss": 0.1528,
+      "step": 49320
+    },
+    {
+      "epoch": 297.83,
+      "learning_rate": 8.235279807847223e-05,
+      "loss": 0.1536,
+      "step": 49340
+    },
+    {
+      "epoch": 297.95,
+      "learning_rate": 8.231856671015772e-05,
+      "loss": 0.1589,
+      "step": 49360
+    },
+    {
+      "epoch": 298.07,
+      "learning_rate": 8.228430930335775e-05,
+      "loss": 0.1374,
+      "step": 49380
+    },
+    {
+      "epoch": 298.19,
+      "learning_rate": 8.225002588567291e-05,
+      "loss": 0.1271,
+      "step": 49400
+    },
+    {
+      "epoch": 298.31,
+      "learning_rate": 8.221571648472472e-05,
+      "loss": 0.132,
+      "step": 49420
+    },
+    {
+      "epoch": 298.43,
+      "learning_rate": 8.218138112815564e-05,
+      "loss": 0.141,
+      "step": 49440
+    },
+    {
+      "epoch": 298.55,
+      "learning_rate": 8.214701984362906e-05,
+      "loss": 0.1447,
+      "step": 49460
+    },
+    {
+      "epoch": 298.67,
+      "learning_rate": 8.211263265882923e-05,
+      "loss": 0.1473,
+      "step": 49480
+    },
+    {
+      "epoch": 298.79,
+      "learning_rate": 8.207821960146128e-05,
+      "loss": 0.1529,
+      "step": 49500
+    },
+    {
+      "epoch": 298.91,
+      "learning_rate": 8.20437806992512e-05,
+      "loss": 0.1556,
+      "step": 49520
+    },
+    {
+      "epoch": 299.03,
+      "learning_rate": 8.200931597994582e-05,
+      "loss": 0.1437,
+      "step": 49540
+    },
+    {
+      "epoch": 299.15,
+      "learning_rate": 8.197482547131268e-05,
+      "loss": 0.1192,
+      "step": 49560
+    },
+    {
+      "epoch": 299.28,
+      "learning_rate": 8.194030920114021e-05,
+      "loss": 0.1297,
+      "step": 49580
+    },
+    {
+      "epoch": 299.4,
+      "learning_rate": 8.190576719723752e-05,
+      "loss": 0.1367,
+      "step": 49600
+    },
+    {
+      "epoch": 299.52,
+      "learning_rate": 8.18711994874345e-05,
+      "loss": 0.1429,
+      "step": 49620
+    },
+    {
+      "epoch": 299.64,
+      "learning_rate": 8.183660609958169e-05,
+      "loss": 0.1467,
+      "step": 49640
+    },
+    {
+      "epoch": 299.76,
+      "learning_rate": 8.18019870615504e-05,
+      "loss": 0.1531,
+      "step": 49660
+    },
+    {
+      "epoch": 299.88,
+      "learning_rate": 8.176734240123255e-05,
+      "loss": 0.1564,
+      "step": 49680
+    },
+    {
+      "epoch": 300.0,
+      "learning_rate": 8.17326721465407e-05,
+      "loss": 0.1527,
+      "step": 49700
+    },
+    {
+      "epoch": 300.12,
+      "learning_rate": 8.16979763254081e-05,
+      "loss": 0.1109,
+      "step": 49720
+    },
+    {
+      "epoch": 300.24,
+      "learning_rate": 8.166325496578847e-05,
+      "loss": 0.1241,
+      "step": 49740
+    },
+    {
+      "epoch": 300.36,
+      "learning_rate": 8.162850809565623e-05,
+      "loss": 0.1373,
+      "step": 49760
+    },
+    {
+      "epoch": 300.48,
+      "learning_rate": 8.159373574300629e-05,
+      "loss": 0.141,
+      "step": 49780
+    },
+    {
+      "epoch": 300.6,
+      "learning_rate": 8.155893793585413e-05,
+      "loss": 0.1433,
+      "step": 49800
+    },
+    {
+      "epoch": 300.72,
+      "learning_rate": 8.152411470223569e-05,
+      "loss": 0.1458,
+      "step": 49820
+    },
+    {
+      "epoch": 300.85,
+      "learning_rate": 8.148926607020743e-05,
+      "loss": 0.1525,
+      "step": 49840
+    },
+    {
+      "epoch": 300.97,
+      "learning_rate": 8.145439206784626e-05,
+      "loss": 0.1573,
+      "step": 49860
+    },
+    {
+      "epoch": 301.09,
+      "learning_rate": 8.141949272324953e-05,
+      "loss": 0.1243,
+      "step": 49880
+    },
+    {
+      "epoch": 301.21,
+      "learning_rate": 8.138456806453503e-05,
+      "loss": 0.1229,
+      "step": 49900
+    },
+    {
+      "epoch": 301.33,
+      "learning_rate": 8.13496181198409e-05,
+      "loss": 0.1313,
+      "step": 49920
+    },
+    {
+      "epoch": 301.45,
+      "learning_rate": 8.131464291732572e-05,
+      "loss": 0.1363,
+      "step": 49940
+    },
+    {
+      "epoch": 301.57,
+      "learning_rate": 8.127964248516832e-05,
+      "loss": 0.1392,
+      "step": 49960
+    },
+    {
+      "epoch": 301.69,
+      "learning_rate": 8.124461685156795e-05,
+      "loss": 0.1404,
+      "step": 49980
+    },
+    {
+      "epoch": 301.81,
+      "learning_rate": 8.120956604474415e-05,
+      "loss": 0.1491,
+      "step": 50000
+    },
+    {
+      "epoch": 301.81,
+      "eval_accuracy": 0.0012639408876727485,
+      "eval_loss": 19.587871551513672,
+      "eval_runtime": 18.2116,
+      "eval_samples_per_second": 112.016,
+      "eval_steps_per_second": 2.361,
+      "step": 50000
+    },
+    {
+      "epoch": 301.93,
+      "learning_rate": 8.117449009293668e-05,
+      "loss": 0.1501,
+      "step": 50020
+    },
+    {
+      "epoch": 302.05,
+      "learning_rate": 8.113938902440564e-05,
+      "loss": 0.1327,
+      "step": 50040
+    },
+    {
+      "epoch": 302.17,
+      "learning_rate": 8.110426286743129e-05,
+      "loss": 0.1179,
+      "step": 50060
+    },
+    {
+      "epoch": 302.29,
+      "learning_rate": 8.106911165031415e-05,
+      "loss": 0.125,
+      "step": 50080
+    },
+    {
+      "epoch": 302.41,
+      "learning_rate": 8.103393540137496e-05,
+      "loss": 0.1345,
+      "step": 50100
+    },
+    {
+      "epoch": 302.54,
+      "learning_rate": 8.099873414895453e-05,
+      "loss": 0.1386,
+      "step": 50120
+    },
+    {
+      "epoch": 302.66,
+      "learning_rate": 8.096350792141392e-05,
+      "loss": 0.1422,
+      "step": 50140
+    },
+    {
+      "epoch": 302.78,
+      "learning_rate": 8.092825674713425e-05,
+      "loss": 0.1463,
+      "step": 50160
+    },
+    {
+      "epoch": 302.9,
+      "learning_rate": 8.089298065451672e-05,
+      "loss": 0.1481,
+      "step": 50180
+    },
+    {
+      "epoch": 303.02,
+      "learning_rate": 8.085767967198269e-05,
+      "loss": 0.1403,
+      "step": 50200
+    },
+    {
+      "epoch": 303.14,
+      "learning_rate": 8.082235382797349e-05,
+      "loss": 0.11,
+      "step": 50220
+    },
+    {
+      "epoch": 303.26,
+      "learning_rate": 8.078700315095055e-05,
+      "loss": 0.1223,
+      "step": 50240
+    },
+    {
+      "epoch": 303.38,
+      "learning_rate": 8.075162766939526e-05,
+      "loss": 0.1302,
+      "step": 50260
+    },
+    {
+      "epoch": 303.5,
+      "learning_rate": 8.071622741180898e-05,
+      "loss": 0.1352,
+      "step": 50280
+    },
+    {
+      "epoch": 303.62,
+      "learning_rate": 8.068080240671308e-05,
+      "loss": 0.1387,
+      "step": 50300
+    },
+    {
+      "epoch": 303.74,
+      "learning_rate": 8.064535268264883e-05,
+      "loss": 0.1357,
+      "step": 50320
+    },
+    {
+      "epoch": 303.86,
+      "learning_rate": 8.060987826817745e-05,
+      "loss": 0.1428,
+      "step": 50340
+    },
+    {
+      "epoch": 303.98,
+      "learning_rate": 8.057437919188005e-05,
+      "loss": 0.1516,
+      "step": 50360
+    },
+    {
+      "epoch": 304.1,
+      "learning_rate": 8.053885548235755e-05,
+      "loss": 0.1121,
+      "step": 50380
+    },
+    {
+      "epoch": 304.23,
+      "learning_rate": 8.05033071682308e-05,
+      "loss": 0.1177,
+      "step": 50400
+    },
+    {
+      "epoch": 304.35,
+      "learning_rate": 8.046773427814042e-05,
+      "loss": 0.1241,
+      "step": 50420
+    },
+    {
+      "epoch": 304.47,
+      "learning_rate": 8.043213684074684e-05,
+      "loss": 0.1314,
+      "step": 50440
+    },
+    {
+      "epoch": 304.59,
+      "learning_rate": 8.039651488473028e-05,
+      "loss": 0.1347,
+      "step": 50460
+    },
+    {
+      "epoch": 304.71,
+      "learning_rate": 8.03608684387907e-05,
+      "loss": 0.1367,
+      "step": 50480
+    },
+    {
+      "epoch": 304.83,
+      "learning_rate": 8.03251975316478e-05,
+      "loss": 0.1446,
+      "step": 50500
+    },
+    {
+      "epoch": 304.95,
+      "learning_rate": 8.0289502192041e-05,
+      "loss": 0.1429,
+      "step": 50520
+    },
+    {
+      "epoch": 305.07,
+      "learning_rate": 8.025378244872936e-05,
+      "loss": 0.1184,
+      "step": 50540
+    },
+    {
+      "epoch": 305.19,
+      "learning_rate": 8.021803833049166e-05,
+      "loss": 0.1131,
+      "step": 50560
+    },
+    {
+      "epoch": 305.31,
+      "learning_rate": 8.01822698661263e-05,
+      "loss": 0.1242,
+      "step": 50580
+    },
+    {
+      "epoch": 305.43,
+      "learning_rate": 8.014647708445124e-05,
+      "loss": 0.13,
+      "step": 50600
+    },
+    {
+      "epoch": 305.55,
+      "learning_rate": 8.011066001430412e-05,
+      "loss": 0.1324,
+      "step": 50620
+    },
+    {
+      "epoch": 305.67,
+      "learning_rate": 8.007481868454208e-05,
+      "loss": 0.1361,
+      "step": 50640
+    },
+    {
+      "epoch": 305.79,
+      "learning_rate": 8.003895312404183e-05,
+      "loss": 0.1416,
+      "step": 50660
+    },
+    {
+      "epoch": 305.92,
+      "learning_rate": 8.000306336169963e-05,
+      "loss": 0.1446,
+      "step": 50680
+    },
+    {
+      "epoch": 306.04,
+      "learning_rate": 7.99671494264312e-05,
+      "loss": 0.1321,
+      "step": 50700
+    },
+    {
+      "epoch": 306.16,
+      "learning_rate": 7.993121134717177e-05,
+      "loss": 0.1086,
+      "step": 50720
+    },
+    {
+      "epoch": 306.28,
+      "learning_rate": 7.989524915287595e-05,
+      "loss": 0.1181,
+      "step": 50740
+    },
+    {
+      "epoch": 306.4,
+      "learning_rate": 7.985926287251787e-05,
+      "loss": 0.1277,
+      "step": 50760
+    },
+    {
+      "epoch": 306.52,
+      "learning_rate": 7.982325253509102e-05,
+      "loss": 0.1331,
+      "step": 50780
+    },
+    {
+      "epoch": 306.64,
+      "learning_rate": 7.978721816960826e-05,
+      "loss": 0.1349,
+      "step": 50800
+    },
+    {
+      "epoch": 306.76,
+      "learning_rate": 7.975115980510187e-05,
+      "loss": 0.1399,
+      "step": 50820
+    },
+    {
+      "epoch": 306.88,
+      "learning_rate": 7.971507747062337e-05,
+      "loss": 0.1395,
+      "step": 50840
+    },
+    {
+      "epoch": 307.0,
+      "learning_rate": 7.967897119524368e-05,
+      "loss": 0.1431,
+      "step": 50860
+    },
+    {
+      "epoch": 307.12,
+      "learning_rate": 7.964284100805297e-05,
+      "loss": 0.1063,
+      "step": 50880
+    },
+    {
+      "epoch": 307.24,
+      "learning_rate": 7.960668693816067e-05,
+      "loss": 0.119,
+      "step": 50900
+    },
+    {
+      "epoch": 307.36,
+      "learning_rate": 7.957050901469545e-05,
+      "loss": 0.127,
+      "step": 50920
+    },
+    {
+      "epoch": 307.48,
+      "learning_rate": 7.953430726680524e-05,
+      "loss": 0.1279,
+      "step": 50940
+    },
+    {
+      "epoch": 307.61,
+      "learning_rate": 7.949808172365713e-05,
+      "loss": 0.1325,
+      "step": 50960
+    },
+    {
+      "epoch": 307.73,
+      "learning_rate": 7.946183241443736e-05,
+      "loss": 0.1374,
+      "step": 50980
+    },
+    {
+      "epoch": 307.85,
+      "learning_rate": 7.942555936835135e-05,
+      "loss": 0.1387,
+      "step": 51000
+    },
+    {
+      "epoch": 307.85,
+      "eval_accuracy": 0.001271843605219745,
+      "eval_loss": 19.7823429107666,
+      "eval_runtime": 18.1904,
+      "eval_samples_per_second": 112.147,
+      "eval_steps_per_second": 2.364,
+      "step": 51000
+    },
+    {
+      "epoch": 307.97,
+      "learning_rate": 7.938926261462366e-05,
+      "loss": 0.1433,
+      "step": 51020
+    },
+    {
+      "epoch": 308.09,
+      "learning_rate": 7.935294218249791e-05,
+      "loss": 0.1137,
+      "step": 51040
+    },
+    {
+      "epoch": 308.21,
+      "learning_rate": 7.931659810123683e-05,
+      "loss": 0.113,
+      "step": 51060
+    },
+    {
+      "epoch": 308.33,
+      "learning_rate": 7.928204934569717e-05,
+      "loss": 0.1231,
+      "step": 51080
+    },
+    {
+      "epoch": 308.45,
+      "learning_rate": 7.92456592328612e-05,
+      "loss": 0.1247,
+      "step": 51100
+    },
+    {
+      "epoch": 308.57,
+      "learning_rate": 7.920924555732582e-05,
+      "loss": 0.132,
+      "step": 51120
+    },
+    {
+      "epoch": 308.69,
+      "learning_rate": 7.917280834842888e-05,
+      "loss": 0.1351,
+      "step": 51140
+    },
+    {
+      "epoch": 308.81,
+      "learning_rate": 7.913634763552717e-05,
+      "loss": 0.1395,
+      "step": 51160
+    },
+    {
+      "epoch": 308.93,
+      "learning_rate": 7.909986344799645e-05,
+      "loss": 0.1372,
+      "step": 51180
+    },
+    {
+      "epoch": 309.05,
+      "learning_rate": 7.906335581523135e-05,
+      "loss": 0.1199,
+      "step": 51200
+    },
+    {
+      "epoch": 309.18,
+      "learning_rate": 7.90268247666454e-05,
+      "loss": 0.1085,
+      "step": 51220
+    },
+    {
+      "epoch": 309.3,
+      "learning_rate": 7.899027033167104e-05,
+      "loss": 0.1169,
+      "step": 51240
+    },
+    {
+      "epoch": 309.42,
+      "learning_rate": 7.895369253975951e-05,
+      "loss": 0.1223,
+      "step": 51260
+    },
+    {
+      "epoch": 309.54,
+      "learning_rate": 7.891709142038084e-05,
+      "loss": 0.129,
+      "step": 51280
+    },
+    {
+      "epoch": 309.66,
+      "learning_rate": 7.888046700302394e-05,
+      "loss": 0.1346,
+      "step": 51300
+    },
+    {
+      "epoch": 309.78,
+      "learning_rate": 7.884381931719642e-05,
+      "loss": 0.1356,
+      "step": 51320
+    },
+    {
+      "epoch": 309.9,
+      "learning_rate": 7.880714839242464e-05,
+      "loss": 0.1366,
+      "step": 51340
+    },
+    {
+      "epoch": 310.02,
+      "learning_rate": 7.877045425825374e-05,
+      "loss": 0.133,
+      "step": 51360
+    },
+    {
+      "epoch": 310.14,
+      "learning_rate": 7.873373694424751e-05,
+      "loss": 0.1043,
+      "step": 51380
+    },
+    {
+      "epoch": 310.26,
+      "learning_rate": 7.869699647998839e-05,
+      "loss": 0.1133,
+      "step": 51400
+    },
+    {
+      "epoch": 310.38,
+      "learning_rate": 7.866023289507754e-05,
+      "loss": 0.1199,
+      "step": 51420
+    },
+    {
+      "epoch": 310.5,
+      "learning_rate": 7.862344621913472e-05,
+      "loss": 0.1252,
+      "step": 51440
+    },
+    {
+      "epoch": 310.62,
+      "learning_rate": 7.858663648179828e-05,
+      "loss": 0.1313,
+      "step": 51460
+    },
+    {
+      "epoch": 310.74,
+      "learning_rate": 7.854980371272517e-05,
+      "loss": 0.1312,
+      "step": 51480
+    },
+    {
+      "epoch": 310.87,
+      "learning_rate": 7.851294794159087e-05,
+      "loss": 0.1364,
+      "step": 51500
+    },
+    {
+      "epoch": 310.99,
+      "learning_rate": 7.847606919808944e-05,
+      "loss": 0.1399,
+      "step": 51520
+    },
+    {
+      "epoch": 311.11,
+      "learning_rate": 7.843916751193336e-05,
+      "loss": 0.1065,
+      "step": 51540
+    },
+    {
+      "epoch": 311.23,
+      "learning_rate": 7.840224291285372e-05,
+      "loss": 0.1115,
+      "step": 51560
+    },
+    {
+      "epoch": 311.35,
+      "learning_rate": 7.836529543059995e-05,
+      "loss": 0.119,
+      "step": 51580
+    },
+    {
+      "epoch": 311.47,
+      "learning_rate": 7.832832509494001e-05,
+      "loss": 0.1225,
+      "step": 51600
+    },
+    {
+      "epoch": 311.59,
+      "learning_rate": 7.82913319356602e-05,
+      "loss": 0.1261,
+      "step": 51620
+    },
+    {
+      "epoch": 311.71,
+      "learning_rate": 7.825431598256525e-05,
+      "loss": 0.129,
+      "step": 51640
+    },
+    {
+      "epoch": 311.83,
+      "learning_rate": 7.821727726547825e-05,
+      "loss": 0.131,
+      "step": 51660
+    },
+    {
+      "epoch": 311.95,
+      "learning_rate": 7.818021581424064e-05,
+      "loss": 0.1364,
+      "step": 51680
+    },
+    {
+      "epoch": 312.07,
+      "learning_rate": 7.814313165871212e-05,
+      "loss": 0.1154,
+      "step": 51700
+    },
+    {
+      "epoch": 312.19,
+      "learning_rate": 7.810602482877075e-05,
+      "loss": 0.1043,
+      "step": 51720
+    },
+    {
+      "epoch": 312.31,
+      "learning_rate": 7.806889535431284e-05,
+      "loss": 0.1127,
+      "step": 51740
+    },
+    {
+      "epoch": 312.43,
+      "learning_rate": 7.803174326525293e-05,
+      "loss": 0.1205,
+      "step": 51760
+    },
+    {
+      "epoch": 312.56,
+      "learning_rate": 7.799456859152378e-05,
+      "loss": 0.1243,
+      "step": 51780
+    },
+    {
+      "epoch": 312.68,
+      "learning_rate": 7.795737136307636e-05,
+      "loss": 0.1318,
+      "step": 51800
+    },
+    {
+      "epoch": 312.8,
+      "learning_rate": 7.79201516098798e-05,
+      "loss": 0.1335,
+      "step": 51820
+    },
+    {
+      "epoch": 312.92,
+      "learning_rate": 7.78829093619214e-05,
+      "loss": 0.1371,
+      "step": 51840
+    },
+    {
+      "epoch": 313.04,
+      "learning_rate": 7.784564464920654e-05,
+      "loss": 0.1249,
+      "step": 51860
+    },
+    {
+      "epoch": 313.16,
+      "learning_rate": 7.780835750175874e-05,
+      "loss": 0.103,
+      "step": 51880
+    },
+    {
+      "epoch": 313.28,
+      "learning_rate": 7.777104794961957e-05,
+      "loss": 0.113,
+      "step": 51900
+    },
+    {
+      "epoch": 313.4,
+      "learning_rate": 7.773371602284869e-05,
+      "loss": 0.1204,
+      "step": 51920
+    },
+    {
+      "epoch": 313.52,
+      "learning_rate": 7.769636175152374e-05,
+      "loss": 0.1231,
+      "step": 51940
+    },
+    {
+      "epoch": 313.64,
+      "learning_rate": 7.765898516574038e-05,
+      "loss": 0.1306,
+      "step": 51960
+    },
+    {
+      "epoch": 313.76,
+      "learning_rate": 7.762158629561225e-05,
+      "loss": 0.1301,
+      "step": 51980
+    },
+    {
+      "epoch": 313.88,
+      "learning_rate": 7.758416517127094e-05,
+      "loss": 0.1332,
+      "step": 52000
+    },
+    {
+      "epoch": 313.88,
+      "eval_accuracy": 0.0012550503304323774,
+      "eval_loss": 19.966257095336914,
+      "eval_runtime": 18.2423,
+      "eval_samples_per_second": 111.828,
+      "eval_steps_per_second": 2.357,
+      "step": 52000
+    },
+    {
+      "epoch": 314.0,
+      "learning_rate": 7.7546721822866e-05,
+      "loss": 0.1348,
+      "step": 52020
+    },
+    {
+      "epoch": 314.12,
+      "learning_rate": 7.750925628056482e-05,
+      "loss": 0.0974,
+      "step": 52040
+    },
+    {
+      "epoch": 314.25,
+      "learning_rate": 7.747176857455275e-05,
+      "loss": 0.1108,
+      "step": 52060
+    },
+    {
+      "epoch": 314.37,
+      "learning_rate": 7.743425873503294e-05,
+      "loss": 0.118,
+      "step": 52080
+    },
+    {
+      "epoch": 314.49,
+      "learning_rate": 7.739672679222638e-05,
+      "loss": 0.1192,
+      "step": 52100
+    },
+    {
+      "epoch": 314.61,
+      "learning_rate": 7.735917277637189e-05,
+      "loss": 0.1247,
+      "step": 52120
+    },
+    {
+      "epoch": 314.73,
+      "learning_rate": 7.732159671772605e-05,
+      "loss": 0.1268,
+      "step": 52140
+    },
+    {
+      "epoch": 314.85,
+      "learning_rate": 7.728399864656324e-05,
+      "loss": 0.129,
+      "step": 52160
+    },
+    {
+      "epoch": 314.97,
+      "learning_rate": 7.724637859317551e-05,
+      "loss": 0.1327,
+      "step": 52180
+    },
+    {
+      "epoch": 315.09,
+      "learning_rate": 7.720873658787268e-05,
+      "loss": 0.1058,
+      "step": 52200
+    },
+    {
+      "epoch": 315.21,
+      "learning_rate": 7.717107266098225e-05,
+      "loss": 0.105,
+      "step": 52220
+    },
+    {
+      "epoch": 315.33,
+      "learning_rate": 7.713338684284932e-05,
+      "loss": 0.1141,
+      "step": 52240
+    },
+    {
+      "epoch": 315.45,
+      "learning_rate": 7.709567916383672e-05,
+      "loss": 0.1154,
+      "step": 52260
+    },
+    {
+      "epoch": 315.57,
+      "learning_rate": 7.705794965432481e-05,
+      "loss": 0.1197,
+      "step": 52280
+    },
+    {
+      "epoch": 315.69,
+      "learning_rate": 7.702019834471159e-05,
+      "loss": 0.1265,
+      "step": 52300
+    },
+    {
+      "epoch": 315.81,
+      "learning_rate": 7.698242526541262e-05,
+      "loss": 0.1258,
+      "step": 52320
+    },
+    {
+      "epoch": 315.94,
+      "learning_rate": 7.694463044686095e-05,
+      "loss": 0.128,
+      "step": 52340
+    },
+    {
+      "epoch": 316.06,
+      "learning_rate": 7.690681391950723e-05,
+      "loss": 0.1119,
+      "step": 52360
+    },
+    {
+      "epoch": 316.18,
+      "learning_rate": 7.686897571381952e-05,
+      "loss": 0.1039,
+      "step": 52380
+    },
+    {
+      "epoch": 316.3,
+      "learning_rate": 7.68311158602834e-05,
+      "loss": 0.1102,
+      "step": 52400
+    },
+    {
+      "epoch": 316.42,
+      "learning_rate": 7.679323438940184e-05,
+      "loss": 0.1139,
+      "step": 52420
+    },
+    {
+      "epoch": 316.54,
+      "learning_rate": 7.67553313316953e-05,
+      "loss": 0.1171,
+      "step": 52440
+    },
+    {
+      "epoch": 316.66,
+      "learning_rate": 7.671740671770153e-05,
+      "loss": 0.1205,
+      "step": 52460
+    },
+    {
+      "epoch": 316.78,
+      "learning_rate": 7.667946057797578e-05,
+      "loss": 0.1236,
+      "step": 52480
+    },
+    {
+      "epoch": 316.9,
+      "learning_rate": 7.664149294309051e-05,
+      "loss": 0.129,
+      "step": 52500
+    },
+    {
+      "epoch": 317.02,
+      "learning_rate": 7.66035038436356e-05,
+      "loss": 0.1206,
+      "step": 52520
+    },
+    {
+      "epoch": 317.14,
+      "learning_rate": 7.656549331021814e-05,
+      "loss": 0.0967,
+      "step": 52540
+    },
+    {
+      "epoch": 317.26,
+      "learning_rate": 7.652746137346255e-05,
+      "loss": 0.1072,
+      "step": 52560
+    },
+    {
+      "epoch": 317.38,
+      "learning_rate": 7.648940806401048e-05,
+      "loss": 0.1102,
+      "step": 52580
+    },
+    {
+      "epoch": 317.51,
+      "learning_rate": 7.645133341252078e-05,
+      "loss": 0.1166,
+      "step": 52600
+    },
+    {
+      "epoch": 317.63,
+      "learning_rate": 7.641323744966953e-05,
+      "loss": 0.1216,
+      "step": 52620
+    },
+    {
+      "epoch": 317.75,
+      "learning_rate": 7.637512020614995e-05,
+      "loss": 0.1245,
+      "step": 52640
+    },
+    {
+      "epoch": 317.87,
+      "learning_rate": 7.633698171267241e-05,
+      "loss": 0.1263,
+      "step": 52660
+    },
+    {
+      "epoch": 317.99,
+      "learning_rate": 7.629882199996441e-05,
+      "loss": 0.1286,
+      "step": 52680
+    },
+    {
+      "epoch": 318.11,
+      "learning_rate": 7.626064109877054e-05,
+      "loss": 0.0974,
+      "step": 52700
+    },
+    {
+      "epoch": 318.23,
+      "learning_rate": 7.622243903985245e-05,
+      "loss": 0.1049,
+      "step": 52720
+    },
+    {
+      "epoch": 318.35,
+      "learning_rate": 7.618421585398885e-05,
+      "loss": 0.1115,
+      "step": 52740
+    },
+    {
+      "epoch": 318.47,
+      "learning_rate": 7.61459715719755e-05,
+      "loss": 0.1154,
+      "step": 52760
+    },
+    {
+      "epoch": 318.59,
+      "learning_rate": 7.610770622462508e-05,
+      "loss": 0.1164,
+      "step": 52780
+    },
+    {
+      "epoch": 318.71,
+      "learning_rate": 7.606941984276734e-05,
+      "loss": 0.1205,
+      "step": 52800
+    },
+    {
+      "epoch": 318.83,
+      "learning_rate": 7.60311124572489e-05,
+      "loss": 0.1233,
+      "step": 52820
+    },
+    {
+      "epoch": 318.95,
+      "learning_rate": 7.599278409893334e-05,
+      "loss": 0.1259,
+      "step": 52840
+    },
+    {
+      "epoch": 319.07,
+      "learning_rate": 7.59544347987011e-05,
+      "loss": 0.1046,
+      "step": 52860
+    },
+    {
+      "epoch": 319.2,
+      "learning_rate": 7.591606458744955e-05,
+      "loss": 0.0991,
+      "step": 52880
+    },
+    {
+      "epoch": 319.32,
+      "learning_rate": 7.587767349609284e-05,
+      "loss": 0.1062,
+      "step": 52900
+    },
+    {
+      "epoch": 319.44,
+      "learning_rate": 7.583926155556203e-05,
+      "loss": 0.1112,
+      "step": 52920
+    },
+    {
+      "epoch": 319.56,
+      "learning_rate": 7.580082879680488e-05,
+      "loss": 0.1148,
+      "step": 52940
+    },
+    {
+      "epoch": 319.68,
+      "learning_rate": 7.5762375250786e-05,
+      "loss": 0.1182,
+      "step": 52960
+    },
+    {
+      "epoch": 319.8,
+      "learning_rate": 7.572390094848669e-05,
+      "loss": 0.1205,
+      "step": 52980
+    },
+    {
+      "epoch": 319.92,
+      "learning_rate": 7.568540592090503e-05,
+      "loss": 0.1256,
+      "step": 53000
+    },
+    {
+      "epoch": 319.92,
+      "eval_accuracy": 0.0012787584830733667,
+      "eval_loss": 20.19066619873047,
+      "eval_runtime": 18.1725,
+      "eval_samples_per_second": 112.257,
+      "eval_steps_per_second": 2.366,
+      "step": 53000
+    },
+    {
+      "epoch": 320.04,
+      "learning_rate": 7.564689019905575e-05,
+      "loss": 0.1129,
+      "step": 53020
+    },
+    {
+      "epoch": 320.16,
+      "learning_rate": 7.560835381397027e-05,
+      "loss": 0.0971,
+      "step": 53040
+    },
+    {
+      "epoch": 320.28,
+      "learning_rate": 7.556979679669666e-05,
+      "loss": 0.1016,
+      "step": 53060
+    },
+    {
+      "epoch": 320.4,
+      "learning_rate": 7.553314854801641e-05,
+      "loss": 0.1067,
+      "step": 53080
+    },
+    {
+      "epoch": 320.52,
+      "learning_rate": 7.549455138734089e-05,
+      "loss": 0.1125,
+      "step": 53100
+    },
+    {
+      "epoch": 320.64,
+      "learning_rate": 7.545593368616582e-05,
+      "loss": 0.115,
+      "step": 53120
+    },
+    {
+      "epoch": 320.76,
+      "learning_rate": 7.541729547560477e-05,
+      "loss": 0.1155,
+      "step": 53140
+    },
+    {
+      "epoch": 320.89,
+      "learning_rate": 7.537863678678783e-05,
+      "loss": 0.1214,
+      "step": 53160
+    },
+    {
+      "epoch": 321.01,
+      "learning_rate": 7.533995765086165e-05,
+      "loss": 0.121,
+      "step": 53180
+    },
+    {
+      "epoch": 321.13,
+      "learning_rate": 7.530125809898927e-05,
+      "loss": 0.0912,
+      "step": 53200
+    },
+    {
+      "epoch": 321.25,
+      "learning_rate": 7.526253816235023e-05,
+      "loss": 0.1005,
+      "step": 53220
+    },
+    {
+      "epoch": 321.37,
+      "learning_rate": 7.522379787214049e-05,
+      "loss": 0.1048,
+      "step": 53240
+    },
+    {
+      "epoch": 321.49,
+      "learning_rate": 7.51850372595724e-05,
+      "loss": 0.1052,
+      "step": 53260
+    },
+    {
+      "epoch": 321.61,
+      "learning_rate": 7.51462563558747e-05,
+      "loss": 0.1141,
+      "step": 53280
+    },
+    {
+      "epoch": 321.73,
+      "learning_rate": 7.510745519229243e-05,
+      "loss": 0.1162,
+      "step": 53300
+    },
+    {
+      "epoch": 321.85,
+      "learning_rate": 7.506863380008702e-05,
+      "loss": 0.118,
+      "step": 53320
+    },
+    {
+      "epoch": 321.97,
+      "learning_rate": 7.502979221053615e-05,
+      "loss": 0.1209,
+      "step": 53340
+    },
+    {
+      "epoch": 322.09,
+      "learning_rate": 7.499093045493379e-05,
+      "loss": 0.0971,
+      "step": 53360
+    },
+    {
+      "epoch": 322.21,
+      "learning_rate": 7.495204856459014e-05,
+      "loss": 0.0977,
+      "step": 53380
+    },
+    {
+      "epoch": 322.33,
+      "learning_rate": 7.491314657083166e-05,
+      "loss": 0.1014,
+      "step": 53400
+    },
+    {
+      "epoch": 322.45,
+      "learning_rate": 7.487422450500096e-05,
+      "loss": 0.1085,
+      "step": 53420
+    },
+    {
+      "epoch": 322.58,
+      "learning_rate": 7.483528239845688e-05,
+      "loss": 0.1129,
+      "step": 53440
+    },
+    {
+      "epoch": 322.7,
+      "learning_rate": 7.479632028257435e-05,
+      "loss": 0.1163,
+      "step": 53460
+    },
+    {
+      "epoch": 322.82,
+      "learning_rate": 7.475733818874443e-05,
+      "loss": 0.1189,
+      "step": 53480
+    },
+    {
+      "epoch": 322.94,
+      "learning_rate": 7.471833614837431e-05,
+      "loss": 0.1211,
+      "step": 53500
+    },
+    {
+      "epoch": 323.06,
+      "learning_rate": 7.467931419288721e-05,
+      "loss": 0.1047,
+      "step": 53520
+    },
+    {
+      "epoch": 323.18,
+      "learning_rate": 7.464027235372244e-05,
+      "loss": 0.0945,
+      "step": 53540
+    },
+    {
+      "epoch": 323.3,
+      "learning_rate": 7.460121066233529e-05,
+      "loss": 0.1007,
+      "step": 53560
+    },
+    {
+      "epoch": 323.42,
+      "learning_rate": 7.456212915019706e-05,
+      "loss": 0.1046,
+      "step": 53580
+    },
+    {
+      "epoch": 323.54,
+      "learning_rate": 7.452302784879501e-05,
+      "loss": 0.1108,
+      "step": 53600
+    },
+    {
+      "epoch": 323.66,
+      "learning_rate": 7.448390678963238e-05,
+      "loss": 0.1148,
+      "step": 53620
+    },
+    {
+      "epoch": 323.78,
+      "learning_rate": 7.444476600422828e-05,
+      "loss": 0.1164,
+      "step": 53640
+    },
+    {
+      "epoch": 323.9,
+      "learning_rate": 7.440560552411771e-05,
+      "loss": 0.1192,
+      "step": 53660
+    },
+    {
+      "epoch": 324.02,
+      "learning_rate": 7.43664253808516e-05,
+      "loss": 0.1115,
+      "step": 53680
+    },
+    {
+      "epoch": 324.14,
+      "learning_rate": 7.432722560599669e-05,
+      "loss": 0.0896,
+      "step": 53700
+    },
+    {
+      "epoch": 324.27,
+      "learning_rate": 7.428800623113548e-05,
+      "loss": 0.0974,
+      "step": 53720
+    },
+    {
+      "epoch": 324.39,
+      "learning_rate": 7.424876728786637e-05,
+      "loss": 0.1004,
+      "step": 53740
+    },
+    {
+      "epoch": 324.51,
+      "learning_rate": 7.420950880780343e-05,
+      "loss": 0.1065,
+      "step": 53760
+    },
+    {
+      "epoch": 324.63,
+      "learning_rate": 7.417023082257652e-05,
+      "loss": 0.1095,
+      "step": 53780
+    },
+    {
+      "epoch": 324.75,
+      "learning_rate": 7.413093336383121e-05,
+      "loss": 0.1108,
+      "step": 53800
+    },
+    {
+      "epoch": 324.87,
+      "learning_rate": 7.409161646322874e-05,
+      "loss": 0.1144,
+      "step": 53820
+    },
+    {
+      "epoch": 324.99,
+      "learning_rate": 7.405228015244603e-05,
+      "loss": 0.1161,
+      "step": 53840
+    },
+    {
+      "epoch": 325.11,
+      "learning_rate": 7.401292446317564e-05,
+      "loss": 0.0876,
+      "step": 53860
+    },
+    {
+      "epoch": 325.23,
+      "learning_rate": 7.397354942712574e-05,
+      "loss": 0.0939,
+      "step": 53880
+    },
+    {
+      "epoch": 325.35,
+      "learning_rate": 7.393415507602007e-05,
+      "loss": 0.1007,
+      "step": 53900
+    },
+    {
+      "epoch": 325.47,
+      "learning_rate": 7.389474144159796e-05,
+      "loss": 0.1036,
+      "step": 53920
+    },
+    {
+      "epoch": 325.59,
+      "learning_rate": 7.385530855561426e-05,
+      "loss": 0.107,
+      "step": 53940
+    },
+    {
+      "epoch": 325.71,
+      "learning_rate": 7.381585644983935e-05,
+      "loss": 0.1091,
+      "step": 53960
+    },
+    {
+      "epoch": 325.84,
+      "learning_rate": 7.377638515605904e-05,
+      "loss": 0.1124,
+      "step": 53980
+    },
+    {
+      "epoch": 325.96,
+      "learning_rate": 7.373689470607469e-05,
+      "loss": 0.1154,
+      "step": 54000
+    },
+    {
+      "epoch": 325.96,
+      "eval_accuracy": 0.0012624591281326866,
+      "eval_loss": 20.39387321472168,
+      "eval_runtime": 18.3547,
+      "eval_samples_per_second": 111.143,
+      "eval_steps_per_second": 2.343,
+      "step": 54000
+    },
+    {
+      "epoch": 326.08,
+      "learning_rate": 7.369738513170301e-05,
+      "loss": 0.0942,
+      "step": 54020
+    },
+    {
+      "epoch": 326.2,
+      "learning_rate": 7.365785646477614e-05,
+      "loss": 0.091,
+      "step": 54040
+    },
+    {
+      "epoch": 326.32,
+      "learning_rate": 7.361830873714165e-05,
+      "loss": 0.0967,
+      "step": 54060
+    },
+    {
+      "epoch": 326.44,
+      "learning_rate": 7.35787419806624e-05,
+      "loss": 0.1024,
+      "step": 54080
+    },
+    {
+      "epoch": 326.56,
+      "learning_rate": 7.35391562272166e-05,
+      "loss": 0.1043,
+      "step": 54100
+    },
+    {
+      "epoch": 326.68,
+      "learning_rate": 7.349955150869781e-05,
+      "loss": 0.1083,
+      "step": 54120
+    },
+    {
+      "epoch": 326.8,
+      "learning_rate": 7.345992785701483e-05,
+      "loss": 0.1097,
+      "step": 54140
+    },
+    {
+      "epoch": 326.92,
+      "learning_rate": 7.342028530409171e-05,
+      "loss": 0.1117,
+      "step": 54160
+    },
+    {
+      "epoch": 327.04,
+      "learning_rate": 7.338062388186772e-05,
+      "loss": 0.1025,
+      "step": 54180
+    },
+    {
+      "epoch": 327.16,
+      "learning_rate": 7.33409436222974e-05,
+      "loss": 0.0874,
+      "step": 54200
+    },
+    {
+      "epoch": 327.28,
+      "learning_rate": 7.330124455735039e-05,
+      "loss": 0.0968,
+      "step": 54220
+    },
+    {
+      "epoch": 327.4,
+      "learning_rate": 7.32615267190115e-05,
+      "loss": 0.1013,
+      "step": 54240
+    },
+    {
+      "epoch": 327.53,
+      "learning_rate": 7.322179013928067e-05,
+      "loss": 0.104,
+      "step": 54260
+    },
+    {
+      "epoch": 327.65,
+      "learning_rate": 7.318203485017299e-05,
+      "loss": 0.1066,
+      "step": 54280
+    },
+    {
+      "epoch": 327.77,
+      "learning_rate": 7.314226088371854e-05,
+      "loss": 0.1099,
+      "step": 54300
+    },
+    {
+      "epoch": 327.89,
+      "learning_rate": 7.31024682719625e-05,
+      "loss": 0.1097,
+      "step": 54320
+    },
+    {
+      "epoch": 328.01,
+      "learning_rate": 7.306265704696504e-05,
+      "loss": 0.1103,
+      "step": 54340
+    },
+    {
+      "epoch": 328.13,
+      "learning_rate": 7.302282724080138e-05,
+      "loss": 0.0831,
+      "step": 54360
+    },
+    {
+      "epoch": 328.25,
+      "learning_rate": 7.298297888556164e-05,
+      "loss": 0.091,
+      "step": 54380
+    },
+    {
+      "epoch": 328.37,
+      "learning_rate": 7.294311201335093e-05,
+      "loss": 0.0931,
+      "step": 54400
+    },
+    {
+      "epoch": 328.49,
+      "learning_rate": 7.290322665628928e-05,
+      "loss": 0.0992,
+      "step": 54420
+    },
+    {
+      "epoch": 328.61,
+      "learning_rate": 7.286332284651159e-05,
+      "loss": 0.1046,
+      "step": 54440
+    },
+    {
+      "epoch": 328.73,
+      "learning_rate": 7.282340061616766e-05,
+      "loss": 0.1055,
+      "step": 54460
+    },
+    {
+      "epoch": 328.85,
+      "learning_rate": 7.278345999742208e-05,
+      "loss": 0.109,
+      "step": 54480
+    },
+    {
+      "epoch": 328.97,
+      "learning_rate": 7.274350102245431e-05,
+      "loss": 0.1102,
+      "step": 54500
+    },
+    {
+      "epoch": 329.09,
+      "learning_rate": 7.270352372345855e-05,
+      "loss": 0.0873,
+      "step": 54520
+    },
+    {
+      "epoch": 329.22,
+      "learning_rate": 7.266352813264378e-05,
+      "loss": 0.0888,
+      "step": 54540
+    },
+    {
+      "epoch": 329.34,
+      "learning_rate": 7.262351428223378e-05,
+      "loss": 0.0948,
+      "step": 54560
+    },
+    {
+      "epoch": 329.46,
+      "learning_rate": 7.258348220446695e-05,
+      "loss": 0.0994,
+      "step": 54580
+    },
+    {
+      "epoch": 329.58,
+      "learning_rate": 7.25434319315964e-05,
+      "loss": 0.1023,
+      "step": 54600
+    },
+    {
+      "epoch": 329.7,
+      "learning_rate": 7.250336349588994e-05,
+      "loss": 0.1052,
+      "step": 54620
+    },
+    {
+      "epoch": 329.82,
+      "learning_rate": 7.246327692962996e-05,
+      "loss": 0.1095,
+      "step": 54640
+    },
+    {
+      "epoch": 329.94,
+      "learning_rate": 7.24231722651135e-05,
+      "loss": 0.1091,
+      "step": 54660
+    },
+    {
+      "epoch": 330.06,
+      "learning_rate": 7.238304953465217e-05,
+      "loss": 0.0947,
+      "step": 54680
+    },
+    {
+      "epoch": 330.18,
+      "learning_rate": 7.234290877057208e-05,
+      "loss": 0.0863,
+      "step": 54700
+    },
+    {
+      "epoch": 330.3,
+      "learning_rate": 7.230275000521398e-05,
+      "loss": 0.0924,
+      "step": 54720
+    },
+    {
+      "epoch": 330.42,
+      "learning_rate": 7.226257327093304e-05,
+      "loss": 0.0977,
+      "step": 54740
+    },
+    {
+      "epoch": 330.54,
+      "learning_rate": 7.222237860009892e-05,
+      "loss": 0.1017,
+      "step": 54760
+    },
+    {
+      "epoch": 330.66,
+      "learning_rate": 7.218216602509574e-05,
+      "loss": 0.1038,
+      "step": 54780
+    },
+    {
+      "epoch": 330.78,
+      "learning_rate": 7.214193557832206e-05,
+      "loss": 0.1062,
+      "step": 54800
+    },
+    {
+      "epoch": 330.91,
+      "learning_rate": 7.21016872921908e-05,
+      "loss": 0.1092,
+      "step": 54820
+    },
+    {
+      "epoch": 331.03,
+      "learning_rate": 7.206142119912931e-05,
+      "loss": 0.1028,
+      "step": 54840
+    },
+    {
+      "epoch": 331.15,
+      "learning_rate": 7.202113733157923e-05,
+      "loss": 0.0834,
+      "step": 54860
+    },
+    {
+      "epoch": 331.27,
+      "learning_rate": 7.198083572199657e-05,
+      "loss": 0.0871,
+      "step": 54880
+    },
+    {
+      "epoch": 331.39,
+      "learning_rate": 7.194051640285157e-05,
+      "loss": 0.0929,
+      "step": 54900
+    },
+    {
+      "epoch": 331.51,
+      "learning_rate": 7.190017940662878e-05,
+      "loss": 0.0981,
+      "step": 54920
+    },
+    {
+      "epoch": 331.63,
+      "learning_rate": 7.185982476582705e-05,
+      "loss": 0.1029,
+      "step": 54940
+    },
+    {
+      "epoch": 331.75,
+      "learning_rate": 7.181945251295931e-05,
+      "loss": 0.1041,
+      "step": 54960
+    },
+    {
+      "epoch": 331.87,
+      "learning_rate": 7.17790626805528e-05,
+      "loss": 0.1068,
+      "step": 54980
+    },
+    {
+      "epoch": 331.99,
+      "learning_rate": 7.173865530114886e-05,
+      "loss": 0.1091,
+      "step": 55000
+    },
+    {
+      "epoch": 331.99,
+      "eval_accuracy": 0.001258013849512501,
+      "eval_loss": 20.59257698059082,
+      "eval_runtime": 18.2239,
+      "eval_samples_per_second": 111.941,
+      "eval_steps_per_second": 2.36,
+      "step": 55000
+    },
+    {
+      "epoch": 332.11,
+      "learning_rate": 7.1698230407303e-05,
+      "loss": 0.081,
+      "step": 55020
+    },
+    {
+      "epoch": 332.23,
+      "learning_rate": 7.16577880315848e-05,
+      "loss": 0.0891,
+      "step": 55040
+    },
+    {
+      "epoch": 332.35,
+      "learning_rate": 7.161732820657799e-05,
+      "loss": 0.0929,
+      "step": 55060
+    },
+    {
+      "epoch": 332.47,
+      "learning_rate": 7.157685096488029e-05,
+      "loss": 0.0964,
+      "step": 55080
+    },
+    {
+      "epoch": 332.6,
+      "learning_rate": 7.15383814827606e-05,
+      "loss": 0.0999,
+      "step": 55100
+    },
+    {
+      "epoch": 332.72,
+      "learning_rate": 7.14978703723281e-05,
+      "loss": 0.1031,
+      "step": 55120
+    },
+    {
+      "epoch": 332.84,
+      "learning_rate": 7.14573419414497e-05,
+      "loss": 0.1065,
+      "step": 55140
+    },
+    {
+      "epoch": 332.96,
+      "learning_rate": 7.14167962227785e-05,
+      "loss": 0.1079,
+      "step": 55160
+    },
+    {
+      "epoch": 333.08,
+      "learning_rate": 7.13762332489814e-05,
+      "loss": 0.0871,
+      "step": 55180
+    },
+    {
+      "epoch": 333.2,
+      "learning_rate": 7.133565305273927e-05,
+      "loss": 0.0879,
+      "step": 55200
+    },
+    {
+      "epoch": 333.32,
+      "learning_rate": 7.129505566674685e-05,
+      "loss": 0.0927,
+      "step": 55220
+    },
+    {
+      "epoch": 333.44,
+      "learning_rate": 7.125444112371272e-05,
+      "loss": 0.096,
+      "step": 55240
+    },
+    {
+      "epoch": 333.56,
+      "learning_rate": 7.121380945635927e-05,
+      "loss": 0.0951,
+      "step": 55260
+    },
+    {
+      "epoch": 333.68,
+      "learning_rate": 7.11731606974227e-05,
+      "loss": 0.0994,
+      "step": 55280
+    },
+    {
+      "epoch": 333.8,
+      "learning_rate": 7.1132494879653e-05,
+      "loss": 0.1019,
+      "step": 55300
+    },
+    {
+      "epoch": 333.92,
+      "learning_rate": 7.109181203581386e-05,
+      "loss": 0.1061,
+      "step": 55320
+    },
+    {
+      "epoch": 334.04,
+      "learning_rate": 7.105111219868274e-05,
+      "loss": 0.0956,
+      "step": 55340
+    },
+    {
+      "epoch": 334.16,
+      "learning_rate": 7.101039540105074e-05,
+      "loss": 0.0816,
+      "step": 55360
+    },
+    {
+      "epoch": 334.29,
+      "learning_rate": 7.096966167572268e-05,
+      "loss": 0.088,
+      "step": 55380
+    },
+    {
+      "epoch": 334.41,
+      "learning_rate": 7.092891105551694e-05,
+      "loss": 0.0918,
+      "step": 55400
+    },
+    {
+      "epoch": 334.53,
+      "learning_rate": 7.088814357326561e-05,
+      "loss": 0.0952,
+      "step": 55420
+    },
+    {
+      "epoch": 334.65,
+      "learning_rate": 7.084735926181428e-05,
+      "loss": 0.0985,
+      "step": 55440
+    },
+    {
+      "epoch": 334.77,
+      "learning_rate": 7.080655815402213e-05,
+      "loss": 0.1021,
+      "step": 55460
+    },
+    {
+      "epoch": 334.89,
+      "learning_rate": 7.076574028276188e-05,
+      "loss": 0.1049,
+      "step": 55480
+    },
+    {
+      "epoch": 335.01,
+      "learning_rate": 7.072490568091978e-05,
+      "loss": 0.1021,
+      "step": 55500
+    },
+    {
+      "epoch": 335.13,
+      "learning_rate": 7.06840543813955e-05,
+      "loss": 0.0766,
+      "step": 55520
+    },
+    {
+      "epoch": 335.25,
+      "learning_rate": 7.064318641710221e-05,
+      "loss": 0.0853,
+      "step": 55540
+    },
+    {
+      "epoch": 335.37,
+      "learning_rate": 7.060230182096648e-05,
+      "loss": 0.0908,
+      "step": 55560
+    },
+    {
+      "epoch": 335.49,
+      "learning_rate": 7.056140062592828e-05,
+      "loss": 0.0939,
+      "step": 55580
+    },
+    {
+      "epoch": 335.61,
+      "learning_rate": 7.052048286494101e-05,
+      "loss": 0.0962,
+      "step": 55600
+    },
+    {
+      "epoch": 335.73,
+      "learning_rate": 7.04795485709713e-05,
+      "loss": 0.1002,
+      "step": 55620
+    },
+    {
+      "epoch": 335.86,
+      "learning_rate": 7.043859777699925e-05,
+      "loss": 0.1016,
+      "step": 55640
+    },
+    {
+      "epoch": 335.98,
+      "learning_rate": 7.039763051601811e-05,
+      "loss": 0.1019,
+      "step": 55660
+    },
+    {
+      "epoch": 336.1,
+      "learning_rate": 7.03566468210345e-05,
+      "loss": 0.0824,
+      "step": 55680
+    },
+    {
+      "epoch": 336.22,
+      "learning_rate": 7.031564672506824e-05,
+      "loss": 0.0839,
+      "step": 55700
+    },
+    {
+      "epoch": 336.34,
+      "learning_rate": 7.027463026115235e-05,
+      "loss": 0.0895,
+      "step": 55720
+    },
+    {
+      "epoch": 336.46,
+      "learning_rate": 7.023359746233307e-05,
+      "loss": 0.0913,
+      "step": 55740
+    },
+    {
+      "epoch": 336.58,
+      "learning_rate": 7.019254836166976e-05,
+      "loss": 0.0952,
+      "step": 55760
+    },
+    {
+      "epoch": 336.7,
+      "learning_rate": 7.015148299223497e-05,
+      "loss": 0.0973,
+      "step": 55780
+    },
+    {
+      "epoch": 336.82,
+      "learning_rate": 7.01104013871143e-05,
+      "loss": 0.0999,
+      "step": 55800
+    },
+    {
+      "epoch": 336.94,
+      "learning_rate": 7.006930357940648e-05,
+      "loss": 0.1008,
+      "step": 55820
+    },
+    {
+      "epoch": 337.06,
+      "learning_rate": 7.002818960222326e-05,
+      "loss": 0.0874,
+      "step": 55840
+    },
+    {
+      "epoch": 337.18,
+      "learning_rate": 6.998705948868944e-05,
+      "loss": 0.08,
+      "step": 55860
+    },
+    {
+      "epoch": 337.3,
+      "learning_rate": 6.994591327194278e-05,
+      "loss": 0.0838,
+      "step": 55880
+    },
+    {
+      "epoch": 337.42,
+      "learning_rate": 6.990475098513406e-05,
+      "loss": 0.0882,
+      "step": 55900
+    },
+    {
+      "epoch": 337.55,
+      "learning_rate": 6.986357266142702e-05,
+      "loss": 0.0928,
+      "step": 55920
+    },
+    {
+      "epoch": 337.67,
+      "learning_rate": 6.982237833399825e-05,
+      "loss": 0.0946,
+      "step": 55940
+    },
+    {
+      "epoch": 337.79,
+      "learning_rate": 6.978116803603729e-05,
+      "loss": 0.0989,
+      "step": 55960
+    },
+    {
+      "epoch": 337.91,
+      "learning_rate": 6.973994180074654e-05,
+      "loss": 0.0996,
+      "step": 55980
+    },
+    {
+      "epoch": 338.03,
+      "learning_rate": 6.969869966134123e-05,
+      "loss": 0.0928,
+      "step": 56000
+    },
+    {
+      "epoch": 338.03,
+      "eval_accuracy": 0.0012570260098191265,
+      "eval_loss": 20.80438232421875,
+      "eval_runtime": 18.1697,
+      "eval_samples_per_second": 112.275,
+      "eval_steps_per_second": 2.367,
+      "step": 56000
+    },
+    {
+      "epoch": 338.15,
+      "learning_rate": 6.965744165104939e-05,
+      "loss": 0.0757,
+      "step": 56020
+    },
+    {
+      "epoch": 338.27,
+      "learning_rate": 6.961616780311188e-05,
+      "loss": 0.0818,
+      "step": 56040
+    },
+    {
+      "epoch": 338.39,
+      "learning_rate": 6.957487815078227e-05,
+      "loss": 0.0867,
+      "step": 56060
+    },
+    {
+      "epoch": 338.51,
+      "learning_rate": 6.953357272732691e-05,
+      "loss": 0.0897,
+      "step": 56080
+    },
+    {
+      "epoch": 338.63,
+      "learning_rate": 6.949225156602481e-05,
+      "loss": 0.0923,
+      "step": 56100
+    },
+    {
+      "epoch": 338.75,
+      "learning_rate": 6.945091470016771e-05,
+      "loss": 0.0941,
+      "step": 56120
+    },
+    {
+      "epoch": 338.87,
+      "learning_rate": 6.940956216305996e-05,
+      "loss": 0.0971,
+      "step": 56140
+    },
+    {
+      "epoch": 338.99,
+      "learning_rate": 6.936819398801856e-05,
+      "loss": 0.0988,
+      "step": 56160
+    },
+    {
+      "epoch": 339.11,
+      "learning_rate": 6.932681020837307e-05,
+      "loss": 0.0749,
+      "step": 56180
+    },
+    {
+      "epoch": 339.24,
+      "learning_rate": 6.928541085746571e-05,
+      "loss": 0.0812,
+      "step": 56200
+    },
+    {
+      "epoch": 339.36,
+      "learning_rate": 6.924399596865113e-05,
+      "loss": 0.0852,
+      "step": 56220
+    },
+    {
+      "epoch": 339.48,
+      "learning_rate": 6.92025655752966e-05,
+      "loss": 0.0877,
+      "step": 56240
+    },
+    {
+      "epoch": 339.6,
+      "learning_rate": 6.916111971078183e-05,
+      "loss": 0.0926,
+      "step": 56260
+    },
+    {
+      "epoch": 339.72,
+      "learning_rate": 6.9119658408499e-05,
+      "loss": 0.094,
+      "step": 56280
+    },
+    {
+      "epoch": 339.84,
+      "learning_rate": 6.907818170185273e-05,
+      "loss": 0.0972,
+      "step": 56300
+    },
+    {
+      "epoch": 339.96,
+      "learning_rate": 6.903668962426005e-05,
+      "loss": 0.097,
+      "step": 56320
+    },
+    {
+      "epoch": 340.08,
+      "learning_rate": 6.899518220915038e-05,
+      "loss": 0.0808,
+      "step": 56340
+    },
+    {
+      "epoch": 340.2,
+      "learning_rate": 6.895365948996552e-05,
+      "loss": 0.0772,
+      "step": 56360
+    },
+    {
+      "epoch": 340.32,
+      "learning_rate": 6.891212150015955e-05,
+      "loss": 0.0845,
+      "step": 56380
+    },
+    {
+      "epoch": 340.44,
+      "learning_rate": 6.887056827319885e-05,
+      "loss": 0.0884,
+      "step": 56400
+    },
+    {
+      "epoch": 340.56,
+      "learning_rate": 6.882899984256216e-05,
+      "loss": 0.0918,
+      "step": 56420
+    },
+    {
+      "epoch": 340.68,
+      "learning_rate": 6.878741624174039e-05,
+      "loss": 0.092,
+      "step": 56440
+    },
+    {
+      "epoch": 340.8,
+      "learning_rate": 6.87458175042367e-05,
+      "loss": 0.0932,
+      "step": 56460
+    },
+    {
+      "epoch": 340.93,
+      "learning_rate": 6.870420366356642e-05,
+      "loss": 0.0965,
+      "step": 56480
+    },
+    {
+      "epoch": 341.05,
+      "learning_rate": 6.86625747532571e-05,
+      "loss": 0.0863,
+      "step": 56500
+    },
+    {
+      "epoch": 341.17,
+      "learning_rate": 6.862093080684838e-05,
+      "loss": 0.0756,
+      "step": 56520
+    },
+    {
+      "epoch": 341.29,
+      "learning_rate": 6.857927185789204e-05,
+      "loss": 0.0826,
+      "step": 56540
+    },
+    {
+      "epoch": 341.41,
+      "learning_rate": 6.853759793995196e-05,
+      "loss": 0.0862,
+      "step": 56560
+    },
+    {
+      "epoch": 341.53,
+      "learning_rate": 6.849590908660404e-05,
+      "loss": 0.0896,
+      "step": 56580
+    },
+    {
+      "epoch": 341.65,
+      "learning_rate": 6.845420533143627e-05,
+      "loss": 0.0913,
+      "step": 56600
+    },
+    {
+      "epoch": 341.77,
+      "learning_rate": 6.841248670804853e-05,
+      "loss": 0.0915,
+      "step": 56620
+    },
+    {
+      "epoch": 341.89,
+      "learning_rate": 6.837075325005286e-05,
+      "loss": 0.0938,
+      "step": 56640
+    },
+    {
+      "epoch": 342.01,
+      "learning_rate": 6.832900499107311e-05,
+      "loss": 0.0932,
+      "step": 56660
+    },
+    {
+      "epoch": 342.13,
+      "learning_rate": 6.82872419647451e-05,
+      "loss": 0.0713,
+      "step": 56680
+    },
+    {
+      "epoch": 342.25,
+      "learning_rate": 6.824546420471653e-05,
+      "loss": 0.0782,
+      "step": 56700
+    },
+    {
+      "epoch": 342.37,
+      "learning_rate": 6.820367174464703e-05,
+      "loss": 0.0839,
+      "step": 56720
+    },
+    {
+      "epoch": 342.49,
+      "learning_rate": 6.816186461820798e-05,
+      "loss": 0.0853,
+      "step": 56740
+    },
+    {
+      "epoch": 342.62,
+      "learning_rate": 6.812004285908266e-05,
+      "loss": 0.0841,
+      "step": 56760
+    },
+    {
+      "epoch": 342.74,
+      "learning_rate": 6.807820650096609e-05,
+      "loss": 0.0881,
+      "step": 56780
+    },
+    {
+      "epoch": 342.86,
+      "learning_rate": 6.803635557756507e-05,
+      "loss": 0.0902,
+      "step": 56800
+    },
+    {
+      "epoch": 342.98,
+      "learning_rate": 6.799449012259816e-05,
+      "loss": 0.093,
+      "step": 56820
+    },
+    {
+      "epoch": 343.1,
+      "learning_rate": 6.795261016979555e-05,
+      "loss": 0.0726,
+      "step": 56840
+    },
+    {
+      "epoch": 343.22,
+      "learning_rate": 6.791071575289922e-05,
+      "loss": 0.0756,
+      "step": 56860
+    },
+    {
+      "epoch": 343.34,
+      "learning_rate": 6.786880690566268e-05,
+      "loss": 0.0801,
+      "step": 56880
+    },
+    {
+      "epoch": 343.46,
+      "learning_rate": 6.78268836618512e-05,
+      "loss": 0.0826,
+      "step": 56900
+    },
+    {
+      "epoch": 343.58,
+      "learning_rate": 6.778494605524151e-05,
+      "loss": 0.0865,
+      "step": 56920
+    },
+    {
+      "epoch": 343.7,
+      "learning_rate": 6.774299411962203e-05,
+      "loss": 0.0899,
+      "step": 56940
+    },
+    {
+      "epoch": 343.82,
+      "learning_rate": 6.770102788879267e-05,
+      "loss": 0.0901,
+      "step": 56960
+    },
+    {
+      "epoch": 343.94,
+      "learning_rate": 6.765904739656486e-05,
+      "loss": 0.0931,
+      "step": 56980
+    },
+    {
+      "epoch": 344.06,
+      "learning_rate": 6.761705267676153e-05,
+      "loss": 0.0812,
+      "step": 57000
+    },
+    {
+      "epoch": 344.06,
+      "eval_accuracy": 0.0012797463227667415,
+      "eval_loss": 20.987285614013672,
+      "eval_runtime": 18.1769,
+      "eval_samples_per_second": 112.23,
+      "eval_steps_per_second": 2.366,
+      "step": 57000
+    },
+    {
+      "epoch": 344.19,
+      "learning_rate": 6.757504376321704e-05,
+      "loss": 0.0736,
+      "step": 57020
+    },
+    {
+      "epoch": 344.31,
+      "learning_rate": 6.753302068977725e-05,
+      "loss": 0.0799,
+      "step": 57040
+    },
+    {
+      "epoch": 344.43,
+      "learning_rate": 6.749098349029935e-05,
+      "loss": 0.0823,
+      "step": 57060
+    },
+    {
+      "epoch": 344.55,
+      "learning_rate": 6.7448932198652e-05,
+      "loss": 0.0857,
+      "step": 57080
+    },
+    {
+      "epoch": 344.67,
+      "learning_rate": 6.740686684871515e-05,
+      "loss": 0.0896,
+      "step": 57100
+    },
+    {
+      "epoch": 344.79,
+      "learning_rate": 6.736478747438007e-05,
+      "loss": 0.0906,
+      "step": 57120
+    },
+    {
+      "epoch": 344.91,
+      "learning_rate": 6.732269410954938e-05,
+      "loss": 0.0916,
+      "step": 57140
+    },
+    {
+      "epoch": 345.03,
+      "learning_rate": 6.728058678813694e-05,
+      "loss": 0.0864,
+      "step": 57160
+    },
+    {
+      "epoch": 345.15,
+      "learning_rate": 6.723846554406782e-05,
+      "loss": 0.0712,
+      "step": 57180
+    },
+    {
+      "epoch": 345.27,
+      "learning_rate": 6.719633041127839e-05,
+      "loss": 0.0783,
+      "step": 57200
+    },
+    {
+      "epoch": 345.39,
+      "learning_rate": 6.715418142371614e-05,
+      "loss": 0.0822,
+      "step": 57220
+    },
+    {
+      "epoch": 345.51,
+      "learning_rate": 6.711201861533978e-05,
+      "loss": 0.0831,
+      "step": 57240
+    },
+    {
+      "epoch": 345.63,
+      "learning_rate": 6.707195117679314e-05,
+      "loss": 0.0857,
+      "step": 57260
+    },
+    {
+      "epoch": 345.75,
+      "learning_rate": 6.7029761515545e-05,
+      "loss": 0.0866,
+      "step": 57280
+    },
+    {
+      "epoch": 345.88,
+      "learning_rate": 6.698755813372561e-05,
+      "loss": 0.0893,
+      "step": 57300
+    },
+    {
+      "epoch": 346.0,
+      "learning_rate": 6.694534106533745e-05,
+      "loss": 0.0919,
+      "step": 57320
+    },
+    {
+      "epoch": 346.12,
+      "learning_rate": 6.690311034439408e-05,
+      "loss": 0.0691,
+      "step": 57340
+    },
+    {
+      "epoch": 346.24,
+      "learning_rate": 6.686086600492e-05,
+      "loss": 0.0745,
+      "step": 57360
+    },
+    {
+      "epoch": 346.36,
+      "learning_rate": 6.681860808095074e-05,
+      "loss": 0.0775,
+      "step": 57380
+    },
+    {
+      "epoch": 346.48,
+      "learning_rate": 6.677633660653271e-05,
+      "loss": 0.0811,
+      "step": 57400
+    },
+    {
+      "epoch": 346.6,
+      "learning_rate": 6.673405161572331e-05,
+      "loss": 0.0844,
+      "step": 57420
+    },
+    {
+      "epoch": 346.72,
+      "learning_rate": 6.669175314259076e-05,
+      "loss": 0.0858,
+      "step": 57440
+    },
+    {
+      "epoch": 346.84,
+      "learning_rate": 6.66494412212142e-05,
+      "loss": 0.0876,
+      "step": 57460
+    },
+    {
+      "epoch": 346.96,
+      "learning_rate": 6.660711588568358e-05,
+      "loss": 0.0886,
+      "step": 57480
+    },
+    {
+      "epoch": 347.08,
+      "learning_rate": 6.656477717009965e-05,
+      "loss": 0.0728,
+      "step": 57500
+    },
+    {
+      "epoch": 347.2,
+      "learning_rate": 6.652242510857395e-05,
+      "loss": 0.0716,
+      "step": 57520
+    },
+    {
+      "epoch": 347.32,
+      "learning_rate": 6.648005973522877e-05,
+      "loss": 0.0758,
+      "step": 57540
+    },
+    {
+      "epoch": 347.44,
+      "learning_rate": 6.64376810841971e-05,
+      "loss": 0.0783,
+      "step": 57560
+    },
+    {
+      "epoch": 347.57,
+      "learning_rate": 6.63952891896227e-05,
+      "loss": 0.0818,
+      "step": 57580
+    },
+    {
+      "epoch": 347.69,
+      "learning_rate": 6.635288408565995e-05,
+      "loss": 0.0867,
+      "step": 57600
+    },
+    {
+      "epoch": 347.81,
+      "learning_rate": 6.631046580647384e-05,
+      "loss": 0.0856,
+      "step": 57620
+    },
+    {
+      "epoch": 347.93,
+      "learning_rate": 6.626803438624005e-05,
+      "loss": 0.0866,
+      "step": 57640
+    },
+    {
+      "epoch": 348.05,
+      "learning_rate": 6.622558985914477e-05,
+      "loss": 0.0786,
+      "step": 57660
+    },
+    {
+      "epoch": 348.17,
+      "learning_rate": 6.618313225938483e-05,
+      "loss": 0.0692,
+      "step": 57680
+    },
+    {
+      "epoch": 348.29,
+      "learning_rate": 6.614066162116752e-05,
+      "loss": 0.0728,
+      "step": 57700
+    },
+    {
+      "epoch": 348.41,
+      "learning_rate": 6.609817797871067e-05,
+      "loss": 0.0777,
+      "step": 57720
+    },
+    {
+      "epoch": 348.53,
+      "learning_rate": 6.605568136624262e-05,
+      "loss": 0.0802,
+      "step": 57740
+    },
+    {
+      "epoch": 348.65,
+      "learning_rate": 6.601317181800207e-05,
+      "loss": 0.0828,
+      "step": 57760
+    },
+    {
+      "epoch": 348.77,
+      "learning_rate": 6.597064936823822e-05,
+      "loss": 0.0843,
+      "step": 57780
+    },
+    {
+      "epoch": 348.89,
+      "learning_rate": 6.592811405121064e-05,
+      "loss": 0.0857,
+      "step": 57800
+    },
+    {
+      "epoch": 349.01,
+      "learning_rate": 6.588556590118927e-05,
+      "loss": 0.0846,
+      "step": 57820
+    },
+    {
+      "epoch": 349.13,
+      "learning_rate": 6.584300495245437e-05,
+      "loss": 0.0665,
+      "step": 57840
+    },
+    {
+      "epoch": 349.26,
+      "learning_rate": 6.580043123929655e-05,
+      "loss": 0.0718,
+      "step": 57860
+    },
+    {
+      "epoch": 349.38,
+      "learning_rate": 6.575784479601667e-05,
+      "loss": 0.076,
+      "step": 57880
+    },
+    {
+      "epoch": 349.5,
+      "learning_rate": 6.571524565692586e-05,
+      "loss": 0.0778,
+      "step": 57900
+    },
+    {
+      "epoch": 349.62,
+      "learning_rate": 6.567263385634547e-05,
+      "loss": 0.0808,
+      "step": 57920
+    },
+    {
+      "epoch": 349.74,
+      "learning_rate": 6.563000942860706e-05,
+      "loss": 0.083,
+      "step": 57940
+    },
+    {
+      "epoch": 349.86,
+      "learning_rate": 6.558737240805237e-05,
+      "loss": 0.085,
+      "step": 57960
+    },
+    {
+      "epoch": 349.98,
+      "learning_rate": 6.554472282903328e-05,
+      "loss": 0.0865,
+      "step": 57980
+    },
+    {
+      "epoch": 350.1,
+      "learning_rate": 6.550206072591176e-05,
+      "loss": 0.0677,
+      "step": 58000
+    },
+    {
+      "epoch": 350.1,
+      "eval_accuracy": 0.001262953047979374,
+      "eval_loss": 21.19306755065918,
+      "eval_runtime": 18.1733,
+      "eval_samples_per_second": 112.252,
+      "eval_steps_per_second": 2.366,
+      "step": 58000
+    },
+    {
+      "epoch": 350.22,
+      "learning_rate": 6.545938613305993e-05,
+      "loss": 0.0676,
+      "step": 58020
+    },
+    {
+      "epoch": 350.34,
+      "learning_rate": 6.541669908485992e-05,
+      "loss": 0.0738,
+      "step": 58040
+    },
+    {
+      "epoch": 350.46,
+      "learning_rate": 6.53739996157039e-05,
+      "loss": 0.0772,
+      "step": 58060
+    },
+    {
+      "epoch": 350.58,
+      "learning_rate": 6.533128775999411e-05,
+      "loss": 0.0774,
+      "step": 58080
+    },
+    {
+      "epoch": 350.7,
+      "learning_rate": 6.528856355214267e-05,
+      "loss": 0.0799,
+      "step": 58100
+    },
+    {
+      "epoch": 350.82,
+      "learning_rate": 6.524582702657171e-05,
+      "loss": 0.0817,
+      "step": 58120
+    },
+    {
+      "epoch": 350.95,
+      "learning_rate": 6.52030782177133e-05,
+      "loss": 0.0831,
+      "step": 58140
+    },
+    {
+      "epoch": 351.07,
+      "learning_rate": 6.516031716000936e-05,
+      "loss": 0.0708,
+      "step": 58160
+    },
+    {
+      "epoch": 351.19,
+      "learning_rate": 6.511754388791173e-05,
+      "loss": 0.0686,
+      "step": 58180
+    },
+    {
+      "epoch": 351.31,
+      "learning_rate": 6.507475843588202e-05,
+      "loss": 0.0719,
+      "step": 58200
+    },
+    {
+      "epoch": 351.43,
+      "learning_rate": 6.503196083839174e-05,
+      "loss": 0.0742,
+      "step": 58220
+    },
+    {
+      "epoch": 351.55,
+      "learning_rate": 6.498915112992212e-05,
+      "loss": 0.0771,
+      "step": 58240
+    },
+    {
+      "epoch": 351.67,
+      "learning_rate": 6.494632934496418e-05,
+      "loss": 0.0786,
+      "step": 58260
+    },
+    {
+      "epoch": 351.79,
+      "learning_rate": 6.490349551801864e-05,
+      "loss": 0.081,
+      "step": 58280
+    },
+    {
+      "epoch": 351.91,
+      "learning_rate": 6.486064968359596e-05,
+      "loss": 0.0835,
+      "step": 58300
+    },
+    {
+      "epoch": 352.03,
+      "learning_rate": 6.481779187621621e-05,
+      "loss": 0.0775,
+      "step": 58320
+    },
+    {
+      "epoch": 352.15,
+      "learning_rate": 6.477492213040924e-05,
+      "loss": 0.0634,
+      "step": 58340
+    },
+    {
+      "epoch": 352.27,
+      "learning_rate": 6.473204048071432e-05,
+      "loss": 0.0693,
+      "step": 58360
+    },
+    {
+      "epoch": 352.39,
+      "learning_rate": 6.468914696168053e-05,
+      "loss": 0.0718,
+      "step": 58380
+    },
+    {
+      "epoch": 352.52,
+      "learning_rate": 6.464624160786633e-05,
+      "loss": 0.0758,
+      "step": 58400
+    },
+    {
+      "epoch": 352.64,
+      "learning_rate": 6.46033244538398e-05,
+      "loss": 0.0774,
+      "step": 58420
+    },
+    {
+      "epoch": 352.76,
+      "learning_rate": 6.456039553417852e-05,
+      "loss": 0.0814,
+      "step": 58440
+    },
+    {
+      "epoch": 352.88,
+      "learning_rate": 6.451745488346956e-05,
+      "loss": 0.0812,
+      "step": 58460
+    },
+    {
+      "epoch": 353.0,
+      "learning_rate": 6.447450253630942e-05,
+      "loss": 0.0845,
+      "step": 58480
+    },
+    {
+      "epoch": 353.12,
+      "learning_rate": 6.443153852730404e-05,
+      "loss": 0.062,
+      "step": 58500
+    },
+    {
+      "epoch": 353.24,
+      "learning_rate": 6.438856289106871e-05,
+      "loss": 0.0674,
+      "step": 58520
+    },
+    {
+      "epoch": 353.36,
+      "learning_rate": 6.434557566222817e-05,
+      "loss": 0.0721,
+      "step": 58540
+    },
+    {
+      "epoch": 353.48,
+      "learning_rate": 6.430257687541641e-05,
+      "loss": 0.0739,
+      "step": 58560
+    },
+    {
+      "epoch": 353.6,
+      "learning_rate": 6.425956656527681e-05,
+      "loss": 0.0757,
+      "step": 58580
+    },
+    {
+      "epoch": 353.72,
+      "learning_rate": 6.421654476646199e-05,
+      "loss": 0.0778,
+      "step": 58600
+    },
+    {
+      "epoch": 353.84,
+      "learning_rate": 6.417351151363381e-05,
+      "loss": 0.0806,
+      "step": 58620
+    },
+    {
+      "epoch": 353.96,
+      "learning_rate": 6.413046684146343e-05,
+      "loss": 0.0811,
+      "step": 58640
+    },
+    {
+      "epoch": 354.08,
+      "learning_rate": 6.408741078463114e-05,
+      "loss": 0.0667,
+      "step": 58660
+    },
+    {
+      "epoch": 354.21,
+      "learning_rate": 6.404434337782642e-05,
+      "loss": 0.065,
+      "step": 58680
+    },
+    {
+      "epoch": 354.33,
+      "learning_rate": 6.400126465574793e-05,
+      "loss": 0.0699,
+      "step": 58700
+    },
+    {
+      "epoch": 354.45,
+      "learning_rate": 6.395817465310338e-05,
+      "loss": 0.0726,
+      "step": 58720
+    },
+    {
+      "epoch": 354.57,
+      "learning_rate": 6.391507340460963e-05,
+      "loss": 0.0744,
+      "step": 58740
+    },
+    {
+      "epoch": 354.69,
+      "learning_rate": 6.387196094499258e-05,
+      "loss": 0.0774,
+      "step": 58760
+    },
+    {
+      "epoch": 354.81,
+      "learning_rate": 6.382883730898717e-05,
+      "loss": 0.0778,
+      "step": 58780
+    },
+    {
+      "epoch": 354.93,
+      "learning_rate": 6.378570253133729e-05,
+      "loss": 0.0795,
+      "step": 58800
+    },
+    {
+      "epoch": 355.05,
+      "learning_rate": 6.37425566467959e-05,
+      "loss": 0.0724,
+      "step": 58820
+    },
+    {
+      "epoch": 355.17,
+      "learning_rate": 6.369939969012483e-05,
+      "loss": 0.0633,
+      "step": 58840
+    },
+    {
+      "epoch": 355.29,
+      "learning_rate": 6.365623169609487e-05,
+      "loss": 0.068,
+      "step": 58860
+    },
+    {
+      "epoch": 355.41,
+      "learning_rate": 6.36130526994857e-05,
+      "loss": 0.0704,
+      "step": 58880
+    },
+    {
+      "epoch": 355.53,
+      "learning_rate": 6.356986273508585e-05,
+      "loss": 0.0718,
+      "step": 58900
+    },
+    {
+      "epoch": 355.65,
+      "learning_rate": 6.35266618376927e-05,
+      "loss": 0.0759,
+      "step": 58920
+    },
+    {
+      "epoch": 355.77,
+      "learning_rate": 6.34834500421124e-05,
+      "loss": 0.0765,
+      "step": 58940
+    },
+    {
+      "epoch": 355.9,
+      "learning_rate": 6.344022738315998e-05,
+      "loss": 0.0778,
+      "step": 58960
+    },
+    {
+      "epoch": 356.02,
+      "learning_rate": 6.33969938956591e-05,
+      "loss": 0.0761,
+      "step": 58980
+    },
+    {
+      "epoch": 356.14,
+      "learning_rate": 6.335374961444222e-05,
+      "loss": 0.0609,
+      "step": 59000
+    },
+    {
+      "epoch": 356.14,
+      "eval_accuracy": 0.0012891307998537998,
+      "eval_loss": 21.364957809448242,
+      "eval_runtime": 18.1818,
+      "eval_samples_per_second": 112.2,
+      "eval_steps_per_second": 2.365,
+      "step": 59000
+    },
+    {
+      "epoch": 356.26,
+      "learning_rate": 6.331049457435046e-05,
+      "loss": 0.0655,
+      "step": 59020
+    },
+    {
+      "epoch": 356.38,
+      "learning_rate": 6.326722881023366e-05,
+      "loss": 0.0695,
+      "step": 59040
+    },
+    {
+      "epoch": 356.5,
+      "learning_rate": 6.322395235695022e-05,
+      "loss": 0.072,
+      "step": 59060
+    },
+    {
+      "epoch": 356.62,
+      "learning_rate": 6.318066524936725e-05,
+      "loss": 0.0735,
+      "step": 59080
+    },
+    {
+      "epoch": 356.74,
+      "learning_rate": 6.313736752236033e-05,
+      "loss": 0.0754,
+      "step": 59100
+    },
+    {
+      "epoch": 356.86,
+      "learning_rate": 6.309405921081369e-05,
+      "loss": 0.0771,
+      "step": 59120
+    },
+    {
+      "epoch": 356.98,
+      "learning_rate": 6.305074034962006e-05,
+      "loss": 0.0799,
+      "step": 59140
+    },
+    {
+      "epoch": 357.1,
+      "learning_rate": 6.300741097368065e-05,
+      "loss": 0.0621,
+      "step": 59160
+    },
+    {
+      "epoch": 357.22,
+      "learning_rate": 6.296407111790516e-05,
+      "loss": 0.0634,
+      "step": 59180
+    },
+    {
+      "epoch": 357.34,
+      "learning_rate": 6.292072081721173e-05,
+      "loss": 0.0662,
+      "step": 59200
+    },
+    {
+      "epoch": 357.46,
+      "learning_rate": 6.287736010652693e-05,
+      "loss": 0.0706,
+      "step": 59220
+    },
+    {
+      "epoch": 357.59,
+      "learning_rate": 6.283398902078567e-05,
+      "loss": 0.0729,
+      "step": 59240
+    },
+    {
+      "epoch": 357.71,
+      "learning_rate": 6.279060759493128e-05,
+      "loss": 0.0734,
+      "step": 59260
+    },
+    {
+      "epoch": 357.83,
+      "learning_rate": 6.274938569467407e-05,
+      "loss": 0.0742,
+      "step": 59280
+    },
+    {
+      "epoch": 357.95,
+      "learning_rate": 6.270598420613623e-05,
+      "loss": 0.0771,
+      "step": 59300
+    },
+    {
+      "epoch": 358.07,
+      "learning_rate": 6.266257248061641e-05,
+      "loss": 0.0666,
+      "step": 59320
+    },
+    {
+      "epoch": 358.19,
+      "learning_rate": 6.261915055309066e-05,
+      "loss": 0.0616,
+      "step": 59340
+    },
+    {
+      "epoch": 358.31,
+      "learning_rate": 6.257571845854323e-05,
+      "loss": 0.0649,
+      "step": 59360
+    },
+    {
+      "epoch": 358.43,
+      "learning_rate": 6.253227623196658e-05,
+      "loss": 0.0685,
+      "step": 59380
+    },
+    {
+      "epoch": 358.55,
+      "learning_rate": 6.248882390836135e-05,
+      "loss": 0.0696,
+      "step": 59400
+    },
+    {
+      "epoch": 358.67,
+      "learning_rate": 6.244536152273626e-05,
+      "loss": 0.0713,
+      "step": 59420
+    },
+    {
+      "epoch": 358.79,
+      "learning_rate": 6.240188911010818e-05,
+      "loss": 0.0739,
+      "step": 59440
+    },
+    {
+      "epoch": 358.91,
+      "learning_rate": 6.235840670550204e-05,
+      "loss": 0.0753,
+      "step": 59460
+    },
+    {
+      "epoch": 359.03,
+      "learning_rate": 6.231491434395087e-05,
+      "loss": 0.0715,
+      "step": 59480
+    },
+    {
+      "epoch": 359.15,
+      "learning_rate": 6.22714120604956e-05,
+      "loss": 0.0591,
+      "step": 59500
+    },
+    {
+      "epoch": 359.28,
+      "learning_rate": 6.222789989018532e-05,
+      "loss": 0.0642,
+      "step": 59520
+    },
+    {
+      "epoch": 359.4,
+      "learning_rate": 6.218437786807694e-05,
+      "loss": 0.0675,
+      "step": 59540
+    },
+    {
+      "epoch": 359.52,
+      "learning_rate": 6.21408460292354e-05,
+      "loss": 0.0701,
+      "step": 59560
+    },
+    {
+      "epoch": 359.64,
+      "learning_rate": 6.20973044087335e-05,
+      "loss": 0.0718,
+      "step": 59580
+    },
+    {
+      "epoch": 359.76,
+      "learning_rate": 6.205375304165194e-05,
+      "loss": 0.0729,
+      "step": 59600
+    },
+    {
+      "epoch": 359.88,
+      "learning_rate": 6.201019196307927e-05,
+      "loss": 0.0748,
+      "step": 59620
+    },
+    {
+      "epoch": 360.0,
+      "learning_rate": 6.196662120811186e-05,
+      "loss": 0.0779,
+      "step": 59640
+    },
+    {
+      "epoch": 360.12,
+      "learning_rate": 6.192304081185389e-05,
+      "loss": 0.058,
+      "step": 59660
+    },
+    {
+      "epoch": 360.24,
+      "learning_rate": 6.187945080941731e-05,
+      "loss": 0.0629,
+      "step": 59680
+    },
+    {
+      "epoch": 360.36,
+      "learning_rate": 6.183585123592178e-05,
+      "loss": 0.0668,
+      "step": 59700
+    },
+    {
+      "epoch": 360.48,
+      "learning_rate": 6.179224212649466e-05,
+      "loss": 0.0669,
+      "step": 59720
+    },
+    {
+      "epoch": 360.6,
+      "learning_rate": 6.174862351627108e-05,
+      "loss": 0.0683,
+      "step": 59740
+    },
+    {
+      "epoch": 360.72,
+      "learning_rate": 6.170499544039372e-05,
+      "loss": 0.0713,
+      "step": 59760
+    },
+    {
+      "epoch": 360.85,
+      "learning_rate": 6.166135793401296e-05,
+      "loss": 0.0733,
+      "step": 59780
+    },
+    {
+      "epoch": 360.97,
+      "learning_rate": 6.161771103228674e-05,
+      "loss": 0.0747,
+      "step": 59800
+    },
+    {
+      "epoch": 361.09,
+      "learning_rate": 6.157405477038059e-05,
+      "loss": 0.061,
+      "step": 59820
+    },
+    {
+      "epoch": 361.21,
+      "learning_rate": 6.153038918346756e-05,
+      "loss": 0.0595,
+      "step": 59840
+    },
+    {
+      "epoch": 361.33,
+      "learning_rate": 6.148671430672821e-05,
+      "loss": 0.0635,
+      "step": 59860
+    },
+    {
+      "epoch": 361.45,
+      "learning_rate": 6.144303017535066e-05,
+      "loss": 0.0664,
+      "step": 59880
+    },
+    {
+      "epoch": 361.57,
+      "learning_rate": 6.139933682453036e-05,
+      "loss": 0.0681,
+      "step": 59900
+    },
+    {
+      "epoch": 361.69,
+      "learning_rate": 6.135563428947027e-05,
+      "loss": 0.0702,
+      "step": 59920
+    },
+    {
+      "epoch": 361.81,
+      "learning_rate": 6.131192260538073e-05,
+      "loss": 0.0721,
+      "step": 59940
+    },
+    {
+      "epoch": 361.93,
+      "learning_rate": 6.12682018074795e-05,
+      "loss": 0.0725,
+      "step": 59960
+    },
+    {
+      "epoch": 362.05,
+      "learning_rate": 6.122447193099158e-05,
+      "loss": 0.065,
+      "step": 59980
+    },
+    {
+      "epoch": 362.17,
+      "learning_rate": 6.118073301114937e-05,
+      "loss": 0.058,
+      "step": 60000
+    },
+    {
+      "epoch": 362.17,
+      "eval_accuracy": 0.001262953047979374,
+      "eval_loss": 21.586803436279297,
+      "eval_runtime": 18.2045,
+      "eval_samples_per_second": 112.06,
+      "eval_steps_per_second": 2.362,
+      "step": 60000
+    },
+    {
+      "epoch": 362.29,
+      "learning_rate": 6.113698508319251e-05,
+      "loss": 0.0637,
+      "step": 60020
+    },
+    {
+      "epoch": 362.41,
+      "learning_rate": 6.109322818236793e-05,
+      "loss": 0.0646,
+      "step": 60040
+    },
+    {
+      "epoch": 362.54,
+      "learning_rate": 6.104946234392979e-05,
+      "loss": 0.0672,
+      "step": 60060
+    },
+    {
+      "epoch": 362.66,
+      "learning_rate": 6.1005687603139394e-05,
+      "loss": 0.0685,
+      "step": 60080
+    },
+    {
+      "epoch": 362.78,
+      "learning_rate": 6.096190399526529e-05,
+      "loss": 0.0703,
+      "step": 60100
+    },
+    {
+      "epoch": 362.9,
+      "learning_rate": 6.091811155558313e-05,
+      "loss": 0.0718,
+      "step": 60120
+    },
+    {
+      "epoch": 363.02,
+      "learning_rate": 6.0874310319375706e-05,
+      "loss": 0.0694,
+      "step": 60140
+    },
+    {
+      "epoch": 363.14,
+      "learning_rate": 6.083050032193286e-05,
+      "loss": 0.0545,
+      "step": 60160
+    },
+    {
+      "epoch": 363.26,
+      "learning_rate": 6.078668159855154e-05,
+      "loss": 0.0605,
+      "step": 60180
+    },
+    {
+      "epoch": 363.38,
+      "learning_rate": 6.074285418453567e-05,
+      "loss": 0.0622,
+      "step": 60200
+    },
+    {
+      "epoch": 363.5,
+      "learning_rate": 6.069901811519623e-05,
+      "loss": 0.0648,
+      "step": 60220
+    },
+    {
+      "epoch": 363.62,
+      "learning_rate": 6.065517342585114e-05,
+      "loss": 0.0659,
+      "step": 60240
+    },
+    {
+      "epoch": 363.74,
+      "learning_rate": 6.061132015182528e-05,
+      "loss": 0.0681,
+      "step": 60260
+    },
+    {
+      "epoch": 363.86,
+      "learning_rate": 6.056745832845043e-05,
+      "loss": 0.0703,
+      "step": 60280
+    },
+    {
+      "epoch": 363.98,
+      "learning_rate": 6.052358799106528e-05,
+      "loss": 0.0702,
+      "step": 60300
+    },
+    {
+      "epoch": 364.1,
+      "learning_rate": 6.047970917501538e-05,
+      "loss": 0.0556,
+      "step": 60320
+    },
+    {
+      "epoch": 364.23,
+      "learning_rate": 6.043582191565307e-05,
+      "loss": 0.057,
+      "step": 60340
+    },
+    {
+      "epoch": 364.35,
+      "learning_rate": 6.0391926248337536e-05,
+      "loss": 0.0602,
+      "step": 60360
+    },
+    {
+      "epoch": 364.47,
+      "learning_rate": 6.0348022208434716e-05,
+      "loss": 0.0633,
+      "step": 60380
+    },
+    {
+      "epoch": 364.59,
+      "learning_rate": 6.0304109831317324e-05,
+      "loss": 0.0652,
+      "step": 60400
+    },
+    {
+      "epoch": 364.71,
+      "learning_rate": 6.026018915236474e-05,
+      "loss": 0.0675,
+      "step": 60420
+    },
+    {
+      "epoch": 364.83,
+      "learning_rate": 6.021626020696311e-05,
+      "loss": 0.0677,
+      "step": 60440
+    },
+    {
+      "epoch": 364.95,
+      "learning_rate": 6.017232303050513e-05,
+      "loss": 0.0697,
+      "step": 60460
+    },
+    {
+      "epoch": 365.07,
+      "learning_rate": 6.012837765839022e-05,
+      "loss": 0.0609,
+      "step": 60480
+    },
+    {
+      "epoch": 365.19,
+      "learning_rate": 6.008442412602438e-05,
+      "loss": 0.0564,
+      "step": 60500
+    },
+    {
+      "epoch": 365.31,
+      "learning_rate": 6.004046246882016e-05,
+      "loss": 0.0598,
+      "step": 60520
+    },
+    {
+      "epoch": 365.43,
+      "learning_rate": 5.999649272219668e-05,
+      "loss": 0.0632,
+      "step": 60540
+    },
+    {
+      "epoch": 365.55,
+      "learning_rate": 5.995251492157958e-05,
+      "loss": 0.0652,
+      "step": 60560
+    },
+    {
+      "epoch": 365.67,
+      "learning_rate": 5.990852910240098e-05,
+      "loss": 0.0671,
+      "step": 60580
+    },
+    {
+      "epoch": 365.79,
+      "learning_rate": 5.9864535300099435e-05,
+      "loss": 0.0678,
+      "step": 60600
+    },
+    {
+      "epoch": 365.92,
+      "learning_rate": 5.9820533550119986e-05,
+      "loss": 0.0679,
+      "step": 60620
+    },
+    {
+      "epoch": 366.04,
+      "learning_rate": 5.9776523887914026e-05,
+      "loss": 0.0634,
+      "step": 60640
+    },
+    {
+      "epoch": 366.16,
+      "learning_rate": 5.973250634893936e-05,
+      "loss": 0.0533,
+      "step": 60660
+    },
+    {
+      "epoch": 366.28,
+      "learning_rate": 5.968848096866011e-05,
+      "loss": 0.058,
+      "step": 60680
+    },
+    {
+      "epoch": 366.4,
+      "learning_rate": 5.964444778254675e-05,
+      "loss": 0.0602,
+      "step": 60700
+    },
+    {
+      "epoch": 366.52,
+      "learning_rate": 5.9600406826076006e-05,
+      "loss": 0.0621,
+      "step": 60720
+    },
+    {
+      "epoch": 366.64,
+      "learning_rate": 5.955635813473089e-05,
+      "loss": 0.0642,
+      "step": 60740
+    },
+    {
+      "epoch": 366.76,
+      "learning_rate": 5.951230174400063e-05,
+      "loss": 0.067,
+      "step": 60760
+    },
+    {
+      "epoch": 366.88,
+      "learning_rate": 5.946823768938067e-05,
+      "loss": 0.067,
+      "step": 60780
+    },
+    {
+      "epoch": 367.0,
+      "learning_rate": 5.9424166006372615e-05,
+      "loss": 0.0687,
+      "step": 60800
+    },
+    {
+      "epoch": 367.12,
+      "learning_rate": 5.938008673048423e-05,
+      "loss": 0.0515,
+      "step": 60820
+    },
+    {
+      "epoch": 367.24,
+      "learning_rate": 5.93359998972294e-05,
+      "loss": 0.0552,
+      "step": 60840
+    },
+    {
+      "epoch": 367.36,
+      "learning_rate": 5.929190554212807e-05,
+      "loss": 0.0594,
+      "step": 60860
+    },
+    {
+      "epoch": 367.48,
+      "learning_rate": 5.924780370070629e-05,
+      "loss": 0.0621,
+      "step": 60880
+    },
+    {
+      "epoch": 367.61,
+      "learning_rate": 5.920369440849609e-05,
+      "loss": 0.0645,
+      "step": 60900
+    },
+    {
+      "epoch": 367.73,
+      "learning_rate": 5.915957770103556e-05,
+      "loss": 0.0656,
+      "step": 60920
+    },
+    {
+      "epoch": 367.85,
+      "learning_rate": 5.91154536138687e-05,
+      "loss": 0.0661,
+      "step": 60940
+    },
+    {
+      "epoch": 367.97,
+      "learning_rate": 5.907132218254552e-05,
+      "loss": 0.0682,
+      "step": 60960
+    },
+    {
+      "epoch": 368.09,
+      "learning_rate": 5.90271834426219e-05,
+      "loss": 0.0542,
+      "step": 60980
+    },
+    {
+      "epoch": 368.21,
+      "learning_rate": 5.898303742965964e-05,
+      "loss": 0.0532,
+      "step": 61000
+    },
+    {
+      "epoch": 368.21,
+      "eval_accuracy": 0.0012604834487459375,
+      "eval_loss": 21.773988723754883,
+      "eval_runtime": 18.2904,
+      "eval_samples_per_second": 111.534,
+      "eval_steps_per_second": 2.351,
+      "step": 61000
+    },
+    {
+      "epoch": 368.33,
+      "learning_rate": 5.893888417922636e-05,
+      "loss": 0.0562,
+      "step": 61020
+    },
+    {
+      "epoch": 368.45,
+      "learning_rate": 5.889472372689556e-05,
+      "loss": 0.059,
+      "step": 61040
+    },
+    {
+      "epoch": 368.57,
+      "learning_rate": 5.885055610824652e-05,
+      "loss": 0.0618,
+      "step": 61060
+    },
+    {
+      "epoch": 368.69,
+      "learning_rate": 5.880638135886427e-05,
+      "loss": 0.0632,
+      "step": 61080
+    },
+    {
+      "epoch": 368.81,
+      "learning_rate": 5.8762199514339624e-05,
+      "loss": 0.0652,
+      "step": 61100
+    },
+    {
+      "epoch": 368.93,
+      "learning_rate": 5.8718010610269095e-05,
+      "loss": 0.067,
+      "step": 61120
+    },
+    {
+      "epoch": 369.05,
+      "learning_rate": 5.867381468225489e-05,
+      "loss": 0.06,
+      "step": 61140
+    },
+    {
+      "epoch": 369.18,
+      "learning_rate": 5.862961176590486e-05,
+      "loss": 0.0581,
+      "step": 61160
+    },
+    {
+      "epoch": 369.3,
+      "learning_rate": 5.85854018968325e-05,
+      "loss": 0.0553,
+      "step": 61180
+    },
+    {
+      "epoch": 369.42,
+      "learning_rate": 5.8541185110656906e-05,
+      "loss": 0.0586,
+      "step": 61200
+    },
+    {
+      "epoch": 369.54,
+      "learning_rate": 5.849696144300273e-05,
+      "loss": 0.0606,
+      "step": 61220
+    },
+    {
+      "epoch": 369.66,
+      "learning_rate": 5.845273092950019e-05,
+      "loss": 0.0622,
+      "step": 61240
+    },
+    {
+      "epoch": 369.78,
+      "learning_rate": 5.840849360578503e-05,
+      "loss": 0.0627,
+      "step": 61260
+    },
+    {
+      "epoch": 369.9,
+      "learning_rate": 5.8364249507498435e-05,
+      "loss": 0.0649,
+      "step": 61280
+    },
+    {
+      "epoch": 370.02,
+      "learning_rate": 5.83199986702871e-05,
+      "loss": 0.0633,
+      "step": 61300
+    },
+    {
+      "epoch": 370.14,
+      "learning_rate": 5.827574112980311e-05,
+      "loss": 0.0502,
+      "step": 61320
+    },
+    {
+      "epoch": 370.26,
+      "learning_rate": 5.8231476921703964e-05,
+      "loss": 0.0523,
+      "step": 61340
+    },
+    {
+      "epoch": 370.38,
+      "learning_rate": 5.8187206081652554e-05,
+      "loss": 0.0553,
+      "step": 61360
+    },
+    {
+      "epoch": 370.5,
+      "learning_rate": 5.814292864531705e-05,
+      "loss": 0.0591,
+      "step": 61380
+    },
+    {
+      "epoch": 370.62,
+      "learning_rate": 5.8098644648371046e-05,
+      "loss": 0.0611,
+      "step": 61400
+    },
+    {
+      "epoch": 370.74,
+      "learning_rate": 5.8054354126493324e-05,
+      "loss": 0.0608,
+      "step": 61420
+    },
+    {
+      "epoch": 370.87,
+      "learning_rate": 5.8010057115367966e-05,
+      "loss": 0.0639,
+      "step": 61440
+    },
+    {
+      "epoch": 370.99,
+      "learning_rate": 5.796575365068425e-05,
+      "loss": 0.0644,
+      "step": 61460
+    },
+    {
+      "epoch": 371.11,
+      "learning_rate": 5.792144376813671e-05,
+      "loss": 0.0504,
+      "step": 61480
+    },
+    {
+      "epoch": 371.23,
+      "learning_rate": 5.787712750342501e-05,
+      "loss": 0.0506,
+      "step": 61500
+    },
+    {
+      "epoch": 371.35,
+      "learning_rate": 5.7832804892253946e-05,
+      "loss": 0.0532,
+      "step": 61520
+    },
+    {
+      "epoch": 371.47,
+      "learning_rate": 5.7788475970333455e-05,
+      "loss": 0.0562,
+      "step": 61540
+    },
+    {
+      "epoch": 371.59,
+      "learning_rate": 5.774414077337855e-05,
+      "loss": 0.0588,
+      "step": 61560
+    },
+    {
+      "epoch": 371.71,
+      "learning_rate": 5.769979933710932e-05,
+      "loss": 0.061,
+      "step": 61580
+    },
+    {
+      "epoch": 371.83,
+      "learning_rate": 5.765766922602742e-05,
+      "loss": 0.0618,
+      "step": 61600
+    },
+    {
+      "epoch": 371.95,
+      "learning_rate": 5.7613315725854077e-05,
+      "loss": 0.0632,
+      "step": 61620
+    },
+    {
+      "epoch": 372.07,
+      "learning_rate": 5.756895609176974e-05,
+      "loss": 0.0546,
+      "step": 61640
+    },
+    {
+      "epoch": 372.19,
+      "learning_rate": 5.752459035951421e-05,
+      "loss": 0.0487,
+      "step": 61660
+    },
+    {
+      "epoch": 372.31,
+      "learning_rate": 5.7480218564832125e-05,
+      "loss": 0.0506,
+      "step": 61680
+    },
+    {
+      "epoch": 372.43,
+      "learning_rate": 5.743584074347306e-05,
+      "loss": 0.0537,
+      "step": 61700
+    },
+    {
+      "epoch": 372.56,
+      "learning_rate": 5.739145693119141e-05,
+      "loss": 0.0572,
+      "step": 61720
+    },
+    {
+      "epoch": 372.68,
+      "learning_rate": 5.7347067163746405e-05,
+      "loss": 0.0601,
+      "step": 61740
+    },
+    {
+      "epoch": 372.8,
+      "learning_rate": 5.730267147690212e-05,
+      "loss": 0.0614,
+      "step": 61760
+    },
+    {
+      "epoch": 372.92,
+      "learning_rate": 5.72582699064273e-05,
+      "loss": 0.0624,
+      "step": 61780
+    },
+    {
+      "epoch": 373.04,
+      "learning_rate": 5.7213862488095526e-05,
+      "loss": 0.0578,
+      "step": 61800
+    },
+    {
+      "epoch": 373.16,
+      "learning_rate": 5.716944925768505e-05,
+      "loss": 0.0485,
+      "step": 61820
+    },
+    {
+      "epoch": 373.28,
+      "learning_rate": 5.71250302509788e-05,
+      "loss": 0.0506,
+      "step": 61840
+    },
+    {
+      "epoch": 373.4,
+      "learning_rate": 5.708060550376436e-05,
+      "loss": 0.053,
+      "step": 61860
+    },
+    {
+      "epoch": 373.52,
+      "learning_rate": 5.7036175051833964e-05,
+      "loss": 0.0556,
+      "step": 61880
+    },
+    {
+      "epoch": 373.64,
+      "learning_rate": 5.699173893098442e-05,
+      "loss": 0.0584,
+      "step": 61900
+    },
+    {
+      "epoch": 373.76,
+      "learning_rate": 5.69472971770171e-05,
+      "loss": 0.0597,
+      "step": 61920
+    },
+    {
+      "epoch": 373.88,
+      "learning_rate": 5.690284982573792e-05,
+      "loss": 0.0612,
+      "step": 61940
+    },
+    {
+      "epoch": 374.0,
+      "learning_rate": 5.685839691295733e-05,
+      "loss": 0.0625,
+      "step": 61960
+    },
+    {
+      "epoch": 374.12,
+      "learning_rate": 5.681393847449024e-05,
+      "loss": 0.0473,
+      "step": 61980
+    },
+    {
+      "epoch": 374.25,
+      "learning_rate": 5.676947454615596e-05,
+      "loss": 0.0481,
+      "step": 62000
+    },
+    {
+      "epoch": 374.25,
+      "eval_accuracy": 0.0012693740059863084,
+      "eval_loss": 21.933935165405273,
+      "eval_runtime": 18.3333,
+      "eval_samples_per_second": 111.273,
+      "eval_steps_per_second": 2.345,
+      "step": 62000
+    },
+    {
+      "epoch": 374.37,
+      "learning_rate": 5.672500516377832e-05,
+      "loss": 0.0499,
+      "step": 62020
+    },
+    {
+      "epoch": 374.49,
+      "learning_rate": 5.668053036318549e-05,
+      "loss": 0.0524,
+      "step": 62040
+    },
+    {
+      "epoch": 374.61,
+      "learning_rate": 5.6636050180210035e-05,
+      "loss": 0.055,
+      "step": 62060
+    },
+    {
+      "epoch": 374.73,
+      "learning_rate": 5.6591564650688776e-05,
+      "loss": 0.0575,
+      "step": 62080
+    },
+    {
+      "epoch": 374.85,
+      "learning_rate": 5.6547073810462956e-05,
+      "loss": 0.0589,
+      "step": 62100
+    },
+    {
+      "epoch": 374.97,
+      "learning_rate": 5.650257769537802e-05,
+      "loss": 0.0616,
+      "step": 62120
+    },
+    {
+      "epoch": 375.09,
+      "learning_rate": 5.6458076341283684e-05,
+      "loss": 0.0494,
+      "step": 62140
+    },
+    {
+      "epoch": 375.21,
+      "learning_rate": 5.641356978403388e-05,
+      "loss": 0.0453,
+      "step": 62160
+    },
+    {
+      "epoch": 375.33,
+      "learning_rate": 5.636905805948676e-05,
+      "loss": 0.0467,
+      "step": 62180
+    },
+    {
+      "epoch": 375.45,
+      "learning_rate": 5.6324541203504576e-05,
+      "loss": 0.0478,
+      "step": 62200
+    },
+    {
+      "epoch": 375.57,
+      "learning_rate": 5.628001925195379e-05,
+      "loss": 0.0502,
+      "step": 62220
+    },
+    {
+      "epoch": 375.69,
+      "learning_rate": 5.6235492240704936e-05,
+      "loss": 0.053,
+      "step": 62240
+    },
+    {
+      "epoch": 375.81,
+      "learning_rate": 5.619096020563259e-05,
+      "loss": 0.0557,
+      "step": 62260
+    },
+    {
+      "epoch": 375.94,
+      "learning_rate": 5.614642318261545e-05,
+      "loss": 0.0592,
+      "step": 62280
+    },
+    {
+      "epoch": 376.06,
+      "learning_rate": 5.610188120753616e-05,
+      "loss": 0.0535,
+      "step": 62300
+    },
+    {
+      "epoch": 376.18,
+      "learning_rate": 5.605733431628139e-05,
+      "loss": 0.0471,
+      "step": 62320
+    },
+    {
+      "epoch": 376.3,
+      "learning_rate": 5.601278254474176e-05,
+      "loss": 0.0479,
+      "step": 62340
+    },
+    {
+      "epoch": 376.42,
+      "learning_rate": 5.5968225928811856e-05,
+      "loss": 0.0486,
+      "step": 62360
+    },
+    {
+      "epoch": 376.54,
+      "learning_rate": 5.592366450439012e-05,
+      "loss": 0.0513,
+      "step": 62380
+    },
+    {
+      "epoch": 376.66,
+      "learning_rate": 5.587909830737887e-05,
+      "loss": 0.0536,
+      "step": 62400
+    },
+    {
+      "epoch": 376.78,
+      "learning_rate": 5.583452737368432e-05,
+      "loss": 0.0553,
+      "step": 62420
+    },
+    {
+      "epoch": 376.9,
+      "learning_rate": 5.578995173921644e-05,
+      "loss": 0.0576,
+      "step": 62440
+    },
+    {
+      "epoch": 377.02,
+      "learning_rate": 5.574537143988903e-05,
+      "loss": 0.0566,
+      "step": 62460
+    },
+    {
+      "epoch": 377.14,
+      "learning_rate": 5.5700786511619605e-05,
+      "loss": 0.0453,
+      "step": 62480
+    },
+    {
+      "epoch": 377.26,
+      "learning_rate": 5.565619699032949e-05,
+      "loss": 0.0452,
+      "step": 62500
+    },
+    {
+      "epoch": 377.38,
+      "learning_rate": 5.5611602911943596e-05,
+      "loss": 0.0469,
+      "step": 62520
+    },
+    {
+      "epoch": 377.51,
+      "learning_rate": 5.556700431239061e-05,
+      "loss": 0.0483,
+      "step": 62540
+    },
+    {
+      "epoch": 377.63,
+      "learning_rate": 5.552240122760279e-05,
+      "loss": 0.0505,
+      "step": 62560
+    },
+    {
+      "epoch": 377.75,
+      "learning_rate": 5.547779369351606e-05,
+      "loss": 0.0536,
+      "step": 62580
+    },
+    {
+      "epoch": 377.87,
+      "learning_rate": 5.543318174606987e-05,
+      "loss": 0.0553,
+      "step": 62600
+    },
+    {
+      "epoch": 377.99,
+      "learning_rate": 5.5388565421207304e-05,
+      "loss": 0.0562,
+      "step": 62620
+    },
+    {
+      "epoch": 378.11,
+      "learning_rate": 5.53439447548749e-05,
+      "loss": 0.0455,
+      "step": 62640
+    },
+    {
+      "epoch": 378.23,
+      "learning_rate": 5.529931978302272e-05,
+      "loss": 0.0432,
+      "step": 62660
+    },
+    {
+      "epoch": 378.35,
+      "learning_rate": 5.525469054160433e-05,
+      "loss": 0.0434,
+      "step": 62680
+    },
+    {
+      "epoch": 378.47,
+      "learning_rate": 5.521005706657666e-05,
+      "loss": 0.0441,
+      "step": 62700
+    },
+    {
+      "epoch": 378.59,
+      "learning_rate": 5.516541939390013e-05,
+      "loss": 0.0459,
+      "step": 62720
+    },
+    {
+      "epoch": 378.71,
+      "learning_rate": 5.5120777559538486e-05,
+      "loss": 0.047,
+      "step": 62740
+    },
+    {
+      "epoch": 378.83,
+      "learning_rate": 5.5076131599458855e-05,
+      "loss": 0.0489,
+      "step": 62760
+    },
+    {
+      "epoch": 378.95,
+      "learning_rate": 5.503148154963168e-05,
+      "loss": 0.0513,
+      "step": 62780
+    },
+    {
+      "epoch": 379.07,
+      "learning_rate": 5.498682744603071e-05,
+      "loss": 0.0459,
+      "step": 62800
+    },
+    {
+      "epoch": 379.2,
+      "learning_rate": 5.494216932463293e-05,
+      "loss": 0.0407,
+      "step": 62820
+    },
+    {
+      "epoch": 379.32,
+      "learning_rate": 5.48975072214186e-05,
+      "loss": 0.0405,
+      "step": 62840
+    },
+    {
+      "epoch": 379.44,
+      "learning_rate": 5.485284117237115e-05,
+      "loss": 0.041,
+      "step": 62860
+    },
+    {
+      "epoch": 379.56,
+      "learning_rate": 5.480817121347722e-05,
+      "loss": 0.0417,
+      "step": 62880
+    },
+    {
+      "epoch": 379.68,
+      "learning_rate": 5.476349738072659e-05,
+      "loss": 0.0432,
+      "step": 62900
+    },
+    {
+      "epoch": 379.8,
+      "learning_rate": 5.4718819710112166e-05,
+      "loss": 0.0449,
+      "step": 62920
+    },
+    {
+      "epoch": 379.92,
+      "learning_rate": 5.467413823762993e-05,
+      "loss": 0.0469,
+      "step": 62940
+    },
+    {
+      "epoch": 380.04,
+      "learning_rate": 5.462945299927894e-05,
+      "loss": 0.0444,
+      "step": 62960
+    },
+    {
+      "epoch": 380.16,
+      "learning_rate": 5.4584764031061296e-05,
+      "loss": 0.0368,
+      "step": 62980
+    },
+    {
+      "epoch": 380.28,
+      "learning_rate": 5.454007136898207e-05,
+      "loss": 0.0358,
+      "step": 63000
+    },
+    {
+      "epoch": 380.28,
+      "eval_accuracy": 0.0012466536930386936,
+      "eval_loss": 22.16596221923828,
+      "eval_runtime": 18.1736,
+      "eval_samples_per_second": 112.25,
+      "eval_steps_per_second": 2.366,
+      "step": 63000
+    },
+    {
+      "epoch": 380.4,
+      "learning_rate": 5.4495375049049366e-05,
+      "loss": 0.0357,
+      "step": 63020
+    },
+    {
+      "epoch": 380.52,
+      "learning_rate": 5.445067510727415e-05,
+      "loss": 0.0361,
+      "step": 63040
+    },
+    {
+      "epoch": 380.64,
+      "learning_rate": 5.4405971579670414e-05,
+      "loss": 0.0369,
+      "step": 63060
+    },
+    {
+      "epoch": 380.76,
+      "learning_rate": 5.436126450225495e-05,
+      "loss": 0.0377,
+      "step": 63080
+    },
+    {
+      "epoch": 380.89,
+      "learning_rate": 5.4316553911047455e-05,
+      "loss": 0.0388,
+      "step": 63100
+    },
+    {
+      "epoch": 381.01,
+      "learning_rate": 5.427183984207043e-05,
+      "loss": 0.0395,
+      "step": 63120
+    },
+    {
+      "epoch": 381.13,
+      "learning_rate": 5.42271223313492e-05,
+      "loss": 0.031,
+      "step": 63140
+    },
+    {
+      "epoch": 381.25,
+      "learning_rate": 5.4182401414911873e-05,
+      "loss": 0.0305,
+      "step": 63160
+    },
+    {
+      "epoch": 381.37,
+      "learning_rate": 5.413767712878926e-05,
+      "loss": 0.0296,
+      "step": 63180
+    },
+    {
+      "epoch": 381.49,
+      "learning_rate": 5.409294950901492e-05,
+      "loss": 0.029,
+      "step": 63200
+    },
+    {
+      "epoch": 381.61,
+      "learning_rate": 5.404821859162509e-05,
+      "loss": 0.0285,
+      "step": 63220
+    },
+    {
+      "epoch": 381.73,
+      "learning_rate": 5.4003484412658665e-05,
+      "loss": 0.0287,
+      "step": 63240
+    },
+    {
+      "epoch": 381.85,
+      "learning_rate": 5.395874700815715e-05,
+      "loss": 0.0286,
+      "step": 63260
+    },
+    {
+      "epoch": 381.97,
+      "learning_rate": 5.39140064141647e-05,
+      "loss": 0.0283,
+      "step": 63280
+    },
+    {
+      "epoch": 382.09,
+      "learning_rate": 5.386926266672795e-05,
+      "loss": 0.0231,
+      "step": 63300
+    },
+    {
+      "epoch": 382.21,
+      "learning_rate": 5.382451580189616e-05,
+      "loss": 0.0213,
+      "step": 63320
+    },
+    {
+      "epoch": 382.33,
+      "learning_rate": 5.3779765855721066e-05,
+      "loss": 0.0205,
+      "step": 63340
+    },
+    {
+      "epoch": 382.45,
+      "learning_rate": 5.373501286425691e-05,
+      "loss": 0.0198,
+      "step": 63360
+    },
+    {
+      "epoch": 382.58,
+      "learning_rate": 5.369025686356033e-05,
+      "loss": 0.0193,
+      "step": 63380
+    },
+    {
+      "epoch": 382.7,
+      "learning_rate": 5.364549788969045e-05,
+      "loss": 0.0191,
+      "step": 63400
+    },
+    {
+      "epoch": 382.82,
+      "learning_rate": 5.360073597870879e-05,
+      "loss": 0.0186,
+      "step": 63420
+    },
+    {
+      "epoch": 382.94,
+      "learning_rate": 5.355597116667918e-05,
+      "loss": 0.0185,
+      "step": 63440
+    },
+    {
+      "epoch": 383.06,
+      "learning_rate": 5.351120348966783e-05,
+      "loss": 0.0163,
+      "step": 63460
+    },
+    {
+      "epoch": 383.18,
+      "learning_rate": 5.346643298374322e-05,
+      "loss": 0.0145,
+      "step": 63480
+    },
+    {
+      "epoch": 383.3,
+      "learning_rate": 5.3421659684976197e-05,
+      "loss": 0.0143,
+      "step": 63500
+    },
+    {
+      "epoch": 383.42,
+      "learning_rate": 5.337688362943976e-05,
+      "loss": 0.0141,
+      "step": 63520
+    },
+    {
+      "epoch": 383.54,
+      "learning_rate": 5.333210485320919e-05,
+      "loss": 0.0142,
+      "step": 63540
+    },
+    {
+      "epoch": 383.66,
+      "learning_rate": 5.32873233923619e-05,
+      "loss": 0.0145,
+      "step": 63560
+    },
+    {
+      "epoch": 383.78,
+      "learning_rate": 5.3242539282977543e-05,
+      "loss": 0.0142,
+      "step": 63580
+    },
+    {
+      "epoch": 383.9,
+      "learning_rate": 5.319775256113785e-05,
+      "loss": 0.0141,
+      "step": 63600
+    },
+    {
+      "epoch": 384.02,
+      "learning_rate": 5.315296326292667e-05,
+      "loss": 0.0135,
+      "step": 63620
+    },
+    {
+      "epoch": 384.14,
+      "learning_rate": 5.310817142442994e-05,
+      "loss": 0.0122,
+      "step": 63640
+    },
+    {
+      "epoch": 384.27,
+      "learning_rate": 5.3063377081735645e-05,
+      "loss": 0.0125,
+      "step": 63660
+    },
+    {
+      "epoch": 384.39,
+      "learning_rate": 5.301858027093376e-05,
+      "loss": 0.0125,
+      "step": 63680
+    },
+    {
+      "epoch": 384.51,
+      "learning_rate": 5.2973781028116275e-05,
+      "loss": 0.0125,
+      "step": 63700
+    },
+    {
+      "epoch": 384.63,
+      "learning_rate": 5.292897938937713e-05,
+      "loss": 0.0129,
+      "step": 63720
+    },
+    {
+      "epoch": 384.75,
+      "learning_rate": 5.2884175390812195e-05,
+      "loss": 0.0127,
+      "step": 63740
+    },
+    {
+      "epoch": 384.87,
+      "learning_rate": 5.283936906851923e-05,
+      "loss": 0.0128,
+      "step": 63760
+    },
+    {
+      "epoch": 384.99,
+      "learning_rate": 5.2794560458597897e-05,
+      "loss": 0.0128,
+      "step": 63780
+    },
+    {
+      "epoch": 385.11,
+      "learning_rate": 5.274974959714969e-05,
+      "loss": 0.0118,
+      "step": 63800
+    },
+    {
+      "epoch": 385.23,
+      "learning_rate": 5.2704936520277895e-05,
+      "loss": 0.0118,
+      "step": 63820
+    },
+    {
+      "epoch": 385.35,
+      "learning_rate": 5.2660121264087605e-05,
+      "loss": 0.0119,
+      "step": 63840
+    },
+    {
+      "epoch": 385.47,
+      "learning_rate": 5.261530386468565e-05,
+      "loss": 0.0121,
+      "step": 63860
+    },
+    {
+      "epoch": 385.59,
+      "learning_rate": 5.2570484358180605e-05,
+      "loss": 0.0122,
+      "step": 63880
+    },
+    {
+      "epoch": 385.71,
+      "learning_rate": 5.252566278068275e-05,
+      "loss": 0.0124,
+      "step": 63900
+    },
+    {
+      "epoch": 385.84,
+      "learning_rate": 5.2480839168304e-05,
+      "loss": 0.0126,
+      "step": 63920
+    },
+    {
+      "epoch": 385.96,
+      "learning_rate": 5.243601355715796e-05,
+      "loss": 0.0124,
+      "step": 63940
+    },
+    {
+      "epoch": 386.08,
+      "learning_rate": 5.239118598335978e-05,
+      "loss": 0.0117,
+      "step": 63960
+    },
+    {
+      "epoch": 386.2,
+      "learning_rate": 5.234635648302626e-05,
+      "loss": 0.0115,
+      "step": 63980
+    },
+    {
+      "epoch": 386.32,
+      "learning_rate": 5.23015250922757e-05,
+      "loss": 0.0117,
+      "step": 64000
+    },
+    {
+      "epoch": 386.32,
+      "eval_accuracy": 0.0012555442502790648,
+      "eval_loss": 22.422649383544922,
+      "eval_runtime": 18.1838,
+      "eval_samples_per_second": 112.188,
+      "eval_steps_per_second": 2.365,
+      "step": 64000
+    },
+    {
+      "epoch": 386.44,
+      "learning_rate": 5.225669184722795e-05,
+      "loss": 0.0118,
+      "step": 64020
+    },
+    {
+      "epoch": 386.56,
+      "learning_rate": 5.221185678400431e-05,
+      "loss": 0.012,
+      "step": 64040
+    },
+    {
+      "epoch": 386.68,
+      "learning_rate": 5.216701993872762e-05,
+      "loss": 0.0121,
+      "step": 64060
+    },
+    {
+      "epoch": 386.8,
+      "learning_rate": 5.2122181347522116e-05,
+      "loss": 0.012,
+      "step": 64080
+    },
+    {
+      "epoch": 386.92,
+      "learning_rate": 5.207734104651344e-05,
+      "loss": 0.0124,
+      "step": 64100
+    },
+    {
+      "epoch": 387.04,
+      "learning_rate": 5.203249907182859e-05,
+      "loss": 0.0119,
+      "step": 64120
+    },
+    {
+      "epoch": 387.16,
+      "learning_rate": 5.1987655459595944e-05,
+      "loss": 0.0115,
+      "step": 64140
+    },
+    {
+      "epoch": 387.28,
+      "learning_rate": 5.194281024594522e-05,
+      "loss": 0.0121,
+      "step": 64160
+    },
+    {
+      "epoch": 387.4,
+      "learning_rate": 5.189796346700735e-05,
+      "loss": 0.012,
+      "step": 64180
+    },
+    {
+      "epoch": 387.53,
+      "learning_rate": 5.185311515891459e-05,
+      "loss": 0.0118,
+      "step": 64200
+    },
+    {
+      "epoch": 387.65,
+      "learning_rate": 5.180826535780043e-05,
+      "loss": 0.012,
+      "step": 64220
+    },
+    {
+      "epoch": 387.77,
+      "learning_rate": 5.1763414099799534e-05,
+      "loss": 0.012,
+      "step": 64240
+    },
+    {
+      "epoch": 387.89,
+      "learning_rate": 5.171856142104775e-05,
+      "loss": 0.0121,
+      "step": 64260
+    },
+    {
+      "epoch": 388.01,
+      "learning_rate": 5.1673707357682077e-05,
+      "loss": 0.0121,
+      "step": 64280
+    },
+    {
+      "epoch": 388.13,
+      "learning_rate": 5.1628851945840596e-05,
+      "loss": 0.0111,
+      "step": 64300
+    },
+    {
+      "epoch": 388.25,
+      "learning_rate": 5.158399522166253e-05,
+      "loss": 0.0114,
+      "step": 64320
+    },
+    {
+      "epoch": 388.37,
+      "learning_rate": 5.153913722128813e-05,
+      "loss": 0.0123,
+      "step": 64340
+    },
+    {
+      "epoch": 388.49,
+      "learning_rate": 5.1494277980858674e-05,
+      "loss": 0.0125,
+      "step": 64360
+    },
+    {
+      "epoch": 388.61,
+      "learning_rate": 5.144941753651643e-05,
+      "loss": 0.014,
+      "step": 64380
+    },
+    {
+      "epoch": 388.73,
+      "learning_rate": 5.140679903218679e-05,
+      "loss": 0.0268,
+      "step": 64400
+    },
+    {
+      "epoch": 388.85,
+      "learning_rate": 5.1361936344172525e-05,
+      "loss": 0.0354,
+      "step": 64420
+    },
+    {
+      "epoch": 388.97,
+      "learning_rate": 5.1317072558870726e-05,
+      "loss": 0.0412,
+      "step": 64440
+    },
+    {
+      "epoch": 389.09,
+      "learning_rate": 5.127220771242737e-05,
+      "loss": 0.0621,
+      "step": 64460
+    },
+    {
+      "epoch": 389.22,
+      "learning_rate": 5.122734184098922e-05,
+      "loss": 0.0829,
+      "step": 64480
+    },
+    {
+      "epoch": 389.34,
+      "learning_rate": 5.118247498070393e-05,
+      "loss": 0.0872,
+      "step": 64500
+    },
+    {
+      "epoch": 389.46,
+      "learning_rate": 5.1137607167719905e-05,
+      "loss": 0.0932,
+      "step": 64520
+    },
+    {
+      "epoch": 389.58,
+      "learning_rate": 5.109273843818633e-05,
+      "loss": 0.1013,
+      "step": 64540
+    },
+    {
+      "epoch": 389.7,
+      "learning_rate": 5.104786882825313e-05,
+      "loss": 0.1057,
+      "step": 64560
+    },
+    {
+      "epoch": 389.82,
+      "learning_rate": 5.100299837407094e-05,
+      "loss": 0.1033,
+      "step": 64580
+    },
+    {
+      "epoch": 389.94,
+      "learning_rate": 5.0958127111791086e-05,
+      "loss": 0.1057,
+      "step": 64600
+    },
+    {
+      "epoch": 390.06,
+      "learning_rate": 5.0913255077565514e-05,
+      "loss": 0.0964,
+      "step": 64620
+    },
+    {
+      "epoch": 390.18,
+      "learning_rate": 5.0868382307546834e-05,
+      "loss": 0.099,
+      "step": 64640
+    },
+    {
+      "epoch": 390.3,
+      "learning_rate": 5.08235088378882e-05,
+      "loss": 0.1026,
+      "step": 64660
+    },
+    {
+      "epoch": 390.42,
+      "learning_rate": 5.077863470474339e-05,
+      "loss": 0.0996,
+      "step": 64680
+    },
+    {
+      "epoch": 390.54,
+      "learning_rate": 5.073375994426664e-05,
+      "loss": 0.1021,
+      "step": 64700
+    },
+    {
+      "epoch": 390.66,
+      "learning_rate": 5.068888459261276e-05,
+      "loss": 0.1036,
+      "step": 64720
+    },
+    {
+      "epoch": 390.78,
+      "learning_rate": 5.0644008685937005e-05,
+      "loss": 0.1073,
+      "step": 64740
+    },
+    {
+      "epoch": 390.91,
+      "learning_rate": 5.05991322603951e-05,
+      "loss": 0.1049,
+      "step": 64760
+    },
+    {
+      "epoch": 391.03,
+      "learning_rate": 5.055425535214314e-05,
+      "loss": 0.1001,
+      "step": 64780
+    },
+    {
+      "epoch": 391.15,
+      "learning_rate": 5.050937799733766e-05,
+      "loss": 0.0803,
+      "step": 64800
+    },
+    {
+      "epoch": 391.27,
+      "learning_rate": 5.0464500232135537e-05,
+      "loss": 0.0844,
+      "step": 64820
+    },
+    {
+      "epoch": 391.39,
+      "learning_rate": 5.041962209269393e-05,
+      "loss": 0.0872,
+      "step": 64840
+    },
+    {
+      "epoch": 391.51,
+      "learning_rate": 5.037474361517042e-05,
+      "loss": 0.0898,
+      "step": 64860
+    },
+    {
+      "epoch": 391.63,
+      "learning_rate": 5.032986483572273e-05,
+      "loss": 0.0907,
+      "step": 64880
+    },
+    {
+      "epoch": 391.75,
+      "learning_rate": 5.0284985790508924e-05,
+      "loss": 0.0911,
+      "step": 64900
+    },
+    {
+      "epoch": 391.87,
+      "learning_rate": 5.0240106515687194e-05,
+      "loss": 0.0916,
+      "step": 64920
+    },
+    {
+      "epoch": 391.99,
+      "learning_rate": 5.019522704741598e-05,
+      "loss": 0.0913,
+      "step": 64940
+    },
+    {
+      "epoch": 392.11,
+      "learning_rate": 5.015034742185386e-05,
+      "loss": 0.0692,
+      "step": 64960
+    },
+    {
+      "epoch": 392.23,
+      "learning_rate": 5.010546767515953e-05,
+      "loss": 0.0741,
+      "step": 64980
+    },
+    {
+      "epoch": 392.35,
+      "learning_rate": 5.006058784349181e-05,
+      "loss": 0.0768,
+      "step": 65000
+    },
+    {
+      "epoch": 392.35,
+      "eval_accuracy": 0.0012575199296658139,
+      "eval_loss": 22.219345092773438,
+      "eval_runtime": 18.2617,
+      "eval_samples_per_second": 111.709,
+      "eval_steps_per_second": 2.355,
+      "step": 65000
+    },
+    {
+      "epoch": 392.47,
+      "learning_rate": 5.0015707963009575e-05,
+      "loss": 0.0777,
+      "step": 65020
+    },
+    {
+      "epoch": 392.6,
+      "learning_rate": 4.99708280698717e-05,
+      "loss": 0.0808,
+      "step": 65040
+    },
+    {
+      "epoch": 392.72,
+      "learning_rate": 4.992594820023712e-05,
+      "loss": 0.0818,
+      "step": 65060
+    },
+    {
+      "epoch": 392.84,
+      "learning_rate": 4.988106839026474e-05,
+      "loss": 0.0829,
+      "step": 65080
+    },
+    {
+      "epoch": 392.96,
+      "learning_rate": 4.983618867611342e-05,
+      "loss": 0.0821,
+      "step": 65100
+    },
+    {
+      "epoch": 393.08,
+      "learning_rate": 4.979130909394191e-05,
+      "loss": 0.0686,
+      "step": 65120
+    },
+    {
+      "epoch": 393.2,
+      "learning_rate": 4.97464296799089e-05,
+      "loss": 0.0649,
+      "step": 65140
+    },
+    {
+      "epoch": 393.32,
+      "learning_rate": 4.9701550470172914e-05,
+      "loss": 0.0684,
+      "step": 65160
+    },
+    {
+      "epoch": 393.44,
+      "learning_rate": 4.965667150089234e-05,
+      "loss": 0.0699,
+      "step": 65180
+    },
+    {
+      "epoch": 393.56,
+      "learning_rate": 4.9611792808225316e-05,
+      "loss": 0.0699,
+      "step": 65200
+    },
+    {
+      "epoch": 393.68,
+      "learning_rate": 4.956691442832981e-05,
+      "loss": 0.0717,
+      "step": 65220
+    },
+    {
+      "epoch": 393.8,
+      "learning_rate": 4.952203639736354e-05,
+      "loss": 0.0734,
+      "step": 65240
+    },
+    {
+      "epoch": 393.92,
+      "learning_rate": 4.947715875148391e-05,
+      "loss": 0.0741,
+      "step": 65260
+    },
+    {
+      "epoch": 394.04,
+      "learning_rate": 4.9432281526848034e-05,
+      "loss": 0.0675,
+      "step": 65280
+    },
+    {
+      "epoch": 394.16,
+      "learning_rate": 4.9387404759612674e-05,
+      "loss": 0.0563,
+      "step": 65300
+    },
+    {
+      "epoch": 394.29,
+      "learning_rate": 4.934252848593424e-05,
+      "loss": 0.0609,
+      "step": 65320
+    },
+    {
+      "epoch": 394.41,
+      "learning_rate": 4.929765274196872e-05,
+      "loss": 0.0634,
+      "step": 65340
+    },
+    {
+      "epoch": 394.53,
+      "learning_rate": 4.925277756387169e-05,
+      "loss": 0.0642,
+      "step": 65360
+    },
+    {
+      "epoch": 394.65,
+      "learning_rate": 4.920790298779831e-05,
+      "loss": 0.0662,
+      "step": 65380
+    },
+    {
+      "epoch": 394.77,
+      "learning_rate": 4.916302904990315e-05,
+      "loss": 0.067,
+      "step": 65400
+    },
+    {
+      "epoch": 394.89,
+      "learning_rate": 4.911815578634041e-05,
+      "loss": 0.0673,
+      "step": 65420
+    },
+    {
+      "epoch": 395.01,
+      "learning_rate": 4.907328323326359e-05,
+      "loss": 0.0652,
+      "step": 65440
+    },
+    {
+      "epoch": 395.13,
+      "learning_rate": 4.9028411426825746e-05,
+      "loss": 0.0507,
+      "step": 65460
+    },
+    {
+      "epoch": 395.25,
+      "learning_rate": 4.898354040317926e-05,
+      "loss": 0.0547,
+      "step": 65480
+    },
+    {
+      "epoch": 395.37,
+      "learning_rate": 4.8938670198475904e-05,
+      "loss": 0.0574,
+      "step": 65500
+    },
+    {
+      "epoch": 395.49,
+      "learning_rate": 4.8893800848866824e-05,
+      "loss": 0.0586,
+      "step": 65520
+    },
+    {
+      "epoch": 395.61,
+      "learning_rate": 4.8848932390502396e-05,
+      "loss": 0.0593,
+      "step": 65540
+    },
+    {
+      "epoch": 395.73,
+      "learning_rate": 4.8804064859532366e-05,
+      "loss": 0.0607,
+      "step": 65560
+    },
+    {
+      "epoch": 395.86,
+      "learning_rate": 4.8759198292105655e-05,
+      "loss": 0.063,
+      "step": 65580
+    },
+    {
+      "epoch": 395.98,
+      "learning_rate": 4.8714332724370504e-05,
+      "loss": 0.0637,
+      "step": 65600
+    },
+    {
+      "epoch": 396.1,
+      "learning_rate": 4.866946819247422e-05,
+      "loss": 0.0489,
+      "step": 65620
+    },
+    {
+      "epoch": 396.22,
+      "learning_rate": 4.8624604732563396e-05,
+      "loss": 0.0468,
+      "step": 65640
+    },
+    {
+      "epoch": 396.34,
+      "learning_rate": 4.85797423807837e-05,
+      "loss": 0.0474,
+      "step": 65660
+    },
+    {
+      "epoch": 396.46,
+      "learning_rate": 4.8534881173279905e-05,
+      "loss": 0.0491,
+      "step": 65680
+    },
+    {
+      "epoch": 396.58,
+      "learning_rate": 4.8490021146195904e-05,
+      "loss": 0.0516,
+      "step": 65700
+    },
+    {
+      "epoch": 396.7,
+      "learning_rate": 4.8445162335674574e-05,
+      "loss": 0.0532,
+      "step": 65720
+    },
+    {
+      "epoch": 396.82,
+      "learning_rate": 4.840030477785788e-05,
+      "loss": 0.0552,
+      "step": 65740
+    },
+    {
+      "epoch": 396.94,
+      "learning_rate": 4.835544850888671e-05,
+      "loss": 0.0573,
+      "step": 65760
+    },
+    {
+      "epoch": 397.06,
+      "learning_rate": 4.831059356490098e-05,
+      "loss": 0.0502,
+      "step": 65780
+    },
+    {
+      "epoch": 397.18,
+      "learning_rate": 4.826573998203952e-05,
+      "loss": 0.0423,
+      "step": 65800
+    },
+    {
+      "epoch": 397.3,
+      "learning_rate": 4.822088779644001e-05,
+      "loss": 0.0409,
+      "step": 65820
+    },
+    {
+      "epoch": 397.42,
+      "learning_rate": 4.817603704423909e-05,
+      "loss": 0.0411,
+      "step": 65840
+    },
+    {
+      "epoch": 397.55,
+      "learning_rate": 4.813118776157216e-05,
+      "loss": 0.0403,
+      "step": 65860
+    },
+    {
+      "epoch": 397.67,
+      "learning_rate": 4.808633998457351e-05,
+      "loss": 0.0416,
+      "step": 65880
+    },
+    {
+      "epoch": 397.79,
+      "learning_rate": 4.804149374937615e-05,
+      "loss": 0.0447,
+      "step": 65900
+    },
+    {
+      "epoch": 397.91,
+      "learning_rate": 4.79966490921119e-05,
+      "loss": 0.0455,
+      "step": 65920
+    },
+    {
+      "epoch": 398.03,
+      "learning_rate": 4.795180604891132e-05,
+      "loss": 0.0456,
+      "step": 65940
+    },
+    {
+      "epoch": 398.15,
+      "learning_rate": 4.790696465590358e-05,
+      "loss": 0.0395,
+      "step": 65960
+    },
+    {
+      "epoch": 398.27,
+      "learning_rate": 4.786212494921664e-05,
+      "loss": 0.0377,
+      "step": 65980
+    },
+    {
+      "epoch": 398.39,
+      "learning_rate": 4.7817286964976995e-05,
+      "loss": 0.0339,
+      "step": 66000
+    },
+    {
+      "epoch": 398.39,
+      "eval_accuracy": 0.0012530746510456283,
+      "eval_loss": 22.383317947387695,
+      "eval_runtime": 18.182,
+      "eval_samples_per_second": 112.199,
+      "eval_steps_per_second": 2.365,
+      "step": 66000
+    },
+    {
+      "epoch": 398.51,
+      "learning_rate": 4.777245073930984e-05,
+      "loss": 0.0323,
+      "step": 66020
+    },
+    {
+      "epoch": 398.63,
+      "learning_rate": 4.772761630833886e-05,
+      "loss": 0.031,
+      "step": 66040
+    },
+    {
+      "epoch": 398.75,
+      "learning_rate": 4.76827837081864e-05,
+      "loss": 0.0301,
+      "step": 66060
+    },
+    {
+      "epoch": 398.87,
+      "learning_rate": 4.7637952974973244e-05,
+      "loss": 0.0333,
+      "step": 66080
+    },
+    {
+      "epoch": 398.99,
+      "learning_rate": 4.7593124144818696e-05,
+      "loss": 0.0308,
+      "step": 66100
+    },
+    {
+      "epoch": 399.11,
+      "learning_rate": 4.754829725384059e-05,
+      "loss": 0.023,
+      "step": 66120
+    },
+    {
+      "epoch": 399.24,
+      "learning_rate": 4.750347233815506e-05,
+      "loss": 0.021,
+      "step": 66140
+    },
+    {
+      "epoch": 399.36,
+      "learning_rate": 4.74586494338768e-05,
+      "loss": 0.0216,
+      "step": 66160
+    },
+    {
+      "epoch": 399.48,
+      "learning_rate": 4.741382857711878e-05,
+      "loss": 0.0238,
+      "step": 66180
+    },
+    {
+      "epoch": 399.6,
+      "learning_rate": 4.736900980399237e-05,
+      "loss": 0.0214,
+      "step": 66200
+    },
+    {
+      "epoch": 399.72,
+      "learning_rate": 4.732419315060723e-05,
+      "loss": 0.0206,
+      "step": 66220
+    },
+    {
+      "epoch": 399.84,
+      "learning_rate": 4.727937865307134e-05,
+      "loss": 0.0237,
+      "step": 66240
+    },
+    {
+      "epoch": 399.96,
+      "learning_rate": 4.723456634749095e-05,
+      "loss": 0.0205,
+      "step": 66260
+    },
+    {
+      "epoch": 400.08,
+      "learning_rate": 4.718975626997047e-05,
+      "loss": 0.017,
+      "step": 66280
+    },
+    {
+      "epoch": 400.2,
+      "learning_rate": 4.7144948456612634e-05,
+      "loss": 0.017,
+      "step": 66300
+    },
+    {
+      "epoch": 400.32,
+      "learning_rate": 4.7100142943518235e-05,
+      "loss": 0.0177,
+      "step": 66320
+    },
+    {
+      "epoch": 400.44,
+      "learning_rate": 4.705533976678631e-05,
+      "loss": 0.0158,
+      "step": 66340
+    },
+    {
+      "epoch": 400.56,
+      "learning_rate": 4.701053896251392e-05,
+      "loss": 0.0195,
+      "step": 66360
+    },
+    {
+      "epoch": 400.68,
+      "learning_rate": 4.6965740566796304e-05,
+      "loss": 0.0234,
+      "step": 66380
+    },
+    {
+      "epoch": 400.8,
+      "learning_rate": 4.692094461572672e-05,
+      "loss": 0.0198,
+      "step": 66400
+    },
+    {
+      "epoch": 400.93,
+      "learning_rate": 4.687615114539644e-05,
+      "loss": 0.0211,
+      "step": 66420
+    },
+    {
+      "epoch": 401.05,
+      "learning_rate": 4.683136019189476e-05,
+      "loss": 0.0174,
+      "step": 66440
+    },
+    {
+      "epoch": 401.17,
+      "learning_rate": 4.6788811150149395e-05,
+      "loss": 0.0221,
+      "step": 66460
+    },
+    {
+      "epoch": 401.29,
+      "learning_rate": 4.674402520825759e-05,
+      "loss": 0.0155,
+      "step": 66480
+    },
+    {
+      "epoch": 401.41,
+      "learning_rate": 4.6699241889645874e-05,
+      "loss": 0.0175,
+      "step": 66500
+    },
+    {
+      "epoch": 401.53,
+      "learning_rate": 4.665446123039534e-05,
+      "loss": 0.0172,
+      "step": 66520
+    },
+    {
+      "epoch": 401.65,
+      "learning_rate": 4.6609683266584974e-05,
+      "loss": 0.0156,
+      "step": 66540
+    },
+    {
+      "epoch": 401.77,
+      "learning_rate": 4.6564908034291553e-05,
+      "loss": 0.0184,
+      "step": 66560
+    },
+    {
+      "epoch": 401.89,
+      "learning_rate": 4.6520135569589675e-05,
+      "loss": 0.0167,
+      "step": 66580
+    },
+    {
+      "epoch": 402.01,
+      "learning_rate": 4.647536590855174e-05,
+      "loss": 0.0152,
+      "step": 66600
+    },
+    {
+      "epoch": 402.13,
+      "learning_rate": 4.643059908724781e-05,
+      "loss": 0.0143,
+      "step": 66620
+    },
+    {
+      "epoch": 402.25,
+      "learning_rate": 4.6385835141745734e-05,
+      "loss": 0.013,
+      "step": 66640
+    },
+    {
+      "epoch": 402.37,
+      "learning_rate": 4.6341074108111e-05,
+      "loss": 0.0212,
+      "step": 66660
+    },
+    {
+      "epoch": 402.49,
+      "learning_rate": 4.6296316022406776e-05,
+      "loss": 0.0212,
+      "step": 66680
+    },
+    {
+      "epoch": 402.62,
+      "learning_rate": 4.625156092069383e-05,
+      "loss": 0.0166,
+      "step": 66700
+    },
+    {
+      "epoch": 402.74,
+      "learning_rate": 4.620680883903054e-05,
+      "loss": 0.0207,
+      "step": 66720
+    },
+    {
+      "epoch": 402.86,
+      "learning_rate": 4.6162059813472876e-05,
+      "loss": 0.0158,
+      "step": 66740
+    },
+    {
+      "epoch": 402.98,
+      "learning_rate": 4.611731388007426e-05,
+      "loss": 0.0153,
+      "step": 66760
+    },
+    {
+      "epoch": 403.1,
+      "learning_rate": 4.6072571074885774e-05,
+      "loss": 0.0129,
+      "step": 66780
+    },
+    {
+      "epoch": 403.22,
+      "learning_rate": 4.602783143395579e-05,
+      "loss": 0.0155,
+      "step": 66800
+    },
+    {
+      "epoch": 403.34,
+      "learning_rate": 4.5983094993330296e-05,
+      "loss": 0.0158,
+      "step": 66820
+    },
+    {
+      "epoch": 403.46,
+      "learning_rate": 4.59383617890526e-05,
+      "loss": 0.0194,
+      "step": 66840
+    },
+    {
+      "epoch": 403.58,
+      "learning_rate": 4.5893631857163457e-05,
+      "loss": 0.0187,
+      "step": 66860
+    },
+    {
+      "epoch": 403.7,
+      "learning_rate": 4.584890523370096e-05,
+      "loss": 0.016,
+      "step": 66880
+    },
+    {
+      "epoch": 403.82,
+      "learning_rate": 4.5804181954700524e-05,
+      "loss": 0.021,
+      "step": 66900
+    },
+    {
+      "epoch": 403.94,
+      "learning_rate": 4.575946205619492e-05,
+      "loss": 0.0194,
+      "step": 66920
+    },
+    {
+      "epoch": 404.06,
+      "learning_rate": 4.5714745574214126e-05,
+      "loss": 0.0163,
+      "step": 66940
+    },
+    {
+      "epoch": 404.19,
+      "learning_rate": 4.5670032544785444e-05,
+      "loss": 0.0153,
+      "step": 66960
+    },
+    {
+      "epoch": 404.31,
+      "learning_rate": 4.5625323003933304e-05,
+      "loss": 0.027,
+      "step": 66980
+    },
+    {
+      "epoch": 404.43,
+      "learning_rate": 4.558061698767941e-05,
+      "loss": 0.0191,
+      "step": 67000
+    },
+    {
+      "epoch": 404.43,
+      "eval_accuracy": 0.0012545564105856902,
+      "eval_loss": 22.592748641967773,
+      "eval_runtime": 18.1967,
+      "eval_samples_per_second": 112.108,
+      "eval_steps_per_second": 2.363,
+      "step": 67000
+    },
+    {
+      "epoch": 404.55,
+      "learning_rate": 4.553591453204261e-05,
+      "loss": 0.0186,
+      "step": 67020
+    },
+    {
+      "epoch": 404.67,
+      "learning_rate": 4.549121567303881e-05,
+      "loss": 0.0181,
+      "step": 67040
+    },
+    {
+      "epoch": 404.79,
+      "learning_rate": 4.544652044668113e-05,
+      "loss": 0.0183,
+      "step": 67060
+    },
+    {
+      "epoch": 404.91,
+      "learning_rate": 4.5401828888979665e-05,
+      "loss": 0.02,
+      "step": 67080
+    },
+    {
+      "epoch": 405.03,
+      "learning_rate": 4.535714103594162e-05,
+      "loss": 0.0195,
+      "step": 67100
+    },
+    {
+      "epoch": 405.15,
+      "learning_rate": 4.531245692357117e-05,
+      "loss": 0.0172,
+      "step": 67120
+    },
+    {
+      "epoch": 405.27,
+      "learning_rate": 4.5267776587869503e-05,
+      "loss": 0.0164,
+      "step": 67140
+    },
+    {
+      "epoch": 405.39,
+      "learning_rate": 4.5223100064834786e-05,
+      "loss": 0.0172,
+      "step": 67160
+    },
+    {
+      "epoch": 405.51,
+      "learning_rate": 4.517842739046205e-05,
+      "loss": 0.0305,
+      "step": 67180
+    },
+    {
+      "epoch": 405.63,
+      "learning_rate": 4.5133758600743295e-05,
+      "loss": 0.0237,
+      "step": 67200
+    },
+    {
+      "epoch": 405.75,
+      "learning_rate": 4.5089093731667333e-05,
+      "loss": 0.0231,
+      "step": 67220
+    },
+    {
+      "epoch": 405.88,
+      "learning_rate": 4.504443281921985e-05,
+      "loss": 0.0239,
+      "step": 67240
+    },
+    {
+      "epoch": 406.0,
+      "learning_rate": 4.499977589938333e-05,
+      "loss": 0.0247,
+      "step": 67260
+    },
+    {
+      "epoch": 406.12,
+      "learning_rate": 4.495512300813707e-05,
+      "loss": 0.0231,
+      "step": 67280
+    },
+    {
+      "epoch": 406.24,
+      "learning_rate": 4.4910474181457055e-05,
+      "loss": 0.0243,
+      "step": 67300
+    },
+    {
+      "epoch": 406.36,
+      "learning_rate": 4.486582945531607e-05,
+      "loss": 0.0259,
+      "step": 67320
+    },
+    {
+      "epoch": 406.48,
+      "learning_rate": 4.482118886568357e-05,
+      "loss": 0.0348,
+      "step": 67340
+    },
+    {
+      "epoch": 406.6,
+      "learning_rate": 4.4776552448525646e-05,
+      "loss": 0.0428,
+      "step": 67360
+    },
+    {
+      "epoch": 406.72,
+      "learning_rate": 4.473192023980509e-05,
+      "loss": 0.0538,
+      "step": 67380
+    },
+    {
+      "epoch": 406.84,
+      "learning_rate": 4.468729227548122e-05,
+      "loss": 0.067,
+      "step": 67400
+    },
+    {
+      "epoch": 406.96,
+      "learning_rate": 4.464266859151003e-05,
+      "loss": 0.0714,
+      "step": 67420
+    },
+    {
+      "epoch": 407.08,
+      "learning_rate": 4.459804922384398e-05,
+      "loss": 0.0662,
+      "step": 67440
+    },
+    {
+      "epoch": 407.2,
+      "learning_rate": 4.455343420843207e-05,
+      "loss": 0.0684,
+      "step": 67460
+    },
+    {
+      "epoch": 407.32,
+      "learning_rate": 4.4508823581219886e-05,
+      "loss": 0.0719,
+      "step": 67480
+    },
+    {
+      "epoch": 407.44,
+      "learning_rate": 4.446421737814933e-05,
+      "loss": 0.0727,
+      "step": 67500
+    },
+    {
+      "epoch": 407.57,
+      "learning_rate": 4.4419615635158875e-05,
+      "loss": 0.0756,
+      "step": 67520
+    },
+    {
+      "epoch": 407.69,
+      "learning_rate": 4.437501838818327e-05,
+      "loss": 0.0771,
+      "step": 67540
+    },
+    {
+      "epoch": 407.81,
+      "learning_rate": 4.433042567315378e-05,
+      "loss": 0.0765,
+      "step": 67560
+    },
+    {
+      "epoch": 407.93,
+      "learning_rate": 4.42858375259979e-05,
+      "loss": 0.0769,
+      "step": 67580
+    },
+    {
+      "epoch": 408.05,
+      "learning_rate": 4.4241253982639525e-05,
+      "loss": 0.0689,
+      "step": 67600
+    },
+    {
+      "epoch": 408.17,
+      "learning_rate": 4.4196675078998814e-05,
+      "loss": 0.0606,
+      "step": 67620
+    },
+    {
+      "epoch": 408.29,
+      "learning_rate": 4.415210085099215e-05,
+      "loss": 0.0647,
+      "step": 67640
+    },
+    {
+      "epoch": 408.41,
+      "learning_rate": 4.410753133453222e-05,
+      "loss": 0.0667,
+      "step": 67660
+    },
+    {
+      "epoch": 408.53,
+      "learning_rate": 4.4062966565527844e-05,
+      "loss": 0.0672,
+      "step": 67680
+    },
+    {
+      "epoch": 408.65,
+      "learning_rate": 4.40184065798841e-05,
+      "loss": 0.0675,
+      "step": 67700
+    },
+    {
+      "epoch": 408.77,
+      "learning_rate": 4.39738514135021e-05,
+      "loss": 0.0684,
+      "step": 67720
+    },
+    {
+      "epoch": 408.89,
+      "learning_rate": 4.3929301102279183e-05,
+      "loss": 0.0696,
+      "step": 67740
+    },
+    {
+      "epoch": 409.01,
+      "learning_rate": 4.388475568210873e-05,
+      "loss": 0.067,
+      "step": 67760
+    },
+    {
+      "epoch": 409.13,
+      "learning_rate": 4.3840215188880157e-05,
+      "loss": 0.052,
+      "step": 67780
+    },
+    {
+      "epoch": 409.26,
+      "learning_rate": 4.379567965847896e-05,
+      "loss": 0.0557,
+      "step": 67800
+    },
+    {
+      "epoch": 409.38,
+      "learning_rate": 4.3751149126786584e-05,
+      "loss": 0.0573,
+      "step": 67820
+    },
+    {
+      "epoch": 409.5,
+      "learning_rate": 4.370662362968051e-05,
+      "loss": 0.0599,
+      "step": 67840
+    },
+    {
+      "epoch": 409.62,
+      "learning_rate": 4.366210320303407e-05,
+      "loss": 0.0615,
+      "step": 67860
+    },
+    {
+      "epoch": 409.74,
+      "learning_rate": 4.361758788271662e-05,
+      "loss": 0.062,
+      "step": 67880
+    },
+    {
+      "epoch": 409.86,
+      "learning_rate": 4.357307770459335e-05,
+      "loss": 0.0625,
+      "step": 67900
+    },
+    {
+      "epoch": 409.98,
+      "learning_rate": 4.352857270452527e-05,
+      "loss": 0.0627,
+      "step": 67920
+    },
+    {
+      "epoch": 410.1,
+      "learning_rate": 4.348407291836928e-05,
+      "loss": 0.0482,
+      "step": 67940
+    },
+    {
+      "epoch": 410.22,
+      "learning_rate": 4.343957838197805e-05,
+      "loss": 0.0469,
+      "step": 67960
+    },
+    {
+      "epoch": 410.34,
+      "learning_rate": 4.339508913120002e-05,
+      "loss": 0.0474,
+      "step": 67980
+    },
+    {
+      "epoch": 410.46,
+      "learning_rate": 4.335060520187938e-05,
+      "loss": 0.0493,
+      "step": 68000
+    },
+    {
+      "epoch": 410.46,
+      "eval_accuracy": 0.0012644348075194357,
+      "eval_loss": 22.606916427612305,
+      "eval_runtime": 18.1884,
+      "eval_samples_per_second": 112.159,
+      "eval_steps_per_second": 2.364,
+      "step": 68000
+    },
+    {
+      "epoch": 410.58,
+      "learning_rate": 4.3306126629856e-05,
+      "loss": 0.0515,
+      "step": 68020
+    },
+    {
+      "epoch": 410.7,
+      "learning_rate": 4.3261653450965536e-05,
+      "loss": 0.0526,
+      "step": 68040
+    },
+    {
+      "epoch": 410.82,
+      "learning_rate": 4.3217185701039144e-05,
+      "loss": 0.055,
+      "step": 68060
+    },
+    {
+      "epoch": 410.95,
+      "learning_rate": 4.317272341590373e-05,
+      "loss": 0.0548,
+      "step": 68080
+    },
+    {
+      "epoch": 411.07,
+      "learning_rate": 4.312826663138174e-05,
+      "loss": 0.0479,
+      "step": 68100
+    },
+    {
+      "epoch": 411.19,
+      "learning_rate": 4.308381538329123e-05,
+      "loss": 0.0404,
+      "step": 68120
+    },
+    {
+      "epoch": 411.31,
+      "learning_rate": 4.3039369707445714e-05,
+      "loss": 0.0391,
+      "step": 68140
+    },
+    {
+      "epoch": 411.43,
+      "learning_rate": 4.2994929639654315e-05,
+      "loss": 0.0384,
+      "step": 68160
+    },
+    {
+      "epoch": 411.55,
+      "learning_rate": 4.295049521572157e-05,
+      "loss": 0.0377,
+      "step": 68180
+    },
+    {
+      "epoch": 411.67,
+      "learning_rate": 4.29060664714475e-05,
+      "loss": 0.0376,
+      "step": 68200
+    },
+    {
+      "epoch": 411.79,
+      "learning_rate": 4.2861643442627564e-05,
+      "loss": 0.0374,
+      "step": 68220
+    },
+    {
+      "epoch": 411.91,
+      "learning_rate": 4.281722616505254e-05,
+      "loss": 0.0371,
+      "step": 68240
+    },
+    {
+      "epoch": 412.03,
+      "learning_rate": 4.277281467450869e-05,
+      "loss": 0.0345,
+      "step": 68260
+    },
+    {
+      "epoch": 412.15,
+      "learning_rate": 4.27284090067775e-05,
+      "loss": 0.0258,
+      "step": 68280
+    },
+    {
+      "epoch": 412.27,
+      "learning_rate": 4.2684009197635824e-05,
+      "loss": 0.0242,
+      "step": 68300
+    },
+    {
+      "epoch": 412.39,
+      "learning_rate": 4.2639615282855816e-05,
+      "loss": 0.0229,
+      "step": 68320
+    },
+    {
+      "epoch": 412.52,
+      "learning_rate": 4.259522729820482e-05,
+      "loss": 0.0223,
+      "step": 68340
+    },
+    {
+      "epoch": 412.64,
+      "learning_rate": 4.255084527944545e-05,
+      "loss": 0.0217,
+      "step": 68360
+    },
+    {
+      "epoch": 412.76,
+      "learning_rate": 4.250646926233547e-05,
+      "loss": 0.021,
+      "step": 68380
+    },
+    {
+      "epoch": 412.88,
+      "learning_rate": 4.246209928262789e-05,
+      "loss": 0.0206,
+      "step": 68400
+    },
+    {
+      "epoch": 413.0,
+      "learning_rate": 4.241773537607073e-05,
+      "loss": 0.0201,
+      "step": 68420
+    },
+    {
+      "epoch": 413.12,
+      "learning_rate": 4.237337757840726e-05,
+      "loss": 0.0153,
+      "step": 68440
+    },
+    {
+      "epoch": 413.24,
+      "learning_rate": 4.232902592537569e-05,
+      "loss": 0.0145,
+      "step": 68460
+    },
+    {
+      "epoch": 413.36,
+      "learning_rate": 4.2284680452709394e-05,
+      "loss": 0.0141,
+      "step": 68480
+    },
+    {
+      "epoch": 413.48,
+      "learning_rate": 4.22403411961367e-05,
+      "loss": 0.014,
+      "step": 68500
+    },
+    {
+      "epoch": 413.6,
+      "learning_rate": 4.219600819138095e-05,
+      "loss": 0.0139,
+      "step": 68520
+    },
+    {
+      "epoch": 413.72,
+      "learning_rate": 4.215168147416045e-05,
+      "loss": 0.0141,
+      "step": 68540
+    },
+    {
+      "epoch": 413.84,
+      "learning_rate": 4.2107361080188415e-05,
+      "loss": 0.0139,
+      "step": 68560
+    },
+    {
+      "epoch": 413.96,
+      "learning_rate": 4.206304704517301e-05,
+      "loss": 0.0139,
+      "step": 68580
+    },
+    {
+      "epoch": 414.08,
+      "learning_rate": 4.201873940481722e-05,
+      "loss": 0.0124,
+      "step": 68600
+    },
+    {
+      "epoch": 414.21,
+      "learning_rate": 4.1974438194818946e-05,
+      "loss": 0.012,
+      "step": 68620
+    },
+    {
+      "epoch": 414.33,
+      "learning_rate": 4.193014345087087e-05,
+      "loss": 0.012,
+      "step": 68640
+    },
+    {
+      "epoch": 414.45,
+      "learning_rate": 4.188585520866045e-05,
+      "loss": 0.0121,
+      "step": 68660
+    },
+    {
+      "epoch": 414.57,
+      "learning_rate": 4.184157350386994e-05,
+      "loss": 0.0122,
+      "step": 68680
+    },
+    {
+      "epoch": 414.69,
+      "learning_rate": 4.179729837217628e-05,
+      "loss": 0.0122,
+      "step": 68700
+    },
+    {
+      "epoch": 414.81,
+      "learning_rate": 4.1753029849251175e-05,
+      "loss": 0.0124,
+      "step": 68720
+    },
+    {
+      "epoch": 414.93,
+      "learning_rate": 4.170876797076095e-05,
+      "loss": 0.0124,
+      "step": 68740
+    },
+    {
+      "epoch": 415.05,
+      "learning_rate": 4.166451277236661e-05,
+      "loss": 0.0121,
+      "step": 68760
+    },
+    {
+      "epoch": 415.17,
+      "learning_rate": 4.16202642897238e-05,
+      "loss": 0.0114,
+      "step": 68780
+    },
+    {
+      "epoch": 415.29,
+      "learning_rate": 4.1576022558482666e-05,
+      "loss": 0.0115,
+      "step": 68800
+    },
+    {
+      "epoch": 415.41,
+      "learning_rate": 4.153178761428802e-05,
+      "loss": 0.0115,
+      "step": 68820
+    },
+    {
+      "epoch": 415.53,
+      "learning_rate": 4.148755949277913e-05,
+      "loss": 0.0116,
+      "step": 68840
+    },
+    {
+      "epoch": 415.65,
+      "learning_rate": 4.144333822958981e-05,
+      "loss": 0.0118,
+      "step": 68860
+    },
+    {
+      "epoch": 415.77,
+      "learning_rate": 4.139912386034833e-05,
+      "loss": 0.0119,
+      "step": 68880
+    },
+    {
+      "epoch": 415.9,
+      "learning_rate": 4.135491642067739e-05,
+      "loss": 0.0119,
+      "step": 68900
+    },
+    {
+      "epoch": 416.02,
+      "learning_rate": 4.131071594619416e-05,
+      "loss": 0.0119,
+      "step": 68920
+    },
+    {
+      "epoch": 416.14,
+      "learning_rate": 4.1266522472510115e-05,
+      "loss": 0.0111,
+      "step": 68940
+    },
+    {
+      "epoch": 416.26,
+      "learning_rate": 4.122233603523121e-05,
+      "loss": 0.0112,
+      "step": 68960
+    },
+    {
+      "epoch": 416.38,
+      "learning_rate": 4.117815666995758e-05,
+      "loss": 0.0113,
+      "step": 68980
+    },
+    {
+      "epoch": 416.5,
+      "learning_rate": 4.11339844122838e-05,
+      "loss": 0.0115,
+      "step": 69000
+    },
+    {
+      "epoch": 416.5,
+      "eval_accuracy": 0.001247147612885381,
+      "eval_loss": 22.865169525146484,
+      "eval_runtime": 18.2075,
+      "eval_samples_per_second": 112.042,
+      "eval_steps_per_second": 2.362,
+      "step": 69000
+    },
+    {
+      "epoch": 416.62,
+      "learning_rate": 4.108981929779863e-05,
+      "loss": 0.0115,
+      "step": 69020
+    },
+    {
+      "epoch": 416.74,
+      "learning_rate": 4.104566136208512e-05,
+      "loss": 0.0117,
+      "step": 69040
+    },
+    {
+      "epoch": 416.86,
+      "learning_rate": 4.100151064072052e-05,
+      "loss": 0.0118,
+      "step": 69060
+    },
+    {
+      "epoch": 416.98,
+      "learning_rate": 4.095736716927625e-05,
+      "loss": 0.0118,
+      "step": 69080
+    },
+    {
+      "epoch": 417.1,
+      "learning_rate": 4.091323098331794e-05,
+      "loss": 0.0109,
+      "step": 69100
+    },
+    {
+      "epoch": 417.22,
+      "learning_rate": 4.0869102118405276e-05,
+      "loss": 0.011,
+      "step": 69120
+    },
+    {
+      "epoch": 417.34,
+      "learning_rate": 4.082498061009215e-05,
+      "loss": 0.0112,
+      "step": 69140
+    },
+    {
+      "epoch": 417.46,
+      "learning_rate": 4.078086649392639e-05,
+      "loss": 0.0113,
+      "step": 69160
+    },
+    {
+      "epoch": 417.59,
+      "learning_rate": 4.0736759805449986e-05,
+      "loss": 0.0114,
+      "step": 69180
+    },
+    {
+      "epoch": 417.71,
+      "learning_rate": 4.06926605801989e-05,
+      "loss": 0.0116,
+      "step": 69200
+    },
+    {
+      "epoch": 417.83,
+      "learning_rate": 4.0648568853703076e-05,
+      "loss": 0.0115,
+      "step": 69220
+    },
+    {
+      "epoch": 417.95,
+      "learning_rate": 4.060448466148642e-05,
+      "loss": 0.0117,
+      "step": 69240
+    },
+    {
+      "epoch": 418.07,
+      "learning_rate": 4.056040803906675e-05,
+      "loss": 0.0111,
+      "step": 69260
+    },
+    {
+      "epoch": 418.19,
+      "learning_rate": 4.051633902195585e-05,
+      "loss": 0.0109,
+      "step": 69280
+    },
+    {
+      "epoch": 418.31,
+      "learning_rate": 4.047227764565929e-05,
+      "loss": 0.0112,
+      "step": 69300
+    },
+    {
+      "epoch": 418.43,
+      "learning_rate": 4.0428223945676524e-05,
+      "loss": 0.0112,
+      "step": 69320
+    },
+    {
+      "epoch": 418.55,
+      "learning_rate": 4.0384177957500866e-05,
+      "loss": 0.0114,
+      "step": 69340
+    },
+    {
+      "epoch": 418.67,
+      "learning_rate": 4.034013971661933e-05,
+      "loss": 0.0113,
+      "step": 69360
+    },
+    {
+      "epoch": 418.79,
+      "learning_rate": 4.0296109258512755e-05,
+      "loss": 0.0115,
+      "step": 69380
+    },
+    {
+      "epoch": 418.91,
+      "learning_rate": 4.025208661865567e-05,
+      "loss": 0.0116,
+      "step": 69400
+    },
+    {
+      "epoch": 419.03,
+      "learning_rate": 4.0208071832516333e-05,
+      "loss": 0.0112,
+      "step": 69420
+    },
+    {
+      "epoch": 419.15,
+      "learning_rate": 4.016406493555666e-05,
+      "loss": 0.0108,
+      "step": 69440
+    },
+    {
+      "epoch": 419.28,
+      "learning_rate": 4.0120065963232186e-05,
+      "loss": 0.011,
+      "step": 69460
+    },
+    {
+      "epoch": 419.4,
+      "learning_rate": 4.007607495099215e-05,
+      "loss": 0.0112,
+      "step": 69480
+    },
+    {
+      "epoch": 419.52,
+      "learning_rate": 4.003209193427924e-05,
+      "loss": 0.0112,
+      "step": 69500
+    },
+    {
+      "epoch": 419.64,
+      "learning_rate": 3.9988116948529844e-05,
+      "loss": 0.0114,
+      "step": 69520
+    },
+    {
+      "epoch": 419.76,
+      "learning_rate": 3.994415002917378e-05,
+      "loss": 0.0115,
+      "step": 69540
+    },
+    {
+      "epoch": 419.88,
+      "learning_rate": 3.990019121163441e-05,
+      "loss": 0.0115,
+      "step": 69560
+    },
+    {
+      "epoch": 420.0,
+      "learning_rate": 3.985624053132855e-05,
+      "loss": 0.0115,
+      "step": 69580
+    },
+    {
+      "epoch": 420.12,
+      "learning_rate": 3.981229802366647e-05,
+      "loss": 0.0105,
+      "step": 69600
+    },
+    {
+      "epoch": 420.24,
+      "learning_rate": 3.976836372405188e-05,
+      "loss": 0.0108,
+      "step": 69620
+    },
+    {
+      "epoch": 420.36,
+      "learning_rate": 3.9724437667881796e-05,
+      "loss": 0.011,
+      "step": 69640
+    },
+    {
+      "epoch": 420.48,
+      "learning_rate": 3.968051989054673e-05,
+      "loss": 0.0111,
+      "step": 69660
+    },
+    {
+      "epoch": 420.6,
+      "learning_rate": 3.963661042743036e-05,
+      "loss": 0.0112,
+      "step": 69680
+    },
+    {
+      "epoch": 420.72,
+      "learning_rate": 3.9592709313909817e-05,
+      "loss": 0.0113,
+      "step": 69700
+    },
+    {
+      "epoch": 420.85,
+      "learning_rate": 3.9548816585355366e-05,
+      "loss": 0.0114,
+      "step": 69720
+    },
+    {
+      "epoch": 420.97,
+      "learning_rate": 3.950493227713064e-05,
+      "loss": 0.0116,
+      "step": 69740
+    },
+    {
+      "epoch": 421.09,
+      "learning_rate": 3.9461056424592416e-05,
+      "loss": 0.0109,
+      "step": 69760
+    },
+    {
+      "epoch": 421.21,
+      "learning_rate": 3.941718906309066e-05,
+      "loss": 0.0108,
+      "step": 69780
+    },
+    {
+      "epoch": 421.33,
+      "learning_rate": 3.937333022796854e-05,
+      "loss": 0.011,
+      "step": 69800
+    },
+    {
+      "epoch": 421.45,
+      "learning_rate": 3.932947995456231e-05,
+      "loss": 0.0111,
+      "step": 69820
+    },
+    {
+      "epoch": 421.57,
+      "learning_rate": 3.928563827820136e-05,
+      "loss": 0.0112,
+      "step": 69840
+    },
+    {
+      "epoch": 421.69,
+      "learning_rate": 3.92418052342081e-05,
+      "loss": 0.0113,
+      "step": 69860
+    },
+    {
+      "epoch": 421.81,
+      "learning_rate": 3.9197980857898074e-05,
+      "loss": 0.0114,
+      "step": 69880
+    },
+    {
+      "epoch": 421.93,
+      "learning_rate": 3.915416518457974e-05,
+      "loss": 0.0114,
+      "step": 69900
+    },
+    {
+      "epoch": 422.05,
+      "learning_rate": 3.9110358249554625e-05,
+      "loss": 0.0111,
+      "step": 69920
+    },
+    {
+      "epoch": 422.17,
+      "learning_rate": 3.906656008811719e-05,
+      "loss": 0.0106,
+      "step": 69940
+    },
+    {
+      "epoch": 422.29,
+      "learning_rate": 3.902277073555479e-05,
+      "loss": 0.0109,
+      "step": 69960
+    },
+    {
+      "epoch": 422.41,
+      "learning_rate": 3.8978990227147755e-05,
+      "loss": 0.011,
+      "step": 69980
+    },
+    {
+      "epoch": 422.54,
+      "learning_rate": 3.893521859816922e-05,
+      "loss": 0.0111,
+      "step": 70000
+    },
+    {
+      "epoch": 422.54,
+      "eval_accuracy": 0.0012476415327320682,
+      "eval_loss": 22.998226165771484,
+      "eval_runtime": 18.2088,
+      "eval_samples_per_second": 112.033,
+      "eval_steps_per_second": 2.361,
+      "step": 70000
+    },
+    {
+      "epoch": 422.66,
+      "learning_rate": 3.889145588388521e-05,
+      "loss": 0.0114,
+      "step": 70020
+    },
+    {
+      "epoch": 422.78,
+      "learning_rate": 3.884770211955454e-05,
+      "loss": 0.0114,
+      "step": 70040
+    },
+    {
+      "epoch": 422.9,
+      "learning_rate": 3.880395734042884e-05,
+      "loss": 0.0114,
+      "step": 70060
+    },
+    {
+      "epoch": 423.02,
+      "learning_rate": 3.87602215817525e-05,
+      "loss": 0.0114,
+      "step": 70080
+    },
+    {
+      "epoch": 423.14,
+      "learning_rate": 3.87164948787626e-05,
+      "loss": 0.0105,
+      "step": 70100
+    },
+    {
+      "epoch": 423.26,
+      "learning_rate": 3.867277726668899e-05,
+      "loss": 0.0108,
+      "step": 70120
+    },
+    {
+      "epoch": 423.38,
+      "learning_rate": 3.862906878075413e-05,
+      "loss": 0.011,
+      "step": 70140
+    },
+    {
+      "epoch": 423.5,
+      "learning_rate": 3.858536945617318e-05,
+      "loss": 0.0113,
+      "step": 70160
+    },
+    {
+      "epoch": 423.62,
+      "learning_rate": 3.854167932815387e-05,
+      "loss": 0.0113,
+      "step": 70180
+    },
+    {
+      "epoch": 423.74,
+      "learning_rate": 3.849799843189657e-05,
+      "loss": 0.0113,
+      "step": 70200
+    },
+    {
+      "epoch": 423.86,
+      "learning_rate": 3.845432680259419e-05,
+      "loss": 0.0114,
+      "step": 70220
+    },
+    {
+      "epoch": 423.98,
+      "learning_rate": 3.841066447543214e-05,
+      "loss": 0.0114,
+      "step": 70240
+    },
+    {
+      "epoch": 424.1,
+      "learning_rate": 3.836701148558843e-05,
+      "loss": 0.0107,
+      "step": 70260
+    },
+    {
+      "epoch": 424.23,
+      "learning_rate": 3.8323367868233426e-05,
+      "loss": 0.0108,
+      "step": 70280
+    },
+    {
+      "epoch": 424.35,
+      "learning_rate": 3.8279733658530036e-05,
+      "loss": 0.0109,
+      "step": 70300
+    },
+    {
+      "epoch": 424.47,
+      "learning_rate": 3.8236108891633535e-05,
+      "loss": 0.0111,
+      "step": 70320
+    },
+    {
+      "epoch": 424.59,
+      "learning_rate": 3.819249360269162e-05,
+      "loss": 0.0111,
+      "step": 70340
+    },
+    {
+      "epoch": 424.71,
+      "learning_rate": 3.814888782684435e-05,
+      "loss": 0.0113,
+      "step": 70360
+    },
+    {
+      "epoch": 424.83,
+      "learning_rate": 3.810529159922408e-05,
+      "loss": 0.0115,
+      "step": 70380
+    },
+    {
+      "epoch": 424.95,
+      "learning_rate": 3.8061704954955556e-05,
+      "loss": 0.0116,
+      "step": 70400
+    },
+    {
+      "epoch": 425.07,
+      "learning_rate": 3.8018127929155687e-05,
+      "loss": 0.0108,
+      "step": 70420
+    },
+    {
+      "epoch": 425.19,
+      "learning_rate": 3.797456055693375e-05,
+      "loss": 0.0107,
+      "step": 70440
+    },
+    {
+      "epoch": 425.31,
+      "learning_rate": 3.7931002873391154e-05,
+      "loss": 0.0109,
+      "step": 70460
+    },
+    {
+      "epoch": 425.43,
+      "learning_rate": 3.788745491362155e-05,
+      "loss": 0.011,
+      "step": 70480
+    },
+    {
+      "epoch": 425.55,
+      "learning_rate": 3.7843916712710756e-05,
+      "loss": 0.0113,
+      "step": 70500
+    },
+    {
+      "epoch": 425.67,
+      "learning_rate": 3.780038830573668e-05,
+      "loss": 0.0113,
+      "step": 70520
+    },
+    {
+      "epoch": 425.79,
+      "learning_rate": 3.775686972776942e-05,
+      "loss": 0.0115,
+      "step": 70540
+    },
+    {
+      "epoch": 425.92,
+      "learning_rate": 3.771336101387105e-05,
+      "loss": 0.0116,
+      "step": 70560
+    },
+    {
+      "epoch": 426.04,
+      "learning_rate": 3.766986219909583e-05,
+      "loss": 0.0115,
+      "step": 70580
+    },
+    {
+      "epoch": 426.16,
+      "learning_rate": 3.7626373318489886e-05,
+      "loss": 0.0109,
+      "step": 70600
+    },
+    {
+      "epoch": 426.28,
+      "learning_rate": 3.758289440709149e-05,
+      "loss": 0.0114,
+      "step": 70620
+    },
+    {
+      "epoch": 426.4,
+      "learning_rate": 3.753942549993079e-05,
+      "loss": 0.0117,
+      "step": 70640
+    },
+    {
+      "epoch": 426.52,
+      "learning_rate": 3.749596663202989e-05,
+      "loss": 0.0118,
+      "step": 70660
+    },
+    {
+      "epoch": 426.64,
+      "learning_rate": 3.7452517838402834e-05,
+      "loss": 0.0118,
+      "step": 70680
+    },
+    {
+      "epoch": 426.76,
+      "learning_rate": 3.74090791540555e-05,
+      "loss": 0.0123,
+      "step": 70700
+    },
+    {
+      "epoch": 426.88,
+      "learning_rate": 3.7365650613985706e-05,
+      "loss": 0.0126,
+      "step": 70720
+    },
+    {
+      "epoch": 427.0,
+      "learning_rate": 3.732223225318298e-05,
+      "loss": 0.0135,
+      "step": 70740
+    },
+    {
+      "epoch": 427.12,
+      "learning_rate": 3.727882410662874e-05,
+      "loss": 0.0146,
+      "step": 70760
+    },
+    {
+      "epoch": 427.24,
+      "learning_rate": 3.723542620929616e-05,
+      "loss": 0.0192,
+      "step": 70780
+    },
+    {
+      "epoch": 427.36,
+      "learning_rate": 3.7194207732018324e-05,
+      "loss": 0.0373,
+      "step": 70800
+    },
+    {
+      "epoch": 427.48,
+      "learning_rate": 3.715299856619e-05,
+      "loss": 0.0973,
+      "step": 70820
+    },
+    {
+      "epoch": 427.61,
+      "learning_rate": 3.710963058929701e-05,
+      "loss": 0.1284,
+      "step": 70840
+    },
+    {
+      "epoch": 427.73,
+      "learning_rate": 3.706844063031376e-05,
+      "loss": 0.1367,
+      "step": 70860
+    },
+    {
+      "epoch": 427.85,
+      "learning_rate": 3.702509293757578e-05,
+      "loss": 0.1379,
+      "step": 70880
+    },
+    {
+      "epoch": 427.97,
+      "learning_rate": 3.698175569848608e-05,
+      "loss": 0.1376,
+      "step": 70900
+    },
+    {
+      "epoch": 428.09,
+      "learning_rate": 3.6938428947960645e-05,
+      "loss": 0.1232,
+      "step": 70920
+    },
+    {
+      "epoch": 428.21,
+      "learning_rate": 3.6895112720907124e-05,
+      "loss": 0.1227,
+      "step": 70940
+    },
+    {
+      "epoch": 428.33,
+      "learning_rate": 3.6851807052224546e-05,
+      "loss": 0.1246,
+      "step": 70960
+    },
+    {
+      "epoch": 428.45,
+      "learning_rate": 3.680851197680355e-05,
+      "loss": 0.1241,
+      "step": 70980
+    },
+    {
+      "epoch": 428.57,
+      "learning_rate": 3.67652275295262e-05,
+      "loss": 0.1182,
+      "step": 71000
+    },
+    {
+      "epoch": 428.57,
+      "eval_accuracy": 0.0012550503304323774,
+      "eval_loss": 22.662765502929688,
+      "eval_runtime": 18.1641,
+      "eval_samples_per_second": 112.31,
+      "eval_steps_per_second": 2.367,
+      "step": 71000
+    },
+    {
+      "epoch": 428.69,
+      "learning_rate": 3.672195374526598e-05,
+      "loss": 0.1179,
+      "step": 71020
+    },
+    {
+      "epoch": 428.81,
+      "learning_rate": 3.667869065888779e-05,
+      "loss": 0.1156,
+      "step": 71040
+    },
+    {
+      "epoch": 428.93,
+      "learning_rate": 3.663543830524793e-05,
+      "loss": 0.1116,
+      "step": 71060
+    },
+    {
+      "epoch": 429.05,
+      "learning_rate": 3.659219671919404e-05,
+      "loss": 0.0908,
+      "step": 71080
+    },
+    {
+      "epoch": 429.18,
+      "learning_rate": 3.654896593556506e-05,
+      "loss": 0.0793,
+      "step": 71100
+    },
+    {
+      "epoch": 429.3,
+      "learning_rate": 3.6505745989191274e-05,
+      "loss": 0.0815,
+      "step": 71120
+    },
+    {
+      "epoch": 429.42,
+      "learning_rate": 3.6462536914894215e-05,
+      "loss": 0.0813,
+      "step": 71140
+    },
+    {
+      "epoch": 429.54,
+      "learning_rate": 3.6419338747486656e-05,
+      "loss": 0.0819,
+      "step": 71160
+    },
+    {
+      "epoch": 429.66,
+      "learning_rate": 3.6376151521772565e-05,
+      "loss": 0.0826,
+      "step": 71180
+    },
+    {
+      "epoch": 429.78,
+      "learning_rate": 3.633297527254711e-05,
+      "loss": 0.0829,
+      "step": 71200
+    },
+    {
+      "epoch": 429.9,
+      "learning_rate": 3.628981003459664e-05,
+      "loss": 0.0794,
+      "step": 71220
+    },
+    {
+      "epoch": 430.02,
+      "learning_rate": 3.6246655842698575e-05,
+      "loss": 0.0743,
+      "step": 71240
+    },
+    {
+      "epoch": 430.14,
+      "learning_rate": 3.620351273162147e-05,
+      "loss": 0.0547,
+      "step": 71260
+    },
+    {
+      "epoch": 430.26,
+      "learning_rate": 3.6160380736125e-05,
+      "loss": 0.0586,
+      "step": 71280
+    },
+    {
+      "epoch": 430.38,
+      "learning_rate": 3.611725989095976e-05,
+      "loss": 0.0598,
+      "step": 71300
+    },
+    {
+      "epoch": 430.5,
+      "learning_rate": 3.60741502308675e-05,
+      "loss": 0.0622,
+      "step": 71320
+    },
+    {
+      "epoch": 430.62,
+      "learning_rate": 3.603105179058085e-05,
+      "loss": 0.0603,
+      "step": 71340
+    },
+    {
+      "epoch": 430.74,
+      "learning_rate": 3.598796460482348e-05,
+      "loss": 0.0602,
+      "step": 71360
+    },
+    {
+      "epoch": 430.87,
+      "learning_rate": 3.594488870830992e-05,
+      "loss": 0.061,
+      "step": 71380
+    },
+    {
+      "epoch": 430.99,
+      "learning_rate": 3.5901824135745664e-05,
+      "loss": 0.0614,
+      "step": 71400
+    },
+    {
+      "epoch": 431.11,
+      "learning_rate": 3.5858770921827075e-05,
+      "loss": 0.0439,
+      "step": 71420
+    },
+    {
+      "epoch": 431.23,
+      "learning_rate": 3.581572910124131e-05,
+      "loss": 0.0416,
+      "step": 71440
+    },
+    {
+      "epoch": 431.35,
+      "learning_rate": 3.577269870866644e-05,
+      "loss": 0.0403,
+      "step": 71460
+    },
+    {
+      "epoch": 431.47,
+      "learning_rate": 3.572967977877122e-05,
+      "loss": 0.0399,
+      "step": 71480
+    },
+    {
+      "epoch": 431.59,
+      "learning_rate": 3.568667234621527e-05,
+      "loss": 0.0397,
+      "step": 71500
+    },
+    {
+      "epoch": 431.71,
+      "learning_rate": 3.5643676445648855e-05,
+      "loss": 0.0402,
+      "step": 71520
+    },
+    {
+      "epoch": 431.83,
+      "learning_rate": 3.560069211171303e-05,
+      "loss": 0.0394,
+      "step": 71540
+    },
+    {
+      "epoch": 431.95,
+      "learning_rate": 3.55577193790395e-05,
+      "loss": 0.0388,
+      "step": 71560
+    },
+    {
+      "epoch": 432.07,
+      "learning_rate": 3.55147582822506e-05,
+      "loss": 0.0318,
+      "step": 71580
+    },
+    {
+      "epoch": 432.19,
+      "learning_rate": 3.547180885595932e-05,
+      "loss": 0.0254,
+      "step": 71600
+    },
+    {
+      "epoch": 432.31,
+      "learning_rate": 3.5428871134769226e-05,
+      "loss": 0.0237,
+      "step": 71620
+    },
+    {
+      "epoch": 432.43,
+      "learning_rate": 3.5385945153274494e-05,
+      "loss": 0.0225,
+      "step": 71640
+    },
+    {
+      "epoch": 432.56,
+      "learning_rate": 3.534303094605975e-05,
+      "loss": 0.0218,
+      "step": 71660
+    },
+    {
+      "epoch": 432.68,
+      "learning_rate": 3.530012854770026e-05,
+      "loss": 0.0208,
+      "step": 71680
+    },
+    {
+      "epoch": 432.8,
+      "learning_rate": 3.525723799276169e-05,
+      "loss": 0.0207,
+      "step": 71700
+    },
+    {
+      "epoch": 432.92,
+      "learning_rate": 3.521435931580018e-05,
+      "loss": 0.0199,
+      "step": 71720
+    },
+    {
+      "epoch": 433.04,
+      "learning_rate": 3.517149255136232e-05,
+      "loss": 0.0179,
+      "step": 71740
+    },
+    {
+      "epoch": 433.16,
+      "learning_rate": 3.512863773398506e-05,
+      "loss": 0.0143,
+      "step": 71760
+    },
+    {
+      "epoch": 433.28,
+      "learning_rate": 3.508579489819578e-05,
+      "loss": 0.0137,
+      "step": 71780
+    },
+    {
+      "epoch": 433.4,
+      "learning_rate": 3.5042964078512166e-05,
+      "loss": 0.0135,
+      "step": 71800
+    },
+    {
+      "epoch": 433.52,
+      "learning_rate": 3.500014530944223e-05,
+      "loss": 0.0135,
+      "step": 71820
+    },
+    {
+      "epoch": 433.64,
+      "learning_rate": 3.495733862548428e-05,
+      "loss": 0.0134,
+      "step": 71840
+    },
+    {
+      "epoch": 433.76,
+      "learning_rate": 3.491454406112691e-05,
+      "loss": 0.0134,
+      "step": 71860
+    },
+    {
+      "epoch": 433.88,
+      "learning_rate": 3.4871761650848916e-05,
+      "loss": 0.0133,
+      "step": 71880
+    },
+    {
+      "epoch": 434.0,
+      "learning_rate": 3.482899142911931e-05,
+      "loss": 0.0133,
+      "step": 71900
+    },
+    {
+      "epoch": 434.12,
+      "learning_rate": 3.478623343039729e-05,
+      "loss": 0.0116,
+      "step": 71920
+    },
+    {
+      "epoch": 434.25,
+      "learning_rate": 3.474348768913219e-05,
+      "loss": 0.0116,
+      "step": 71940
+    },
+    {
+      "epoch": 434.37,
+      "learning_rate": 3.470075423976351e-05,
+      "loss": 0.0117,
+      "step": 71960
+    },
+    {
+      "epoch": 434.49,
+      "learning_rate": 3.465803311672079e-05,
+      "loss": 0.0117,
+      "step": 71980
+    },
+    {
+      "epoch": 434.61,
+      "learning_rate": 3.4615324354423663e-05,
+      "loss": 0.0118,
+      "step": 72000
+    },
+    {
+      "epoch": 434.61,
+      "eval_accuracy": 0.0012604834487459375,
+      "eval_loss": 22.90355682373047,
+      "eval_runtime": 18.1657,
+      "eval_samples_per_second": 112.3,
+      "eval_steps_per_second": 2.367,
+      "step": 72000
+    },
+    {
+      "epoch": 434.73,
+      "learning_rate": 3.457262798728186e-05,
+      "loss": 0.0119,
+      "step": 72020
+    },
+    {
+      "epoch": 434.85,
+      "learning_rate": 3.452994404969499e-05,
+      "loss": 0.0119,
+      "step": 72040
+    },
+    {
+      "epoch": 434.97,
+      "learning_rate": 3.448727257605282e-05,
+      "loss": 0.012,
+      "step": 72060
+    },
+    {
+      "epoch": 435.09,
+      "learning_rate": 3.4444613600734924e-05,
+      "loss": 0.0113,
+      "step": 72080
+    },
+    {
+      "epoch": 435.21,
+      "learning_rate": 3.4401967158110894e-05,
+      "loss": 0.0111,
+      "step": 72100
+    },
+    {
+      "epoch": 435.33,
+      "learning_rate": 3.435933328254017e-05,
+      "loss": 0.0113,
+      "step": 72120
+    },
+    {
+      "epoch": 435.45,
+      "learning_rate": 3.431671200837213e-05,
+      "loss": 0.0113,
+      "step": 72140
+    },
+    {
+      "epoch": 435.57,
+      "learning_rate": 3.4274103369945955e-05,
+      "loss": 0.0114,
+      "step": 72160
+    },
+    {
+      "epoch": 435.69,
+      "learning_rate": 3.423150740159064e-05,
+      "loss": 0.0116,
+      "step": 72180
+    },
+    {
+      "epoch": 435.81,
+      "learning_rate": 3.418892413762501e-05,
+      "loss": 0.0115,
+      "step": 72200
+    },
+    {
+      "epoch": 435.94,
+      "learning_rate": 3.4146353612357604e-05,
+      "loss": 0.0116,
+      "step": 72220
+    },
+    {
+      "epoch": 436.06,
+      "learning_rate": 3.4103795860086765e-05,
+      "loss": 0.0113,
+      "step": 72240
+    },
+    {
+      "epoch": 436.18,
+      "learning_rate": 3.406125091510046e-05,
+      "loss": 0.0108,
+      "step": 72260
+    },
+    {
+      "epoch": 436.3,
+      "learning_rate": 3.4018718811676406e-05,
+      "loss": 0.011,
+      "step": 72280
+    },
+    {
+      "epoch": 436.42,
+      "learning_rate": 3.397619958408196e-05,
+      "loss": 0.0112,
+      "step": 72300
+    },
+    {
+      "epoch": 436.54,
+      "learning_rate": 3.393369326657408e-05,
+      "loss": 0.0112,
+      "step": 72320
+    },
+    {
+      "epoch": 436.66,
+      "learning_rate": 3.389119989339935e-05,
+      "loss": 0.0113,
+      "step": 72340
+    },
+    {
+      "epoch": 436.78,
+      "learning_rate": 3.384871949879388e-05,
+      "loss": 0.0115,
+      "step": 72360
+    },
+    {
+      "epoch": 436.9,
+      "learning_rate": 3.380625211698342e-05,
+      "loss": 0.0115,
+      "step": 72380
+    },
+    {
+      "epoch": 437.02,
+      "learning_rate": 3.3763797782183104e-05,
+      "loss": 0.0114,
+      "step": 72400
+    },
+    {
+      "epoch": 437.14,
+      "learning_rate": 3.372135652859766e-05,
+      "loss": 0.0107,
+      "step": 72420
+    },
+    {
+      "epoch": 437.26,
+      "learning_rate": 3.367892839042125e-05,
+      "loss": 0.0109,
+      "step": 72440
+    },
+    {
+      "epoch": 437.38,
+      "learning_rate": 3.363651340183743e-05,
+      "loss": 0.011,
+      "step": 72460
+    },
+    {
+      "epoch": 437.51,
+      "learning_rate": 3.359411159701921e-05,
+      "loss": 0.0112,
+      "step": 72480
+    },
+    {
+      "epoch": 437.63,
+      "learning_rate": 3.355172301012893e-05,
+      "loss": 0.0112,
+      "step": 72500
+    },
+    {
+      "epoch": 437.75,
+      "learning_rate": 3.350934767531836e-05,
+      "loss": 0.0113,
+      "step": 72520
+    },
+    {
+      "epoch": 437.87,
+      "learning_rate": 3.346698562672849e-05,
+      "loss": 0.0113,
+      "step": 72540
+    },
+    {
+      "epoch": 437.99,
+      "learning_rate": 3.342463689848967e-05,
+      "loss": 0.0114,
+      "step": 72560
+    },
+    {
+      "epoch": 438.11,
+      "learning_rate": 3.3382301524721525e-05,
+      "loss": 0.0107,
+      "step": 72580
+    },
+    {
+      "epoch": 438.23,
+      "learning_rate": 3.3339979539532894e-05,
+      "loss": 0.0107,
+      "step": 72600
+    },
+    {
+      "epoch": 438.35,
+      "learning_rate": 3.329767097702183e-05,
+      "loss": 0.011,
+      "step": 72620
+    },
+    {
+      "epoch": 438.47,
+      "learning_rate": 3.325537587127558e-05,
+      "loss": 0.0111,
+      "step": 72640
+    },
+    {
+      "epoch": 438.59,
+      "learning_rate": 3.3213094256370546e-05,
+      "loss": 0.0111,
+      "step": 72660
+    },
+    {
+      "epoch": 438.71,
+      "learning_rate": 3.317082616637226e-05,
+      "loss": 0.0112,
+      "step": 72680
+    },
+    {
+      "epoch": 438.83,
+      "learning_rate": 3.3128571635335345e-05,
+      "loss": 0.0113,
+      "step": 72700
+    },
+    {
+      "epoch": 438.95,
+      "learning_rate": 3.308633069730357e-05,
+      "loss": 0.0114,
+      "step": 72720
+    },
+    {
+      "epoch": 439.07,
+      "learning_rate": 3.304410338630961e-05,
+      "loss": 0.0108,
+      "step": 72740
+    },
+    {
+      "epoch": 439.2,
+      "learning_rate": 3.3001889736375326e-05,
+      "loss": 0.0107,
+      "step": 72760
+    },
+    {
+      "epoch": 439.32,
+      "learning_rate": 3.295968978151144e-05,
+      "loss": 0.0108,
+      "step": 72780
+    },
+    {
+      "epoch": 439.44,
+      "learning_rate": 3.291750355571772e-05,
+      "loss": 0.0109,
+      "step": 72800
+    },
+    {
+      "epoch": 439.56,
+      "learning_rate": 3.287533109298283e-05,
+      "loss": 0.0111,
+      "step": 72820
+    },
+    {
+      "epoch": 439.68,
+      "learning_rate": 3.283317242728437e-05,
+      "loss": 0.0111,
+      "step": 72840
+    },
+    {
+      "epoch": 439.8,
+      "learning_rate": 3.279102759258882e-05,
+      "loss": 0.0112,
+      "step": 72860
+    },
+    {
+      "epoch": 439.92,
+      "learning_rate": 3.27488966228515e-05,
+      "loss": 0.0113,
+      "step": 72880
+    },
+    {
+      "epoch": 440.04,
+      "learning_rate": 3.2706779552016595e-05,
+      "loss": 0.011,
+      "step": 72900
+    },
+    {
+      "epoch": 440.16,
+      "learning_rate": 3.266467641401703e-05,
+      "loss": 0.0106,
+      "step": 72920
+    },
+    {
+      "epoch": 440.28,
+      "learning_rate": 3.262258724277459e-05,
+      "loss": 0.0108,
+      "step": 72940
+    },
+    {
+      "epoch": 440.4,
+      "learning_rate": 3.258051207219971e-05,
+      "loss": 0.0109,
+      "step": 72960
+    },
+    {
+      "epoch": 440.52,
+      "learning_rate": 3.253845093619163e-05,
+      "loss": 0.011,
+      "step": 72980
+    },
+    {
+      "epoch": 440.64,
+      "learning_rate": 3.249640386863825e-05,
+      "loss": 0.0111,
+      "step": 73000
+    },
+    {
+      "epoch": 440.64,
+      "eval_accuracy": 0.0012664104869061848,
+      "eval_loss": 23.069204330444336,
+      "eval_runtime": 18.1572,
+      "eval_samples_per_second": 112.352,
+      "eval_steps_per_second": 2.368,
+      "step": 73000
+    },
+    {
+      "epoch": 440.76,
+      "learning_rate": 3.24543709034161e-05,
+      "loss": 0.0112,
+      "step": 73020
+    },
+    {
+      "epoch": 440.89,
+      "learning_rate": 3.2412352074390415e-05,
+      "loss": 0.0112,
+      "step": 73040
+    },
+    {
+      "epoch": 441.01,
+      "learning_rate": 3.237034741541499e-05,
+      "loss": 0.0112,
+      "step": 73060
+    },
+    {
+      "epoch": 441.13,
+      "learning_rate": 3.232835696033222e-05,
+      "loss": 0.0104,
+      "step": 73080
+    },
+    {
+      "epoch": 441.25,
+      "learning_rate": 3.228638074297304e-05,
+      "loss": 0.0106,
+      "step": 73100
+    },
+    {
+      "epoch": 441.37,
+      "learning_rate": 3.224441879715699e-05,
+      "loss": 0.0109,
+      "step": 73120
+    },
+    {
+      "epoch": 441.49,
+      "learning_rate": 3.220247115669198e-05,
+      "loss": 0.0109,
+      "step": 73140
+    },
+    {
+      "epoch": 441.61,
+      "learning_rate": 3.216053785537449e-05,
+      "loss": 0.011,
+      "step": 73160
+    },
+    {
+      "epoch": 441.73,
+      "learning_rate": 3.2118618926989466e-05,
+      "loss": 0.0111,
+      "step": 73180
+    },
+    {
+      "epoch": 441.85,
+      "learning_rate": 3.2076714405310174e-05,
+      "loss": 0.0112,
+      "step": 73200
+    },
+    {
+      "epoch": 441.97,
+      "learning_rate": 3.203482432409836e-05,
+      "loss": 0.0112,
+      "step": 73220
+    },
+    {
+      "epoch": 442.09,
+      "learning_rate": 3.1992948717104085e-05,
+      "loss": 0.0106,
+      "step": 73240
+    },
+    {
+      "epoch": 442.21,
+      "learning_rate": 3.19510876180658e-05,
+      "loss": 0.0105,
+      "step": 73260
+    },
+    {
+      "epoch": 442.33,
+      "learning_rate": 3.1909241060710194e-05,
+      "loss": 0.0108,
+      "step": 73280
+    },
+    {
+      "epoch": 442.45,
+      "learning_rate": 3.1867409078752316e-05,
+      "loss": 0.0109,
+      "step": 73300
+    },
+    {
+      "epoch": 442.58,
+      "learning_rate": 3.182559170589544e-05,
+      "loss": 0.0109,
+      "step": 73320
+    },
+    {
+      "epoch": 442.7,
+      "learning_rate": 3.178378897583105e-05,
+      "loss": 0.0111,
+      "step": 73340
+    },
+    {
+      "epoch": 442.82,
+      "learning_rate": 3.174200092223887e-05,
+      "loss": 0.0113,
+      "step": 73360
+    },
+    {
+      "epoch": 442.94,
+      "learning_rate": 3.1700227578786746e-05,
+      "loss": 0.0111,
+      "step": 73380
+    },
+    {
+      "epoch": 443.06,
+      "learning_rate": 3.165846897913075e-05,
+      "loss": 0.0108,
+      "step": 73400
+    },
+    {
+      "epoch": 443.18,
+      "learning_rate": 3.1616725156914995e-05,
+      "loss": 0.0106,
+      "step": 73420
+    },
+    {
+      "epoch": 443.3,
+      "learning_rate": 3.157499614577173e-05,
+      "loss": 0.0107,
+      "step": 73440
+    },
+    {
+      "epoch": 443.42,
+      "learning_rate": 3.1533281979321306e-05,
+      "loss": 0.0108,
+      "step": 73460
+    },
+    {
+      "epoch": 443.54,
+      "learning_rate": 3.149158269117202e-05,
+      "loss": 0.0108,
+      "step": 73480
+    },
+    {
+      "epoch": 443.66,
+      "learning_rate": 3.144989831492026e-05,
+      "loss": 0.011,
+      "step": 73500
+    },
+    {
+      "epoch": 443.78,
+      "learning_rate": 3.140822888415038e-05,
+      "loss": 0.0111,
+      "step": 73520
+    },
+    {
+      "epoch": 443.9,
+      "learning_rate": 3.136657443243468e-05,
+      "loss": 0.0112,
+      "step": 73540
+    },
+    {
+      "epoch": 444.02,
+      "learning_rate": 3.132493499333339e-05,
+      "loss": 0.011,
+      "step": 73560
+    },
+    {
+      "epoch": 444.14,
+      "learning_rate": 3.1283310600394666e-05,
+      "loss": 0.0103,
+      "step": 73580
+    },
+    {
+      "epoch": 444.27,
+      "learning_rate": 3.1241701287154536e-05,
+      "loss": 0.0106,
+      "step": 73600
+    },
+    {
+      "epoch": 444.39,
+      "learning_rate": 3.120010708713684e-05,
+      "loss": 0.0107,
+      "step": 73620
+    },
+    {
+      "epoch": 444.51,
+      "learning_rate": 3.115852803385334e-05,
+      "loss": 0.0109,
+      "step": 73640
+    },
+    {
+      "epoch": 444.63,
+      "learning_rate": 3.111696416080344e-05,
+      "loss": 0.011,
+      "step": 73660
+    },
+    {
+      "epoch": 444.75,
+      "learning_rate": 3.107541550147448e-05,
+      "loss": 0.0111,
+      "step": 73680
+    },
+    {
+      "epoch": 444.87,
+      "learning_rate": 3.1033882089341415e-05,
+      "loss": 0.0111,
+      "step": 73700
+    },
+    {
+      "epoch": 444.99,
+      "learning_rate": 3.0992363957866975e-05,
+      "loss": 0.0112,
+      "step": 73720
+    },
+    {
+      "epoch": 445.11,
+      "learning_rate": 3.09508611405016e-05,
+      "loss": 0.0104,
+      "step": 73740
+    },
+    {
+      "epoch": 445.23,
+      "learning_rate": 3.090937367068331e-05,
+      "loss": 0.0105,
+      "step": 73760
+    },
+    {
+      "epoch": 445.35,
+      "learning_rate": 3.0867901581837854e-05,
+      "loss": 0.0107,
+      "step": 73780
+    },
+    {
+      "epoch": 445.47,
+      "learning_rate": 3.08264449073785e-05,
+      "loss": 0.0108,
+      "step": 73800
+    },
+    {
+      "epoch": 445.59,
+      "learning_rate": 3.078500368070621e-05,
+      "loss": 0.0109,
+      "step": 73820
+    },
+    {
+      "epoch": 445.71,
+      "learning_rate": 3.0743577935209353e-05,
+      "loss": 0.011,
+      "step": 73840
+    },
+    {
+      "epoch": 445.84,
+      "learning_rate": 3.0702167704263954e-05,
+      "loss": 0.0111,
+      "step": 73860
+    },
+    {
+      "epoch": 445.96,
+      "learning_rate": 3.066077302123348e-05,
+      "loss": 0.0111,
+      "step": 73880
+    },
+    {
+      "epoch": 446.08,
+      "learning_rate": 3.061939391946886e-05,
+      "loss": 0.0107,
+      "step": 73900
+    },
+    {
+      "epoch": 446.2,
+      "learning_rate": 3.0578030432308526e-05,
+      "loss": 0.0104,
+      "step": 73920
+    },
+    {
+      "epoch": 446.32,
+      "learning_rate": 3.0536682593078246e-05,
+      "loss": 0.0106,
+      "step": 73940
+    },
+    {
+      "epoch": 446.44,
+      "learning_rate": 3.0495350435091273e-05,
+      "loss": 0.0108,
+      "step": 73960
+    },
+    {
+      "epoch": 446.56,
+      "learning_rate": 3.0454033991648147e-05,
+      "loss": 0.0108,
+      "step": 73980
+    },
+    {
+      "epoch": 446.68,
+      "learning_rate": 3.0412733296036776e-05,
+      "loss": 0.011,
+      "step": 74000
+    },
+    {
+      "epoch": 446.68,
+      "eval_accuracy": 0.0012619652082859994,
+      "eval_loss": 23.18569564819336,
+      "eval_runtime": 18.1369,
+      "eval_samples_per_second": 112.478,
+      "eval_steps_per_second": 2.371,
+      "step": 74000
+    },
+    {
+      "epoch": 446.8,
+      "learning_rate": 3.0371448381532418e-05,
+      "loss": 0.0111,
+      "step": 74020
+    },
+    {
+      "epoch": 446.92,
+      "learning_rate": 3.0330179281397565e-05,
+      "loss": 0.0112,
+      "step": 74040
+    },
+    {
+      "epoch": 447.04,
+      "learning_rate": 3.0288926028881986e-05,
+      "loss": 0.011,
+      "step": 74060
+    },
+    {
+      "epoch": 447.16,
+      "learning_rate": 3.024768865722267e-05,
+      "loss": 0.0104,
+      "step": 74080
+    },
+    {
+      "epoch": 447.28,
+      "learning_rate": 3.0206467199643852e-05,
+      "loss": 0.0106,
+      "step": 74100
+    },
+    {
+      "epoch": 447.4,
+      "learning_rate": 3.0165261689356884e-05,
+      "loss": 0.0108,
+      "step": 74120
+    },
+    {
+      "epoch": 447.53,
+      "learning_rate": 3.01240721595603e-05,
+      "loss": 0.0109,
+      "step": 74140
+    },
+    {
+      "epoch": 447.65,
+      "learning_rate": 3.0082898643439804e-05,
+      "loss": 0.011,
+      "step": 74160
+    },
+    {
+      "epoch": 447.77,
+      "learning_rate": 3.0041741174168103e-05,
+      "loss": 0.011,
+      "step": 74180
+    },
+    {
+      "epoch": 447.89,
+      "learning_rate": 3.0000599784905065e-05,
+      "loss": 0.0113,
+      "step": 74200
+    },
+    {
+      "epoch": 448.01,
+      "learning_rate": 2.9959474508797535e-05,
+      "loss": 0.0112,
+      "step": 74220
+    },
+    {
+      "epoch": 448.13,
+      "learning_rate": 2.991836537897943e-05,
+      "loss": 0.0104,
+      "step": 74240
+    },
+    {
+      "epoch": 448.25,
+      "learning_rate": 2.9877272428571613e-05,
+      "loss": 0.0107,
+      "step": 74260
+    },
+    {
+      "epoch": 448.37,
+      "learning_rate": 2.9836195690681935e-05,
+      "loss": 0.0108,
+      "step": 74280
+    },
+    {
+      "epoch": 448.49,
+      "learning_rate": 2.97951351984052e-05,
+      "loss": 0.0109,
+      "step": 74300
+    },
+    {
+      "epoch": 448.61,
+      "learning_rate": 2.9754090984823073e-05,
+      "loss": 0.011,
+      "step": 74320
+    },
+    {
+      "epoch": 448.73,
+      "learning_rate": 2.9713063083004188e-05,
+      "loss": 0.0113,
+      "step": 74340
+    },
+    {
+      "epoch": 448.85,
+      "learning_rate": 2.9672051526003912e-05,
+      "loss": 0.0113,
+      "step": 74360
+    },
+    {
+      "epoch": 448.97,
+      "learning_rate": 2.9631056346864594e-05,
+      "loss": 0.0114,
+      "step": 74380
+    },
+    {
+      "epoch": 449.09,
+      "learning_rate": 2.9590077578615234e-05,
+      "loss": 0.0109,
+      "step": 74400
+    },
+    {
+      "epoch": 449.22,
+      "learning_rate": 2.954911525427174e-05,
+      "loss": 0.0112,
+      "step": 74420
+    },
+    {
+      "epoch": 449.34,
+      "learning_rate": 2.9508169406836696e-05,
+      "loss": 0.0116,
+      "step": 74440
+    },
+    {
+      "epoch": 449.46,
+      "learning_rate": 2.946724006929943e-05,
+      "loss": 0.0124,
+      "step": 74460
+    },
+    {
+      "epoch": 449.58,
+      "learning_rate": 2.942632727463598e-05,
+      "loss": 0.0133,
+      "step": 74480
+    },
+    {
+      "epoch": 449.7,
+      "learning_rate": 2.9385431055809032e-05,
+      "loss": 0.0143,
+      "step": 74500
+    },
+    {
+      "epoch": 449.82,
+      "learning_rate": 2.9344551445767954e-05,
+      "loss": 0.0169,
+      "step": 74520
+    },
+    {
+      "epoch": 449.94,
+      "learning_rate": 2.9303688477448667e-05,
+      "loss": 0.0311,
+      "step": 74540
+    },
+    {
+      "epoch": 450.06,
+      "learning_rate": 2.9262842183773788e-05,
+      "loss": 0.0887,
+      "step": 74560
+    },
+    {
+      "epoch": 450.18,
+      "learning_rate": 2.9222012597652372e-05,
+      "loss": 0.1177,
+      "step": 74580
+    },
+    {
+      "epoch": 450.3,
+      "learning_rate": 2.9181199751980114e-05,
+      "loss": 0.1211,
+      "step": 74600
+    },
+    {
+      "epoch": 450.42,
+      "learning_rate": 2.914040367963921e-05,
+      "loss": 0.1186,
+      "step": 74620
+    },
+    {
+      "epoch": 450.54,
+      "learning_rate": 2.909962441349825e-05,
+      "loss": 0.1154,
+      "step": 74640
+    },
+    {
+      "epoch": 450.66,
+      "learning_rate": 2.905886198641244e-05,
+      "loss": 0.1104,
+      "step": 74660
+    },
+    {
+      "epoch": 450.78,
+      "learning_rate": 2.9018116431223257e-05,
+      "loss": 0.1068,
+      "step": 74680
+    },
+    {
+      "epoch": 450.91,
+      "learning_rate": 2.897738778075869e-05,
+      "loss": 0.0997,
+      "step": 74700
+    },
+    {
+      "epoch": 451.03,
+      "learning_rate": 2.893667606783308e-05,
+      "loss": 0.0908,
+      "step": 74720
+    },
+    {
+      "epoch": 451.15,
+      "learning_rate": 2.88959813252471e-05,
+      "loss": 0.0712,
+      "step": 74740
+    },
+    {
+      "epoch": 451.27,
+      "learning_rate": 2.8855303585787813e-05,
+      "loss": 0.0733,
+      "step": 74760
+    },
+    {
+      "epoch": 451.39,
+      "learning_rate": 2.881464288222846e-05,
+      "loss": 0.0725,
+      "step": 74780
+    },
+    {
+      "epoch": 451.51,
+      "learning_rate": 2.877399924732873e-05,
+      "loss": 0.0716,
+      "step": 74800
+    },
+    {
+      "epoch": 451.63,
+      "learning_rate": 2.8733372713834396e-05,
+      "loss": 0.0697,
+      "step": 74820
+    },
+    {
+      "epoch": 451.75,
+      "learning_rate": 2.8692763314477566e-05,
+      "loss": 0.0685,
+      "step": 74840
+    },
+    {
+      "epoch": 451.87,
+      "learning_rate": 2.8652171081976435e-05,
+      "loss": 0.0668,
+      "step": 74860
+    },
+    {
+      "epoch": 451.99,
+      "learning_rate": 2.86115960490355e-05,
+      "loss": 0.0644,
+      "step": 74880
+    },
+    {
+      "epoch": 452.11,
+      "learning_rate": 2.857103824834533e-05,
+      "loss": 0.0438,
+      "step": 74900
+    },
+    {
+      "epoch": 452.23,
+      "learning_rate": 2.853049771258256e-05,
+      "loss": 0.0421,
+      "step": 74920
+    },
+    {
+      "epoch": 452.35,
+      "learning_rate": 2.8489974474410003e-05,
+      "loss": 0.0412,
+      "step": 74940
+    },
+    {
+      "epoch": 452.47,
+      "learning_rate": 2.844946856647649e-05,
+      "loss": 0.0404,
+      "step": 74960
+    },
+    {
+      "epoch": 452.6,
+      "learning_rate": 2.840898002141693e-05,
+      "loss": 0.0396,
+      "step": 74980
+    },
+    {
+      "epoch": 452.72,
+      "learning_rate": 2.836850887185214e-05,
+      "loss": 0.0386,
+      "step": 75000
+    },
+    {
+      "epoch": 452.72,
+      "eval_accuracy": 0.0012659165670594976,
+      "eval_loss": 22.926258087158203,
+      "eval_runtime": 18.1625,
+      "eval_samples_per_second": 112.319,
+      "eval_steps_per_second": 2.368,
+      "step": 75000
+    },
+    {
+      "epoch": 452.84,
+      "learning_rate": 2.8328055150389064e-05,
+      "loss": 0.0373,
+      "step": 75020
+    },
+    {
+      "epoch": 452.96,
+      "learning_rate": 2.8287618889620537e-05,
+      "loss": 0.037,
+      "step": 75040
+    },
+    {
+      "epoch": 453.08,
+      "learning_rate": 2.824720012212527e-05,
+      "loss": 0.0286,
+      "step": 75060
+    },
+    {
+      "epoch": 453.2,
+      "learning_rate": 2.8206798880467967e-05,
+      "loss": 0.0226,
+      "step": 75080
+    },
+    {
+      "epoch": 453.32,
+      "learning_rate": 2.816641519719917e-05,
+      "loss": 0.0212,
+      "step": 75100
+    },
+    {
+      "epoch": 453.44,
+      "learning_rate": 2.81260491048553e-05,
+      "loss": 0.0199,
+      "step": 75120
+    },
+    {
+      "epoch": 453.56,
+      "learning_rate": 2.808570063595857e-05,
+      "loss": 0.0194,
+      "step": 75140
+    },
+    {
+      "epoch": 453.68,
+      "learning_rate": 2.8045369823016988e-05,
+      "loss": 0.0186,
+      "step": 75160
+    },
+    {
+      "epoch": 453.8,
+      "learning_rate": 2.800505669852443e-05,
+      "loss": 0.018,
+      "step": 75180
+    },
+    {
+      "epoch": 453.92,
+      "learning_rate": 2.7964761294960396e-05,
+      "loss": 0.0177,
+      "step": 75200
+    },
+    {
+      "epoch": 454.04,
+      "learning_rate": 2.7924483644790184e-05,
+      "loss": 0.0158,
+      "step": 75220
+    },
+    {
+      "epoch": 454.16,
+      "learning_rate": 2.788422378046477e-05,
+      "loss": 0.0129,
+      "step": 75240
+    },
+    {
+      "epoch": 454.29,
+      "learning_rate": 2.784398173442081e-05,
+      "loss": 0.0127,
+      "step": 75260
+    },
+    {
+      "epoch": 454.41,
+      "learning_rate": 2.7803757539080565e-05,
+      "loss": 0.0125,
+      "step": 75280
+    },
+    {
+      "epoch": 454.53,
+      "learning_rate": 2.7763551226851943e-05,
+      "loss": 0.0126,
+      "step": 75300
+    },
+    {
+      "epoch": 454.65,
+      "learning_rate": 2.7723362830128496e-05,
+      "loss": 0.0124,
+      "step": 75320
+    },
+    {
+      "epoch": 454.77,
+      "learning_rate": 2.7683192381289226e-05,
+      "loss": 0.0124,
+      "step": 75340
+    },
+    {
+      "epoch": 454.89,
+      "learning_rate": 2.7643039912698786e-05,
+      "loss": 0.0125,
+      "step": 75360
+    },
+    {
+      "epoch": 455.01,
+      "learning_rate": 2.760290545670724e-05,
+      "loss": 0.0123,
+      "step": 75380
+    },
+    {
+      "epoch": 455.13,
+      "learning_rate": 2.7562789045650268e-05,
+      "loss": 0.0112,
+      "step": 75400
+    },
+    {
+      "epoch": 455.25,
+      "learning_rate": 2.7522690711848877e-05,
+      "loss": 0.0112,
+      "step": 75420
+    },
+    {
+      "epoch": 455.37,
+      "learning_rate": 2.7482610487609584e-05,
+      "loss": 0.0113,
+      "step": 75440
+    },
+    {
+      "epoch": 455.49,
+      "learning_rate": 2.744254840522431e-05,
+      "loss": 0.0113,
+      "step": 75460
+    },
+    {
+      "epoch": 455.61,
+      "learning_rate": 2.740250449697035e-05,
+      "loss": 0.0114,
+      "step": 75480
+    },
+    {
+      "epoch": 455.73,
+      "learning_rate": 2.736247879511038e-05,
+      "loss": 0.0115,
+      "step": 75500
+    },
+    {
+      "epoch": 455.86,
+      "learning_rate": 2.732247133189232e-05,
+      "loss": 0.0114,
+      "step": 75520
+    },
+    {
+      "epoch": 455.98,
+      "learning_rate": 2.7282482139549538e-05,
+      "loss": 0.0116,
+      "step": 75540
+    },
+    {
+      "epoch": 456.1,
+      "learning_rate": 2.7242511250300552e-05,
+      "loss": 0.0109,
+      "step": 75560
+    },
+    {
+      "epoch": 456.22,
+      "learning_rate": 2.7202558696349206e-05,
+      "loss": 0.0109,
+      "step": 75580
+    },
+    {
+      "epoch": 456.34,
+      "learning_rate": 2.7162624509884543e-05,
+      "loss": 0.0109,
+      "step": 75600
+    },
+    {
+      "epoch": 456.46,
+      "learning_rate": 2.7122708723080825e-05,
+      "loss": 0.011,
+      "step": 75620
+    },
+    {
+      "epoch": 456.58,
+      "learning_rate": 2.7082811368097487e-05,
+      "loss": 0.0113,
+      "step": 75640
+    },
+    {
+      "epoch": 456.7,
+      "learning_rate": 2.704293247707908e-05,
+      "loss": 0.0112,
+      "step": 75660
+    },
+    {
+      "epoch": 456.82,
+      "learning_rate": 2.7003072082155324e-05,
+      "loss": 0.0113,
+      "step": 75680
+    },
+    {
+      "epoch": 456.94,
+      "learning_rate": 2.6963230215441016e-05,
+      "loss": 0.0112,
+      "step": 75700
+    },
+    {
+      "epoch": 457.06,
+      "learning_rate": 2.692340690903604e-05,
+      "loss": 0.011,
+      "step": 75720
+    },
+    {
+      "epoch": 457.18,
+      "learning_rate": 2.688360219502528e-05,
+      "loss": 0.0106,
+      "step": 75740
+    },
+    {
+      "epoch": 457.3,
+      "learning_rate": 2.6843816105478715e-05,
+      "loss": 0.0108,
+      "step": 75760
+    },
+    {
+      "epoch": 457.42,
+      "learning_rate": 2.6804048672451275e-05,
+      "loss": 0.0108,
+      "step": 75780
+    },
+    {
+      "epoch": 457.55,
+      "learning_rate": 2.6764299927982833e-05,
+      "loss": 0.0111,
+      "step": 75800
+    },
+    {
+      "epoch": 457.67,
+      "learning_rate": 2.672456990409825e-05,
+      "loss": 0.0111,
+      "step": 75820
+    },
+    {
+      "epoch": 457.79,
+      "learning_rate": 2.6684858632807298e-05,
+      "loss": 0.0112,
+      "step": 75840
+    },
+    {
+      "epoch": 457.91,
+      "learning_rate": 2.664516614610464e-05,
+      "loss": 0.0112,
+      "step": 75860
+    },
+    {
+      "epoch": 458.03,
+      "learning_rate": 2.660549247596976e-05,
+      "loss": 0.011,
+      "step": 75880
+    },
+    {
+      "epoch": 458.15,
+      "learning_rate": 2.6565837654367016e-05,
+      "loss": 0.0104,
+      "step": 75900
+    },
+    {
+      "epoch": 458.27,
+      "learning_rate": 2.652620171324564e-05,
+      "loss": 0.0107,
+      "step": 75920
+    },
+    {
+      "epoch": 458.39,
+      "learning_rate": 2.648658468453954e-05,
+      "loss": 0.0108,
+      "step": 75940
+    },
+    {
+      "epoch": 458.51,
+      "learning_rate": 2.6446986600167456e-05,
+      "loss": 0.0109,
+      "step": 75960
+    },
+    {
+      "epoch": 458.63,
+      "learning_rate": 2.6407407492032843e-05,
+      "loss": 0.011,
+      "step": 75980
+    },
+    {
+      "epoch": 458.75,
+      "learning_rate": 2.6367847392023904e-05,
+      "loss": 0.0109,
+      "step": 76000
+    },
+    {
+      "epoch": 458.75,
+      "eval_accuracy": 0.0012575199296658139,
+      "eval_loss": 23.15482521057129,
+      "eval_runtime": 18.205,
+      "eval_samples_per_second": 112.057,
+      "eval_steps_per_second": 2.362,
+      "step": 76000
+    },
+    {
+      "epoch": 458.87,
+      "learning_rate": 2.632830633201346e-05,
+      "loss": 0.0111,
+      "step": 76020
+    },
+    {
+      "epoch": 458.99,
+      "learning_rate": 2.6288784343859013e-05,
+      "loss": 0.0112,
+      "step": 76040
+    },
+    {
+      "epoch": 459.11,
+      "learning_rate": 2.6249281459402787e-05,
+      "loss": 0.0104,
+      "step": 76060
+    },
+    {
+      "epoch": 459.24,
+      "learning_rate": 2.6209797710471483e-05,
+      "loss": 0.0106,
+      "step": 76080
+    },
+    {
+      "epoch": 459.36,
+      "learning_rate": 2.6170333128876455e-05,
+      "loss": 0.0106,
+      "step": 76100
+    },
+    {
+      "epoch": 459.48,
+      "learning_rate": 2.6130887746413614e-05,
+      "loss": 0.0108,
+      "step": 76120
+    },
+    {
+      "epoch": 459.6,
+      "learning_rate": 2.6091461594863408e-05,
+      "loss": 0.0108,
+      "step": 76140
+    },
+    {
+      "epoch": 459.72,
+      "learning_rate": 2.6052054705990735e-05,
+      "loss": 0.0109,
+      "step": 76160
+    },
+    {
+      "epoch": 459.84,
+      "learning_rate": 2.601266711154503e-05,
+      "loss": 0.011,
+      "step": 76180
+    },
+    {
+      "epoch": 459.96,
+      "learning_rate": 2.5973298843260173e-05,
+      "loss": 0.0111,
+      "step": 76200
+    },
+    {
+      "epoch": 460.08,
+      "learning_rate": 2.593394993285446e-05,
+      "loss": 0.0106,
+      "step": 76220
+    },
+    {
+      "epoch": 460.2,
+      "learning_rate": 2.5894620412030628e-05,
+      "loss": 0.0105,
+      "step": 76240
+    },
+    {
+      "epoch": 460.32,
+      "learning_rate": 2.585531031247569e-05,
+      "loss": 0.0106,
+      "step": 76260
+    },
+    {
+      "epoch": 460.44,
+      "learning_rate": 2.581601966586117e-05,
+      "loss": 0.0107,
+      "step": 76280
+    },
+    {
+      "epoch": 460.56,
+      "learning_rate": 2.5776748503842774e-05,
+      "loss": 0.0109,
+      "step": 76300
+    },
+    {
+      "epoch": 460.68,
+      "learning_rate": 2.5737496858060572e-05,
+      "loss": 0.0109,
+      "step": 76320
+    },
+    {
+      "epoch": 460.8,
+      "learning_rate": 2.569826476013893e-05,
+      "loss": 0.0109,
+      "step": 76340
+    },
+    {
+      "epoch": 460.93,
+      "learning_rate": 2.565905224168642e-05,
+      "loss": 0.0111,
+      "step": 76360
+    },
+    {
+      "epoch": 461.05,
+      "learning_rate": 2.5619859334295894e-05,
+      "loss": 0.0107,
+      "step": 76380
+    },
+    {
+      "epoch": 461.17,
+      "learning_rate": 2.5580686069544324e-05,
+      "loss": 0.0104,
+      "step": 76400
+    },
+    {
+      "epoch": 461.29,
+      "learning_rate": 2.554153247899292e-05,
+      "loss": 0.0105,
+      "step": 76420
+    },
+    {
+      "epoch": 461.41,
+      "learning_rate": 2.5502398594187025e-05,
+      "loss": 0.0107,
+      "step": 76440
+    },
+    {
+      "epoch": 461.53,
+      "learning_rate": 2.54632844466561e-05,
+      "loss": 0.0108,
+      "step": 76460
+    },
+    {
+      "epoch": 461.65,
+      "learning_rate": 2.542419006791371e-05,
+      "loss": 0.0107,
+      "step": 76480
+    },
+    {
+      "epoch": 461.77,
+      "learning_rate": 2.53851154894575e-05,
+      "loss": 0.0109,
+      "step": 76500
+    },
+    {
+      "epoch": 461.89,
+      "learning_rate": 2.534606074276915e-05,
+      "loss": 0.011,
+      "step": 76520
+    },
+    {
+      "epoch": 462.01,
+      "learning_rate": 2.5307025859314344e-05,
+      "loss": 0.011,
+      "step": 76540
+    },
+    {
+      "epoch": 462.13,
+      "learning_rate": 2.526801087054279e-05,
+      "loss": 0.0103,
+      "step": 76560
+    },
+    {
+      "epoch": 462.25,
+      "learning_rate": 2.5229015807888166e-05,
+      "loss": 0.0105,
+      "step": 76580
+    },
+    {
+      "epoch": 462.37,
+      "learning_rate": 2.5190040702768085e-05,
+      "loss": 0.0106,
+      "step": 76600
+    },
+    {
+      "epoch": 462.49,
+      "learning_rate": 2.5151085586584095e-05,
+      "loss": 0.0107,
+      "step": 76620
+    },
+    {
+      "epoch": 462.62,
+      "learning_rate": 2.511215049072162e-05,
+      "loss": 0.0107,
+      "step": 76640
+    },
+    {
+      "epoch": 462.74,
+      "learning_rate": 2.507323544654998e-05,
+      "loss": 0.0108,
+      "step": 76660
+    },
+    {
+      "epoch": 462.86,
+      "learning_rate": 2.503434048542229e-05,
+      "loss": 0.011,
+      "step": 76680
+    },
+    {
+      "epoch": 462.98,
+      "learning_rate": 2.4995465638675547e-05,
+      "loss": 0.011,
+      "step": 76700
+    },
+    {
+      "epoch": 463.1,
+      "learning_rate": 2.4956610937630492e-05,
+      "loss": 0.0104,
+      "step": 76720
+    },
+    {
+      "epoch": 463.22,
+      "learning_rate": 2.4917776413591675e-05,
+      "loss": 0.0103,
+      "step": 76740
+    },
+    {
+      "epoch": 463.34,
+      "learning_rate": 2.4878962097847375e-05,
+      "loss": 0.0107,
+      "step": 76760
+    },
+    {
+      "epoch": 463.46,
+      "learning_rate": 2.4840168021669534e-05,
+      "loss": 0.0108,
+      "step": 76780
+    },
+    {
+      "epoch": 463.58,
+      "learning_rate": 2.4801394216313922e-05,
+      "loss": 0.0107,
+      "step": 76800
+    },
+    {
+      "epoch": 463.7,
+      "learning_rate": 2.4762640713019824e-05,
+      "loss": 0.0107,
+      "step": 76820
+    },
+    {
+      "epoch": 463.82,
+      "learning_rate": 2.4723907543010266e-05,
+      "loss": 0.011,
+      "step": 76840
+    },
+    {
+      "epoch": 463.94,
+      "learning_rate": 2.4685194737491857e-05,
+      "loss": 0.011,
+      "step": 76860
+    },
+    {
+      "epoch": 464.06,
+      "learning_rate": 2.4646502327654825e-05,
+      "loss": 0.0106,
+      "step": 76880
+    },
+    {
+      "epoch": 464.19,
+      "learning_rate": 2.460783034467291e-05,
+      "loss": 0.0103,
+      "step": 76900
+    },
+    {
+      "epoch": 464.31,
+      "learning_rate": 2.4569178819703448e-05,
+      "loss": 0.0104,
+      "step": 76920
+    },
+    {
+      "epoch": 464.43,
+      "learning_rate": 2.453054778388728e-05,
+      "loss": 0.0106,
+      "step": 76940
+    },
+    {
+      "epoch": 464.55,
+      "learning_rate": 2.4491937268348724e-05,
+      "loss": 0.0107,
+      "step": 76960
+    },
+    {
+      "epoch": 464.67,
+      "learning_rate": 2.44533473041956e-05,
+      "loss": 0.0108,
+      "step": 76980
+    },
+    {
+      "epoch": 464.79,
+      "learning_rate": 2.441477792251909e-05,
+      "loss": 0.0109,
+      "step": 77000
+    },
+    {
+      "epoch": 464.79,
+      "eval_accuracy": 0.0012496172121188173,
+      "eval_loss": 23.27609634399414,
+      "eval_runtime": 18.2071,
+      "eval_samples_per_second": 112.044,
+      "eval_steps_per_second": 2.362,
+      "step": 77000
+    },
+    {
+      "epoch": 464.91,
+      "learning_rate": 2.437622915439392e-05,
+      "loss": 0.0109,
+      "step": 77020
+    },
+    {
+      "epoch": 465.03,
+      "learning_rate": 2.433770103087808e-05,
+      "loss": 0.0108,
+      "step": 77040
+    },
+    {
+      "epoch": 465.15,
+      "learning_rate": 2.4299193583013e-05,
+      "loss": 0.0103,
+      "step": 77060
+    },
+    {
+      "epoch": 465.27,
+      "learning_rate": 2.426070684182344e-05,
+      "loss": 0.0104,
+      "step": 77080
+    },
+    {
+      "epoch": 465.39,
+      "learning_rate": 2.4222240838317472e-05,
+      "loss": 0.0106,
+      "step": 77100
+    },
+    {
+      "epoch": 465.51,
+      "learning_rate": 2.4183795603486475e-05,
+      "loss": 0.0106,
+      "step": 77120
+    },
+    {
+      "epoch": 465.63,
+      "learning_rate": 2.4145371168305025e-05,
+      "loss": 0.0109,
+      "step": 77140
+    },
+    {
+      "epoch": 465.75,
+      "learning_rate": 2.4106967563731077e-05,
+      "loss": 0.0107,
+      "step": 77160
+    },
+    {
+      "epoch": 465.88,
+      "learning_rate": 2.4068584820705654e-05,
+      "loss": 0.0108,
+      "step": 77180
+    },
+    {
+      "epoch": 466.0,
+      "learning_rate": 2.4030222970153065e-05,
+      "loss": 0.0109,
+      "step": 77200
+    },
+    {
+      "epoch": 466.12,
+      "learning_rate": 2.3991882042980756e-05,
+      "loss": 0.0102,
+      "step": 77220
+    },
+    {
+      "epoch": 466.24,
+      "learning_rate": 2.395356207007933e-05,
+      "loss": 0.0103,
+      "step": 77240
+    },
+    {
+      "epoch": 466.36,
+      "learning_rate": 2.39152630823225e-05,
+      "loss": 0.0104,
+      "step": 77260
+    },
+    {
+      "epoch": 466.48,
+      "learning_rate": 2.387698511056704e-05,
+      "loss": 0.0106,
+      "step": 77280
+    },
+    {
+      "epoch": 466.6,
+      "learning_rate": 2.3838728185652843e-05,
+      "loss": 0.0107,
+      "step": 77300
+    },
+    {
+      "epoch": 466.72,
+      "learning_rate": 2.3800492338402813e-05,
+      "loss": 0.0108,
+      "step": 77320
+    },
+    {
+      "epoch": 466.84,
+      "learning_rate": 2.3762277599622885e-05,
+      "loss": 0.0109,
+      "step": 77340
+    },
+    {
+      "epoch": 466.96,
+      "learning_rate": 2.3724084000101988e-05,
+      "loss": 0.011,
+      "step": 77360
+    },
+    {
+      "epoch": 467.08,
+      "learning_rate": 2.3685911570612012e-05,
+      "loss": 0.0105,
+      "step": 77380
+    },
+    {
+      "epoch": 467.2,
+      "learning_rate": 2.364776034190781e-05,
+      "loss": 0.0103,
+      "step": 77400
+    },
+    {
+      "epoch": 467.32,
+      "learning_rate": 2.3609630344727106e-05,
+      "loss": 0.0105,
+      "step": 77420
+    },
+    {
+      "epoch": 467.44,
+      "learning_rate": 2.3571521609790566e-05,
+      "loss": 0.0106,
+      "step": 77440
+    },
+    {
+      "epoch": 467.57,
+      "learning_rate": 2.353343416780171e-05,
+      "loss": 0.0107,
+      "step": 77460
+    },
+    {
+      "epoch": 467.69,
+      "learning_rate": 2.3495368049446898e-05,
+      "loss": 0.0107,
+      "step": 77480
+    },
+    {
+      "epoch": 467.81,
+      "learning_rate": 2.3457323285395337e-05,
+      "loss": 0.0108,
+      "step": 77500
+    },
+    {
+      "epoch": 467.93,
+      "learning_rate": 2.3419299906298953e-05,
+      "loss": 0.0109,
+      "step": 77520
+    },
+    {
+      "epoch": 468.05,
+      "learning_rate": 2.3381297942792564e-05,
+      "loss": 0.0106,
+      "step": 77540
+    },
+    {
+      "epoch": 468.17,
+      "learning_rate": 2.334331742549361e-05,
+      "loss": 0.0103,
+      "step": 77560
+    },
+    {
+      "epoch": 468.29,
+      "learning_rate": 2.3305358385002333e-05,
+      "loss": 0.0103,
+      "step": 77580
+    },
+    {
+      "epoch": 468.41,
+      "learning_rate": 2.326742085190164e-05,
+      "loss": 0.0106,
+      "step": 77600
+    },
+    {
+      "epoch": 468.53,
+      "learning_rate": 2.322950485675712e-05,
+      "loss": 0.0107,
+      "step": 77620
+    },
+    {
+      "epoch": 468.65,
+      "learning_rate": 2.3191610430117023e-05,
+      "loss": 0.0108,
+      "step": 77640
+    },
+    {
+      "epoch": 468.77,
+      "learning_rate": 2.3153737602512143e-05,
+      "loss": 0.0108,
+      "step": 77660
+    },
+    {
+      "epoch": 468.89,
+      "learning_rate": 2.3115886404456015e-05,
+      "loss": 0.0108,
+      "step": 77680
+    },
+    {
+      "epoch": 469.01,
+      "learning_rate": 2.307805686644461e-05,
+      "loss": 0.0108,
+      "step": 77700
+    },
+    {
+      "epoch": 469.13,
+      "learning_rate": 2.3040249018956527e-05,
+      "loss": 0.0101,
+      "step": 77720
+    },
+    {
+      "epoch": 469.26,
+      "learning_rate": 2.300246289245283e-05,
+      "loss": 0.0104,
+      "step": 77740
+    },
+    {
+      "epoch": 469.38,
+      "learning_rate": 2.2964698517377165e-05,
+      "loss": 0.0104,
+      "step": 77760
+    },
+    {
+      "epoch": 469.5,
+      "learning_rate": 2.2926955924155608e-05,
+      "loss": 0.0106,
+      "step": 77780
+    },
+    {
+      "epoch": 469.62,
+      "learning_rate": 2.2889235143196642e-05,
+      "loss": 0.0107,
+      "step": 77800
+    },
+    {
+      "epoch": 469.74,
+      "learning_rate": 2.2851536204891244e-05,
+      "loss": 0.0108,
+      "step": 77820
+    },
+    {
+      "epoch": 469.86,
+      "learning_rate": 2.2813859139612764e-05,
+      "loss": 0.0109,
+      "step": 77840
+    },
+    {
+      "epoch": 469.98,
+      "learning_rate": 2.2776203977716953e-05,
+      "loss": 0.0109,
+      "step": 77860
+    },
+    {
+      "epoch": 470.1,
+      "learning_rate": 2.2738570749541844e-05,
+      "loss": 0.0103,
+      "step": 77880
+    },
+    {
+      "epoch": 470.22,
+      "learning_rate": 2.2700959485407897e-05,
+      "loss": 0.0104,
+      "step": 77900
+    },
+    {
+      "epoch": 470.34,
+      "learning_rate": 2.266337021561783e-05,
+      "loss": 0.0104,
+      "step": 77920
+    },
+    {
+      "epoch": 470.46,
+      "learning_rate": 2.26258029704566e-05,
+      "loss": 0.0106,
+      "step": 77940
+    },
+    {
+      "epoch": 470.58,
+      "learning_rate": 2.2588257780191484e-05,
+      "loss": 0.0106,
+      "step": 77960
+    },
+    {
+      "epoch": 470.7,
+      "learning_rate": 2.255073467507196e-05,
+      "loss": 0.0108,
+      "step": 77980
+    },
+    {
+      "epoch": 470.82,
+      "learning_rate": 2.2513233685329733e-05,
+      "loss": 0.0108,
+      "step": 78000
+    },
+    {
+      "epoch": 470.82,
+      "eval_accuracy": 0.0012604834487459375,
+      "eval_loss": 23.376338958740234,
+      "eval_runtime": 18.1777,
+      "eval_samples_per_second": 112.225,
+      "eval_steps_per_second": 2.366,
+      "step": 78000
+    },
+    {
+      "epoch": 470.95,
+      "learning_rate": 2.247575484117864e-05,
+      "loss": 0.011,
+      "step": 78020
+    },
+    {
+      "epoch": 471.07,
+      "learning_rate": 2.2438298172814714e-05,
+      "loss": 0.0105,
+      "step": 78040
+    },
+    {
+      "epoch": 471.19,
+      "learning_rate": 2.240086371041613e-05,
+      "loss": 0.0103,
+      "step": 78060
+    },
+    {
+      "epoch": 471.31,
+      "learning_rate": 2.2363451484143155e-05,
+      "loss": 0.0105,
+      "step": 78080
+    },
+    {
+      "epoch": 471.43,
+      "learning_rate": 2.2326061524138146e-05,
+      "loss": 0.0106,
+      "step": 78100
+    },
+    {
+      "epoch": 471.55,
+      "learning_rate": 2.228869386052551e-05,
+      "loss": 0.0107,
+      "step": 78120
+    },
+    {
+      "epoch": 471.67,
+      "learning_rate": 2.225134852341173e-05,
+      "loss": 0.0108,
+      "step": 78140
+    },
+    {
+      "epoch": 471.79,
+      "learning_rate": 2.2214025542885218e-05,
+      "loss": 0.0108,
+      "step": 78160
+    },
+    {
+      "epoch": 471.91,
+      "learning_rate": 2.217672494901646e-05,
+      "loss": 0.0109,
+      "step": 78180
+    },
+    {
+      "epoch": 472.03,
+      "learning_rate": 2.2139446771857858e-05,
+      "loss": 0.0108,
+      "step": 78200
+    },
+    {
+      "epoch": 472.15,
+      "learning_rate": 2.2102191041443766e-05,
+      "loss": 0.0102,
+      "step": 78220
+    },
+    {
+      "epoch": 472.27,
+      "learning_rate": 2.206495778779048e-05,
+      "loss": 0.0106,
+      "step": 78240
+    },
+    {
+      "epoch": 472.39,
+      "learning_rate": 2.2027747040896097e-05,
+      "loss": 0.0108,
+      "step": 78260
+    },
+    {
+      "epoch": 472.52,
+      "learning_rate": 2.199055883074072e-05,
+      "loss": 0.0109,
+      "step": 78280
+    },
+    {
+      "epoch": 472.64,
+      "learning_rate": 2.195339318728617e-05,
+      "loss": 0.0112,
+      "step": 78300
+    },
+    {
+      "epoch": 472.76,
+      "learning_rate": 2.191625014047614e-05,
+      "loss": 0.0111,
+      "step": 78320
+    },
+    {
+      "epoch": 472.88,
+      "learning_rate": 2.1879129720236118e-05,
+      "loss": 0.0114,
+      "step": 78340
+    },
+    {
+      "epoch": 473.0,
+      "learning_rate": 2.1842031956473365e-05,
+      "loss": 0.0114,
+      "step": 78360
+    },
+    {
+      "epoch": 473.12,
+      "learning_rate": 2.1804956879076876e-05,
+      "loss": 0.0113,
+      "step": 78380
+    },
+    {
+      "epoch": 473.24,
+      "learning_rate": 2.1767904517917336e-05,
+      "loss": 0.0119,
+      "step": 78400
+    },
+    {
+      "epoch": 473.36,
+      "learning_rate": 2.1730874902847232e-05,
+      "loss": 0.0134,
+      "step": 78420
+    },
+    {
+      "epoch": 473.48,
+      "learning_rate": 2.1693868063700596e-05,
+      "loss": 0.0163,
+      "step": 78440
+    },
+    {
+      "epoch": 473.6,
+      "learning_rate": 2.165873268986719e-05,
+      "loss": 0.0263,
+      "step": 78460
+    },
+    {
+      "epoch": 473.72,
+      "learning_rate": 2.162177034951215e-05,
+      "loss": 0.0626,
+      "step": 78480
+    },
+    {
+      "epoch": 473.84,
+      "learning_rate": 2.158667730333516e-05,
+      "loss": 0.0854,
+      "step": 78500
+    },
+    {
+      "epoch": 473.96,
+      "learning_rate": 2.1549759575009732e-05,
+      "loss": 0.0855,
+      "step": 78520
+    },
+    {
+      "epoch": 474.08,
+      "learning_rate": 2.1512864768529134e-05,
+      "loss": 0.0768,
+      "step": 78540
+    },
+    {
+      "epoch": 474.21,
+      "learning_rate": 2.1475992913618858e-05,
+      "loss": 0.0756,
+      "step": 78560
+    },
+    {
+      "epoch": 474.33,
+      "learning_rate": 2.1439144039985898e-05,
+      "loss": 0.07,
+      "step": 78580
+    },
+    {
+      "epoch": 474.45,
+      "learning_rate": 2.1402318177318703e-05,
+      "loss": 0.0656,
+      "step": 78600
+    },
+    {
+      "epoch": 474.57,
+      "learning_rate": 2.1365515355287192e-05,
+      "loss": 0.0627,
+      "step": 78620
+    },
+    {
+      "epoch": 474.69,
+      "learning_rate": 2.13287356035428e-05,
+      "loss": 0.0611,
+      "step": 78640
+    },
+    {
+      "epoch": 474.81,
+      "learning_rate": 2.1291978951718246e-05,
+      "loss": 0.0562,
+      "step": 78660
+    },
+    {
+      "epoch": 474.93,
+      "learning_rate": 2.1255245429427723e-05,
+      "loss": 0.0539,
+      "step": 78680
+    },
+    {
+      "epoch": 475.05,
+      "learning_rate": 2.1218535066266758e-05,
+      "loss": 0.0452,
+      "step": 78700
+    },
+    {
+      "epoch": 475.17,
+      "learning_rate": 2.1181847891812262e-05,
+      "loss": 0.0338,
+      "step": 78720
+    },
+    {
+      "epoch": 475.29,
+      "learning_rate": 2.1145183935622377e-05,
+      "loss": 0.0311,
+      "step": 78740
+    },
+    {
+      "epoch": 475.41,
+      "learning_rate": 2.1108543227236616e-05,
+      "loss": 0.0285,
+      "step": 78760
+    },
+    {
+      "epoch": 475.53,
+      "learning_rate": 2.107192579617574e-05,
+      "loss": 0.0265,
+      "step": 78780
+    },
+    {
+      "epoch": 475.65,
+      "learning_rate": 2.1035331671941737e-05,
+      "loss": 0.0254,
+      "step": 78800
+    },
+    {
+      "epoch": 475.77,
+      "learning_rate": 2.0998760884017865e-05,
+      "loss": 0.0237,
+      "step": 78820
+    },
+    {
+      "epoch": 475.9,
+      "learning_rate": 2.0962213461868485e-05,
+      "loss": 0.0228,
+      "step": 78840
+    },
+    {
+      "epoch": 476.02,
+      "learning_rate": 2.092568943493927e-05,
+      "loss": 0.0212,
+      "step": 78860
+    },
+    {
+      "epoch": 476.14,
+      "learning_rate": 2.0889188832656918e-05,
+      "loss": 0.015,
+      "step": 78880
+    },
+    {
+      "epoch": 476.26,
+      "learning_rate": 2.0852711684429314e-05,
+      "loss": 0.0141,
+      "step": 78900
+    },
+    {
+      "epoch": 476.38,
+      "learning_rate": 2.0816258019645446e-05,
+      "loss": 0.0139,
+      "step": 78920
+    },
+    {
+      "epoch": 476.5,
+      "learning_rate": 2.077982786767537e-05,
+      "loss": 0.0133,
+      "step": 78940
+    },
+    {
+      "epoch": 476.62,
+      "learning_rate": 2.0743421257870215e-05,
+      "loss": 0.0133,
+      "step": 78960
+    },
+    {
+      "epoch": 476.74,
+      "learning_rate": 2.07070382195621e-05,
+      "loss": 0.013,
+      "step": 78980
+    },
+    {
+      "epoch": 476.86,
+      "learning_rate": 2.06706787820642e-05,
+      "loss": 0.0131,
+      "step": 79000
+    },
+    {
+      "epoch": 476.86,
+      "eval_accuracy": 0.0012590016892058756,
+      "eval_loss": 23.20476722717285,
+      "eval_runtime": 18.1597,
+      "eval_samples_per_second": 112.336,
+      "eval_steps_per_second": 2.368,
+      "step": 79000
+    },
+    {
+      "epoch": 476.98,
+      "learning_rate": 2.063434297467065e-05,
+      "loss": 0.0131,
+      "step": 79020
+    },
+    {
+      "epoch": 477.1,
+      "learning_rate": 2.0598030826656577e-05,
+      "loss": 0.0114,
+      "step": 79040
+    },
+    {
+      "epoch": 477.22,
+      "learning_rate": 2.056174236727797e-05,
+      "loss": 0.0113,
+      "step": 79060
+    },
+    {
+      "epoch": 477.34,
+      "learning_rate": 2.052547762577184e-05,
+      "loss": 0.0111,
+      "step": 79080
+    },
+    {
+      "epoch": 477.46,
+      "learning_rate": 2.0489236631356023e-05,
+      "loss": 0.0112,
+      "step": 79100
+    },
+    {
+      "epoch": 477.59,
+      "learning_rate": 2.045301941322921e-05,
+      "loss": 0.0113,
+      "step": 79120
+    },
+    {
+      "epoch": 477.71,
+      "learning_rate": 2.041682600057097e-05,
+      "loss": 0.0114,
+      "step": 79140
+    },
+    {
+      "epoch": 477.83,
+      "learning_rate": 2.0380656422541687e-05,
+      "loss": 0.0113,
+      "step": 79160
+    },
+    {
+      "epoch": 477.95,
+      "learning_rate": 2.0344510708282556e-05,
+      "loss": 0.0115,
+      "step": 79180
+    },
+    {
+      "epoch": 478.07,
+      "learning_rate": 2.0308388886915464e-05,
+      "loss": 0.011,
+      "step": 79200
+    },
+    {
+      "epoch": 478.19,
+      "learning_rate": 2.027229098754317e-05,
+      "loss": 0.0106,
+      "step": 79220
+    },
+    {
+      "epoch": 478.31,
+      "learning_rate": 2.02362170392491e-05,
+      "loss": 0.0108,
+      "step": 79240
+    },
+    {
+      "epoch": 478.43,
+      "learning_rate": 2.0200167071097343e-05,
+      "loss": 0.0108,
+      "step": 79260
+    },
+    {
+      "epoch": 478.55,
+      "learning_rate": 2.016414111213273e-05,
+      "loss": 0.0111,
+      "step": 79280
+    },
+    {
+      "epoch": 478.67,
+      "learning_rate": 2.0128139191380723e-05,
+      "loss": 0.011,
+      "step": 79300
+    },
+    {
+      "epoch": 478.79,
+      "learning_rate": 2.0092161337847433e-05,
+      "loss": 0.011,
+      "step": 79320
+    },
+    {
+      "epoch": 478.91,
+      "learning_rate": 2.0056207580519538e-05,
+      "loss": 0.0111,
+      "step": 79340
+    },
+    {
+      "epoch": 479.03,
+      "learning_rate": 2.0020277948364325e-05,
+      "loss": 0.0109,
+      "step": 79360
+    },
+    {
+      "epoch": 479.15,
+      "learning_rate": 1.998437247032971e-05,
+      "loss": 0.0105,
+      "step": 79380
+    },
+    {
+      "epoch": 479.28,
+      "learning_rate": 1.9948491175344025e-05,
+      "loss": 0.0106,
+      "step": 79400
+    },
+    {
+      "epoch": 479.4,
+      "learning_rate": 1.9912634092316206e-05,
+      "loss": 0.0107,
+      "step": 79420
+    },
+    {
+      "epoch": 479.52,
+      "learning_rate": 1.9876801250135662e-05,
+      "loss": 0.0108,
+      "step": 79440
+    },
+    {
+      "epoch": 479.64,
+      "learning_rate": 1.984099267767227e-05,
+      "loss": 0.0108,
+      "step": 79460
+    },
+    {
+      "epoch": 479.76,
+      "learning_rate": 1.9805208403776333e-05,
+      "loss": 0.0109,
+      "step": 79480
+    },
+    {
+      "epoch": 479.88,
+      "learning_rate": 1.976944845727859e-05,
+      "loss": 0.0109,
+      "step": 79500
+    },
+    {
+      "epoch": 480.0,
+      "learning_rate": 1.97337128669902e-05,
+      "loss": 0.0111,
+      "step": 79520
+    },
+    {
+      "epoch": 480.12,
+      "learning_rate": 1.969800166170268e-05,
+      "loss": 0.0103,
+      "step": 79540
+    },
+    {
+      "epoch": 480.24,
+      "learning_rate": 1.966231487018792e-05,
+      "loss": 0.0105,
+      "step": 79560
+    },
+    {
+      "epoch": 480.36,
+      "learning_rate": 1.962665252119807e-05,
+      "loss": 0.0106,
+      "step": 79580
+    },
+    {
+      "epoch": 480.48,
+      "learning_rate": 1.9591014643465706e-05,
+      "loss": 0.0106,
+      "step": 79600
+    },
+    {
+      "epoch": 480.6,
+      "learning_rate": 1.955540126570358e-05,
+      "loss": 0.0107,
+      "step": 79620
+    },
+    {
+      "epoch": 480.72,
+      "learning_rate": 1.9519812416604748e-05,
+      "loss": 0.0108,
+      "step": 79640
+    },
+    {
+      "epoch": 480.85,
+      "learning_rate": 1.9484248124842526e-05,
+      "loss": 0.0108,
+      "step": 79660
+    },
+    {
+      "epoch": 480.97,
+      "learning_rate": 1.94487084190704e-05,
+      "loss": 0.0108,
+      "step": 79680
+    },
+    {
+      "epoch": 481.09,
+      "learning_rate": 1.9413193327922104e-05,
+      "loss": 0.0105,
+      "step": 79700
+    },
+    {
+      "epoch": 481.21,
+      "learning_rate": 1.9377702880011434e-05,
+      "loss": 0.0103,
+      "step": 79720
+    },
+    {
+      "epoch": 481.33,
+      "learning_rate": 1.934223710393249e-05,
+      "loss": 0.0105,
+      "step": 79740
+    },
+    {
+      "epoch": 481.45,
+      "learning_rate": 1.930679602825935e-05,
+      "loss": 0.0106,
+      "step": 79760
+    },
+    {
+      "epoch": 481.57,
+      "learning_rate": 1.927137968154626e-05,
+      "loss": 0.0107,
+      "step": 79780
+    },
+    {
+      "epoch": 481.69,
+      "learning_rate": 1.923598809232754e-05,
+      "loss": 0.0106,
+      "step": 79800
+    },
+    {
+      "epoch": 481.81,
+      "learning_rate": 1.9200621289117544e-05,
+      "loss": 0.0107,
+      "step": 79820
+    },
+    {
+      "epoch": 481.93,
+      "learning_rate": 1.9165279300410694e-05,
+      "loss": 0.0108,
+      "step": 79840
+    },
+    {
+      "epoch": 482.05,
+      "learning_rate": 1.9129962154681346e-05,
+      "loss": 0.0106,
+      "step": 79860
+    },
+    {
+      "epoch": 482.17,
+      "learning_rate": 1.9094669880383915e-05,
+      "loss": 0.0103,
+      "step": 79880
+    },
+    {
+      "epoch": 482.29,
+      "learning_rate": 1.9059402505952744e-05,
+      "loss": 0.0103,
+      "step": 79900
+    },
+    {
+      "epoch": 482.41,
+      "learning_rate": 1.9024160059802122e-05,
+      "loss": 0.0105,
+      "step": 79920
+    },
+    {
+      "epoch": 482.54,
+      "learning_rate": 1.8988942570326245e-05,
+      "loss": 0.0106,
+      "step": 79940
+    },
+    {
+      "epoch": 482.66,
+      "learning_rate": 1.8953750065899212e-05,
+      "loss": 0.0106,
+      "step": 79960
+    },
+    {
+      "epoch": 482.78,
+      "learning_rate": 1.8918582574875003e-05,
+      "loss": 0.0107,
+      "step": 79980
+    },
+    {
+      "epoch": 482.9,
+      "learning_rate": 1.8883440125587402e-05,
+      "loss": 0.0108,
+      "step": 80000
+    },
+    {
+      "epoch": 482.9,
+      "eval_accuracy": 0.001258013849512501,
+      "eval_loss": 23.377166748046875,
+      "eval_runtime": 18.2373,
+      "eval_samples_per_second": 111.859,
+      "eval_steps_per_second": 2.358,
+      "step": 80000
+    },
+    {
+      "epoch": 483.02,
+      "learning_rate": 1.884832274635006e-05,
+      "loss": 0.0107,
+      "step": 80020
+    },
+    {
+      "epoch": 483.14,
+      "learning_rate": 1.8813230465456416e-05,
+      "loss": 0.0102,
+      "step": 80040
+    },
+    {
+      "epoch": 483.26,
+      "learning_rate": 1.877816331117968e-05,
+      "loss": 0.0103,
+      "step": 80060
+    },
+    {
+      "epoch": 483.38,
+      "learning_rate": 1.8743121311772856e-05,
+      "loss": 0.0104,
+      "step": 80080
+    },
+    {
+      "epoch": 483.5,
+      "learning_rate": 1.870810449546858e-05,
+      "loss": 0.0106,
+      "step": 80100
+    },
+    {
+      "epoch": 483.62,
+      "learning_rate": 1.867311289047935e-05,
+      "loss": 0.0106,
+      "step": 80120
+    },
+    {
+      "epoch": 483.74,
+      "learning_rate": 1.8638146524997217e-05,
+      "loss": 0.0107,
+      "step": 80140
+    },
+    {
+      "epoch": 483.86,
+      "learning_rate": 1.8603205427193964e-05,
+      "loss": 0.0106,
+      "step": 80160
+    },
+    {
+      "epoch": 483.98,
+      "learning_rate": 1.856828962522101e-05,
+      "loss": 0.0107,
+      "step": 80180
+    },
+    {
+      "epoch": 484.1,
+      "learning_rate": 1.8533399147209375e-05,
+      "loss": 0.0101,
+      "step": 80200
+    },
+    {
+      "epoch": 484.23,
+      "learning_rate": 1.8498534021269708e-05,
+      "loss": 0.0103,
+      "step": 80220
+    },
+    {
+      "epoch": 484.35,
+      "learning_rate": 1.846369427549216e-05,
+      "loss": 0.0103,
+      "step": 80240
+    },
+    {
+      "epoch": 484.47,
+      "learning_rate": 1.8428879937946557e-05,
+      "loss": 0.0105,
+      "step": 80260
+    },
+    {
+      "epoch": 484.59,
+      "learning_rate": 1.8394091036682133e-05,
+      "loss": 0.0105,
+      "step": 80280
+    },
+    {
+      "epoch": 484.71,
+      "learning_rate": 1.8359327599727698e-05,
+      "loss": 0.0106,
+      "step": 80300
+    },
+    {
+      "epoch": 484.83,
+      "learning_rate": 1.832458965509149e-05,
+      "loss": 0.0107,
+      "step": 80320
+    },
+    {
+      "epoch": 484.95,
+      "learning_rate": 1.82898772307613e-05,
+      "loss": 0.0107,
+      "step": 80340
+    },
+    {
+      "epoch": 485.07,
+      "learning_rate": 1.8255190354704256e-05,
+      "loss": 0.0104,
+      "step": 80360
+    },
+    {
+      "epoch": 485.19,
+      "learning_rate": 1.8220529054866958e-05,
+      "loss": 0.0102,
+      "step": 80380
+    },
+    {
+      "epoch": 485.31,
+      "learning_rate": 1.81858933591754e-05,
+      "loss": 0.0103,
+      "step": 80400
+    },
+    {
+      "epoch": 485.43,
+      "learning_rate": 1.8151283295534932e-05,
+      "loss": 0.0104,
+      "step": 80420
+    },
+    {
+      "epoch": 485.55,
+      "learning_rate": 1.811669889183027e-05,
+      "loss": 0.0104,
+      "step": 80440
+    },
+    {
+      "epoch": 485.67,
+      "learning_rate": 1.8082140175925393e-05,
+      "loss": 0.0105,
+      "step": 80460
+    },
+    {
+      "epoch": 485.79,
+      "learning_rate": 1.8047607175663712e-05,
+      "loss": 0.0107,
+      "step": 80480
+    },
+    {
+      "epoch": 485.92,
+      "learning_rate": 1.8013099918867778e-05,
+      "loss": 0.0106,
+      "step": 80500
+    },
+    {
+      "epoch": 486.04,
+      "learning_rate": 1.7978618433339484e-05,
+      "loss": 0.0106,
+      "step": 80520
+    },
+    {
+      "epoch": 486.16,
+      "learning_rate": 1.794416274685995e-05,
+      "loss": 0.0101,
+      "step": 80540
+    },
+    {
+      "epoch": 486.28,
+      "learning_rate": 1.7909732887189485e-05,
+      "loss": 0.0102,
+      "step": 80560
+    },
+    {
+      "epoch": 486.4,
+      "learning_rate": 1.7875328882067634e-05,
+      "loss": 0.0104,
+      "step": 80580
+    },
+    {
+      "epoch": 486.52,
+      "learning_rate": 1.7840950759213043e-05,
+      "loss": 0.0105,
+      "step": 80600
+    },
+    {
+      "epoch": 486.64,
+      "learning_rate": 1.7806598546323567e-05,
+      "loss": 0.0105,
+      "step": 80620
+    },
+    {
+      "epoch": 486.76,
+      "learning_rate": 1.7772272271076157e-05,
+      "loss": 0.0106,
+      "step": 80640
+    },
+    {
+      "epoch": 486.88,
+      "learning_rate": 1.7737971961126875e-05,
+      "loss": 0.0106,
+      "step": 80660
+    },
+    {
+      "epoch": 487.0,
+      "learning_rate": 1.7703697644110863e-05,
+      "loss": 0.0107,
+      "step": 80680
+    },
+    {
+      "epoch": 487.12,
+      "learning_rate": 1.7669449347642313e-05,
+      "loss": 0.01,
+      "step": 80700
+    },
+    {
+      "epoch": 487.24,
+      "learning_rate": 1.763522709931449e-05,
+      "loss": 0.0102,
+      "step": 80720
+    },
+    {
+      "epoch": 487.36,
+      "learning_rate": 1.760103092669959e-05,
+      "loss": 0.0103,
+      "step": 80740
+    },
+    {
+      "epoch": 487.48,
+      "learning_rate": 1.7566860857348884e-05,
+      "loss": 0.0104,
+      "step": 80760
+    },
+    {
+      "epoch": 487.61,
+      "learning_rate": 1.7532716918792565e-05,
+      "loss": 0.0105,
+      "step": 80780
+    },
+    {
+      "epoch": 487.73,
+      "learning_rate": 1.7498599138539797e-05,
+      "loss": 0.0105,
+      "step": 80800
+    },
+    {
+      "epoch": 487.85,
+      "learning_rate": 1.7464507544078672e-05,
+      "loss": 0.0107,
+      "step": 80820
+    },
+    {
+      "epoch": 487.97,
+      "learning_rate": 1.743044216287613e-05,
+      "loss": 0.0107,
+      "step": 80840
+    },
+    {
+      "epoch": 488.09,
+      "learning_rate": 1.7396403022378095e-05,
+      "loss": 0.0101,
+      "step": 80860
+    },
+    {
+      "epoch": 488.21,
+      "learning_rate": 1.7362390150009228e-05,
+      "loss": 0.0102,
+      "step": 80880
+    },
+    {
+      "epoch": 488.33,
+      "learning_rate": 1.7328403573173118e-05,
+      "loss": 0.0102,
+      "step": 80900
+    },
+    {
+      "epoch": 488.45,
+      "learning_rate": 1.7294443319252125e-05,
+      "loss": 0.0103,
+      "step": 80920
+    },
+    {
+      "epoch": 488.57,
+      "learning_rate": 1.726050941560742e-05,
+      "loss": 0.0105,
+      "step": 80940
+    },
+    {
+      "epoch": 488.69,
+      "learning_rate": 1.7226601889578946e-05,
+      "loss": 0.0105,
+      "step": 80960
+    },
+    {
+      "epoch": 488.81,
+      "learning_rate": 1.7192720768485336e-05,
+      "loss": 0.0106,
+      "step": 80980
+    },
+    {
+      "epoch": 488.93,
+      "learning_rate": 1.7158866079624075e-05,
+      "loss": 0.0106,
+      "step": 81000
+    },
+    {
+      "epoch": 488.93,
+      "eval_accuracy": 0.0012565320899724393,
+      "eval_loss": 23.473268508911133,
+      "eval_runtime": 18.1503,
+      "eval_samples_per_second": 112.395,
+      "eval_steps_per_second": 2.369,
+      "step": 81000
+    },
+    {
+      "epoch": 489.05,
+      "learning_rate": 1.712503785027121e-05,
+      "loss": 0.0104,
+      "step": 81020
+    },
+    {
+      "epoch": 489.18,
+      "learning_rate": 1.7091236107681552e-05,
+      "loss": 0.0101,
+      "step": 81040
+    },
+    {
+      "epoch": 489.3,
+      "learning_rate": 1.705746087908856e-05,
+      "loss": 0.0102,
+      "step": 81060
+    },
+    {
+      "epoch": 489.42,
+      "learning_rate": 1.702371219170433e-05,
+      "loss": 0.0104,
+      "step": 81080
+    },
+    {
+      "epoch": 489.54,
+      "learning_rate": 1.698999007271957e-05,
+      "loss": 0.0104,
+      "step": 81100
+    },
+    {
+      "epoch": 489.66,
+      "learning_rate": 1.6956294549303562e-05,
+      "loss": 0.0106,
+      "step": 81120
+    },
+    {
+      "epoch": 489.78,
+      "learning_rate": 1.69226256486042e-05,
+      "loss": 0.0105,
+      "step": 81140
+    },
+    {
+      "epoch": 489.9,
+      "learning_rate": 1.6888983397747902e-05,
+      "loss": 0.0106,
+      "step": 81160
+    },
+    {
+      "epoch": 490.02,
+      "learning_rate": 1.685536782383964e-05,
+      "loss": 0.0105,
+      "step": 81180
+    },
+    {
+      "epoch": 490.14,
+      "learning_rate": 1.6821778953962826e-05,
+      "loss": 0.01,
+      "step": 81200
+    },
+    {
+      "epoch": 490.26,
+      "learning_rate": 1.6788216815179453e-05,
+      "loss": 0.0102,
+      "step": 81220
+    },
+    {
+      "epoch": 490.38,
+      "learning_rate": 1.6754681434529934e-05,
+      "loss": 0.0102,
+      "step": 81240
+    },
+    {
+      "epoch": 490.5,
+      "learning_rate": 1.6721172839033082e-05,
+      "loss": 0.0103,
+      "step": 81260
+    },
+    {
+      "epoch": 490.62,
+      "learning_rate": 1.6687691055686187e-05,
+      "loss": 0.0105,
+      "step": 81280
+    },
+    {
+      "epoch": 490.74,
+      "learning_rate": 1.6654236111464917e-05,
+      "loss": 0.0105,
+      "step": 81300
+    },
+    {
+      "epoch": 490.87,
+      "learning_rate": 1.662080803332333e-05,
+      "loss": 0.0107,
+      "step": 81320
+    },
+    {
+      "epoch": 490.99,
+      "learning_rate": 1.6587406848193792e-05,
+      "loss": 0.0106,
+      "step": 81340
+    },
+    {
+      "epoch": 491.11,
+      "learning_rate": 1.655403258298704e-05,
+      "loss": 0.0101,
+      "step": 81360
+    },
+    {
+      "epoch": 491.23,
+      "learning_rate": 1.6520685264592162e-05,
+      "loss": 0.01,
+      "step": 81380
+    },
+    {
+      "epoch": 491.35,
+      "learning_rate": 1.6487364919876435e-05,
+      "loss": 0.0103,
+      "step": 81400
+    },
+    {
+      "epoch": 491.47,
+      "learning_rate": 1.6454071575685488e-05,
+      "loss": 0.0104,
+      "step": 81420
+    },
+    {
+      "epoch": 491.59,
+      "learning_rate": 1.642080525884316e-05,
+      "loss": 0.0104,
+      "step": 81440
+    },
+    {
+      "epoch": 491.71,
+      "learning_rate": 1.6387565996151532e-05,
+      "loss": 0.0105,
+      "step": 81460
+    },
+    {
+      "epoch": 491.83,
+      "learning_rate": 1.6354353814390854e-05,
+      "loss": 0.0105,
+      "step": 81480
+    },
+    {
+      "epoch": 491.95,
+      "learning_rate": 1.632116874031958e-05,
+      "loss": 0.0107,
+      "step": 81500
+    },
+    {
+      "epoch": 492.07,
+      "learning_rate": 1.6288010800674368e-05,
+      "loss": 0.0103,
+      "step": 81520
+    },
+    {
+      "epoch": 492.19,
+      "learning_rate": 1.6254880022169933e-05,
+      "loss": 0.0101,
+      "step": 81540
+    },
+    {
+      "epoch": 492.31,
+      "learning_rate": 1.6221776431499148e-05,
+      "loss": 0.0101,
+      "step": 81560
+    },
+    {
+      "epoch": 492.43,
+      "learning_rate": 1.6188700055332983e-05,
+      "loss": 0.0103,
+      "step": 81580
+    },
+    {
+      "epoch": 492.56,
+      "learning_rate": 1.6155650920320496e-05,
+      "loss": 0.0104,
+      "step": 81600
+    },
+    {
+      "epoch": 492.68,
+      "learning_rate": 1.6122629053088752e-05,
+      "loss": 0.0105,
+      "step": 81620
+    },
+    {
+      "epoch": 492.8,
+      "learning_rate": 1.6089634480242872e-05,
+      "loss": 0.0106,
+      "step": 81640
+    },
+    {
+      "epoch": 492.92,
+      "learning_rate": 1.6056667228365996e-05,
+      "loss": 0.0106,
+      "step": 81660
+    },
+    {
+      "epoch": 493.04,
+      "learning_rate": 1.602372732401925e-05,
+      "loss": 0.0104,
+      "step": 81680
+    },
+    {
+      "epoch": 493.16,
+      "learning_rate": 1.5990814793741725e-05,
+      "loss": 0.01,
+      "step": 81700
+    },
+    {
+      "epoch": 493.28,
+      "learning_rate": 1.595792966405041e-05,
+      "loss": 0.0101,
+      "step": 81720
+    },
+    {
+      "epoch": 493.4,
+      "learning_rate": 1.5925071961440324e-05,
+      "loss": 0.0103,
+      "step": 81740
+    },
+    {
+      "epoch": 493.52,
+      "learning_rate": 1.5892241712384276e-05,
+      "loss": 0.0105,
+      "step": 81760
+    },
+    {
+      "epoch": 493.64,
+      "learning_rate": 1.585943894333302e-05,
+      "loss": 0.0104,
+      "step": 81780
+    },
+    {
+      "epoch": 493.76,
+      "learning_rate": 1.5826663680715152e-05,
+      "loss": 0.0105,
+      "step": 81800
+    },
+    {
+      "epoch": 493.88,
+      "learning_rate": 1.5793915950937116e-05,
+      "loss": 0.0106,
+      "step": 81820
+    },
+    {
+      "epoch": 494.0,
+      "learning_rate": 1.5761195780383174e-05,
+      "loss": 0.0107,
+      "step": 81840
+    },
+    {
+      "epoch": 494.12,
+      "learning_rate": 1.5728503195415355e-05,
+      "loss": 0.0101,
+      "step": 81860
+    },
+    {
+      "epoch": 494.25,
+      "learning_rate": 1.5695838222373495e-05,
+      "loss": 0.0101,
+      "step": 81880
+    },
+    {
+      "epoch": 494.37,
+      "learning_rate": 1.5663200887575162e-05,
+      "loss": 0.0102,
+      "step": 81900
+    },
+    {
+      "epoch": 494.49,
+      "learning_rate": 1.56305912173157e-05,
+      "loss": 0.0104,
+      "step": 81920
+    },
+    {
+      "epoch": 494.61,
+      "learning_rate": 1.559800923786807e-05,
+      "loss": 0.0105,
+      "step": 81940
+    },
+    {
+      "epoch": 494.73,
+      "learning_rate": 1.5565454975483047e-05,
+      "loss": 0.0105,
+      "step": 81960
+    },
+    {
+      "epoch": 494.85,
+      "learning_rate": 1.5532928456389e-05,
+      "loss": 0.0105,
+      "step": 81980
+    },
+    {
+      "epoch": 494.97,
+      "learning_rate": 1.550042970679193e-05,
+      "loss": 0.0106,
+      "step": 82000
+    },
+    {
+      "epoch": 494.97,
+      "eval_accuracy": 0.0012590016892058756,
+      "eval_loss": 23.56543731689453,
+      "eval_runtime": 18.1587,
+      "eval_samples_per_second": 112.343,
+      "eval_steps_per_second": 2.368,
+      "step": 82000
+    },
+    {
+      "epoch": 495.09,
+      "learning_rate": 1.5467958752875505e-05,
+      "loss": 0.0101,
+      "step": 82020
+    },
+    {
+      "epoch": 495.21,
+      "learning_rate": 1.543551562080099e-05,
+      "loss": 0.01,
+      "step": 82040
+    },
+    {
+      "epoch": 495.33,
+      "learning_rate": 1.540310033670725e-05,
+      "loss": 0.0102,
+      "step": 82060
+    },
+    {
+      "epoch": 495.45,
+      "learning_rate": 1.537071292671064e-05,
+      "loss": 0.0103,
+      "step": 82080
+    },
+    {
+      "epoch": 495.57,
+      "learning_rate": 1.533835341690516e-05,
+      "loss": 0.0104,
+      "step": 82100
+    },
+    {
+      "epoch": 495.69,
+      "learning_rate": 1.5306021833362282e-05,
+      "loss": 0.0105,
+      "step": 82120
+    },
+    {
+      "epoch": 495.81,
+      "learning_rate": 1.527371820213095e-05,
+      "loss": 0.0106,
+      "step": 82140
+    },
+    {
+      "epoch": 495.94,
+      "learning_rate": 1.5241442549237628e-05,
+      "loss": 0.0106,
+      "step": 82160
+    },
+    {
+      "epoch": 496.06,
+      "learning_rate": 1.5209194900686235e-05,
+      "loss": 0.0102,
+      "step": 82180
+    },
+    {
+      "epoch": 496.18,
+      "learning_rate": 1.5176975282458122e-05,
+      "loss": 0.01,
+      "step": 82200
+    },
+    {
+      "epoch": 496.3,
+      "learning_rate": 1.5144783720512035e-05,
+      "loss": 0.0101,
+      "step": 82220
+    },
+    {
+      "epoch": 496.42,
+      "learning_rate": 1.5112620240784126e-05,
+      "loss": 0.0103,
+      "step": 82240
+    },
+    {
+      "epoch": 496.54,
+      "learning_rate": 1.5080484869187977e-05,
+      "loss": 0.0104,
+      "step": 82260
+    },
+    {
+      "epoch": 496.66,
+      "learning_rate": 1.504837763161443e-05,
+      "loss": 0.0105,
+      "step": 82280
+    },
+    {
+      "epoch": 496.78,
+      "learning_rate": 1.501629855393173e-05,
+      "loss": 0.0106,
+      "step": 82300
+    },
+    {
+      "epoch": 496.9,
+      "learning_rate": 1.4984247661985396e-05,
+      "loss": 0.0106,
+      "step": 82320
+    },
+    {
+      "epoch": 497.02,
+      "learning_rate": 1.4952224981598278e-05,
+      "loss": 0.0105,
+      "step": 82340
+    },
+    {
+      "epoch": 497.14,
+      "learning_rate": 1.492023053857044e-05,
+      "loss": 0.01,
+      "step": 82360
+    },
+    {
+      "epoch": 497.26,
+      "learning_rate": 1.4888264358679238e-05,
+      "loss": 0.0101,
+      "step": 82380
+    },
+    {
+      "epoch": 497.38,
+      "learning_rate": 1.4856326467679248e-05,
+      "loss": 0.0103,
+      "step": 82400
+    },
+    {
+      "epoch": 497.51,
+      "learning_rate": 1.4824416891302256e-05,
+      "loss": 0.0104,
+      "step": 82420
+    },
+    {
+      "epoch": 497.63,
+      "learning_rate": 1.479253565525724e-05,
+      "loss": 0.0104,
+      "step": 82440
+    },
+    {
+      "epoch": 497.75,
+      "learning_rate": 1.476068278523029e-05,
+      "loss": 0.0106,
+      "step": 82460
+    },
+    {
+      "epoch": 497.87,
+      "learning_rate": 1.4728858306884757e-05,
+      "loss": 0.0105,
+      "step": 82480
+    },
+    {
+      "epoch": 497.99,
+      "learning_rate": 1.4697062245860992e-05,
+      "loss": 0.0106,
+      "step": 82500
+    },
+    {
+      "epoch": 498.11,
+      "learning_rate": 1.4665294627776533e-05,
+      "loss": 0.01,
+      "step": 82520
+    },
+    {
+      "epoch": 498.23,
+      "learning_rate": 1.463355547822597e-05,
+      "loss": 0.0102,
+      "step": 82540
+    },
+    {
+      "epoch": 498.35,
+      "learning_rate": 1.4601844822780964e-05,
+      "loss": 0.0102,
+      "step": 82560
+    },
+    {
+      "epoch": 498.47,
+      "learning_rate": 1.457016268699023e-05,
+      "loss": 0.0103,
+      "step": 82580
+    },
+    {
+      "epoch": 498.59,
+      "learning_rate": 1.4538509096379449e-05,
+      "loss": 0.0105,
+      "step": 82600
+    },
+    {
+      "epoch": 498.71,
+      "learning_rate": 1.450688407645141e-05,
+      "loss": 0.0105,
+      "step": 82620
+    },
+    {
+      "epoch": 498.83,
+      "learning_rate": 1.4475287652685775e-05,
+      "loss": 0.0107,
+      "step": 82640
+    },
+    {
+      "epoch": 498.95,
+      "learning_rate": 1.4443719850539222e-05,
+      "loss": 0.0106,
+      "step": 82660
+    },
+    {
+      "epoch": 499.07,
+      "learning_rate": 1.441218069544536e-05,
+      "loss": 0.0102,
+      "step": 82680
+    },
+    {
+      "epoch": 499.2,
+      "learning_rate": 1.4380670212814718e-05,
+      "loss": 0.0101,
+      "step": 82700
+    },
+    {
+      "epoch": 499.32,
+      "learning_rate": 1.4349188428034742e-05,
+      "loss": 0.0107,
+      "step": 82720
+    },
+    {
+      "epoch": 499.44,
+      "learning_rate": 1.4317735366469703e-05,
+      "loss": 0.011,
+      "step": 82740
+    },
+    {
+      "epoch": 499.56,
+      "learning_rate": 1.4286311053460777e-05,
+      "loss": 0.0108,
+      "step": 82760
+    },
+    {
+      "epoch": 499.68,
+      "learning_rate": 1.4254915514325972e-05,
+      "loss": 0.0108,
+      "step": 82780
+    },
+    {
+      "epoch": 499.8,
+      "learning_rate": 1.4223548774360117e-05,
+      "loss": 0.0109,
+      "step": 82800
+    },
+    {
+      "epoch": 499.92,
+      "learning_rate": 1.4192210858834786e-05,
+      "loss": 0.0109,
+      "step": 82820
+    },
+    {
+      "epoch": 500.04,
+      "learning_rate": 1.416090179299841e-05,
+      "loss": 0.011,
+      "step": 82840
+    },
+    {
+      "epoch": 500.16,
+      "learning_rate": 1.4129621602076149e-05,
+      "loss": 0.0106,
+      "step": 82860
+    },
+    {
+      "epoch": 500.28,
+      "learning_rate": 1.4098370311269848e-05,
+      "loss": 0.0113,
+      "step": 82880
+    },
+    {
+      "epoch": 500.4,
+      "learning_rate": 1.4067147945758125e-05,
+      "loss": 0.0121,
+      "step": 82900
+    },
+    {
+      "epoch": 500.52,
+      "learning_rate": 1.4035954530696277e-05,
+      "loss": 0.0124,
+      "step": 82920
+    },
+    {
+      "epoch": 500.64,
+      "learning_rate": 1.4004790091216291e-05,
+      "loss": 0.0131,
+      "step": 82940
+    },
+    {
+      "epoch": 500.76,
+      "learning_rate": 1.3973654652426766e-05,
+      "loss": 0.0146,
+      "step": 82960
+    },
+    {
+      "epoch": 500.89,
+      "learning_rate": 1.394254823941295e-05,
+      "loss": 0.0171,
+      "step": 82980
+    },
+    {
+      "epoch": 501.01,
+      "learning_rate": 1.3911470877236777e-05,
+      "loss": 0.0242,
+      "step": 83000
+    },
+    {
+      "epoch": 501.01,
+      "eval_accuracy": 0.0012713496853730577,
+      "eval_loss": 23.54586410522461,
+      "eval_runtime": 18.1613,
+      "eval_samples_per_second": 112.327,
+      "eval_steps_per_second": 2.368,
+      "step": 83000
+    },
+    {
+      "epoch": 501.13,
+      "learning_rate": 1.3881974314313356e-05,
+      "loss": 0.0352,
+      "step": 83020
+    },
+    {
+      "epoch": 501.25,
+      "learning_rate": 1.3850953673266066e-05,
+      "loss": 0.0441,
+      "step": 83040
+    },
+    {
+      "epoch": 501.37,
+      "learning_rate": 1.3819962156852468e-05,
+      "loss": 0.0434,
+      "step": 83060
+    },
+    {
+      "epoch": 501.49,
+      "learning_rate": 1.3788999790041867e-05,
+      "loss": 0.0399,
+      "step": 83080
+    },
+    {
+      "epoch": 501.61,
+      "learning_rate": 1.3758066597780106e-05,
+      "loss": 0.0358,
+      "step": 83100
+    },
+    {
+      "epoch": 501.73,
+      "learning_rate": 1.3727162604989451e-05,
+      "loss": 0.0313,
+      "step": 83120
+    },
+    {
+      "epoch": 501.85,
+      "learning_rate": 1.3696287836568744e-05,
+      "loss": 0.0278,
+      "step": 83140
+    },
+    {
+      "epoch": 501.97,
+      "learning_rate": 1.3665442317393196e-05,
+      "loss": 0.0249,
+      "step": 83160
+    },
+    {
+      "epoch": 502.09,
+      "learning_rate": 1.3634626072314482e-05,
+      "loss": 0.019,
+      "step": 83180
+    },
+    {
+      "epoch": 502.21,
+      "learning_rate": 1.3603839126160711e-05,
+      "loss": 0.0165,
+      "step": 83200
+    },
+    {
+      "epoch": 502.33,
+      "learning_rate": 1.3573081503736362e-05,
+      "loss": 0.0161,
+      "step": 83220
+    },
+    {
+      "epoch": 502.45,
+      "learning_rate": 1.3542353229822308e-05,
+      "loss": 0.0154,
+      "step": 83240
+    },
+    {
+      "epoch": 502.58,
+      "learning_rate": 1.3511654329175727e-05,
+      "loss": 0.0149,
+      "step": 83260
+    },
+    {
+      "epoch": 502.7,
+      "learning_rate": 1.3480984826530218e-05,
+      "loss": 0.0149,
+      "step": 83280
+    },
+    {
+      "epoch": 502.82,
+      "learning_rate": 1.3450344746595612e-05,
+      "loss": 0.0146,
+      "step": 83300
+    },
+    {
+      "epoch": 502.94,
+      "learning_rate": 1.3419734114058075e-05,
+      "loss": 0.0144,
+      "step": 83320
+    },
+    {
+      "epoch": 503.06,
+      "learning_rate": 1.3389152953580047e-05,
+      "loss": 0.0129,
+      "step": 83340
+    },
+    {
+      "epoch": 503.18,
+      "learning_rate": 1.335860128980021e-05,
+      "loss": 0.0112,
+      "step": 83360
+    },
+    {
+      "epoch": 503.3,
+      "learning_rate": 1.3328079147333505e-05,
+      "loss": 0.0112,
+      "step": 83380
+    },
+    {
+      "epoch": 503.42,
+      "learning_rate": 1.329758655077104e-05,
+      "loss": 0.0112,
+      "step": 83400
+    },
+    {
+      "epoch": 503.54,
+      "learning_rate": 1.326712352468017e-05,
+      "loss": 0.0112,
+      "step": 83420
+    },
+    {
+      "epoch": 503.66,
+      "learning_rate": 1.3236690093604399e-05,
+      "loss": 0.0111,
+      "step": 83440
+    },
+    {
+      "epoch": 503.78,
+      "learning_rate": 1.3206286282063395e-05,
+      "loss": 0.0112,
+      "step": 83460
+    },
+    {
+      "epoch": 503.9,
+      "learning_rate": 1.3175912114552963e-05,
+      "loss": 0.0112,
+      "step": 83480
+    },
+    {
+      "epoch": 504.02,
+      "learning_rate": 1.3145567615545012e-05,
+      "loss": 0.011,
+      "step": 83500
+    },
+    {
+      "epoch": 504.14,
+      "learning_rate": 1.3115252809487577e-05,
+      "loss": 0.0103,
+      "step": 83520
+    },
+    {
+      "epoch": 504.27,
+      "learning_rate": 1.3084967720804725e-05,
+      "loss": 0.0105,
+      "step": 83540
+    },
+    {
+      "epoch": 504.39,
+      "learning_rate": 1.3054712373896611e-05,
+      "loss": 0.0105,
+      "step": 83560
+    },
+    {
+      "epoch": 504.51,
+      "learning_rate": 1.3024486793139423e-05,
+      "loss": 0.0106,
+      "step": 83580
+    },
+    {
+      "epoch": 504.63,
+      "learning_rate": 1.2994291002885373e-05,
+      "loss": 0.0107,
+      "step": 83600
+    },
+    {
+      "epoch": 504.75,
+      "learning_rate": 1.296412502746267e-05,
+      "loss": 0.0106,
+      "step": 83620
+    },
+    {
+      "epoch": 504.87,
+      "learning_rate": 1.2933988891175458e-05,
+      "loss": 0.0106,
+      "step": 83640
+    },
+    {
+      "epoch": 504.99,
+      "learning_rate": 1.290388261830393e-05,
+      "loss": 0.0107,
+      "step": 83660
+    },
+    {
+      "epoch": 505.11,
+      "learning_rate": 1.287380623310413e-05,
+      "loss": 0.0102,
+      "step": 83680
+    },
+    {
+      "epoch": 505.23,
+      "learning_rate": 1.2843759759808061e-05,
+      "loss": 0.0103,
+      "step": 83700
+    },
+    {
+      "epoch": 505.35,
+      "learning_rate": 1.2813743222623637e-05,
+      "loss": 0.0103,
+      "step": 83720
+    },
+    {
+      "epoch": 505.47,
+      "learning_rate": 1.2783756645734635e-05,
+      "loss": 0.0104,
+      "step": 83740
+    },
+    {
+      "epoch": 505.59,
+      "learning_rate": 1.2753800053300707e-05,
+      "loss": 0.0105,
+      "step": 83760
+    },
+    {
+      "epoch": 505.71,
+      "learning_rate": 1.2723873469457304e-05,
+      "loss": 0.0105,
+      "step": 83780
+    },
+    {
+      "epoch": 505.84,
+      "learning_rate": 1.2693976918315786e-05,
+      "loss": 0.0105,
+      "step": 83800
+    },
+    {
+      "epoch": 505.96,
+      "learning_rate": 1.2664110423963221e-05,
+      "loss": 0.0106,
+      "step": 83820
+    },
+    {
+      "epoch": 506.08,
+      "learning_rate": 1.2634274010462538e-05,
+      "loss": 0.0103,
+      "step": 83840
+    },
+    {
+      "epoch": 506.2,
+      "learning_rate": 1.2604467701852351e-05,
+      "loss": 0.0102,
+      "step": 83860
+    },
+    {
+      "epoch": 506.32,
+      "learning_rate": 1.2574691522147103e-05,
+      "loss": 0.0102,
+      "step": 83880
+    },
+    {
+      "epoch": 506.44,
+      "learning_rate": 1.254494549533693e-05,
+      "loss": 0.0103,
+      "step": 83900
+    },
+    {
+      "epoch": 506.56,
+      "learning_rate": 1.2515229645387638e-05,
+      "loss": 0.0104,
+      "step": 83920
+    },
+    {
+      "epoch": 506.68,
+      "learning_rate": 1.2485543996240762e-05,
+      "loss": 0.0103,
+      "step": 83940
+    },
+    {
+      "epoch": 506.8,
+      "learning_rate": 1.2455888571813495e-05,
+      "loss": 0.0104,
+      "step": 83960
+    },
+    {
+      "epoch": 506.92,
+      "learning_rate": 1.2426263395998694e-05,
+      "loss": 0.0105,
+      "step": 83980
+    },
+    {
+      "epoch": 507.04,
+      "learning_rate": 1.2396668492664788e-05,
+      "loss": 0.0104,
+      "step": 84000
+    },
+    {
+      "epoch": 507.04,
+      "eval_accuracy": 0.0012540624907390028,
+      "eval_loss": 23.56952667236328,
+      "eval_runtime": 18.1296,
+      "eval_samples_per_second": 112.523,
+      "eval_steps_per_second": 2.372,
+      "step": 84000
+    },
+    {
+      "epoch": 507.16,
+      "learning_rate": 1.23671038856559e-05,
+      "loss": 0.01,
+      "step": 84020
+    },
+    {
+      "epoch": 507.28,
+      "learning_rate": 1.233756959879167e-05,
+      "loss": 0.0102,
+      "step": 84040
+    },
+    {
+      "epoch": 507.4,
+      "learning_rate": 1.2308065655867346e-05,
+      "loss": 0.0102,
+      "step": 84060
+    },
+    {
+      "epoch": 507.53,
+      "learning_rate": 1.2278592080653728e-05,
+      "loss": 0.0103,
+      "step": 84080
+    },
+    {
+      "epoch": 507.65,
+      "learning_rate": 1.2249148896897139e-05,
+      "loss": 0.0104,
+      "step": 84100
+    },
+    {
+      "epoch": 507.77,
+      "learning_rate": 1.2219736128319436e-05,
+      "loss": 0.0104,
+      "step": 84120
+    },
+    {
+      "epoch": 507.89,
+      "learning_rate": 1.2190353798617925e-05,
+      "loss": 0.0104,
+      "step": 84140
+    },
+    {
+      "epoch": 508.01,
+      "learning_rate": 1.2161001931465437e-05,
+      "loss": 0.0104,
+      "step": 84160
+    },
+    {
+      "epoch": 508.13,
+      "learning_rate": 1.2131680550510238e-05,
+      "loss": 0.0101,
+      "step": 84180
+    },
+    {
+      "epoch": 508.25,
+      "learning_rate": 1.2102389679376036e-05,
+      "loss": 0.0101,
+      "step": 84200
+    },
+    {
+      "epoch": 508.37,
+      "learning_rate": 1.207312934166196e-05,
+      "loss": 0.0101,
+      "step": 84220
+    },
+    {
+      "epoch": 508.49,
+      "learning_rate": 1.2043899560942523e-05,
+      "loss": 0.0102,
+      "step": 84240
+    },
+    {
+      "epoch": 508.61,
+      "learning_rate": 1.2014700360767662e-05,
+      "loss": 0.0102,
+      "step": 84260
+    },
+    {
+      "epoch": 508.73,
+      "learning_rate": 1.1985531764662605e-05,
+      "loss": 0.0103,
+      "step": 84280
+    },
+    {
+      "epoch": 508.85,
+      "learning_rate": 1.1956393796127979e-05,
+      "loss": 0.0104,
+      "step": 84300
+    },
+    {
+      "epoch": 508.97,
+      "learning_rate": 1.1927286478639726e-05,
+      "loss": 0.0104,
+      "step": 84320
+    },
+    {
+      "epoch": 509.09,
+      "learning_rate": 1.1898209835649082e-05,
+      "loss": 0.01,
+      "step": 84340
+    },
+    {
+      "epoch": 509.22,
+      "learning_rate": 1.1869163890582586e-05,
+      "loss": 0.01,
+      "step": 84360
+    },
+    {
+      "epoch": 509.34,
+      "learning_rate": 1.1840148666841989e-05,
+      "loss": 0.0102,
+      "step": 84380
+    },
+    {
+      "epoch": 509.46,
+      "learning_rate": 1.181116418780439e-05,
+      "loss": 0.0101,
+      "step": 84400
+    },
+    {
+      "epoch": 509.58,
+      "learning_rate": 1.1782210476822015e-05,
+      "loss": 0.0103,
+      "step": 84420
+    },
+    {
+      "epoch": 509.7,
+      "learning_rate": 1.1753287557222365e-05,
+      "loss": 0.0103,
+      "step": 84440
+    },
+    {
+      "epoch": 509.82,
+      "learning_rate": 1.1724395452308112e-05,
+      "loss": 0.0103,
+      "step": 84460
+    },
+    {
+      "epoch": 509.94,
+      "learning_rate": 1.1695534185357099e-05,
+      "loss": 0.0104,
+      "step": 84480
+    },
+    {
+      "epoch": 510.06,
+      "learning_rate": 1.1666703779622351e-05,
+      "loss": 0.0102,
+      "step": 84500
+    },
+    {
+      "epoch": 510.18,
+      "learning_rate": 1.1637904258331956e-05,
+      "loss": 0.01,
+      "step": 84520
+    },
+    {
+      "epoch": 510.3,
+      "learning_rate": 1.1609135644689224e-05,
+      "loss": 0.01,
+      "step": 84540
+    },
+    {
+      "epoch": 510.42,
+      "learning_rate": 1.1580397961872463e-05,
+      "loss": 0.0101,
+      "step": 84560
+    },
+    {
+      "epoch": 510.54,
+      "learning_rate": 1.1551691233035144e-05,
+      "loss": 0.0102,
+      "step": 84580
+    },
+    {
+      "epoch": 510.66,
+      "learning_rate": 1.1523015481305704e-05,
+      "loss": 0.0103,
+      "step": 84600
+    },
+    {
+      "epoch": 510.78,
+      "learning_rate": 1.1494370729787728e-05,
+      "loss": 0.0103,
+      "step": 84620
+    },
+    {
+      "epoch": 510.91,
+      "learning_rate": 1.146575700155978e-05,
+      "loss": 0.0104,
+      "step": 84640
+    },
+    {
+      "epoch": 511.03,
+      "learning_rate": 1.1437174319675386e-05,
+      "loss": 0.0103,
+      "step": 84660
+    },
+    {
+      "epoch": 511.15,
+      "learning_rate": 1.1408622707163113e-05,
+      "loss": 0.0098,
+      "step": 84680
+    },
+    {
+      "epoch": 511.27,
+      "learning_rate": 1.1380102187026481e-05,
+      "loss": 0.01,
+      "step": 84700
+    },
+    {
+      "epoch": 511.39,
+      "learning_rate": 1.1351612782243976e-05,
+      "loss": 0.0101,
+      "step": 84720
+    },
+    {
+      "epoch": 511.51,
+      "learning_rate": 1.1323154515768947e-05,
+      "loss": 0.0102,
+      "step": 84740
+    },
+    {
+      "epoch": 511.63,
+      "learning_rate": 1.1294727410529754e-05,
+      "loss": 0.0102,
+      "step": 84760
+    },
+    {
+      "epoch": 511.75,
+      "learning_rate": 1.1266331489429593e-05,
+      "loss": 0.0103,
+      "step": 84780
+    },
+    {
+      "epoch": 511.87,
+      "learning_rate": 1.1237966775346526e-05,
+      "loss": 0.0103,
+      "step": 84800
+    },
+    {
+      "epoch": 511.99,
+      "learning_rate": 1.1209633291133503e-05,
+      "loss": 0.0104,
+      "step": 84820
+    },
+    {
+      "epoch": 512.11,
+      "learning_rate": 1.11813310596183e-05,
+      "loss": 0.0099,
+      "step": 84840
+    },
+    {
+      "epoch": 512.23,
+      "learning_rate": 1.1153060103603524e-05,
+      "loss": 0.0099,
+      "step": 84860
+    },
+    {
+      "epoch": 512.35,
+      "learning_rate": 1.1124820445866551e-05,
+      "loss": 0.0101,
+      "step": 84880
+    },
+    {
+      "epoch": 512.47,
+      "learning_rate": 1.109661210915956e-05,
+      "loss": 0.0101,
+      "step": 84900
+    },
+    {
+      "epoch": 512.6,
+      "learning_rate": 1.106843511620954e-05,
+      "loss": 0.0102,
+      "step": 84920
+    },
+    {
+      "epoch": 512.72,
+      "learning_rate": 1.1040289489718148e-05,
+      "loss": 0.0103,
+      "step": 84940
+    },
+    {
+      "epoch": 512.84,
+      "learning_rate": 1.1012175252361818e-05,
+      "loss": 0.0103,
+      "step": 84960
+    },
+    {
+      "epoch": 512.96,
+      "learning_rate": 1.0984092426791676e-05,
+      "loss": 0.0104,
+      "step": 84980
+    },
+    {
+      "epoch": 513.08,
+      "learning_rate": 1.0956041035633574e-05,
+      "loss": 0.01,
+      "step": 85000
+    },
+    {
+      "epoch": 513.08,
+      "eval_accuracy": 0.0012575199296658139,
+      "eval_loss": 23.66585350036621,
+      "eval_runtime": 18.2862,
+      "eval_samples_per_second": 111.56,
+      "eval_steps_per_second": 2.352,
+      "step": 85000
+    },
+    {
+      "epoch": 513.2,
+      "learning_rate": 1.0928021101487973e-05,
+      "loss": 0.01,
+      "step": 85020
+    },
+    {
+      "epoch": 513.32,
+      "learning_rate": 1.0900032646930031e-05,
+      "loss": 0.0099,
+      "step": 85040
+    },
+    {
+      "epoch": 513.44,
+      "learning_rate": 1.0872075694509587e-05,
+      "loss": 0.0101,
+      "step": 85060
+    },
+    {
+      "epoch": 513.56,
+      "learning_rate": 1.0844150266751007e-05,
+      "loss": 0.0102,
+      "step": 85080
+    },
+    {
+      "epoch": 513.68,
+      "learning_rate": 1.0816256386153328e-05,
+      "loss": 0.0103,
+      "step": 85100
+    },
+    {
+      "epoch": 513.8,
+      "learning_rate": 1.0788394075190112e-05,
+      "loss": 0.0102,
+      "step": 85120
+    },
+    {
+      "epoch": 513.92,
+      "learning_rate": 1.0760563356309566e-05,
+      "loss": 0.0103,
+      "step": 85140
+    },
+    {
+      "epoch": 514.04,
+      "learning_rate": 1.0732764251934363e-05,
+      "loss": 0.0102,
+      "step": 85160
+    },
+    {
+      "epoch": 514.16,
+      "learning_rate": 1.0704996784461752e-05,
+      "loss": 0.0099,
+      "step": 85180
+    },
+    {
+      "epoch": 514.29,
+      "learning_rate": 1.0677260976263476e-05,
+      "loss": 0.01,
+      "step": 85200
+    },
+    {
+      "epoch": 514.41,
+      "learning_rate": 1.0649556849685777e-05,
+      "loss": 0.01,
+      "step": 85220
+    },
+    {
+      "epoch": 514.53,
+      "learning_rate": 1.0621884427049389e-05,
+      "loss": 0.0101,
+      "step": 85240
+    },
+    {
+      "epoch": 514.65,
+      "learning_rate": 1.0594243730649428e-05,
+      "loss": 0.0101,
+      "step": 85260
+    },
+    {
+      "epoch": 514.77,
+      "learning_rate": 1.0566634782755574e-05,
+      "loss": 0.0103,
+      "step": 85280
+    },
+    {
+      "epoch": 514.89,
+      "learning_rate": 1.0539057605611808e-05,
+      "loss": 0.0103,
+      "step": 85300
+    },
+    {
+      "epoch": 515.01,
+      "learning_rate": 1.0511512221436581e-05,
+      "loss": 0.0103,
+      "step": 85320
+    },
+    {
+      "epoch": 515.13,
+      "learning_rate": 1.0483998652422706e-05,
+      "loss": 0.0098,
+      "step": 85340
+    },
+    {
+      "epoch": 515.25,
+      "learning_rate": 1.0456516920737375e-05,
+      "loss": 0.0099,
+      "step": 85360
+    },
+    {
+      "epoch": 515.37,
+      "learning_rate": 1.042906704852213e-05,
+      "loss": 0.01,
+      "step": 85380
+    },
+    {
+      "epoch": 515.49,
+      "learning_rate": 1.0401649057892814e-05,
+      "loss": 0.0101,
+      "step": 85400
+    },
+    {
+      "epoch": 515.61,
+      "learning_rate": 1.0374262970939625e-05,
+      "loss": 0.0102,
+      "step": 85420
+    },
+    {
+      "epoch": 515.73,
+      "learning_rate": 1.0346908809727029e-05,
+      "loss": 0.0102,
+      "step": 85440
+    },
+    {
+      "epoch": 515.86,
+      "learning_rate": 1.03195865962938e-05,
+      "loss": 0.0102,
+      "step": 85460
+    },
+    {
+      "epoch": 515.98,
+      "learning_rate": 1.0292296352652908e-05,
+      "loss": 0.0103,
+      "step": 85480
+    },
+    {
+      "epoch": 516.1,
+      "learning_rate": 1.0265038100791652e-05,
+      "loss": 0.0099,
+      "step": 85500
+    },
+    {
+      "epoch": 516.22,
+      "learning_rate": 1.0237811862671509e-05,
+      "loss": 0.01,
+      "step": 85520
+    },
+    {
+      "epoch": 516.34,
+      "learning_rate": 1.0210617660228145e-05,
+      "loss": 0.01,
+      "step": 85540
+    },
+    {
+      "epoch": 516.46,
+      "learning_rate": 1.0183455515371454e-05,
+      "loss": 0.0101,
+      "step": 85560
+    },
+    {
+      "epoch": 516.58,
+      "learning_rate": 1.0156325449985477e-05,
+      "loss": 0.0101,
+      "step": 85580
+    },
+    {
+      "epoch": 516.7,
+      "learning_rate": 1.0129227485928433e-05,
+      "loss": 0.0103,
+      "step": 85600
+    },
+    {
+      "epoch": 516.82,
+      "learning_rate": 1.0102161645032643e-05,
+      "loss": 0.0102,
+      "step": 85620
+    },
+    {
+      "epoch": 516.94,
+      "learning_rate": 1.0075127949104558e-05,
+      "loss": 0.0103,
+      "step": 85640
+    },
+    {
+      "epoch": 517.06,
+      "learning_rate": 1.004812641992478e-05,
+      "loss": 0.01,
+      "step": 85660
+    },
+    {
+      "epoch": 517.18,
+      "learning_rate": 1.0021157079247917e-05,
+      "loss": 0.0098,
+      "step": 85680
+    },
+    {
+      "epoch": 517.3,
+      "learning_rate": 9.994219948802686e-06,
+      "loss": 0.0099,
+      "step": 85700
+    },
+    {
+      "epoch": 517.42,
+      "learning_rate": 9.96731505029186e-06,
+      "loss": 0.01,
+      "step": 85720
+    },
+    {
+      "epoch": 517.55,
+      "learning_rate": 9.940442405392226e-06,
+      "loss": 0.0102,
+      "step": 85740
+    },
+    {
+      "epoch": 517.67,
+      "learning_rate": 9.913602035754572e-06,
+      "loss": 0.0102,
+      "step": 85760
+    },
+    {
+      "epoch": 517.79,
+      "learning_rate": 9.886793963003698e-06,
+      "loss": 0.0102,
+      "step": 85780
+    },
+    {
+      "epoch": 517.91,
+      "learning_rate": 9.860018208738425e-06,
+      "loss": 0.0103,
+      "step": 85800
+    },
+    {
+      "epoch": 518.03,
+      "learning_rate": 9.83327479453145e-06,
+      "loss": 0.0102,
+      "step": 85820
+    },
+    {
+      "epoch": 518.15,
+      "learning_rate": 9.806563741929475e-06,
+      "loss": 0.0097,
+      "step": 85840
+    },
+    {
+      "epoch": 518.27,
+      "learning_rate": 9.779885072453116e-06,
+      "loss": 0.0099,
+      "step": 85860
+    },
+    {
+      "epoch": 518.39,
+      "learning_rate": 9.753238807596903e-06,
+      "loss": 0.01,
+      "step": 85880
+    },
+    {
+      "epoch": 518.51,
+      "learning_rate": 9.726624968829223e-06,
+      "loss": 0.0101,
+      "step": 85900
+    },
+    {
+      "epoch": 518.63,
+      "learning_rate": 9.700043577592388e-06,
+      "loss": 0.0102,
+      "step": 85920
+    },
+    {
+      "epoch": 518.75,
+      "learning_rate": 9.673494655302533e-06,
+      "loss": 0.0102,
+      "step": 85940
+    },
+    {
+      "epoch": 518.87,
+      "learning_rate": 9.646978223349651e-06,
+      "loss": 0.0103,
+      "step": 85960
+    },
+    {
+      "epoch": 518.99,
+      "learning_rate": 9.620494303097566e-06,
+      "loss": 0.0103,
+      "step": 85980
+    },
+    {
+      "epoch": 519.11,
+      "learning_rate": 9.594042915883849e-06,
+      "loss": 0.0098,
+      "step": 86000
+    },
+    {
+      "epoch": 519.11,
+      "eval_accuracy": 0.0012545564105856902,
+      "eval_loss": 23.733713150024414,
+      "eval_runtime": 18.3445,
+      "eval_samples_per_second": 111.205,
+      "eval_steps_per_second": 2.344,
+      "step": 86000
+    },
+    {
+      "epoch": 519.24,
+      "learning_rate": 9.567624083019966e-06,
+      "loss": 0.0098,
+      "step": 86020
+    },
+    {
+      "epoch": 519.36,
+      "learning_rate": 9.541237825791044e-06,
+      "loss": 0.0099,
+      "step": 86040
+    },
+    {
+      "epoch": 519.48,
+      "learning_rate": 9.51488416545604e-06,
+      "loss": 0.0101,
+      "step": 86060
+    },
+    {
+      "epoch": 519.6,
+      "learning_rate": 9.488563123247612e-06,
+      "loss": 0.0101,
+      "step": 86080
+    },
+    {
+      "epoch": 519.72,
+      "learning_rate": 9.46227472037215e-06,
+      "loss": 0.0102,
+      "step": 86100
+    },
+    {
+      "epoch": 519.84,
+      "learning_rate": 9.436018978009759e-06,
+      "loss": 0.0102,
+      "step": 86120
+    },
+    {
+      "epoch": 519.96,
+      "learning_rate": 9.409795917314195e-06,
+      "loss": 0.0103,
+      "step": 86140
+    },
+    {
+      "epoch": 520.08,
+      "learning_rate": 9.383605559412911e-06,
+      "loss": 0.01,
+      "step": 86160
+    },
+    {
+      "epoch": 520.2,
+      "learning_rate": 9.357447925407008e-06,
+      "loss": 0.0098,
+      "step": 86180
+    },
+    {
+      "epoch": 520.32,
+      "learning_rate": 9.33132303637122e-06,
+      "loss": 0.01,
+      "step": 86200
+    },
+    {
+      "epoch": 520.44,
+      "learning_rate": 9.305230913353896e-06,
+      "loss": 0.0101,
+      "step": 86220
+    },
+    {
+      "epoch": 520.56,
+      "learning_rate": 9.279171577376988e-06,
+      "loss": 0.0102,
+      "step": 86240
+    },
+    {
+      "epoch": 520.68,
+      "learning_rate": 9.253145049436046e-06,
+      "loss": 0.0101,
+      "step": 86260
+    },
+    {
+      "epoch": 520.8,
+      "learning_rate": 9.227151350500151e-06,
+      "loss": 0.0103,
+      "step": 86280
+    },
+    {
+      "epoch": 520.93,
+      "learning_rate": 9.201190501511964e-06,
+      "loss": 0.0102,
+      "step": 86300
+    },
+    {
+      "epoch": 521.05,
+      "learning_rate": 9.175262523387678e-06,
+      "loss": 0.0101,
+      "step": 86320
+    },
+    {
+      "epoch": 521.17,
+      "learning_rate": 9.149367437016992e-06,
+      "loss": 0.0097,
+      "step": 86340
+    },
+    {
+      "epoch": 521.29,
+      "learning_rate": 9.12350526326311e-06,
+      "loss": 0.0099,
+      "step": 86360
+    },
+    {
+      "epoch": 521.41,
+      "learning_rate": 9.09767602296272e-06,
+      "loss": 0.01,
+      "step": 86380
+    },
+    {
+      "epoch": 521.53,
+      "learning_rate": 9.071879736925987e-06,
+      "loss": 0.0101,
+      "step": 86400
+    },
+    {
+      "epoch": 521.65,
+      "learning_rate": 9.046116425936491e-06,
+      "loss": 0.0101,
+      "step": 86420
+    },
+    {
+      "epoch": 521.77,
+      "learning_rate": 9.020386110751266e-06,
+      "loss": 0.0102,
+      "step": 86440
+    },
+    {
+      "epoch": 521.89,
+      "learning_rate": 8.994688812100776e-06,
+      "loss": 0.0102,
+      "step": 86460
+    },
+    {
+      "epoch": 522.01,
+      "learning_rate": 8.969024550688881e-06,
+      "loss": 0.0103,
+      "step": 86480
+    },
+    {
+      "epoch": 522.13,
+      "learning_rate": 8.94339334719278e-06,
+      "loss": 0.0097,
+      "step": 86500
+    },
+    {
+      "epoch": 522.25,
+      "learning_rate": 8.917795222263076e-06,
+      "loss": 0.0099,
+      "step": 86520
+    },
+    {
+      "epoch": 522.37,
+      "learning_rate": 8.892230196523754e-06,
+      "loss": 0.0101,
+      "step": 86540
+    },
+    {
+      "epoch": 522.49,
+      "learning_rate": 8.866698290572051e-06,
+      "loss": 0.01,
+      "step": 86560
+    },
+    {
+      "epoch": 522.62,
+      "learning_rate": 8.841199524978583e-06,
+      "loss": 0.01,
+      "step": 86580
+    },
+    {
+      "epoch": 522.74,
+      "learning_rate": 8.81573392028724e-06,
+      "loss": 0.0101,
+      "step": 86600
+    },
+    {
+      "epoch": 522.86,
+      "learning_rate": 8.790301497015207e-06,
+      "loss": 0.0103,
+      "step": 86620
+    },
+    {
+      "epoch": 522.98,
+      "learning_rate": 8.764902275652914e-06,
+      "loss": 0.0102,
+      "step": 86640
+    },
+    {
+      "epoch": 523.1,
+      "learning_rate": 8.739536276664063e-06,
+      "loss": 0.0098,
+      "step": 86660
+    },
+    {
+      "epoch": 523.22,
+      "learning_rate": 8.714203520485582e-06,
+      "loss": 0.0099,
+      "step": 86680
+    },
+    {
+      "epoch": 523.34,
+      "learning_rate": 8.688904027527605e-06,
+      "loss": 0.0099,
+      "step": 86700
+    },
+    {
+      "epoch": 523.46,
+      "learning_rate": 8.663637818173504e-06,
+      "loss": 0.01,
+      "step": 86720
+    },
+    {
+      "epoch": 523.58,
+      "learning_rate": 8.638404912779763e-06,
+      "loss": 0.0101,
+      "step": 86740
+    },
+    {
+      "epoch": 523.7,
+      "learning_rate": 8.613205331676133e-06,
+      "loss": 0.0101,
+      "step": 86760
+    },
+    {
+      "epoch": 523.82,
+      "learning_rate": 8.588039095165412e-06,
+      "loss": 0.0102,
+      "step": 86780
+    },
+    {
+      "epoch": 523.94,
+      "learning_rate": 8.562906223523603e-06,
+      "loss": 0.0103,
+      "step": 86800
+    },
+    {
+      "epoch": 524.06,
+      "learning_rate": 8.53780673699981e-06,
+      "loss": 0.01,
+      "step": 86820
+    },
+    {
+      "epoch": 524.19,
+      "learning_rate": 8.512740655816232e-06,
+      "loss": 0.0098,
+      "step": 86840
+    },
+    {
+      "epoch": 524.31,
+      "learning_rate": 8.487708000168165e-06,
+      "loss": 0.0099,
+      "step": 86860
+    },
+    {
+      "epoch": 524.43,
+      "learning_rate": 8.46270879022394e-06,
+      "loss": 0.01,
+      "step": 86880
+    },
+    {
+      "epoch": 524.55,
+      "learning_rate": 8.437743046125013e-06,
+      "loss": 0.0101,
+      "step": 86900
+    },
+    {
+      "epoch": 524.67,
+      "learning_rate": 8.412810787985797e-06,
+      "loss": 0.0101,
+      "step": 86920
+    },
+    {
+      "epoch": 524.79,
+      "learning_rate": 8.387912035893774e-06,
+      "loss": 0.0102,
+      "step": 86940
+    },
+    {
+      "epoch": 524.91,
+      "learning_rate": 8.36304680990942e-06,
+      "loss": 0.0102,
+      "step": 86960
+    },
+    {
+      "epoch": 525.03,
+      "learning_rate": 8.338215130066195e-06,
+      "loss": 0.0102,
+      "step": 86980
+    },
+    {
+      "epoch": 525.15,
+      "learning_rate": 8.313417016370556e-06,
+      "loss": 0.0097,
+      "step": 87000
+    },
+    {
+      "epoch": 525.15,
+      "eval_accuracy": 0.001259495609052563,
+      "eval_loss": 23.796077728271484,
+      "eval_runtime": 18.2469,
+      "eval_samples_per_second": 111.8,
+      "eval_steps_per_second": 2.357,
+      "step": 87000
+    },
+    {
+      "epoch": 525.27,
+      "learning_rate": 8.288652488801858e-06,
+      "loss": 0.0098,
+      "step": 87020
+    },
+    {
+      "epoch": 525.39,
+      "learning_rate": 8.263921567312454e-06,
+      "loss": 0.0099,
+      "step": 87040
+    },
+    {
+      "epoch": 525.51,
+      "learning_rate": 8.2392242718276e-06,
+      "loss": 0.01,
+      "step": 87060
+    },
+    {
+      "epoch": 525.63,
+      "learning_rate": 8.21456062224546e-06,
+      "loss": 0.0102,
+      "step": 87080
+    },
+    {
+      "epoch": 525.75,
+      "learning_rate": 8.189930638437094e-06,
+      "loss": 0.0101,
+      "step": 87100
+    },
+    {
+      "epoch": 525.88,
+      "learning_rate": 8.165334340246427e-06,
+      "loss": 0.0103,
+      "step": 87120
+    },
+    {
+      "epoch": 526.0,
+      "learning_rate": 8.140771747490273e-06,
+      "loss": 0.0102,
+      "step": 87140
+    },
+    {
+      "epoch": 526.12,
+      "learning_rate": 8.116242879958236e-06,
+      "loss": 0.0098,
+      "step": 87160
+    },
+    {
+      "epoch": 526.24,
+      "learning_rate": 8.091747757412804e-06,
+      "loss": 0.0098,
+      "step": 87180
+    },
+    {
+      "epoch": 526.36,
+      "learning_rate": 8.067286399589246e-06,
+      "loss": 0.0099,
+      "step": 87200
+    },
+    {
+      "epoch": 526.48,
+      "learning_rate": 8.042858826195648e-06,
+      "loss": 0.01,
+      "step": 87220
+    },
+    {
+      "epoch": 526.6,
+      "learning_rate": 8.01846505691286e-06,
+      "loss": 0.01,
+      "step": 87240
+    },
+    {
+      "epoch": 526.72,
+      "learning_rate": 7.99410511139448e-06,
+      "loss": 0.0102,
+      "step": 87260
+    },
+    {
+      "epoch": 526.84,
+      "learning_rate": 7.969779009266915e-06,
+      "loss": 0.0102,
+      "step": 87280
+    },
+    {
+      "epoch": 526.96,
+      "learning_rate": 7.945486770129234e-06,
+      "loss": 0.0103,
+      "step": 87300
+    },
+    {
+      "epoch": 527.08,
+      "learning_rate": 7.921228413553272e-06,
+      "loss": 0.0098,
+      "step": 87320
+    },
+    {
+      "epoch": 527.2,
+      "learning_rate": 7.897003959083538e-06,
+      "loss": 0.0098,
+      "step": 87340
+    },
+    {
+      "epoch": 527.32,
+      "learning_rate": 7.87281342623724e-06,
+      "loss": 0.0099,
+      "step": 87360
+    },
+    {
+      "epoch": 527.44,
+      "learning_rate": 7.848656834504276e-06,
+      "loss": 0.01,
+      "step": 87380
+    },
+    {
+      "epoch": 527.57,
+      "learning_rate": 7.824534203347122e-06,
+      "loss": 0.01,
+      "step": 87400
+    },
+    {
+      "epoch": 527.69,
+      "learning_rate": 7.800445552201013e-06,
+      "loss": 0.0102,
+      "step": 87420
+    },
+    {
+      "epoch": 527.81,
+      "learning_rate": 7.77639090047369e-06,
+      "loss": 0.0102,
+      "step": 87440
+    },
+    {
+      "epoch": 527.93,
+      "learning_rate": 7.752370267545584e-06,
+      "loss": 0.0103,
+      "step": 87460
+    },
+    {
+      "epoch": 528.05,
+      "learning_rate": 7.728383672769641e-06,
+      "loss": 0.0099,
+      "step": 87480
+    },
+    {
+      "epoch": 528.17,
+      "learning_rate": 7.704431135471473e-06,
+      "loss": 0.0097,
+      "step": 87500
+    },
+    {
+      "epoch": 528.29,
+      "learning_rate": 7.680512674949197e-06,
+      "loss": 0.0099,
+      "step": 87520
+    },
+    {
+      "epoch": 528.41,
+      "learning_rate": 7.656628310473468e-06,
+      "loss": 0.0099,
+      "step": 87540
+    },
+    {
+      "epoch": 528.53,
+      "learning_rate": 7.632778061287493e-06,
+      "loss": 0.01,
+      "step": 87560
+    },
+    {
+      "epoch": 528.65,
+      "learning_rate": 7.608961946606996e-06,
+      "loss": 0.0101,
+      "step": 87580
+    },
+    {
+      "epoch": 528.77,
+      "learning_rate": 7.5851799856201945e-06,
+      "loss": 0.0101,
+      "step": 87600
+    },
+    {
+      "epoch": 528.89,
+      "learning_rate": 7.56143219748775e-06,
+      "loss": 0.0102,
+      "step": 87620
+    },
+    {
+      "epoch": 529.01,
+      "learning_rate": 7.537718601342858e-06,
+      "loss": 0.0102,
+      "step": 87640
+    },
+    {
+      "epoch": 529.13,
+      "learning_rate": 7.514039216291147e-06,
+      "loss": 0.0097,
+      "step": 87660
+    },
+    {
+      "epoch": 529.26,
+      "learning_rate": 7.490394061410638e-06,
+      "loss": 0.0098,
+      "step": 87680
+    },
+    {
+      "epoch": 529.38,
+      "learning_rate": 7.4667831557518165e-06,
+      "loss": 0.0099,
+      "step": 87700
+    },
+    {
+      "epoch": 529.5,
+      "learning_rate": 7.443206518337564e-06,
+      "loss": 0.0101,
+      "step": 87720
+    },
+    {
+      "epoch": 529.62,
+      "learning_rate": 7.419664168163165e-06,
+      "loss": 0.01,
+      "step": 87740
+    },
+    {
+      "epoch": 529.74,
+      "learning_rate": 7.396156124196241e-06,
+      "loss": 0.0101,
+      "step": 87760
+    },
+    {
+      "epoch": 529.86,
+      "learning_rate": 7.372682405376807e-06,
+      "loss": 0.0102,
+      "step": 87780
+    },
+    {
+      "epoch": 529.98,
+      "learning_rate": 7.34924303061722e-06,
+      "loss": 0.0102,
+      "step": 87800
+    },
+    {
+      "epoch": 530.1,
+      "learning_rate": 7.325838018802156e-06,
+      "loss": 0.0097,
+      "step": 87820
+    },
+    {
+      "epoch": 530.22,
+      "learning_rate": 7.302467388788614e-06,
+      "loss": 0.0098,
+      "step": 87840
+    },
+    {
+      "epoch": 530.34,
+      "learning_rate": 7.279131159405888e-06,
+      "loss": 0.0099,
+      "step": 87860
+    },
+    {
+      "epoch": 530.46,
+      "learning_rate": 7.255829349455567e-06,
+      "loss": 0.01,
+      "step": 87880
+    },
+    {
+      "epoch": 530.58,
+      "learning_rate": 7.232561977711472e-06,
+      "loss": 0.01,
+      "step": 87900
+    },
+    {
+      "epoch": 530.7,
+      "learning_rate": 7.209329062919723e-06,
+      "loss": 0.0102,
+      "step": 87920
+    },
+    {
+      "epoch": 530.82,
+      "learning_rate": 7.186130623798648e-06,
+      "loss": 0.0101,
+      "step": 87940
+    },
+    {
+      "epoch": 530.95,
+      "learning_rate": 7.1629666790388236e-06,
+      "loss": 0.0102,
+      "step": 87960
+    },
+    {
+      "epoch": 531.07,
+      "learning_rate": 7.139837247303028e-06,
+      "loss": 0.0099,
+      "step": 87980
+    },
+    {
+      "epoch": 531.19,
+      "learning_rate": 7.11674234722619e-06,
+      "loss": 0.0097,
+      "step": 88000
+    },
+    {
+      "epoch": 531.19,
+      "eval_accuracy": 0.0012550503304323774,
+      "eval_loss": 23.857345581054688,
+      "eval_runtime": 18.1585,
+      "eval_samples_per_second": 112.344,
+      "eval_steps_per_second": 2.368,
+      "step": 88000
+    },
+    {
+      "epoch": 531.31,
+      "learning_rate": 7.093681997415508e-06,
+      "loss": 0.0099,
+      "step": 88020
+    },
+    {
+      "epoch": 531.43,
+      "learning_rate": 7.070656216450239e-06,
+      "loss": 0.0099,
+      "step": 88040
+    },
+    {
+      "epoch": 531.55,
+      "learning_rate": 7.047665022881866e-06,
+      "loss": 0.01,
+      "step": 88060
+    },
+    {
+      "epoch": 531.67,
+      "learning_rate": 7.0247084352339675e-06,
+      "loss": 0.0101,
+      "step": 88080
+    },
+    {
+      "epoch": 531.79,
+      "learning_rate": 7.001786472002259e-06,
+      "loss": 0.0101,
+      "step": 88100
+    },
+    {
+      "epoch": 531.91,
+      "learning_rate": 6.978899151654555e-06,
+      "loss": 0.0102,
+      "step": 88120
+    },
+    {
+      "epoch": 532.03,
+      "learning_rate": 6.95604649263073e-06,
+      "loss": 0.01,
+      "step": 88140
+    },
+    {
+      "epoch": 532.15,
+      "learning_rate": 6.933228513342804e-06,
+      "loss": 0.0096,
+      "step": 88160
+    },
+    {
+      "epoch": 532.27,
+      "learning_rate": 6.910445232174772e-06,
+      "loss": 0.0098,
+      "step": 88180
+    },
+    {
+      "epoch": 532.39,
+      "learning_rate": 6.887696667482729e-06,
+      "loss": 0.0099,
+      "step": 88200
+    },
+    {
+      "epoch": 532.52,
+      "learning_rate": 6.8649828375947745e-06,
+      "loss": 0.01,
+      "step": 88220
+    },
+    {
+      "epoch": 532.64,
+      "learning_rate": 6.8423037608110415e-06,
+      "loss": 0.01,
+      "step": 88240
+    },
+    {
+      "epoch": 532.76,
+      "learning_rate": 6.8196594554036545e-06,
+      "loss": 0.0101,
+      "step": 88260
+    },
+    {
+      "epoch": 532.88,
+      "learning_rate": 6.797049939616701e-06,
+      "loss": 0.0102,
+      "step": 88280
+    },
+    {
+      "epoch": 533.0,
+      "learning_rate": 6.774475231666272e-06,
+      "loss": 0.0102,
+      "step": 88300
+    },
+    {
+      "epoch": 533.12,
+      "learning_rate": 6.751935349740407e-06,
+      "loss": 0.0096,
+      "step": 88320
+    },
+    {
+      "epoch": 533.24,
+      "learning_rate": 6.729430311999085e-06,
+      "loss": 0.0097,
+      "step": 88340
+    },
+    {
+      "epoch": 533.36,
+      "learning_rate": 6.706960136574175e-06,
+      "loss": 0.0099,
+      "step": 88360
+    },
+    {
+      "epoch": 533.48,
+      "learning_rate": 6.684524841569534e-06,
+      "loss": 0.0099,
+      "step": 88380
+    },
+    {
+      "epoch": 533.6,
+      "learning_rate": 6.662124445060863e-06,
+      "loss": 0.01,
+      "step": 88400
+    },
+    {
+      "epoch": 533.72,
+      "learning_rate": 6.639758965095744e-06,
+      "loss": 0.01,
+      "step": 88420
+    },
+    {
+      "epoch": 533.84,
+      "learning_rate": 6.617428419693639e-06,
+      "loss": 0.0102,
+      "step": 88440
+    },
+    {
+      "epoch": 533.96,
+      "learning_rate": 6.595132826845879e-06,
+      "loss": 0.0102,
+      "step": 88460
+    },
+    {
+      "epoch": 534.08,
+      "learning_rate": 6.5728722045156285e-06,
+      "loss": 0.0099,
+      "step": 88480
+    },
+    {
+      "epoch": 534.21,
+      "learning_rate": 6.550646570637836e-06,
+      "loss": 0.0097,
+      "step": 88500
+    },
+    {
+      "epoch": 534.33,
+      "learning_rate": 6.528455943119305e-06,
+      "loss": 0.0098,
+      "step": 88520
+    },
+    {
+      "epoch": 534.45,
+      "learning_rate": 6.506300339838656e-06,
+      "loss": 0.0099,
+      "step": 88540
+    },
+    {
+      "epoch": 534.57,
+      "learning_rate": 6.484179778646216e-06,
+      "loss": 0.0099,
+      "step": 88560
+    },
+    {
+      "epoch": 534.69,
+      "learning_rate": 6.462094277364139e-06,
+      "loss": 0.0101,
+      "step": 88580
+    },
+    {
+      "epoch": 534.81,
+      "learning_rate": 6.440043853786315e-06,
+      "loss": 0.0101,
+      "step": 88600
+    },
+    {
+      "epoch": 534.93,
+      "learning_rate": 6.418028525678382e-06,
+      "loss": 0.0102,
+      "step": 88620
+    },
+    {
+      "epoch": 535.05,
+      "learning_rate": 6.396048310777669e-06,
+      "loss": 0.01,
+      "step": 88640
+    },
+    {
+      "epoch": 535.17,
+      "learning_rate": 6.374103226793243e-06,
+      "loss": 0.0097,
+      "step": 88660
+    },
+    {
+      "epoch": 535.29,
+      "learning_rate": 6.352193291405883e-06,
+      "loss": 0.0097,
+      "step": 88680
+    },
+    {
+      "epoch": 535.41,
+      "learning_rate": 6.330318522268008e-06,
+      "loss": 0.0099,
+      "step": 88700
+    },
+    {
+      "epoch": 535.53,
+      "learning_rate": 6.308478937003731e-06,
+      "loss": 0.01,
+      "step": 88720
+    },
+    {
+      "epoch": 535.65,
+      "learning_rate": 6.28667455320881e-06,
+      "loss": 0.0099,
+      "step": 88740
+    },
+    {
+      "epoch": 535.77,
+      "learning_rate": 6.264905388450659e-06,
+      "loss": 0.0101,
+      "step": 88760
+    },
+    {
+      "epoch": 535.9,
+      "learning_rate": 6.2431714602682714e-06,
+      "loss": 0.0101,
+      "step": 88780
+    },
+    {
+      "epoch": 536.02,
+      "learning_rate": 6.221472786172294e-06,
+      "loss": 0.0101,
+      "step": 88800
+    },
+    {
+      "epoch": 536.14,
+      "learning_rate": 6.199809383644956e-06,
+      "loss": 0.0096,
+      "step": 88820
+    },
+    {
+      "epoch": 536.26,
+      "learning_rate": 6.178181270140077e-06,
+      "loss": 0.0098,
+      "step": 88840
+    },
+    {
+      "epoch": 536.38,
+      "learning_rate": 6.156588463083035e-06,
+      "loss": 0.0098,
+      "step": 88860
+    },
+    {
+      "epoch": 536.5,
+      "learning_rate": 6.135030979870743e-06,
+      "loss": 0.0099,
+      "step": 88880
+    },
+    {
+      "epoch": 536.62,
+      "learning_rate": 6.113508837871718e-06,
+      "loss": 0.01,
+      "step": 88900
+    },
+    {
+      "epoch": 536.74,
+      "learning_rate": 6.092022054425928e-06,
+      "loss": 0.0101,
+      "step": 88920
+    },
+    {
+      "epoch": 536.86,
+      "learning_rate": 6.070570646844886e-06,
+      "loss": 0.01,
+      "step": 88940
+    },
+    {
+      "epoch": 536.98,
+      "learning_rate": 6.049154632411624e-06,
+      "loss": 0.0102,
+      "step": 88960
+    },
+    {
+      "epoch": 537.1,
+      "learning_rate": 6.027774028380623e-06,
+      "loss": 0.0097,
+      "step": 88980
+    },
+    {
+      "epoch": 537.22,
+      "learning_rate": 6.0064288519778635e-06,
+      "loss": 0.0097,
+      "step": 89000
+    },
+    {
+      "epoch": 537.22,
+      "eval_accuracy": 0.0012664104869061848,
+      "eval_loss": 23.905231475830078,
+      "eval_runtime": 18.1439,
+      "eval_samples_per_second": 112.435,
+      "eval_steps_per_second": 2.37,
+      "step": 89000
+    },
+    {
+      "epoch": 537.34,
+      "learning_rate": 5.985119120400745e-06,
+      "loss": 0.0099,
+      "step": 89020
+    },
+    {
+      "epoch": 537.46,
+      "learning_rate": 5.963844850818151e-06,
+      "loss": 0.0099,
+      "step": 89040
+    },
+    {
+      "epoch": 537.59,
+      "learning_rate": 5.94260606037037e-06,
+      "loss": 0.01,
+      "step": 89060
+    },
+    {
+      "epoch": 537.71,
+      "learning_rate": 5.921402766169126e-06,
+      "loss": 0.01,
+      "step": 89080
+    },
+    {
+      "epoch": 537.83,
+      "learning_rate": 5.900234985297498e-06,
+      "loss": 0.0101,
+      "step": 89100
+    },
+    {
+      "epoch": 537.95,
+      "learning_rate": 5.879102734810016e-06,
+      "loss": 0.0102,
+      "step": 89120
+    },
+    {
+      "epoch": 538.07,
+      "learning_rate": 5.858006031732549e-06,
+      "loss": 0.0098,
+      "step": 89140
+    },
+    {
+      "epoch": 538.19,
+      "learning_rate": 5.836944893062318e-06,
+      "loss": 0.0097,
+      "step": 89160
+    },
+    {
+      "epoch": 538.31,
+      "learning_rate": 5.8159193357679e-06,
+      "loss": 0.0097,
+      "step": 89180
+    },
+    {
+      "epoch": 538.43,
+      "learning_rate": 5.794929376789215e-06,
+      "loss": 0.0099,
+      "step": 89200
+    },
+    {
+      "epoch": 538.55,
+      "learning_rate": 5.773975033037499e-06,
+      "loss": 0.0099,
+      "step": 89220
+    },
+    {
+      "epoch": 538.67,
+      "learning_rate": 5.753056321395267e-06,
+      "loss": 0.01,
+      "step": 89240
+    },
+    {
+      "epoch": 538.79,
+      "learning_rate": 5.732173258716366e-06,
+      "loss": 0.0102,
+      "step": 89260
+    },
+    {
+      "epoch": 538.91,
+      "learning_rate": 5.711325861825906e-06,
+      "loss": 0.01,
+      "step": 89280
+    },
+    {
+      "epoch": 539.03,
+      "learning_rate": 5.690514147520243e-06,
+      "loss": 0.01,
+      "step": 89300
+    },
+    {
+      "epoch": 539.15,
+      "learning_rate": 5.66973813256701e-06,
+      "loss": 0.0096,
+      "step": 89320
+    },
+    {
+      "epoch": 539.28,
+      "learning_rate": 5.6489978337050555e-06,
+      "loss": 0.0098,
+      "step": 89340
+    },
+    {
+      "epoch": 539.4,
+      "learning_rate": 5.6282932676444856e-06,
+      "loss": 0.0098,
+      "step": 89360
+    },
+    {
+      "epoch": 539.52,
+      "learning_rate": 5.607624451066568e-06,
+      "loss": 0.0098,
+      "step": 89380
+    },
+    {
+      "epoch": 539.64,
+      "learning_rate": 5.586991400623798e-06,
+      "loss": 0.01,
+      "step": 89400
+    },
+    {
+      "epoch": 539.76,
+      "learning_rate": 5.566394132939884e-06,
+      "loss": 0.01,
+      "step": 89420
+    },
+    {
+      "epoch": 539.88,
+      "learning_rate": 5.54583266460964e-06,
+      "loss": 0.0101,
+      "step": 89440
+    },
+    {
+      "epoch": 540.0,
+      "learning_rate": 5.525307012199077e-06,
+      "loss": 0.0101,
+      "step": 89460
+    },
+    {
+      "epoch": 540.12,
+      "learning_rate": 5.504817192245343e-06,
+      "loss": 0.0096,
+      "step": 89480
+    },
+    {
+      "epoch": 540.24,
+      "learning_rate": 5.484363221256733e-06,
+      "loss": 0.0097,
+      "step": 89500
+    },
+    {
+      "epoch": 540.36,
+      "learning_rate": 5.463945115712609e-06,
+      "loss": 0.0098,
+      "step": 89520
+    },
+    {
+      "epoch": 540.48,
+      "learning_rate": 5.443562892063497e-06,
+      "loss": 0.0099,
+      "step": 89540
+    },
+    {
+      "epoch": 540.6,
+      "learning_rate": 5.423216566730971e-06,
+      "loss": 0.0099,
+      "step": 89560
+    },
+    {
+      "epoch": 540.72,
+      "learning_rate": 5.4029061561077064e-06,
+      "loss": 0.01,
+      "step": 89580
+    },
+    {
+      "epoch": 540.85,
+      "learning_rate": 5.382631676557437e-06,
+      "loss": 0.0101,
+      "step": 89600
+    },
+    {
+      "epoch": 540.97,
+      "learning_rate": 5.3623931444149235e-06,
+      "loss": 0.0101,
+      "step": 89620
+    },
+    {
+      "epoch": 541.09,
+      "learning_rate": 5.342190575986022e-06,
+      "loss": 0.0097,
+      "step": 89640
+    },
+    {
+      "epoch": 541.21,
+      "learning_rate": 5.322023987547547e-06,
+      "loss": 0.0096,
+      "step": 89660
+    },
+    {
+      "epoch": 541.33,
+      "learning_rate": 5.301893395347363e-06,
+      "loss": 0.0098,
+      "step": 89680
+    },
+    {
+      "epoch": 541.45,
+      "learning_rate": 5.281798815604327e-06,
+      "loss": 0.0099,
+      "step": 89700
+    },
+    {
+      "epoch": 541.57,
+      "learning_rate": 5.261740264508275e-06,
+      "loss": 0.0099,
+      "step": 89720
+    },
+    {
+      "epoch": 541.69,
+      "learning_rate": 5.2417177582200325e-06,
+      "loss": 0.01,
+      "step": 89740
+    },
+    {
+      "epoch": 541.81,
+      "learning_rate": 5.2217313128713415e-06,
+      "loss": 0.0099,
+      "step": 89760
+    },
+    {
+      "epoch": 541.93,
+      "learning_rate": 5.20178094456496e-06,
+      "loss": 0.01,
+      "step": 89780
+    },
+    {
+      "epoch": 542.05,
+      "learning_rate": 5.1818666693745076e-06,
+      "loss": 0.0098,
+      "step": 89800
+    },
+    {
+      "epoch": 542.17,
+      "learning_rate": 5.161988503344561e-06,
+      "loss": 0.0096,
+      "step": 89820
+    },
+    {
+      "epoch": 542.29,
+      "learning_rate": 5.1421464624906155e-06,
+      "loss": 0.0097,
+      "step": 89840
+    },
+    {
+      "epoch": 542.41,
+      "learning_rate": 5.122340562799027e-06,
+      "loss": 0.0098,
+      "step": 89860
+    },
+    {
+      "epoch": 542.54,
+      "learning_rate": 5.1025708202270765e-06,
+      "loss": 0.0098,
+      "step": 89880
+    },
+    {
+      "epoch": 542.66,
+      "learning_rate": 5.0828372507028545e-06,
+      "loss": 0.01,
+      "step": 89900
+    },
+    {
+      "epoch": 542.78,
+      "learning_rate": 5.063139870125367e-06,
+      "loss": 0.0101,
+      "step": 89920
+    },
+    {
+      "epoch": 542.9,
+      "learning_rate": 5.043478694364423e-06,
+      "loss": 0.01,
+      "step": 89940
+    },
+    {
+      "epoch": 543.02,
+      "learning_rate": 5.023853739260681e-06,
+      "loss": 0.0099,
+      "step": 89960
+    },
+    {
+      "epoch": 543.14,
+      "learning_rate": 5.0042650206256146e-06,
+      "loss": 0.0096,
+      "step": 89980
+    },
+    {
+      "epoch": 543.26,
+      "learning_rate": 4.9847125542415055e-06,
+      "loss": 0.0097,
+      "step": 90000
+    },
+    {
+      "epoch": 543.26,
+      "eval_accuracy": 0.001261471288439312,
+      "eval_loss": 23.952375411987305,
+      "eval_runtime": 18.3553,
+      "eval_samples_per_second": 111.139,
+      "eval_steps_per_second": 2.343,
+      "step": 90000
+    },
+    {
+      "epoch": 543.38,
+      "learning_rate": 4.965196355861423e-06,
+      "loss": 0.0098,
+      "step": 90020
+    },
+    {
+      "epoch": 543.5,
+      "learning_rate": 4.9457164412092025e-06,
+      "loss": 0.0098,
+      "step": 90040
+    },
+    {
+      "epoch": 543.62,
+      "learning_rate": 4.926272825979466e-06,
+      "loss": 0.0099,
+      "step": 90060
+    },
+    {
+      "epoch": 543.74,
+      "learning_rate": 4.906865525837589e-06,
+      "loss": 0.01,
+      "step": 90080
+    },
+    {
+      "epoch": 543.86,
+      "learning_rate": 4.887494556419675e-06,
+      "loss": 0.0099,
+      "step": 90100
+    },
+    {
+      "epoch": 543.98,
+      "learning_rate": 4.868159933332572e-06,
+      "loss": 0.01,
+      "step": 90120
+    },
+    {
+      "epoch": 544.1,
+      "learning_rate": 4.8488616721538205e-06,
+      "loss": 0.0096,
+      "step": 90140
+    },
+    {
+      "epoch": 544.23,
+      "learning_rate": 4.82959978843171e-06,
+      "loss": 0.0096,
+      "step": 90160
+    },
+    {
+      "epoch": 544.35,
+      "learning_rate": 4.810374297685161e-06,
+      "loss": 0.0097,
+      "step": 90180
+    },
+    {
+      "epoch": 544.47,
+      "learning_rate": 4.791185215403821e-06,
+      "loss": 0.0098,
+      "step": 90200
+    },
+    {
+      "epoch": 544.59,
+      "learning_rate": 4.772032557047984e-06,
+      "loss": 0.0099,
+      "step": 90220
+    },
+    {
+      "epoch": 544.71,
+      "learning_rate": 4.7529163380486074e-06,
+      "loss": 0.01,
+      "step": 90240
+    },
+    {
+      "epoch": 544.83,
+      "learning_rate": 4.7338365738072655e-06,
+      "loss": 0.0099,
+      "step": 90260
+    },
+    {
+      "epoch": 544.95,
+      "learning_rate": 4.714793279696189e-06,
+      "loss": 0.01,
+      "step": 90280
+    },
+    {
+      "epoch": 545.07,
+      "learning_rate": 4.695786471058233e-06,
+      "loss": 0.0097,
+      "step": 90300
+    },
+    {
+      "epoch": 545.19,
+      "learning_rate": 4.676816163206815e-06,
+      "loss": 0.0096,
+      "step": 90320
+    },
+    {
+      "epoch": 545.31,
+      "learning_rate": 4.657882371425987e-06,
+      "loss": 0.0096,
+      "step": 90340
+    },
+    {
+      "epoch": 545.43,
+      "learning_rate": 4.638985110970339e-06,
+      "loss": 0.0097,
+      "step": 90360
+    },
+    {
+      "epoch": 545.55,
+      "learning_rate": 4.62012439706509e-06,
+      "loss": 0.0099,
+      "step": 90380
+    },
+    {
+      "epoch": 545.67,
+      "learning_rate": 4.601300244905943e-06,
+      "loss": 0.0099,
+      "step": 90400
+    },
+    {
+      "epoch": 545.79,
+      "learning_rate": 4.582512669659189e-06,
+      "loss": 0.0099,
+      "step": 90420
+    },
+    {
+      "epoch": 545.92,
+      "learning_rate": 4.563761686461638e-06,
+      "loss": 0.01,
+      "step": 90440
+    },
+    {
+      "epoch": 546.04,
+      "learning_rate": 4.545047310420619e-06,
+      "loss": 0.0098,
+      "step": 90460
+    },
+    {
+      "epoch": 546.16,
+      "learning_rate": 4.52636955661398e-06,
+      "loss": 0.0095,
+      "step": 90480
+    },
+    {
+      "epoch": 546.28,
+      "learning_rate": 4.507728440090014e-06,
+      "loss": 0.0096,
+      "step": 90500
+    },
+    {
+      "epoch": 546.4,
+      "learning_rate": 4.489123975867576e-06,
+      "loss": 0.0097,
+      "step": 90520
+    },
+    {
+      "epoch": 546.52,
+      "learning_rate": 4.47055617893592e-06,
+      "loss": 0.0098,
+      "step": 90540
+    },
+    {
+      "epoch": 546.64,
+      "learning_rate": 4.4520250642547835e-06,
+      "loss": 0.01,
+      "step": 90560
+    },
+    {
+      "epoch": 546.76,
+      "learning_rate": 4.433530646754364e-06,
+      "loss": 0.0099,
+      "step": 90580
+    },
+    {
+      "epoch": 546.88,
+      "learning_rate": 4.415072941335269e-06,
+      "loss": 0.01,
+      "step": 90600
+    },
+    {
+      "epoch": 547.0,
+      "learning_rate": 4.396651962868553e-06,
+      "loss": 0.01,
+      "step": 90620
+    },
+    {
+      "epoch": 547.12,
+      "learning_rate": 4.378267726195645e-06,
+      "loss": 0.0095,
+      "step": 90640
+    },
+    {
+      "epoch": 547.24,
+      "learning_rate": 4.359920246128402e-06,
+      "loss": 0.0096,
+      "step": 90660
+    },
+    {
+      "epoch": 547.36,
+      "learning_rate": 4.34160953744906e-06,
+      "loss": 0.0097,
+      "step": 90680
+    },
+    {
+      "epoch": 547.48,
+      "learning_rate": 4.323335614910224e-06,
+      "loss": 0.0097,
+      "step": 90700
+    },
+    {
+      "epoch": 547.61,
+      "learning_rate": 4.30509849323486e-06,
+      "loss": 0.0099,
+      "step": 90720
+    },
+    {
+      "epoch": 547.73,
+      "learning_rate": 4.286898187116295e-06,
+      "loss": 0.0099,
+      "step": 90740
+    },
+    {
+      "epoch": 547.85,
+      "learning_rate": 4.268734711218192e-06,
+      "loss": 0.0099,
+      "step": 90760
+    },
+    {
+      "epoch": 547.97,
+      "learning_rate": 4.250608080174512e-06,
+      "loss": 0.01,
+      "step": 90780
+    },
+    {
+      "epoch": 548.09,
+      "learning_rate": 4.232518308589573e-06,
+      "loss": 0.0096,
+      "step": 90800
+    },
+    {
+      "epoch": 548.21,
+      "learning_rate": 4.214465411037971e-06,
+      "loss": 0.0095,
+      "step": 90820
+    },
+    {
+      "epoch": 548.33,
+      "learning_rate": 4.196449402064606e-06,
+      "loss": 0.0097,
+      "step": 90840
+    },
+    {
+      "epoch": 548.45,
+      "learning_rate": 4.17847029618465e-06,
+      "loss": 0.0097,
+      "step": 90860
+    },
+    {
+      "epoch": 548.57,
+      "learning_rate": 4.160528107883527e-06,
+      "loss": 0.0098,
+      "step": 90880
+    },
+    {
+      "epoch": 548.69,
+      "learning_rate": 4.142622851616962e-06,
+      "loss": 0.0099,
+      "step": 90900
+    },
+    {
+      "epoch": 548.81,
+      "learning_rate": 4.1247545418108715e-06,
+      "loss": 0.0099,
+      "step": 90920
+    },
+    {
+      "epoch": 548.93,
+      "learning_rate": 4.106923192861445e-06,
+      "loss": 0.01,
+      "step": 90940
+    },
+    {
+      "epoch": 549.05,
+      "learning_rate": 4.089128819135069e-06,
+      "loss": 0.0097,
+      "step": 90960
+    },
+    {
+      "epoch": 549.18,
+      "learning_rate": 4.071371434968352e-06,
+      "loss": 0.0095,
+      "step": 90980
+    },
+    {
+      "epoch": 549.3,
+      "learning_rate": 4.053651054668112e-06,
+      "loss": 0.0096,
+      "step": 91000
+    },
+    {
+      "epoch": 549.3,
+      "eval_accuracy": 0.0012683861662929339,
+      "eval_loss": 23.982318878173828,
+      "eval_runtime": 18.2415,
+      "eval_samples_per_second": 111.833,
+      "eval_steps_per_second": 2.357,
+      "step": 91000
+    },
+    {
+      "epoch": 549.42,
+      "learning_rate": 4.035967692511311e-06,
+      "loss": 0.0097,
+      "step": 91020
+    },
+    {
+      "epoch": 549.54,
+      "learning_rate": 4.018321362745142e-06,
+      "loss": 0.0098,
+      "step": 91040
+    },
+    {
+      "epoch": 549.66,
+      "learning_rate": 4.000712079586916e-06,
+      "loss": 0.0098,
+      "step": 91060
+    },
+    {
+      "epoch": 549.78,
+      "learning_rate": 3.983139857224122e-06,
+      "loss": 0.0099,
+      "step": 91080
+    },
+    {
+      "epoch": 549.9,
+      "learning_rate": 3.96560470981438e-06,
+      "loss": 0.0098,
+      "step": 91100
+    },
+    {
+      "epoch": 550.02,
+      "learning_rate": 3.948106651485439e-06,
+      "loss": 0.0099,
+      "step": 91120
+    },
+    {
+      "epoch": 550.14,
+      "learning_rate": 3.9306456963351835e-06,
+      "loss": 0.0095,
+      "step": 91140
+    },
+    {
+      "epoch": 550.26,
+      "learning_rate": 3.9132218584315704e-06,
+      "loss": 0.0095,
+      "step": 91160
+    },
+    {
+      "epoch": 550.38,
+      "learning_rate": 3.895835151812677e-06,
+      "loss": 0.0097,
+      "step": 91180
+    },
+    {
+      "epoch": 550.5,
+      "learning_rate": 3.8784855904866635e-06,
+      "loss": 0.0097,
+      "step": 91200
+    },
+    {
+      "epoch": 550.62,
+      "learning_rate": 3.86117318843176e-06,
+      "loss": 0.0098,
+      "step": 91220
+    },
+    {
+      "epoch": 550.74,
+      "learning_rate": 3.843897959596238e-06,
+      "loss": 0.0098,
+      "step": 91240
+    },
+    {
+      "epoch": 550.87,
+      "learning_rate": 3.826659917898457e-06,
+      "loss": 0.0099,
+      "step": 91260
+    },
+    {
+      "epoch": 550.99,
+      "learning_rate": 3.8094590772268013e-06,
+      "loss": 0.0099,
+      "step": 91280
+    },
+    {
+      "epoch": 551.11,
+      "learning_rate": 3.7922954514396535e-06,
+      "loss": 0.0095,
+      "step": 91300
+    },
+    {
+      "epoch": 551.23,
+      "learning_rate": 3.775169054365446e-06,
+      "loss": 0.0095,
+      "step": 91320
+    },
+    {
+      "epoch": 551.35,
+      "learning_rate": 3.758079899802619e-06,
+      "loss": 0.0096,
+      "step": 91340
+    },
+    {
+      "epoch": 551.47,
+      "learning_rate": 3.7410280015195866e-06,
+      "loss": 0.0097,
+      "step": 91360
+    },
+    {
+      "epoch": 551.59,
+      "learning_rate": 3.724013373254753e-06,
+      "loss": 0.0098,
+      "step": 91380
+    },
+    {
+      "epoch": 551.71,
+      "learning_rate": 3.707036028716493e-06,
+      "loss": 0.0098,
+      "step": 91400
+    },
+    {
+      "epoch": 551.83,
+      "learning_rate": 3.6900959815831692e-06,
+      "loss": 0.0098,
+      "step": 91420
+    },
+    {
+      "epoch": 551.95,
+      "learning_rate": 3.6731932455030425e-06,
+      "loss": 0.0099,
+      "step": 91440
+    },
+    {
+      "epoch": 552.07,
+      "learning_rate": 3.656327834094364e-06,
+      "loss": 0.0096,
+      "step": 91460
+    },
+    {
+      "epoch": 552.19,
+      "learning_rate": 3.6394997609452752e-06,
+      "loss": 0.0094,
+      "step": 91480
+    },
+    {
+      "epoch": 552.31,
+      "learning_rate": 3.6227090396138697e-06,
+      "loss": 0.0096,
+      "step": 91500
+    },
+    {
+      "epoch": 552.43,
+      "learning_rate": 3.605955683628104e-06,
+      "loss": 0.0097,
+      "step": 91520
+    },
+    {
+      "epoch": 552.56,
+      "learning_rate": 3.5892397064858697e-06,
+      "loss": 0.0098,
+      "step": 91540
+    },
+    {
+      "epoch": 552.68,
+      "learning_rate": 3.572561121654927e-06,
+      "loss": 0.0097,
+      "step": 91560
+    },
+    {
+      "epoch": 552.8,
+      "learning_rate": 3.555919942572905e-06,
+      "loss": 0.0098,
+      "step": 91580
+    },
+    {
+      "epoch": 552.92,
+      "learning_rate": 3.5393161826473066e-06,
+      "loss": 0.0098,
+      "step": 91600
+    },
+    {
+      "epoch": 553.04,
+      "learning_rate": 3.5227498552554807e-06,
+      "loss": 0.0098,
+      "step": 91620
+    },
+    {
+      "epoch": 553.16,
+      "learning_rate": 3.5062209737446226e-06,
+      "loss": 0.0094,
+      "step": 91640
+    },
+    {
+      "epoch": 553.28,
+      "learning_rate": 3.489729551431742e-06,
+      "loss": 0.0095,
+      "step": 91660
+    },
+    {
+      "epoch": 553.4,
+      "learning_rate": 3.4732756016036814e-06,
+      "loss": 0.0097,
+      "step": 91680
+    },
+    {
+      "epoch": 553.52,
+      "learning_rate": 3.4568591375170935e-06,
+      "loss": 0.0097,
+      "step": 91700
+    },
+    {
+      "epoch": 553.64,
+      "learning_rate": 3.440480172398436e-06,
+      "loss": 0.0098,
+      "step": 91720
+    },
+    {
+      "epoch": 553.76,
+      "learning_rate": 3.4241387194439367e-06,
+      "loss": 0.0098,
+      "step": 91740
+    },
+    {
+      "epoch": 553.88,
+      "learning_rate": 3.407834791819603e-06,
+      "loss": 0.0098,
+      "step": 91760
+    },
+    {
+      "epoch": 554.0,
+      "learning_rate": 3.3915684026612327e-06,
+      "loss": 0.0099,
+      "step": 91780
+    },
+    {
+      "epoch": 554.12,
+      "learning_rate": 3.3753395650743423e-06,
+      "loss": 0.0094,
+      "step": 91800
+    },
+    {
+      "epoch": 554.25,
+      "learning_rate": 3.3591482921342234e-06,
+      "loss": 0.0095,
+      "step": 91820
+    },
+    {
+      "epoch": 554.37,
+      "learning_rate": 3.3429945968858956e-06,
+      "loss": 0.0097,
+      "step": 91840
+    },
+    {
+      "epoch": 554.49,
+      "learning_rate": 3.326878492344088e-06,
+      "loss": 0.0096,
+      "step": 91860
+    },
+    {
+      "epoch": 554.61,
+      "learning_rate": 3.3107999914932686e-06,
+      "loss": 0.0097,
+      "step": 91880
+    },
+    {
+      "epoch": 554.73,
+      "learning_rate": 3.2947591072875814e-06,
+      "loss": 0.0098,
+      "step": 91900
+    },
+    {
+      "epoch": 554.85,
+      "learning_rate": 3.2787558526508887e-06,
+      "loss": 0.0097,
+      "step": 91920
+    },
+    {
+      "epoch": 554.97,
+      "learning_rate": 3.2627902404767153e-06,
+      "loss": 0.0099,
+      "step": 91940
+    },
+    {
+      "epoch": 555.09,
+      "learning_rate": 3.246862283628277e-06,
+      "loss": 0.0095,
+      "step": 91960
+    },
+    {
+      "epoch": 555.21,
+      "learning_rate": 3.230971994938414e-06,
+      "loss": 0.0095,
+      "step": 91980
+    },
+    {
+      "epoch": 555.33,
+      "learning_rate": 3.215119387209681e-06,
+      "loss": 0.0096,
+      "step": 92000
+    },
+    {
+      "epoch": 555.33,
+      "eval_accuracy": 0.0012604834487459375,
+      "eval_loss": 24.008394241333008,
+      "eval_runtime": 18.2132,
+      "eval_samples_per_second": 112.007,
+      "eval_steps_per_second": 2.361,
+      "step": 92000
+    },
+    {
+      "epoch": 555.45,
+      "learning_rate": 3.1993044732142206e-06,
+      "loss": 0.0096,
+      "step": 92020
+    },
+    {
+      "epoch": 555.57,
+      "learning_rate": 3.1835272656938197e-06,
+      "loss": 0.0097,
+      "step": 92040
+    },
+    {
+      "epoch": 555.69,
+      "learning_rate": 3.1677877773598984e-06,
+      "loss": 0.0097,
+      "step": 92060
+    },
+    {
+      "epoch": 555.81,
+      "learning_rate": 3.152086020893469e-06,
+      "loss": 0.0098,
+      "step": 92080
+    },
+    {
+      "epoch": 555.94,
+      "learning_rate": 3.1364220089451667e-06,
+      "loss": 0.0098,
+      "step": 92100
+    },
+    {
+      "epoch": 556.06,
+      "learning_rate": 3.120795754135175e-06,
+      "loss": 0.0096,
+      "step": 92120
+    },
+    {
+      "epoch": 556.18,
+      "learning_rate": 3.1052072690533173e-06,
+      "loss": 0.0094,
+      "step": 92140
+    },
+    {
+      "epoch": 556.3,
+      "learning_rate": 3.089656566258942e-06,
+      "loss": 0.0095,
+      "step": 92160
+    },
+    {
+      "epoch": 556.42,
+      "learning_rate": 3.0741436582809703e-06,
+      "loss": 0.0096,
+      "step": 92180
+    },
+    {
+      "epoch": 556.54,
+      "learning_rate": 3.0586685576178608e-06,
+      "loss": 0.0097,
+      "step": 92200
+    },
+    {
+      "epoch": 556.66,
+      "learning_rate": 3.0432312767376402e-06,
+      "loss": 0.0097,
+      "step": 92220
+    },
+    {
+      "epoch": 556.78,
+      "learning_rate": 3.0278318280778426e-06,
+      "loss": 0.0097,
+      "step": 92240
+    },
+    {
+      "epoch": 556.9,
+      "learning_rate": 3.012470224045516e-06,
+      "loss": 0.0098,
+      "step": 92260
+    },
+    {
+      "epoch": 557.02,
+      "learning_rate": 2.997146477017221e-06,
+      "loss": 0.0098,
+      "step": 92280
+    },
+    {
+      "epoch": 557.14,
+      "learning_rate": 2.9818605993390593e-06,
+      "loss": 0.0094,
+      "step": 92300
+    },
+    {
+      "epoch": 557.26,
+      "learning_rate": 2.9666126033265517e-06,
+      "loss": 0.0095,
+      "step": 92320
+    },
+    {
+      "epoch": 557.38,
+      "learning_rate": 2.951402501264744e-06,
+      "loss": 0.0096,
+      "step": 92340
+    },
+    {
+      "epoch": 557.51,
+      "learning_rate": 2.9362303054081387e-06,
+      "loss": 0.0096,
+      "step": 92360
+    },
+    {
+      "epoch": 557.63,
+      "learning_rate": 2.9210960279807133e-06,
+      "loss": 0.0097,
+      "step": 92380
+    },
+    {
+      "epoch": 557.75,
+      "learning_rate": 2.9059996811758637e-06,
+      "loss": 0.0096,
+      "step": 92400
+    },
+    {
+      "epoch": 557.87,
+      "learning_rate": 2.8909412771564502e-06,
+      "loss": 0.0098,
+      "step": 92420
+    },
+    {
+      "epoch": 557.99,
+      "learning_rate": 2.875920828054757e-06,
+      "loss": 0.0098,
+      "step": 92440
+    },
+    {
+      "epoch": 558.11,
+      "learning_rate": 2.8609383459724913e-06,
+      "loss": 0.0094,
+      "step": 92460
+    },
+    {
+      "epoch": 558.23,
+      "learning_rate": 2.8459938429807763e-06,
+      "loss": 0.0094,
+      "step": 92480
+    },
+    {
+      "epoch": 558.35,
+      "learning_rate": 2.8310873311201027e-06,
+      "loss": 0.0096,
+      "step": 92500
+    },
+    {
+      "epoch": 558.47,
+      "learning_rate": 2.816218822400407e-06,
+      "loss": 0.0096,
+      "step": 92520
+    },
+    {
+      "epoch": 558.59,
+      "learning_rate": 2.8013883288009522e-06,
+      "loss": 0.0097,
+      "step": 92540
+    },
+    {
+      "epoch": 558.71,
+      "learning_rate": 2.7865958622704192e-06,
+      "loss": 0.0097,
+      "step": 92560
+    },
+    {
+      "epoch": 558.83,
+      "learning_rate": 2.7718414347268183e-06,
+      "loss": 0.0097,
+      "step": 92580
+    },
+    {
+      "epoch": 558.95,
+      "learning_rate": 2.757125058057536e-06,
+      "loss": 0.0097,
+      "step": 92600
+    },
+    {
+      "epoch": 559.07,
+      "learning_rate": 2.7424467441192927e-06,
+      "loss": 0.0095,
+      "step": 92620
+    },
+    {
+      "epoch": 559.2,
+      "learning_rate": 2.727806504738112e-06,
+      "loss": 0.0094,
+      "step": 92640
+    },
+    {
+      "epoch": 559.32,
+      "learning_rate": 2.713204351709414e-06,
+      "loss": 0.0094,
+      "step": 92660
+    },
+    {
+      "epoch": 559.44,
+      "learning_rate": 2.6986402967978554e-06,
+      "loss": 0.0096,
+      "step": 92680
+    },
+    {
+      "epoch": 559.56,
+      "learning_rate": 2.6841143517374543e-06,
+      "loss": 0.0096,
+      "step": 92700
+    },
+    {
+      "epoch": 559.68,
+      "learning_rate": 2.669626528231478e-06,
+      "loss": 0.0097,
+      "step": 92720
+    },
+    {
+      "epoch": 559.8,
+      "learning_rate": 2.6551768379525277e-06,
+      "loss": 0.0097,
+      "step": 92740
+    },
+    {
+      "epoch": 559.92,
+      "learning_rate": 2.640765292542452e-06,
+      "loss": 0.0098,
+      "step": 92760
+    },
+    {
+      "epoch": 560.04,
+      "learning_rate": 2.6263919036123687e-06,
+      "loss": 0.0097,
+      "step": 92780
+    },
+    {
+      "epoch": 560.16,
+      "learning_rate": 2.612056682742653e-06,
+      "loss": 0.0094,
+      "step": 92800
+    },
+    {
+      "epoch": 560.28,
+      "learning_rate": 2.5977596414829388e-06,
+      "loss": 0.0095,
+      "step": 92820
+    },
+    {
+      "epoch": 560.4,
+      "learning_rate": 2.583500791352106e-06,
+      "loss": 0.0095,
+      "step": 92840
+    },
+    {
+      "epoch": 560.52,
+      "learning_rate": 2.5692801438382265e-06,
+      "loss": 0.0096,
+      "step": 92860
+    },
+    {
+      "epoch": 560.64,
+      "learning_rate": 2.555097710398635e-06,
+      "loss": 0.0096,
+      "step": 92880
+    },
+    {
+      "epoch": 560.76,
+      "learning_rate": 2.5409535024598694e-06,
+      "loss": 0.0097,
+      "step": 92900
+    },
+    {
+      "epoch": 560.89,
+      "learning_rate": 2.5268475314176476e-06,
+      "loss": 0.0097,
+      "step": 92920
+    },
+    {
+      "epoch": 561.01,
+      "learning_rate": 2.512779808636906e-06,
+      "loss": 0.0096,
+      "step": 92940
+    },
+    {
+      "epoch": 561.13,
+      "learning_rate": 2.4987503454517446e-06,
+      "loss": 0.0094,
+      "step": 92960
+    },
+    {
+      "epoch": 561.25,
+      "learning_rate": 2.4847591531654614e-06,
+      "loss": 0.0094,
+      "step": 92980
+    },
+    {
+      "epoch": 561.37,
+      "learning_rate": 2.470806243050494e-06,
+      "loss": 0.0095,
+      "step": 93000
+    },
+    {
+      "epoch": 561.37,
+      "eval_accuracy": 0.0012673983265995593,
+      "eval_loss": 24.03643035888672,
+      "eval_runtime": 18.3059,
+      "eval_samples_per_second": 111.439,
+      "eval_steps_per_second": 2.349,
+      "step": 93000
+    },
+    {
+      "epoch": 561.49,
+      "learning_rate": 2.4568916263484508e-06,
+      "loss": 0.0096,
+      "step": 93020
+    },
+    {
+      "epoch": 561.61,
+      "learning_rate": 2.4430153142701082e-06,
+      "loss": 0.0096,
+      "step": 93040
+    },
+    {
+      "epoch": 561.73,
+      "learning_rate": 2.4291773179953346e-06,
+      "loss": 0.0096,
+      "step": 93060
+    },
+    {
+      "epoch": 561.85,
+      "learning_rate": 2.4153776486731617e-06,
+      "loss": 0.0096,
+      "step": 93080
+    },
+    {
+      "epoch": 561.97,
+      "learning_rate": 2.40161631742174e-06,
+      "loss": 0.0097,
+      "step": 93100
+    },
+    {
+      "epoch": 562.09,
+      "learning_rate": 2.3878933353283283e-06,
+      "loss": 0.0095,
+      "step": 93120
+    },
+    {
+      "epoch": 562.21,
+      "learning_rate": 2.374208713449272e-06,
+      "loss": 0.0094,
+      "step": 93140
+    },
+    {
+      "epoch": 562.33,
+      "learning_rate": 2.3605624628100175e-06,
+      "loss": 0.0095,
+      "step": 93160
+    },
+    {
+      "epoch": 562.45,
+      "learning_rate": 2.346954594405132e-06,
+      "loss": 0.0096,
+      "step": 93180
+    },
+    {
+      "epoch": 562.58,
+      "learning_rate": 2.333385119198206e-06,
+      "loss": 0.0096,
+      "step": 93200
+    },
+    {
+      "epoch": 562.7,
+      "learning_rate": 2.319854048121939e-06,
+      "loss": 0.0096,
+      "step": 93220
+    },
+    {
+      "epoch": 562.82,
+      "learning_rate": 2.306361392078038e-06,
+      "loss": 0.0096,
+      "step": 93240
+    },
+    {
+      "epoch": 562.94,
+      "learning_rate": 2.2929071619373298e-06,
+      "loss": 0.0097,
+      "step": 93260
+    },
+    {
+      "epoch": 563.06,
+      "learning_rate": 2.2794913685396214e-06,
+      "loss": 0.0095,
+      "step": 93280
+    },
+    {
+      "epoch": 563.18,
+      "learning_rate": 2.266114022693777e-06,
+      "loss": 0.0094,
+      "step": 93300
+    },
+    {
+      "epoch": 563.3,
+      "learning_rate": 2.2527751351776927e-06,
+      "loss": 0.0095,
+      "step": 93320
+    },
+    {
+      "epoch": 563.42,
+      "learning_rate": 2.23947471673826e-06,
+      "loss": 0.0095,
+      "step": 93340
+    },
+    {
+      "epoch": 563.54,
+      "learning_rate": 2.2262127780913965e-06,
+      "loss": 0.0096,
+      "step": 93360
+    },
+    {
+      "epoch": 563.66,
+      "learning_rate": 2.2129893299219886e-06,
+      "loss": 0.0096,
+      "step": 93380
+    },
+    {
+      "epoch": 563.78,
+      "learning_rate": 2.1998043828839523e-06,
+      "loss": 0.0096,
+      "step": 93400
+    },
+    {
+      "epoch": 563.9,
+      "learning_rate": 2.1866579476001404e-06,
+      "loss": 0.0096,
+      "step": 93420
+    },
+    {
+      "epoch": 564.02,
+      "learning_rate": 2.173550034662408e-06,
+      "loss": 0.0097,
+      "step": 93440
+    },
+    {
+      "epoch": 564.14,
+      "learning_rate": 2.160480654631569e-06,
+      "loss": 0.0093,
+      "step": 93460
+    },
+    {
+      "epoch": 564.27,
+      "learning_rate": 2.147449818037378e-06,
+      "loss": 0.0095,
+      "step": 93480
+    },
+    {
+      "epoch": 564.39,
+      "learning_rate": 2.1344575353785533e-06,
+      "loss": 0.0094,
+      "step": 93500
+    },
+    {
+      "epoch": 564.51,
+      "learning_rate": 2.1215038171227276e-06,
+      "loss": 0.0096,
+      "step": 93520
+    },
+    {
+      "epoch": 564.63,
+      "learning_rate": 2.108588673706491e-06,
+      "loss": 0.0096,
+      "step": 93540
+    },
+    {
+      "epoch": 564.75,
+      "learning_rate": 2.0957121155353364e-06,
+      "loss": 0.0096,
+      "step": 93560
+    },
+    {
+      "epoch": 564.87,
+      "learning_rate": 2.082874152983677e-06,
+      "loss": 0.0097,
+      "step": 93580
+    },
+    {
+      "epoch": 564.99,
+      "learning_rate": 2.070074796394822e-06,
+      "loss": 0.0097,
+      "step": 93600
+    },
+    {
+      "epoch": 565.11,
+      "learning_rate": 2.057314056080989e-06,
+      "loss": 0.0094,
+      "step": 93620
+    },
+    {
+      "epoch": 565.23,
+      "learning_rate": 2.044591942323276e-06,
+      "loss": 0.0094,
+      "step": 93640
+    },
+    {
+      "epoch": 565.35,
+      "learning_rate": 2.031908465371646e-06,
+      "loss": 0.0095,
+      "step": 93660
+    },
+    {
+      "epoch": 565.47,
+      "learning_rate": 2.0192636354449623e-06,
+      "loss": 0.0094,
+      "step": 93680
+    },
+    {
+      "epoch": 565.59,
+      "learning_rate": 2.006657462730932e-06,
+      "loss": 0.0095,
+      "step": 93700
+    },
+    {
+      "epoch": 565.71,
+      "learning_rate": 1.9940899573861195e-06,
+      "loss": 0.0096,
+      "step": 93720
+    },
+    {
+      "epoch": 565.84,
+      "learning_rate": 1.9815611295359416e-06,
+      "loss": 0.0096,
+      "step": 93740
+    },
+    {
+      "epoch": 565.96,
+      "learning_rate": 1.9690709892746352e-06,
+      "loss": 0.0097,
+      "step": 93760
+    },
+    {
+      "epoch": 566.08,
+      "learning_rate": 1.9566195466653003e-06,
+      "loss": 0.0094,
+      "step": 93780
+    },
+    {
+      "epoch": 566.2,
+      "learning_rate": 1.9442068117398238e-06,
+      "loss": 0.0094,
+      "step": 93800
+    },
+    {
+      "epoch": 566.32,
+      "learning_rate": 1.931832794498928e-06,
+      "loss": 0.0094,
+      "step": 93820
+    },
+    {
+      "epoch": 566.44,
+      "learning_rate": 1.9194975049121324e-06,
+      "loss": 0.0095,
+      "step": 93840
+    },
+    {
+      "epoch": 566.56,
+      "learning_rate": 1.9072009529177648e-06,
+      "loss": 0.0095,
+      "step": 93860
+    },
+    {
+      "epoch": 566.68,
+      "learning_rate": 1.8949431484229285e-06,
+      "loss": 0.0096,
+      "step": 93880
+    },
+    {
+      "epoch": 566.8,
+      "learning_rate": 1.8827241013035068e-06,
+      "loss": 0.0096,
+      "step": 93900
+    },
+    {
+      "epoch": 566.92,
+      "learning_rate": 1.8705438214041859e-06,
+      "loss": 0.0096,
+      "step": 93920
+    },
+    {
+      "epoch": 567.04,
+      "learning_rate": 1.858402318538377e-06,
+      "loss": 0.0095,
+      "step": 93940
+    },
+    {
+      "epoch": 567.16,
+      "learning_rate": 1.8462996024882828e-06,
+      "loss": 0.0093,
+      "step": 93960
+    },
+    {
+      "epoch": 567.28,
+      "learning_rate": 1.8342356830048369e-06,
+      "loss": 0.0094,
+      "step": 93980
+    },
+    {
+      "epoch": 567.4,
+      "learning_rate": 1.8222105698077251e-06,
+      "loss": 0.0095,
+      "step": 94000
+    },
+    {
+      "epoch": 567.4,
+      "eval_accuracy": 0.0012624591281326866,
+      "eval_loss": 24.054536819458008,
+      "eval_runtime": 18.1691,
+      "eval_samples_per_second": 112.278,
+      "eval_steps_per_second": 2.367,
+      "step": 94000
+    },
+    {
+      "epoch": 567.53,
+      "learning_rate": 1.8102242725853646e-06,
+      "loss": 0.0095,
+      "step": 94020
+    },
+    {
+      "epoch": 567.65,
+      "learning_rate": 1.7982768009948858e-06,
+      "loss": 0.0095,
+      "step": 94040
+    },
+    {
+      "epoch": 567.77,
+      "learning_rate": 1.7863681646621666e-06,
+      "loss": 0.0096,
+      "step": 94060
+    },
+    {
+      "epoch": 567.89,
+      "learning_rate": 1.7744983731817765e-06,
+      "loss": 0.0096,
+      "step": 94080
+    },
+    {
+      "epoch": 568.01,
+      "learning_rate": 1.7626674361169938e-06,
+      "loss": 0.0096,
+      "step": 94100
+    },
+    {
+      "epoch": 568.13,
+      "learning_rate": 1.750875362999771e-06,
+      "loss": 0.0094,
+      "step": 94120
+    },
+    {
+      "epoch": 568.25,
+      "learning_rate": 1.739122163330803e-06,
+      "loss": 0.0094,
+      "step": 94140
+    },
+    {
+      "epoch": 568.37,
+      "learning_rate": 1.727407846579404e-06,
+      "loss": 0.0094,
+      "step": 94160
+    },
+    {
+      "epoch": 568.49,
+      "learning_rate": 1.7157324221835912e-06,
+      "loss": 0.0095,
+      "step": 94180
+    },
+    {
+      "epoch": 568.61,
+      "learning_rate": 1.704095899550051e-06,
+      "loss": 0.0095,
+      "step": 94200
+    },
+    {
+      "epoch": 568.73,
+      "learning_rate": 1.6924982880541174e-06,
+      "loss": 0.0095,
+      "step": 94220
+    },
+    {
+      "epoch": 568.85,
+      "learning_rate": 1.6809395970397767e-06,
+      "loss": 0.0095,
+      "step": 94240
+    },
+    {
+      "epoch": 568.97,
+      "learning_rate": 1.6694198358196523e-06,
+      "loss": 0.0096,
+      "step": 94260
+    },
+    {
+      "epoch": 569.09,
+      "learning_rate": 1.6579390136750084e-06,
+      "loss": 0.0095,
+      "step": 94280
+    },
+    {
+      "epoch": 569.22,
+      "learning_rate": 1.6464971398557406e-06,
+      "loss": 0.0093,
+      "step": 94300
+    },
+    {
+      "epoch": 569.34,
+      "learning_rate": 1.6350942235803579e-06,
+      "loss": 0.0094,
+      "step": 94320
+    },
+    {
+      "epoch": 569.46,
+      "learning_rate": 1.6237302740359838e-06,
+      "loss": 0.0095,
+      "step": 94340
+    },
+    {
+      "epoch": 569.58,
+      "learning_rate": 1.61240530037835e-06,
+      "loss": 0.0095,
+      "step": 94360
+    },
+    {
+      "epoch": 569.7,
+      "learning_rate": 1.6011193117317857e-06,
+      "loss": 0.0095,
+      "step": 94380
+    },
+    {
+      "epoch": 569.82,
+      "learning_rate": 1.5898723171891948e-06,
+      "loss": 0.0095,
+      "step": 94400
+    },
+    {
+      "epoch": 569.94,
+      "learning_rate": 1.5786643258120904e-06,
+      "loss": 0.0096,
+      "step": 94420
+    },
+    {
+      "epoch": 570.06,
+      "learning_rate": 1.567495346630543e-06,
+      "loss": 0.0094,
+      "step": 94440
+    },
+    {
+      "epoch": 570.18,
+      "learning_rate": 1.5563653886431994e-06,
+      "loss": 0.0093,
+      "step": 94460
+    },
+    {
+      "epoch": 570.3,
+      "learning_rate": 1.5452744608172586e-06,
+      "loss": 0.0094,
+      "step": 94480
+    },
+    {
+      "epoch": 570.42,
+      "learning_rate": 1.534222572088495e-06,
+      "loss": 0.0094,
+      "step": 94500
+    },
+    {
+      "epoch": 570.54,
+      "learning_rate": 1.5232097313612082e-06,
+      "loss": 0.0095,
+      "step": 94520
+    },
+    {
+      "epoch": 570.66,
+      "learning_rate": 1.512235947508245e-06,
+      "loss": 0.0095,
+      "step": 94540
+    },
+    {
+      "epoch": 570.78,
+      "learning_rate": 1.5013012293709827e-06,
+      "loss": 0.0095,
+      "step": 94560
+    },
+    {
+      "epoch": 570.91,
+      "learning_rate": 1.4904055857593246e-06,
+      "loss": 0.0096,
+      "step": 94580
+    },
+    {
+      "epoch": 571.03,
+      "learning_rate": 1.479549025451704e-06,
+      "loss": 0.0095,
+      "step": 94600
+    },
+    {
+      "epoch": 571.15,
+      "learning_rate": 1.4687315571950465e-06,
+      "loss": 0.0094,
+      "step": 94620
+    },
+    {
+      "epoch": 571.27,
+      "learning_rate": 1.4579531897047915e-06,
+      "loss": 0.0093,
+      "step": 94640
+    },
+    {
+      "epoch": 571.39,
+      "learning_rate": 1.4472139316648814e-06,
+      "loss": 0.0094,
+      "step": 94660
+    },
+    {
+      "epoch": 571.51,
+      "learning_rate": 1.4365137917277337e-06,
+      "loss": 0.0095,
+      "step": 94680
+    },
+    {
+      "epoch": 571.63,
+      "learning_rate": 1.425852778514264e-06,
+      "loss": 0.0095,
+      "step": 94700
+    },
+    {
+      "epoch": 571.75,
+      "learning_rate": 1.4152309006138565e-06,
+      "loss": 0.0095,
+      "step": 94720
+    },
+    {
+      "epoch": 571.87,
+      "learning_rate": 1.4046481665843603e-06,
+      "loss": 0.0095,
+      "step": 94740
+    },
+    {
+      "epoch": 571.99,
+      "learning_rate": 1.3941045849521106e-06,
+      "loss": 0.0095,
+      "step": 94760
+    },
+    {
+      "epoch": 572.11,
+      "learning_rate": 1.3836001642118624e-06,
+      "loss": 0.0093,
+      "step": 94780
+    },
+    {
+      "epoch": 572.23,
+      "learning_rate": 1.3731349128268457e-06,
+      "loss": 0.0093,
+      "step": 94800
+    },
+    {
+      "epoch": 572.35,
+      "learning_rate": 1.3627088392287212e-06,
+      "loss": 0.0094,
+      "step": 94820
+    },
+    {
+      "epoch": 572.47,
+      "learning_rate": 1.3523219518175923e-06,
+      "loss": 0.0094,
+      "step": 94840
+    },
+    {
+      "epoch": 572.6,
+      "learning_rate": 1.3419742589619756e-06,
+      "loss": 0.0095,
+      "step": 94860
+    },
+    {
+      "epoch": 572.72,
+      "learning_rate": 1.3316657689988355e-06,
+      "loss": 0.0096,
+      "step": 94880
+    },
+    {
+      "epoch": 572.84,
+      "learning_rate": 1.3213964902335341e-06,
+      "loss": 0.0095,
+      "step": 94900
+    },
+    {
+      "epoch": 572.96,
+      "learning_rate": 1.3111664309398364e-06,
+      "loss": 0.0095,
+      "step": 94920
+    },
+    {
+      "epoch": 573.08,
+      "learning_rate": 1.300975599359916e-06,
+      "loss": 0.0094,
+      "step": 94940
+    },
+    {
+      "epoch": 573.2,
+      "learning_rate": 1.2908240037043495e-06,
+      "loss": 0.0093,
+      "step": 94960
+    },
+    {
+      "epoch": 573.32,
+      "learning_rate": 1.2807116521520945e-06,
+      "loss": 0.0094,
+      "step": 94980
+    },
+    {
+      "epoch": 573.44,
+      "learning_rate": 1.2706385528504728e-06,
+      "loss": 0.0094,
+      "step": 95000
+    },
+    {
+      "epoch": 573.44,
+      "eval_accuracy": 0.0012634469678260611,
+      "eval_loss": 24.07013702392578,
+      "eval_runtime": 18.2478,
+      "eval_samples_per_second": 111.795,
+      "eval_steps_per_second": 2.356,
+      "step": 95000
+    },
+    {
+      "epoch": 573.56,
+      "learning_rate": 1.2606047139152143e-06,
+      "loss": 0.0094,
+      "step": 95020
+    },
+    {
+      "epoch": 573.68,
+      "learning_rate": 1.2506101434304086e-06,
+      "loss": 0.0094,
+      "step": 95040
+    },
+    {
+      "epoch": 573.8,
+      "learning_rate": 1.2406548494484805e-06,
+      "loss": 0.0095,
+      "step": 95060
+    },
+    {
+      "epoch": 573.92,
+      "learning_rate": 1.230738839990242e-06,
+      "loss": 0.0095,
+      "step": 95080
+    },
+    {
+      "epoch": 574.04,
+      "learning_rate": 1.220862123044847e-06,
+      "loss": 0.0094,
+      "step": 95100
+    },
+    {
+      "epoch": 574.16,
+      "learning_rate": 1.21102470656978e-06,
+      "loss": 0.0093,
+      "step": 95120
+    },
+    {
+      "epoch": 574.29,
+      "learning_rate": 1.2012265984908733e-06,
+      "loss": 0.0093,
+      "step": 95140
+    },
+    {
+      "epoch": 574.41,
+      "learning_rate": 1.1914678067022789e-06,
+      "loss": 0.0095,
+      "step": 95160
+    },
+    {
+      "epoch": 574.53,
+      "learning_rate": 1.1817483390664908e-06,
+      "loss": 0.0094,
+      "step": 95180
+    },
+    {
+      "epoch": 574.65,
+      "learning_rate": 1.1720682034143005e-06,
+      "loss": 0.0094,
+      "step": 95200
+    },
+    {
+      "epoch": 574.77,
+      "learning_rate": 1.1624274075448193e-06,
+      "loss": 0.0095,
+      "step": 95220
+    },
+    {
+      "epoch": 574.89,
+      "learning_rate": 1.152825959225462e-06,
+      "loss": 0.0095,
+      "step": 95240
+    },
+    {
+      "epoch": 575.01,
+      "learning_rate": 1.1432638661919516e-06,
+      "loss": 0.0095,
+      "step": 95260
+    },
+    {
+      "epoch": 575.13,
+      "learning_rate": 1.1337411361482752e-06,
+      "loss": 0.0093,
+      "step": 95280
+    },
+    {
+      "epoch": 575.25,
+      "learning_rate": 1.1242577767667407e-06,
+      "loss": 0.0093,
+      "step": 95300
+    },
+    {
+      "epoch": 575.37,
+      "learning_rate": 1.114813795687908e-06,
+      "loss": 0.0093,
+      "step": 95320
+    },
+    {
+      "epoch": 575.49,
+      "learning_rate": 1.1054092005206351e-06,
+      "loss": 0.0094,
+      "step": 95340
+    },
+    {
+      "epoch": 575.61,
+      "learning_rate": 1.096043998842028e-06,
+      "loss": 0.0095,
+      "step": 95360
+    },
+    {
+      "epoch": 575.73,
+      "learning_rate": 1.0867181981974506e-06,
+      "loss": 0.0095,
+      "step": 95380
+    },
+    {
+      "epoch": 575.86,
+      "learning_rate": 1.0774318061005483e-06,
+      "loss": 0.0095,
+      "step": 95400
+    },
+    {
+      "epoch": 575.98,
+      "learning_rate": 1.0681848300331866e-06,
+      "loss": 0.0095,
+      "step": 95420
+    },
+    {
+      "epoch": 576.1,
+      "learning_rate": 1.0589772774454942e-06,
+      "loss": 0.0093,
+      "step": 95440
+    },
+    {
+      "epoch": 576.22,
+      "learning_rate": 1.049809155755832e-06,
+      "loss": 0.0093,
+      "step": 95460
+    },
+    {
+      "epoch": 576.34,
+      "learning_rate": 1.0406804723507857e-06,
+      "loss": 0.0094,
+      "step": 95480
+    },
+    {
+      "epoch": 576.46,
+      "learning_rate": 1.0315912345851718e-06,
+      "loss": 0.0094,
+      "step": 95500
+    },
+    {
+      "epoch": 576.58,
+      "learning_rate": 1.0225414497820162e-06,
+      "loss": 0.0094,
+      "step": 95520
+    },
+    {
+      "epoch": 576.7,
+      "learning_rate": 1.0135311252325864e-06,
+      "loss": 0.0095,
+      "step": 95540
+    },
+    {
+      "epoch": 576.82,
+      "learning_rate": 1.0045602681963195e-06,
+      "loss": 0.0094,
+      "step": 95560
+    },
+    {
+      "epoch": 576.94,
+      "learning_rate": 9.9562888590089e-07,
+      "loss": 0.0095,
+      "step": 95580
+    },
+    {
+      "epoch": 577.06,
+      "learning_rate": 9.867369855421305e-07,
+      "loss": 0.0094,
+      "step": 95600
+    },
+    {
+      "epoch": 577.18,
+      "learning_rate": 9.778845742840992e-07,
+      "loss": 0.0093,
+      "step": 95620
+    },
+    {
+      "epoch": 577.3,
+      "learning_rate": 9.690716592590298e-07,
+      "loss": 0.0093,
+      "step": 95640
+    },
+    {
+      "epoch": 577.42,
+      "learning_rate": 9.602982475673095e-07,
+      "loss": 0.0094,
+      "step": 95660
+    },
+    {
+      "epoch": 577.55,
+      "learning_rate": 9.515643462775337e-07,
+      "loss": 0.0093,
+      "step": 95680
+    },
+    {
+      "epoch": 577.67,
+      "learning_rate": 9.428699624264403e-07,
+      "loss": 0.0094,
+      "step": 95700
+    },
+    {
+      "epoch": 577.79,
+      "learning_rate": 9.342151030189428e-07,
+      "loss": 0.0095,
+      "step": 95720
+    },
+    {
+      "epoch": 577.91,
+      "learning_rate": 9.255997750280909e-07,
+      "loss": 0.0095,
+      "step": 95740
+    },
+    {
+      "epoch": 578.03,
+      "learning_rate": 9.170239853951212e-07,
+      "loss": 0.0095,
+      "step": 95760
+    },
+    {
+      "epoch": 578.15,
+      "learning_rate": 9.084877410293846e-07,
+      "loss": 0.0093,
+      "step": 95780
+    },
+    {
+      "epoch": 578.27,
+      "learning_rate": 8.999910488083685e-07,
+      "loss": 0.0094,
+      "step": 95800
+    },
+    {
+      "epoch": 578.39,
+      "learning_rate": 8.915339155777136e-07,
+      "loss": 0.0094,
+      "step": 95820
+    },
+    {
+      "epoch": 578.51,
+      "learning_rate": 8.83116348151175e-07,
+      "loss": 0.0094,
+      "step": 95840
+    },
+    {
+      "epoch": 578.63,
+      "learning_rate": 8.747383533106445e-07,
+      "loss": 0.0094,
+      "step": 95860
+    },
+    {
+      "epoch": 578.75,
+      "learning_rate": 8.663999378061116e-07,
+      "loss": 0.0094,
+      "step": 95880
+    },
+    {
+      "epoch": 578.87,
+      "learning_rate": 8.5810110835568e-07,
+      "loss": 0.0094,
+      "step": 95900
+    },
+    {
+      "epoch": 578.99,
+      "learning_rate": 8.498418716455903e-07,
+      "loss": 0.0094,
+      "step": 95920
+    },
+    {
+      "epoch": 579.11,
+      "learning_rate": 8.416222343301361e-07,
+      "loss": 0.0092,
+      "step": 95940
+    },
+    {
+      "epoch": 579.24,
+      "learning_rate": 8.334422030317424e-07,
+      "loss": 0.0093,
+      "step": 95960
+    },
+    {
+      "epoch": 579.36,
+      "learning_rate": 8.25301784340915e-07,
+      "loss": 0.0094,
+      "step": 95980
+    },
+    {
+      "epoch": 579.48,
+      "learning_rate": 8.172009848162409e-07,
+      "loss": 0.0094,
+      "step": 96000
+    },
+    {
+      "epoch": 579.48,
+      "eval_accuracy": 0.0012644348075194357,
+      "eval_loss": 24.082643508911133,
+      "eval_runtime": 18.205,
+      "eval_samples_per_second": 112.057,
+      "eval_steps_per_second": 2.362,
+      "step": 96000
+    },
+    {
+      "epoch": 579.6,
+      "learning_rate": 8.091398109843773e-07,
+      "loss": 0.0094,
+      "step": 96020
+    },
+    {
+      "epoch": 579.72,
+      "learning_rate": 8.011182693400732e-07,
+      "loss": 0.0094,
+      "step": 96040
+    },
+    {
+      "epoch": 579.84,
+      "learning_rate": 7.931363663461477e-07,
+      "loss": 0.0094,
+      "step": 96060
+    },
+    {
+      "epoch": 579.96,
+      "learning_rate": 7.85194108433468e-07,
+      "loss": 0.0094,
+      "step": 96080
+    },
+    {
+      "epoch": 580.08,
+      "learning_rate": 7.772915020009707e-07,
+      "loss": 0.0093,
+      "step": 96100
+    },
+    {
+      "epoch": 580.2,
+      "learning_rate": 7.694285534156354e-07,
+      "loss": 0.0093,
+      "step": 96120
+    },
+    {
+      "epoch": 580.32,
+      "learning_rate": 7.616052690125109e-07,
+      "loss": 0.0093,
+      "step": 96140
+    },
+    {
+      "epoch": 580.44,
+      "learning_rate": 7.538216550946664e-07,
+      "loss": 0.0093,
+      "step": 96160
+    },
+    {
+      "epoch": 580.56,
+      "learning_rate": 7.46077717933219e-07,
+      "loss": 0.0093,
+      "step": 96180
+    },
+    {
+      "epoch": 580.68,
+      "learning_rate": 7.383734637673223e-07,
+      "loss": 0.0094,
+      "step": 96200
+    },
+    {
+      "epoch": 580.8,
+      "learning_rate": 7.307088988041555e-07,
+      "loss": 0.0095,
+      "step": 96220
+    },
+    {
+      "epoch": 580.93,
+      "learning_rate": 7.230840292189178e-07,
+      "loss": 0.0094,
+      "step": 96240
+    },
+    {
+      "epoch": 581.05,
+      "learning_rate": 7.154988611548286e-07,
+      "loss": 0.0094,
+      "step": 96260
+    },
+    {
+      "epoch": 581.17,
+      "learning_rate": 7.07953400723127e-07,
+      "loss": 0.0093,
+      "step": 96280
+    },
+    {
+      "epoch": 581.29,
+      "learning_rate": 7.004476540030447e-07,
+      "loss": 0.0093,
+      "step": 96300
+    },
+    {
+      "epoch": 581.41,
+      "learning_rate": 6.929816270418332e-07,
+      "loss": 0.0093,
+      "step": 96320
+    },
+    {
+      "epoch": 581.53,
+      "learning_rate": 6.855553258547309e-07,
+      "loss": 0.0094,
+      "step": 96340
+    },
+    {
+      "epoch": 581.65,
+      "learning_rate": 6.781687564249739e-07,
+      "loss": 0.0094,
+      "step": 96360
+    },
+    {
+      "epoch": 581.77,
+      "learning_rate": 6.708219247038017e-07,
+      "loss": 0.0094,
+      "step": 96380
+    },
+    {
+      "epoch": 581.89,
+      "learning_rate": 6.635148366104016e-07,
+      "loss": 0.0094,
+      "step": 96400
+    },
+    {
+      "epoch": 582.01,
+      "learning_rate": 6.56247498031981e-07,
+      "loss": 0.0094,
+      "step": 96420
+    },
+    {
+      "epoch": 582.13,
+      "learning_rate": 6.490199148236898e-07,
+      "loss": 0.0093,
+      "step": 96440
+    },
+    {
+      "epoch": 582.25,
+      "learning_rate": 6.418320928086752e-07,
+      "loss": 0.0093,
+      "step": 96460
+    },
+    {
+      "epoch": 582.37,
+      "learning_rate": 6.346840377780216e-07,
+      "loss": 0.0093,
+      "step": 96480
+    },
+    {
+      "epoch": 582.49,
+      "learning_rate": 6.275757554907946e-07,
+      "loss": 0.0093,
+      "step": 96500
+    },
+    {
+      "epoch": 582.62,
+      "learning_rate": 6.205072516740129e-07,
+      "loss": 0.0094,
+      "step": 96520
+    },
+    {
+      "epoch": 582.74,
+      "learning_rate": 6.134785320226322e-07,
+      "loss": 0.0093,
+      "step": 96540
+    },
+    {
+      "epoch": 582.86,
+      "learning_rate": 6.064896021995725e-07,
+      "loss": 0.0094,
+      "step": 96560
+    },
+    {
+      "epoch": 582.98,
+      "learning_rate": 5.995404678356853e-07,
+      "loss": 0.0094,
+      "step": 96580
+    },
+    {
+      "epoch": 583.1,
+      "learning_rate": 5.926311345297697e-07,
+      "loss": 0.0093,
+      "step": 96600
+    },
+    {
+      "epoch": 583.22,
+      "learning_rate": 5.857616078485395e-07,
+      "loss": 0.0093,
+      "step": 96620
+    },
+    {
+      "epoch": 583.34,
+      "learning_rate": 5.789318933266508e-07,
+      "loss": 0.0094,
+      "step": 96640
+    },
+    {
+      "epoch": 583.46,
+      "learning_rate": 5.721419964666908e-07,
+      "loss": 0.0094,
+      "step": 96660
+    },
+    {
+      "epoch": 583.58,
+      "learning_rate": 5.653919227391502e-07,
+      "loss": 0.0093,
+      "step": 96680
+    },
+    {
+      "epoch": 583.7,
+      "learning_rate": 5.586816775824344e-07,
+      "loss": 0.0093,
+      "step": 96700
+    },
+    {
+      "epoch": 583.82,
+      "learning_rate": 5.520112664028799e-07,
+      "loss": 0.0093,
+      "step": 96720
+    },
+    {
+      "epoch": 583.94,
+      "learning_rate": 5.453806945747098e-07,
+      "loss": 0.0094,
+      "step": 96740
+    },
+    {
+      "epoch": 584.06,
+      "learning_rate": 5.387899674400509e-07,
+      "loss": 0.0093,
+      "step": 96760
+    },
+    {
+      "epoch": 584.19,
+      "learning_rate": 5.322390903089392e-07,
+      "loss": 0.0093,
+      "step": 96780
+    },
+    {
+      "epoch": 584.31,
+      "learning_rate": 5.257280684593024e-07,
+      "loss": 0.0093,
+      "step": 96800
+    },
+    {
+      "epoch": 584.43,
+      "learning_rate": 5.192569071369446e-07,
+      "loss": 0.0093,
+      "step": 96820
+    },
+    {
+      "epoch": 584.55,
+      "learning_rate": 5.128256115555675e-07,
+      "loss": 0.0093,
+      "step": 96840
+    },
+    {
+      "epoch": 584.67,
+      "learning_rate": 5.064341868967537e-07,
+      "loss": 0.0093,
+      "step": 96860
+    },
+    {
+      "epoch": 584.79,
+      "learning_rate": 5.000826383099511e-07,
+      "loss": 0.0094,
+      "step": 96880
+    },
+    {
+      "epoch": 584.91,
+      "learning_rate": 4.937709709124937e-07,
+      "loss": 0.0094,
+      "step": 96900
+    },
+    {
+      "epoch": 585.03,
+      "learning_rate": 4.874991897895697e-07,
+      "loss": 0.0093,
+      "step": 96920
+    },
+    {
+      "epoch": 585.15,
+      "learning_rate": 4.81267299994248e-07,
+      "loss": 0.0093,
+      "step": 96940
+    },
+    {
+      "epoch": 585.27,
+      "learning_rate": 4.750753065474456e-07,
+      "loss": 0.0093,
+      "step": 96960
+    },
+    {
+      "epoch": 585.39,
+      "learning_rate": 4.689232144379496e-07,
+      "loss": 0.0093,
+      "step": 96980
+    },
+    {
+      "epoch": 585.51,
+      "learning_rate": 4.628110286223675e-07,
+      "loss": 0.0093,
+      "step": 97000
+    },
+    {
+      "epoch": 585.51,
+      "eval_accuracy": 0.001262953047979374,
+      "eval_loss": 24.08979034423828,
+      "eval_runtime": 18.188,
+      "eval_samples_per_second": 112.162,
+      "eval_steps_per_second": 2.364,
+      "step": 97000
+    },
+    {
+      "epoch": 585.63,
+      "learning_rate": 4.57041419787968e-07,
+      "loss": 0.0093,
+      "step": 97020
+    },
+    {
+      "epoch": 585.75,
+      "learning_rate": 4.5100706538029334e-07,
+      "loss": 0.0094,
+      "step": 97040
+    },
+    {
+      "epoch": 585.88,
+      "learning_rate": 4.450126317012637e-07,
+      "loss": 0.0093,
+      "step": 97060
+    },
+    {
+      "epoch": 586.0,
+      "learning_rate": 4.3905812358048246e-07,
+      "loss": 0.0094,
+      "step": 97080
+    },
+    {
+      "epoch": 586.12,
+      "learning_rate": 4.331435458153954e-07,
+      "loss": 0.0092,
+      "step": 97100
+    },
+    {
+      "epoch": 586.24,
+      "learning_rate": 4.2726890317126845e-07,
+      "loss": 0.0093,
+      "step": 97120
+    },
+    {
+      "epoch": 586.36,
+      "learning_rate": 4.2143420038119884e-07,
+      "loss": 0.0093,
+      "step": 97140
+    },
+    {
+      "epoch": 586.48,
+      "learning_rate": 4.15639442146093e-07,
+      "loss": 0.0093,
+      "step": 97160
+    },
+    {
+      "epoch": 586.6,
+      "learning_rate": 4.0988463313469395e-07,
+      "loss": 0.0093,
+      "step": 97180
+    },
+    {
+      "epoch": 586.72,
+      "learning_rate": 4.0416977798353737e-07,
+      "loss": 0.0093,
+      "step": 97200
+    },
+    {
+      "epoch": 586.84,
+      "learning_rate": 3.984948812969791e-07,
+      "loss": 0.0093,
+      "step": 97220
+    },
+    {
+      "epoch": 586.96,
+      "learning_rate": 3.9285994764718945e-07,
+      "loss": 0.0094,
+      "step": 97240
+    },
+    {
+      "epoch": 587.08,
+      "learning_rate": 3.8726498157412026e-07,
+      "loss": 0.0093,
+      "step": 97260
+    },
+    {
+      "epoch": 587.2,
+      "learning_rate": 3.81709987585549e-07,
+      "loss": 0.0093,
+      "step": 97280
+    },
+    {
+      "epoch": 587.32,
+      "learning_rate": 3.761949701570233e-07,
+      "loss": 0.0093,
+      "step": 97300
+    },
+    {
+      "epoch": 587.44,
+      "learning_rate": 3.7071993373190004e-07,
+      "loss": 0.0093,
+      "step": 97320
+    },
+    {
+      "epoch": 587.57,
+      "learning_rate": 3.652848827213118e-07,
+      "loss": 0.0093,
+      "step": 97340
+    },
+    {
+      "epoch": 587.69,
+      "learning_rate": 3.598898215041835e-07,
+      "loss": 0.0093,
+      "step": 97360
+    },
+    {
+      "epoch": 587.81,
+      "learning_rate": 3.5453475442721597e-07,
+      "loss": 0.0093,
+      "step": 97380
+    },
+    {
+      "epoch": 587.93,
+      "learning_rate": 3.492196858048913e-07,
+      "loss": 0.0094,
+      "step": 97400
+    },
+    {
+      "epoch": 588.05,
+      "learning_rate": 3.4394461991946736e-07,
+      "loss": 0.0093,
+      "step": 97420
+    },
+    {
+      "epoch": 588.17,
+      "learning_rate": 3.3870956102096675e-07,
+      "loss": 0.0093,
+      "step": 97440
+    },
+    {
+      "epoch": 588.29,
+      "learning_rate": 3.335145133271822e-07,
+      "loss": 0.0093,
+      "step": 97460
+    },
+    {
+      "epoch": 588.41,
+      "learning_rate": 3.283594810236712e-07,
+      "loss": 0.0093,
+      "step": 97480
+    },
+    {
+      "epoch": 588.53,
+      "learning_rate": 3.232444682637503e-07,
+      "loss": 0.0093,
+      "step": 97500
+    },
+    {
+      "epoch": 588.65,
+      "learning_rate": 3.181694791684842e-07,
+      "loss": 0.0093,
+      "step": 97520
+    },
+    {
+      "epoch": 588.77,
+      "learning_rate": 3.1313451782671313e-07,
+      "loss": 0.0093,
+      "step": 97540
+    },
+    {
+      "epoch": 588.89,
+      "learning_rate": 3.0813958829500886e-07,
+      "loss": 0.0093,
+      "step": 97560
+    },
+    {
+      "epoch": 589.01,
+      "learning_rate": 3.031846945976913e-07,
+      "loss": 0.0093,
+      "step": 97580
+    },
+    {
+      "epoch": 589.13,
+      "learning_rate": 2.982698407268336e-07,
+      "loss": 0.0093,
+      "step": 97600
+    },
+    {
+      "epoch": 589.26,
+      "learning_rate": 2.9339503064224615e-07,
+      "loss": 0.0093,
+      "step": 97620
+    },
+    {
+      "epoch": 589.38,
+      "learning_rate": 2.885602682714761e-07,
+      "loss": 0.0093,
+      "step": 97640
+    },
+    {
+      "epoch": 589.5,
+      "learning_rate": 2.837655575097964e-07,
+      "loss": 0.0093,
+      "step": 97660
+    },
+    {
+      "epoch": 589.62,
+      "learning_rate": 2.790109022202281e-07,
+      "loss": 0.0093,
+      "step": 97680
+    },
+    {
+      "epoch": 589.74,
+      "learning_rate": 2.742963062335069e-07,
+      "loss": 0.0093,
+      "step": 97700
+    },
+    {
+      "epoch": 589.86,
+      "learning_rate": 2.6962177334809436e-07,
+      "loss": 0.0094,
+      "step": 97720
+    },
+    {
+      "epoch": 589.98,
+      "learning_rate": 2.649873073301834e-07,
+      "loss": 0.0093,
+      "step": 97740
+    },
+    {
+      "epoch": 590.1,
+      "learning_rate": 2.603929119136761e-07,
+      "loss": 0.0093,
+      "step": 97760
+    },
+    {
+      "epoch": 590.22,
+      "learning_rate": 2.558385908001948e-07,
+      "loss": 0.0093,
+      "step": 97780
+    },
+    {
+      "epoch": 590.34,
+      "learning_rate": 2.5132434765907654e-07,
+      "loss": 0.0093,
+      "step": 97800
+    },
+    {
+      "epoch": 590.46,
+      "learning_rate": 2.4685018612735644e-07,
+      "loss": 0.0093,
+      "step": 97820
+    },
+    {
+      "epoch": 590.58,
+      "learning_rate": 2.4241610980979546e-07,
+      "loss": 0.0093,
+      "step": 97840
+    },
+    {
+      "epoch": 590.7,
+      "learning_rate": 2.3802212227884145e-07,
+      "loss": 0.0093,
+      "step": 97860
+    },
+    {
+      "epoch": 590.82,
+      "learning_rate": 2.336682270746571e-07,
+      "loss": 0.0093,
+      "step": 97880
+    },
+    {
+      "epoch": 590.95,
+      "learning_rate": 2.2935442770509187e-07,
+      "loss": 0.0093,
+      "step": 97900
+    },
+    {
+      "epoch": 591.07,
+      "learning_rate": 2.2508072764570454e-07,
+      "loss": 0.0093,
+      "step": 97920
+    },
+    {
+      "epoch": 591.19,
+      "learning_rate": 2.2084713033972415e-07,
+      "loss": 0.0092,
+      "step": 97940
+    },
+    {
+      "epoch": 591.31,
+      "learning_rate": 2.1665363919809445e-07,
+      "loss": 0.0093,
+      "step": 97960
+    },
+    {
+      "epoch": 591.43,
+      "learning_rate": 2.1250025759943504e-07,
+      "loss": 0.0093,
+      "step": 97980
+    },
+    {
+      "epoch": 591.55,
+      "learning_rate": 2.0838698889004693e-07,
+      "loss": 0.0093,
+      "step": 98000
+    },
+    {
+      "epoch": 591.55,
+      "eval_accuracy": 0.0012634469678260611,
+      "eval_loss": 24.093477249145508,
+      "eval_runtime": 18.1793,
+      "eval_samples_per_second": 112.215,
+      "eval_steps_per_second": 2.365,
+      "step": 98000
+    },
+    {
+      "epoch": 591.67,
+      "learning_rate": 2.043138363839181e-07,
+      "loss": 0.0093,
+      "step": 98020
+    },
+    {
+      "epoch": 591.79,
+      "learning_rate": 2.0028080336270683e-07,
+      "loss": 0.0093,
+      "step": 98040
+    },
+    {
+      "epoch": 591.91,
+      "learning_rate": 1.9628789307576945e-07,
+      "loss": 0.0093,
+      "step": 98060
+    },
+    {
+      "epoch": 592.03,
+      "learning_rate": 1.923351087401104e-07,
+      "loss": 0.0093,
+      "step": 98080
+    },
+    {
+      "epoch": 592.15,
+      "learning_rate": 1.884224535404211e-07,
+      "loss": 0.0093,
+      "step": 98100
+    },
+    {
+      "epoch": 592.27,
+      "learning_rate": 1.8454993062905212e-07,
+      "loss": 0.0093,
+      "step": 98120
+    },
+    {
+      "epoch": 592.39,
+      "learning_rate": 1.8071754312602995e-07,
+      "loss": 0.0093,
+      "step": 98140
+    },
+    {
+      "epoch": 592.52,
+      "learning_rate": 1.7692529411904578e-07,
+      "loss": 0.0093,
+      "step": 98160
+    },
+    {
+      "epoch": 592.64,
+      "learning_rate": 1.7317318666342786e-07,
+      "loss": 0.0093,
+      "step": 98180
+    },
+    {
+      "epoch": 592.76,
+      "learning_rate": 1.6946122378220242e-07,
+      "loss": 0.0092,
+      "step": 98200
+    },
+    {
+      "epoch": 592.88,
+      "learning_rate": 1.6578940846602165e-07,
+      "loss": 0.0093,
+      "step": 98220
+    },
+    {
+      "epoch": 593.0,
+      "learning_rate": 1.6215774367319691e-07,
+      "loss": 0.0093,
+      "step": 98240
+    },
+    {
+      "epoch": 593.12,
+      "learning_rate": 1.5856623232969324e-07,
+      "loss": 0.0093,
+      "step": 98260
+    },
+    {
+      "epoch": 593.24,
+      "learning_rate": 1.5501487732914045e-07,
+      "loss": 0.0092,
+      "step": 98280
+    },
+    {
+      "epoch": 593.36,
+      "learning_rate": 1.515036815327886e-07,
+      "loss": 0.0092,
+      "step": 98300
+    },
+    {
+      "epoch": 593.48,
+      "learning_rate": 1.4803264776955262e-07,
+      "loss": 0.0093,
+      "step": 98320
+    },
+    {
+      "epoch": 593.6,
+      "learning_rate": 1.4460177883597327e-07,
+      "loss": 0.0092,
+      "step": 98340
+    },
+    {
+      "epoch": 593.72,
+      "learning_rate": 1.4121107749624497e-07,
+      "loss": 0.0093,
+      "step": 98360
+    },
+    {
+      "epoch": 593.84,
+      "learning_rate": 1.378605464821936e-07,
+      "loss": 0.0093,
+      "step": 98380
+    },
+    {
+      "epoch": 593.96,
+      "learning_rate": 1.3455018849328204e-07,
+      "loss": 0.0093,
+      "step": 98400
+    },
+    {
+      "epoch": 594.08,
+      "learning_rate": 1.3128000619660462e-07,
+      "loss": 0.0093,
+      "step": 98420
+    },
+    {
+      "epoch": 594.21,
+      "learning_rate": 1.2805000222689267e-07,
+      "loss": 0.0093,
+      "step": 98440
+    },
+    {
+      "epoch": 594.33,
+      "learning_rate": 1.2486017918649784e-07,
+      "loss": 0.0093,
+      "step": 98460
+    },
+    {
+      "epoch": 594.45,
+      "learning_rate": 1.217105396453977e-07,
+      "loss": 0.0093,
+      "step": 98480
+    },
+    {
+      "epoch": 594.57,
+      "learning_rate": 1.1860108614121234e-07,
+      "loss": 0.0093,
+      "step": 98500
+    },
+    {
+      "epoch": 594.69,
+      "learning_rate": 1.1553182117916006e-07,
+      "loss": 0.0093,
+      "step": 98520
+    },
+    {
+      "epoch": 594.81,
+      "learning_rate": 1.1250274723210163e-07,
+      "loss": 0.0093,
+      "step": 98540
+    },
+    {
+      "epoch": 594.93,
+      "learning_rate": 1.0951386674050157e-07,
+      "loss": 0.0093,
+      "step": 98560
+    },
+    {
+      "epoch": 595.05,
+      "learning_rate": 1.0656518211245025e-07,
+      "loss": 0.0093,
+      "step": 98580
+    },
+    {
+      "epoch": 595.17,
+      "learning_rate": 1.0365669572364733e-07,
+      "loss": 0.0093,
+      "step": 98600
+    },
+    {
+      "epoch": 595.29,
+      "learning_rate": 1.0078840991740723e-07,
+      "loss": 0.0092,
+      "step": 98620
+    },
+    {
+      "epoch": 595.41,
+      "learning_rate": 9.796032700465918e-08,
+      "loss": 0.0092,
+      "step": 98640
+    },
+    {
+      "epoch": 595.53,
+      "learning_rate": 9.517244926393609e-08,
+      "loss": 0.0093,
+      "step": 98660
+    },
+    {
+      "epoch": 595.65,
+      "learning_rate": 9.242477894138568e-08,
+      "loss": 0.0093,
+      "step": 98680
+    },
+    {
+      "epoch": 595.77,
+      "learning_rate": 8.97173182507427e-08,
+      "loss": 0.0093,
+      "step": 98700
+    },
+    {
+      "epoch": 595.9,
+      "learning_rate": 8.705006937337334e-08,
+      "loss": 0.0093,
+      "step": 98720
+    },
+    {
+      "epoch": 596.02,
+      "learning_rate": 8.442303445823085e-08,
+      "loss": 0.0093,
+      "step": 98740
+    },
+    {
+      "epoch": 596.14,
+      "learning_rate": 8.183621562186106e-08,
+      "loss": 0.0092,
+      "step": 98760
+    },
+    {
+      "epoch": 596.26,
+      "learning_rate": 7.928961494841903e-08,
+      "loss": 0.0092,
+      "step": 98780
+    },
+    {
+      "epoch": 596.38,
+      "learning_rate": 7.678323448966351e-08,
+      "loss": 0.0093,
+      "step": 98800
+    },
+    {
+      "epoch": 596.5,
+      "learning_rate": 7.431707626493478e-08,
+      "loss": 0.0093,
+      "step": 98820
+    },
+    {
+      "epoch": 596.62,
+      "learning_rate": 7.189114226117122e-08,
+      "loss": 0.0093,
+      "step": 98840
+    },
+    {
+      "epoch": 596.74,
+      "learning_rate": 6.950543443290381e-08,
+      "loss": 0.0093,
+      "step": 98860
+    },
+    {
+      "epoch": 596.86,
+      "learning_rate": 6.715995470225056e-08,
+      "loss": 0.0093,
+      "step": 98880
+    },
+    {
+      "epoch": 596.98,
+      "learning_rate": 6.48547049589332e-08,
+      "loss": 0.0093,
+      "step": 98900
+    },
+    {
+      "epoch": 597.1,
+      "learning_rate": 6.258968706023827e-08,
+      "loss": 0.0093,
+      "step": 98920
+    },
+    {
+      "epoch": 597.22,
+      "learning_rate": 6.0364902831056e-08,
+      "loss": 0.0092,
+      "step": 98940
+    },
+    {
+      "epoch": 597.34,
+      "learning_rate": 5.818035406385258e-08,
+      "loss": 0.0092,
+      "step": 98960
+    },
+    {
+      "epoch": 597.46,
+      "learning_rate": 5.603604251868122e-08,
+      "loss": 0.0093,
+      "step": 98980
+    },
+    {
+      "epoch": 597.59,
+      "learning_rate": 5.393196992317662e-08,
+      "loss": 0.0093,
+      "step": 99000
+    },
+    {
+      "epoch": 597.59,
+      "eval_accuracy": 0.001262953047979374,
+      "eval_loss": 24.094446182250977,
+      "eval_runtime": 18.2941,
+      "eval_samples_per_second": 111.512,
+      "eval_steps_per_second": 2.35,
+      "step": 99000
+    },
+    {
+      "epoch": 597.71,
+      "learning_rate": 5.1970373828963416e-08,
+      "loss": 0.0092,
+      "step": 99020
+    },
+    {
+      "epoch": 597.83,
+      "learning_rate": 4.9944772031629284e-08,
+      "loss": 0.0093,
+      "step": 99040
+    },
+    {
+      "epoch": 597.95,
+      "learning_rate": 4.795941409158622e-08,
+      "loss": 0.0093,
+      "step": 99060
+    },
+    {
+      "epoch": 598.07,
+      "learning_rate": 4.601430160839915e-08,
+      "loss": 0.0093,
+      "step": 99080
+    },
+    {
+      "epoch": 598.19,
+      "learning_rate": 4.410943614921448e-08,
+      "loss": 0.0093,
+      "step": 99100
+    },
+    {
+      "epoch": 598.31,
+      "learning_rate": 4.2244819248743465e-08,
+      "loss": 0.0093,
+      "step": 99120
+    },
+    {
+      "epoch": 598.43,
+      "learning_rate": 4.042045240927883e-08,
+      "loss": 0.0092,
+      "step": 99140
+    },
+    {
+      "epoch": 598.55,
+      "learning_rate": 3.863633710067815e-08,
+      "loss": 0.0092,
+      "step": 99160
+    },
+    {
+      "epoch": 598.67,
+      "learning_rate": 3.689247476036939e-08,
+      "loss": 0.0093,
+      "step": 99180
+    },
+    {
+      "epoch": 598.79,
+      "learning_rate": 3.518886679334532e-08,
+      "loss": 0.0093,
+      "step": 99200
+    },
+    {
+      "epoch": 598.91,
+      "learning_rate": 3.352551457218578e-08,
+      "loss": 0.0093,
+      "step": 99220
+    },
+    {
+      "epoch": 599.03,
+      "learning_rate": 3.190241943701322e-08,
+      "loss": 0.0093,
+      "step": 99240
+    },
+    {
+      "epoch": 599.15,
+      "learning_rate": 3.031958269553159e-08,
+      "loss": 0.0093,
+      "step": 99260
+    },
+    {
+      "epoch": 599.28,
+      "learning_rate": 2.8777005622998564e-08,
+      "loss": 0.0093,
+      "step": 99280
+    },
+    {
+      "epoch": 599.4,
+      "learning_rate": 2.727468946224776e-08,
+      "loss": 0.0093,
+      "step": 99300
+    },
+    {
+      "epoch": 599.52,
+      "learning_rate": 2.581263542366097e-08,
+      "loss": 0.0092,
+      "step": 99320
+    },
+    {
+      "epoch": 599.64,
+      "learning_rate": 2.4390844685190372e-08,
+      "loss": 0.0092,
+      "step": 99340
+    },
+    {
+      "epoch": 599.76,
+      "learning_rate": 2.3009318392347435e-08,
+      "loss": 0.0093,
+      "step": 99360
+    },
+    {
+      "epoch": 599.88,
+      "learning_rate": 2.1668057658208453e-08,
+      "loss": 0.0093,
+      "step": 99380
+    },
+    {
+      "epoch": 600.0,
+      "learning_rate": 2.036706356339235e-08,
+      "loss": 0.0092,
+      "step": 99400
+    },
+    {
+      "epoch": 600.12,
+      "learning_rate": 1.910633715609955e-08,
+      "loss": 0.0093,
+      "step": 99420
+    },
+    {
+      "epoch": 600.24,
+      "learning_rate": 1.7885879452061993e-08,
+      "loss": 0.0093,
+      "step": 99440
+    },
+    {
+      "epoch": 600.36,
+      "learning_rate": 1.6705691434587556e-08,
+      "loss": 0.0093,
+      "step": 99460
+    },
+    {
+      "epoch": 600.48,
+      "learning_rate": 1.55657740545323e-08,
+      "loss": 0.0092,
+      "step": 99480
+    },
+    {
+      "epoch": 600.6,
+      "learning_rate": 1.446612823030602e-08,
+      "loss": 0.0092,
+      "step": 99500
+    },
+    {
+      "epoch": 600.72,
+      "learning_rate": 1.3406754847872238e-08,
+      "loss": 0.0093,
+      "step": 99520
+    },
+    {
+      "epoch": 600.85,
+      "learning_rate": 1.2387654760748213e-08,
+      "loss": 0.0093,
+      "step": 99540
+    },
+    {
+      "epoch": 600.97,
+      "learning_rate": 1.1408828790010484e-08,
+      "loss": 0.0092,
+      "step": 99560
+    },
+    {
+      "epoch": 601.09,
+      "learning_rate": 1.047027772427267e-08,
+      "loss": 0.0092,
+      "step": 99580
+    },
+    {
+      "epoch": 601.21,
+      "learning_rate": 9.572002319713225e-09,
+      "loss": 0.0093,
+      "step": 99600
+    },
+    {
+      "epoch": 601.33,
+      "learning_rate": 8.714003300058782e-09,
+      "loss": 0.0093,
+      "step": 99620
+    },
+    {
+      "epoch": 601.45,
+      "learning_rate": 7.896281356578605e-09,
+      "loss": 0.0092,
+      "step": 99640
+    },
+    {
+      "epoch": 601.57,
+      "learning_rate": 7.118837148101243e-09,
+      "loss": 0.0093,
+      "step": 99660
+    },
+    {
+      "epoch": 601.69,
+      "learning_rate": 6.38167130099232e-09,
+      "loss": 0.0093,
+      "step": 99680
+    },
+    {
+      "epoch": 601.81,
+      "learning_rate": 5.684784409182298e-09,
+      "loss": 0.0092,
+      "step": 99700
+    },
+    {
+      "epoch": 601.93,
+      "learning_rate": 5.0281770341331635e-09,
+      "loss": 0.0093,
+      "step": 99720
+    },
+    {
+      "epoch": 602.05,
+      "learning_rate": 4.411849704866189e-09,
+      "loss": 0.0093,
+      "step": 99740
+    },
+    {
+      "epoch": 602.17,
+      "learning_rate": 3.835802917939724e-09,
+      "loss": 0.0093,
+      "step": 99760
+    },
+    {
+      "epoch": 602.29,
+      "learning_rate": 3.3000371374714013e-09,
+      "loss": 0.0093,
+      "step": 99780
+    },
+    {
+      "epoch": 602.41,
+      "learning_rate": 2.8045527951103822e-09,
+      "loss": 0.0093,
+      "step": 99800
+    },
+    {
+      "epoch": 602.54,
+      "learning_rate": 2.349350290070662e-09,
+      "loss": 0.0092,
+      "step": 99820
+    },
+    {
+      "epoch": 602.66,
+      "learning_rate": 1.934429989086661e-09,
+      "loss": 0.0093,
+      "step": 99840
+    },
+    {
+      "epoch": 602.78,
+      "learning_rate": 1.5597922264631859e-09,
+      "loss": 0.0092,
+      "step": 99860
+    },
+    {
+      "epoch": 602.9,
+      "learning_rate": 1.2254373040310185e-09,
+      "loss": 0.0092,
+      "step": 99880
+    },
+    {
+      "epoch": 603.02,
+      "learning_rate": 9.31365491180225e-10,
+      "loss": 0.0092,
+      "step": 99900
+    },
+    {
+      "epoch": 603.14,
+      "learning_rate": 6.775770248379499e-10,
+      "loss": 0.0093,
+      "step": 99920
+    },
+    {
+      "epoch": 603.26,
+      "learning_rate": 4.640721094684164e-10,
+      "loss": 0.0093,
+      "step": 99940
+    },
+    {
+      "epoch": 603.38,
+      "learning_rate": 2.908509171062335e-10,
+      "loss": 0.0092,
+      "step": 99960
+    },
+    {
+      "epoch": 603.5,
+      "learning_rate": 1.57913587295333e-10,
+      "loss": 0.0092,
+      "step": 99980
+    },
+    {
+      "epoch": 603.62,
+      "learning_rate": 6.526022715003244e-11,
+      "loss": 0.0092,
+      "step": 100000
+    },
+    {
+      "epoch": 603.62,
+      "eval_accuracy": 0.0012644348075194357,
+      "eval_loss": 24.09497833251953,
+      "eval_runtime": 18.2893,
+      "eval_samples_per_second": 111.54,
+      "eval_steps_per_second": 2.351,
+      "step": 100000
+    },
+    {
+      "epoch": 603.62,
+      "step": 100000,
+      "total_flos": 1.0802570660564705e+18,
+      "train_loss": 1.1590602387964726,
+      "train_runtime": 91564.1404,
+      "train_samples_per_second": 52.422,
+      "train_steps_per_second": 1.092
+    }
+  ],
+  "logging_steps": 20,
+  "max_steps": 100000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 607,
+  "save_steps": 1000,
+  "total_flos": 1.0802570660564705e+18,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}