diff --git "a/checkpoint-300/trainer_state.json" "b/checkpoint-300/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-300/trainer_state.json"
@@ -0,0 +1,5733 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.75,
+  "eval_steps": 500,
+  "global_step": 300,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "Batch Sum": 52.814453125,
+      "accuracy": 0.546875,
+      "epoch": 0,
+      "step": 0
+    },
+    {
+      "Batch Sum": 52.524169921875,
+      "accuracy": 0.40625,
+      "epoch": 0,
+      "step": 0
+    },
+    {
+      "epoch": 0.0025,
+      "grad_norm": 12.003690719604492,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 0.714,
+      "step": 1
+    },
+    {
+      "Batch Sum": 52.9404296875,
+      "accuracy": 0.515625,
+      "epoch": 0.0025,
+      "step": 1
+    },
+    {
+      "Batch Sum": 48.3984375,
+      "accuracy": 0.625,
+      "epoch": 0.0025,
+      "step": 1
+    },
+    {
+      "epoch": 0.005,
+      "grad_norm": 14.317296981811523,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.6849,
+      "step": 2
+    },
+    {
+      "Batch Sum": 52.79443359375,
+      "accuracy": 0.484375,
+      "epoch": 0.005,
+      "step": 2
+    },
+    {
+      "Batch Sum": 53.980224609375,
+      "accuracy": 0.578125,
+      "epoch": 0.005,
+      "step": 2
+    },
+    {
+      "epoch": 0.0075,
+      "grad_norm": 21.00113296508789,
+      "learning_rate": 7.5e-07,
+      "loss": 0.7106,
+      "step": 3
+    },
+    {
+      "Batch Sum": 53.60833740234375,
+      "accuracy": 0.453125,
+      "epoch": 0.0075,
+      "step": 3
+    },
+    {
+      "Batch Sum": 52.896728515625,
+      "accuracy": 0.578125,
+      "epoch": 0.0075,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.431326866149902,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.6978,
+      "step": 4
+    },
+    {
+      "Batch Sum": 54.1025390625,
+      "accuracy": 0.65625,
+      "epoch": 0.01,
+      "step": 4
+    },
+    {
+      "Batch Sum": 54.279052734375,
+      "accuracy": 0.53125,
+      "epoch": 0.01,
+      "step": 4
+    },
+    {
+      "epoch": 0.0125,
+      "grad_norm": 13.951983451843262,
+      "learning_rate": 1.25e-06,
+      "loss": 0.6883,
+      "step": 5
+    },
+    {
+      "Batch Sum": 53.18408203125,
+      "accuracy": 0.578125,
+      "epoch": 0.0125,
+      "step": 5
+    },
+    {
+      "Batch Sum": 49.6630859375,
+      "accuracy": 0.53125,
+      "epoch": 0.0125,
+      "step": 5
+    },
+    {
+      "epoch": 0.015,
+      "grad_norm": 11.490130424499512,
+      "learning_rate": 1.5e-06,
+      "loss": 0.6818,
+      "step": 6
+    },
+    {
+      "Batch Sum": 54.36865234375,
+      "accuracy": 0.546875,
+      "epoch": 0.015,
+      "step": 6
+    },
+    {
+      "Batch Sum": 53.27685546875,
+      "accuracy": 0.515625,
+      "epoch": 0.015,
+      "step": 6
+    },
+    {
+      "epoch": 0.0175,
+      "grad_norm": 18.882749557495117,
+      "learning_rate": 1.75e-06,
+      "loss": 0.7198,
+      "step": 7
+    },
+    {
+      "Batch Sum": 58.3974609375,
+      "accuracy": 0.421875,
+      "epoch": 0.0175,
+      "step": 7
+    },
+    {
+      "Batch Sum": 56.97265625,
+      "accuracy": 0.6875,
+      "epoch": 0.0175,
+      "step": 7
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.89813232421875,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.6877,
+      "step": 8
+    },
+    {
+      "Batch Sum": 56.07861328125,
+      "accuracy": 0.625,
+      "epoch": 0.02,
+      "step": 8
+    },
+    {
+      "Batch Sum": 57.08154296875,
+      "accuracy": 0.609375,
+      "epoch": 0.02,
+      "step": 8
+    },
+    {
+      "epoch": 0.0225,
+      "grad_norm": 16.434814453125,
+      "learning_rate": 2.25e-06,
+      "loss": 0.6739,
+      "step": 9
+    },
+    {
+      "Batch Sum": 59.09326171875,
+      "accuracy": 0.453125,
+      "epoch": 0.0225,
+      "step": 9
+    },
+    {
+      "Batch Sum": 61.3369140625,
+      "accuracy": 0.546875,
+      "epoch": 0.0225,
+      "step": 9
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 13.62090015411377,
+      "learning_rate": 2.5e-06,
+      "loss": 0.6856,
+      "step": 10
+    },
+    {
+      "Batch Sum": 70.0703125,
+      "accuracy": 0.609375,
+      "epoch": 0.025,
+      "step": 10
+    },
+    {
+      "Batch Sum": 64.751953125,
+      "accuracy": 0.578125,
+      "epoch": 0.025,
+      "step": 10
+    },
+    {
+      "epoch": 0.0275,
+      "grad_norm": 13.072342872619629,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.6546,
+      "step": 11
+    },
+    {
+      "Batch Sum": 73.357421875,
+      "accuracy": 0.609375,
+      "epoch": 0.0275,
+      "step": 11
+    },
+    {
+      "Batch Sum": 72.185546875,
+      "accuracy": 0.578125,
+      "epoch": 0.0275,
+      "step": 11
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.209249496459961,
+      "learning_rate": 3e-06,
+      "loss": 0.6519,
+      "step": 12
+    },
+    {
+      "Batch Sum": 77.0615234375,
+      "accuracy": 0.703125,
+      "epoch": 0.03,
+      "step": 12
+    },
+    {
+      "Batch Sum": 78.0577392578125,
+      "accuracy": 0.703125,
+      "epoch": 0.03,
+      "step": 12
+    },
+    {
+      "epoch": 0.0325,
+      "grad_norm": 12.592546463012695,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.6092,
+      "step": 13
+    },
+    {
+      "Batch Sum": 91.876953125,
+      "accuracy": 0.625,
+      "epoch": 0.0325,
+      "step": 13
+    },
+    {
+      "Batch Sum": 87.7666015625,
+      "accuracy": 0.6875,
+      "epoch": 0.0325,
+      "step": 13
+    },
+    {
+      "epoch": 0.035,
+      "grad_norm": 14.43427848815918,
+      "learning_rate": 3.5e-06,
+      "loss": 0.601,
+      "step": 14
+    },
+    {
+      "Batch Sum": 97.5859375,
+      "accuracy": 0.65625,
+      "epoch": 0.035,
+      "step": 14
+    },
+    {
+      "Batch Sum": 107.59765625,
+      "accuracy": 0.71875,
+      "epoch": 0.035,
+      "step": 14
+    },
+    {
+      "epoch": 0.0375,
+      "grad_norm": 12.276965141296387,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.6045,
+      "step": 15
+    },
+    {
+      "Batch Sum": 111.8828125,
+      "accuracy": 0.59375,
+      "epoch": 0.0375,
+      "step": 15
+    },
+    {
+      "Batch Sum": 116.759765625,
+      "accuracy": 0.6875,
+      "epoch": 0.0375,
+      "step": 15
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.269201278686523,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.6262,
+      "step": 16
+    },
+    {
+      "Batch Sum": 120.9140625,
+      "accuracy": 0.625,
+      "epoch": 0.04,
+      "step": 16
+    },
+    {
+      "Batch Sum": 125.150390625,
+      "accuracy": 0.703125,
+      "epoch": 0.04,
+      "step": 16
+    },
+    {
+      "epoch": 0.0425,
+      "grad_norm": 12.212128639221191,
+      "learning_rate": 4.25e-06,
+      "loss": 0.6265,
+      "step": 17
+    },
+    {
+      "Batch Sum": 136.623046875,
+      "accuracy": 0.71875,
+      "epoch": 0.0425,
+      "step": 17
+    },
+    {
+      "Batch Sum": 133.275390625,
+      "accuracy": 0.640625,
+      "epoch": 0.0425,
+      "step": 17
+    },
+    {
+      "epoch": 0.045,
+      "grad_norm": 17.11771583557129,
+      "learning_rate": 4.5e-06,
+      "loss": 0.596,
+      "step": 18
+    },
+    {
+      "Batch Sum": 155.64453125,
+      "accuracy": 0.71875,
+      "epoch": 0.045,
+      "step": 18
+    },
+    {
+      "Batch Sum": 152.97265625,
+      "accuracy": 0.671875,
+      "epoch": 0.045,
+      "step": 18
+    },
+    {
+      "epoch": 0.0475,
+      "grad_norm": 17.395545959472656,
+      "learning_rate": 4.75e-06,
+      "loss": 0.6295,
+      "step": 19
+    },
+    {
+      "Batch Sum": 173.0703125,
+      "accuracy": 0.65625,
+      "epoch": 0.0475,
+      "step": 19
+    },
+    {
+      "Batch Sum": 173.05078125,
+      "accuracy": 0.671875,
+      "epoch": 0.0475,
+      "step": 19
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 24.041831970214844,
+      "learning_rate": 5e-06,
+      "loss": 0.6095,
+      "step": 20
+    },
+    {
+      "Batch Sum": 191.30859375,
+      "accuracy": 0.765625,
+      "epoch": 0.05,
+      "step": 20
+    },
+    {
+      "Batch Sum": 196.74609375,
+      "accuracy": 0.6875,
+      "epoch": 0.05,
+      "step": 20
+    },
+    {
+      "epoch": 0.0525,
+      "grad_norm": 13.471965789794922,
+      "learning_rate": 4.986842105263158e-06,
+      "loss": 0.527,
+      "step": 21
+    },
+    {
+      "Batch Sum": 205.40625,
+      "accuracy": 0.703125,
+      "epoch": 0.0525,
+      "step": 21
+    },
+    {
+      "Batch Sum": 204.26953125,
+      "accuracy": 0.703125,
+      "epoch": 0.0525,
+      "step": 21
+    },
+    {
+      "epoch": 0.055,
+      "grad_norm": 22.169654846191406,
+      "learning_rate": 4.973684210526316e-06,
+      "loss": 0.6532,
+      "step": 22
+    },
+    {
+      "Batch Sum": 208.29296875,
+      "accuracy": 0.578125,
+      "epoch": 0.055,
+      "step": 22
+    },
+    {
+      "Batch Sum": 208.50390625,
+      "accuracy": 0.75,
+      "epoch": 0.055,
+      "step": 22
+    },
+    {
+      "epoch": 0.0575,
+      "grad_norm": 21.459922790527344,
+      "learning_rate": 4.960526315789474e-06,
+      "loss": 0.6456,
+      "step": 23
+    },
+    {
+      "Batch Sum": 207.7421875,
+      "accuracy": 0.78125,
+      "epoch": 0.0575,
+      "step": 23
+    },
+    {
+      "Batch Sum": 205.7421875,
+      "accuracy": 0.625,
+      "epoch": 0.0575,
+      "step": 23
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.531883239746094,
+      "learning_rate": 4.947368421052632e-06,
+      "loss": 0.5219,
+      "step": 24
+    },
+    {
+      "Batch Sum": 211.6640625,
+      "accuracy": 0.71875,
+      "epoch": 0.06,
+      "step": 24
+    },
+    {
+      "Batch Sum": 208.45703125,
+      "accuracy": 0.703125,
+      "epoch": 0.06,
+      "step": 24
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 16.02347755432129,
+      "learning_rate": 4.9342105263157895e-06,
+      "loss": 0.5895,
+      "step": 25
+    },
+    {
+      "Batch Sum": 213.375,
+      "accuracy": 0.640625,
+      "epoch": 0.0625,
+      "step": 25
+    },
+    {
+      "Batch Sum": 203.58203125,
+      "accuracy": 0.703125,
+      "epoch": 0.0625,
+      "step": 25
+    },
+    {
+      "epoch": 0.065,
+      "grad_norm": 12.315845489501953,
+      "learning_rate": 4.921052631578948e-06,
+      "loss": 0.5732,
+      "step": 26
+    },
+    {
+      "Batch Sum": 202.23828125,
+      "accuracy": 0.734375,
+      "epoch": 0.065,
+      "step": 26
+    },
+    {
+      "Batch Sum": 197.6953125,
+      "accuracy": 0.703125,
+      "epoch": 0.065,
+      "step": 26
+    },
+    {
+      "epoch": 0.0675,
+      "grad_norm": 10.496895790100098,
+      "learning_rate": 4.907894736842106e-06,
+      "loss": 0.5467,
+      "step": 27
+    },
+    {
+      "Batch Sum": 183.37890625,
+      "accuracy": 0.703125,
+      "epoch": 0.0675,
+      "step": 27
+    },
+    {
+      "Batch Sum": 185.86328125,
+      "accuracy": 0.75,
+      "epoch": 0.0675,
+      "step": 27
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.82455825805664,
+      "learning_rate": 4.894736842105264e-06,
+      "loss": 0.5028,
+      "step": 28
+    },
+    {
+      "Batch Sum": 179.2158203125,
+      "accuracy": 0.71875,
+      "epoch": 0.07,
+      "step": 28
+    },
+    {
+      "Batch Sum": 179.328125,
+      "accuracy": 0.734375,
+      "epoch": 0.07,
+      "step": 28
+    },
+    {
+      "epoch": 0.0725,
+      "grad_norm": 13.485640525817871,
+      "learning_rate": 4.881578947368422e-06,
+      "loss": 0.5516,
+      "step": 29
+    },
+    {
+      "Batch Sum": 174.9296875,
+      "accuracy": 0.703125,
+      "epoch": 0.0725,
+      "step": 29
+    },
+    {
+      "Batch Sum": 172.44921875,
+      "accuracy": 0.734375,
+      "epoch": 0.0725,
+      "step": 29
+    },
+    {
+      "epoch": 0.075,
+      "grad_norm": 11.573375701904297,
+      "learning_rate": 4.8684210526315795e-06,
+      "loss": 0.5523,
+      "step": 30
+    },
+    {
+      "Batch Sum": 170.3046875,
+      "accuracy": 0.671875,
+      "epoch": 0.075,
+      "step": 30
+    },
+    {
+      "Batch Sum": 174.54296875,
+      "accuracy": 0.703125,
+      "epoch": 0.075,
+      "step": 30
+    },
+    {
+      "epoch": 0.0775,
+      "grad_norm": 10.271368026733398,
+      "learning_rate": 4.855263157894737e-06,
+      "loss": 0.5549,
+      "step": 31
+    },
+    {
+      "Batch Sum": 168.15625,
+      "accuracy": 0.75,
+      "epoch": 0.0775,
+      "step": 31
+    },
+    {
+      "Batch Sum": 169.49609375,
+      "accuracy": 0.671875,
+      "epoch": 0.0775,
+      "step": 31
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.423661231994629,
+      "learning_rate": 4.842105263157895e-06,
+      "loss": 0.5279,
+      "step": 32
+    },
+    {
+      "Batch Sum": 166.53125,
+      "accuracy": 0.765625,
+      "epoch": 0.08,
+      "step": 32
+    },
+    {
+      "Batch Sum": 169.66015625,
+      "accuracy": 0.75,
+      "epoch": 0.08,
+      "step": 32
+    },
+    {
+      "epoch": 0.0825,
+      "grad_norm": 11.485343933105469,
+      "learning_rate": 4.828947368421053e-06,
+      "loss": 0.5109,
+      "step": 33
+    },
+    {
+      "Batch Sum": 169.38671875,
+      "accuracy": 0.6875,
+      "epoch": 0.0825,
+      "step": 33
+    },
+    {
+      "Batch Sum": 172.9375,
+      "accuracy": 0.796875,
+      "epoch": 0.0825,
+      "step": 33
+    },
+    {
+      "epoch": 0.085,
+      "grad_norm": 10.612264633178711,
+      "learning_rate": 4.815789473684211e-06,
+      "loss": 0.555,
+      "step": 34
+    },
+    {
+      "Batch Sum": 179.78515625,
+      "accuracy": 0.71875,
+      "epoch": 0.085,
+      "step": 34
+    },
+    {
+      "Batch Sum": 173.041015625,
+      "accuracy": 0.75,
+      "epoch": 0.085,
+      "step": 34
+    },
+    {
+      "epoch": 0.0875,
+      "grad_norm": 9.249446868896484,
+      "learning_rate": 4.802631578947369e-06,
+      "loss": 0.528,
+      "step": 35
+    },
+    {
+      "Batch Sum": 169.671875,
+      "accuracy": 0.765625,
+      "epoch": 0.0875,
+      "step": 35
+    },
+    {
+      "Batch Sum": 184.578125,
+      "accuracy": 0.734375,
+      "epoch": 0.0875,
+      "step": 35
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.817283630371094,
+      "learning_rate": 4.789473684210527e-06,
+      "loss": 0.452,
+      "step": 36
+    },
+    {
+      "Batch Sum": 182.33984375,
+      "accuracy": 0.765625,
+      "epoch": 0.09,
+      "step": 36
+    },
+    {
+      "Batch Sum": 190.103515625,
+      "accuracy": 0.75,
+      "epoch": 0.09,
+      "step": 36
+    },
+    {
+      "epoch": 0.0925,
+      "grad_norm": 18.01405906677246,
+      "learning_rate": 4.7763157894736844e-06,
+      "loss": 0.5748,
+      "step": 37
+    },
+    {
+      "Batch Sum": 192.9296875,
+      "accuracy": 0.796875,
+      "epoch": 0.0925,
+      "step": 37
+    },
+    {
+      "Batch Sum": 195.75390625,
+      "accuracy": 0.734375,
+      "epoch": 0.0925,
+      "step": 37
+    },
+    {
+      "epoch": 0.095,
+      "grad_norm": 10.098393440246582,
+      "learning_rate": 4.763157894736842e-06,
+      "loss": 0.4995,
+      "step": 38
+    },
+    {
+      "Batch Sum": 199.21484375,
+      "accuracy": 0.75,
+      "epoch": 0.095,
+      "step": 38
+    },
+    {
+      "Batch Sum": 204.1640625,
+      "accuracy": 0.703125,
+      "epoch": 0.095,
+      "step": 38
+    },
+    {
+      "epoch": 0.0975,
+      "grad_norm": 13.419852256774902,
+      "learning_rate": 4.75e-06,
+      "loss": 0.5398,
+      "step": 39
+    },
+    {
+      "Batch Sum": 209.2578125,
+      "accuracy": 0.71875,
+      "epoch": 0.0975,
+      "step": 39
+    },
+    {
+      "Batch Sum": 213.6796875,
+      "accuracy": 0.640625,
+      "epoch": 0.0975,
+      "step": 39
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.879420280456543,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 0.5389,
+      "step": 40
+    },
+    {
+      "Batch Sum": 220.0625,
+      "accuracy": 0.734375,
+      "epoch": 0.1,
+      "step": 40
+    },
+    {
+      "Batch Sum": 214.107421875,
+      "accuracy": 0.703125,
+      "epoch": 0.1,
+      "step": 40
+    },
+    {
+      "epoch": 0.1025,
+      "grad_norm": 10.593208312988281,
+      "learning_rate": 4.723684210526316e-06,
+      "loss": 0.4675,
+      "step": 41
+    },
+    {
+      "Batch Sum": 225.80078125,
+      "accuracy": 0.796875,
+      "epoch": 0.1025,
+      "step": 41
+    },
+    {
+      "Batch Sum": 218.39453125,
+      "accuracy": 0.765625,
+      "epoch": 0.1025,
+      "step": 41
+    },
+    {
+      "epoch": 0.105,
+      "grad_norm": 10.301755905151367,
+      "learning_rate": 4.710526315789474e-06,
+      "loss": 0.4833,
+      "step": 42
+    },
+    {
+      "Batch Sum": 234.8203125,
+      "accuracy": 0.765625,
+      "epoch": 0.105,
+      "step": 42
+    },
+    {
+      "Batch Sum": 229.140625,
+      "accuracy": 0.734375,
+      "epoch": 0.105,
+      "step": 42
+    },
+    {
+      "epoch": 0.1075,
+      "grad_norm": 11.743901252746582,
+      "learning_rate": 4.697368421052632e-06,
+      "loss": 0.5026,
+      "step": 43
+    },
+    {
+      "Batch Sum": 226.81640625,
+      "accuracy": 0.734375,
+      "epoch": 0.1075,
+      "step": 43
+    },
+    {
+      "Batch Sum": 225.296875,
+      "accuracy": 0.75,
+      "epoch": 0.1075,
+      "step": 43
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.593162536621094,
+      "learning_rate": 4.68421052631579e-06,
+      "loss": 0.5254,
+      "step": 44
+    },
+    {
+      "Batch Sum": 227.609375,
+      "accuracy": 0.703125,
+      "epoch": 0.11,
+      "step": 44
+    },
+    {
+      "Batch Sum": 216.0625,
+      "accuracy": 0.765625,
+      "epoch": 0.11,
+      "step": 44
+    },
+    {
+      "epoch": 0.1125,
+      "grad_norm": 11.014060974121094,
+      "learning_rate": 4.671052631578948e-06,
+      "loss": 0.4679,
+      "step": 45
+    },
+    {
+      "Batch Sum": 227.234375,
+      "accuracy": 0.640625,
+      "epoch": 0.1125,
+      "step": 45
+    },
+    {
+      "Batch Sum": 222.55078125,
+      "accuracy": 0.734375,
+      "epoch": 0.1125,
+      "step": 45
+    },
+    {
+      "epoch": 0.115,
+      "grad_norm": 14.091309547424316,
+      "learning_rate": 4.657894736842106e-06,
+      "loss": 0.5738,
+      "step": 46
+    },
+    {
+      "Batch Sum": 217.4609375,
+      "accuracy": 0.6875,
+      "epoch": 0.115,
+      "step": 46
+    },
+    {
+      "Batch Sum": 224.85546875,
+      "accuracy": 0.671875,
+      "epoch": 0.115,
+      "step": 46
+    },
+    {
+      "epoch": 0.1175,
+      "grad_norm": 12.615903854370117,
+      "learning_rate": 4.6447368421052635e-06,
+      "loss": 0.5464,
+      "step": 47
+    },
+    {
+      "Batch Sum": 211.90625,
+      "accuracy": 0.75,
+      "epoch": 0.1175,
+      "step": 47
+    },
+    {
+      "Batch Sum": 207.115234375,
+      "accuracy": 0.734375,
+      "epoch": 0.1175,
+      "step": 47
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.995285987854004,
+      "learning_rate": 4.631578947368421e-06,
+      "loss": 0.4911,
+      "step": 48
+    },
+    {
+      "Batch Sum": 198.1484375,
+      "accuracy": 0.734375,
+      "epoch": 0.12,
+      "step": 48
+    },
+    {
+      "Batch Sum": 195.65234375,
+      "accuracy": 0.765625,
+      "epoch": 0.12,
+      "step": 48
+    },
+    {
+      "epoch": 0.1225,
+      "grad_norm": 12.166397094726562,
+      "learning_rate": 4.618421052631579e-06,
+      "loss": 0.4656,
+      "step": 49
+    },
+    {
+      "Batch Sum": 192.26953125,
+      "accuracy": 0.703125,
+      "epoch": 0.1225,
+      "step": 49
+    },
+    {
+      "Batch Sum": 187.31640625,
+      "accuracy": 0.78125,
+      "epoch": 0.1225,
+      "step": 49
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 12.223002433776855,
+      "learning_rate": 4.605263157894737e-06,
+      "loss": 0.5363,
+      "step": 50
+    },
+    {
+      "Batch Sum": 192.75390625,
+      "accuracy": 0.796875,
+      "epoch": 0.125,
+      "step": 50
+    },
+    {
+      "Batch Sum": 182.095703125,
+      "accuracy": 0.734375,
+      "epoch": 0.125,
+      "step": 50
+    },
+    {
+      "epoch": 0.1275,
+      "grad_norm": 10.82153606414795,
+      "learning_rate": 4.592105263157895e-06,
+      "loss": 0.4713,
+      "step": 51
+    },
+    {
+      "Batch Sum": 187.1015625,
+      "accuracy": 0.71875,
+      "epoch": 0.1275,
+      "step": 51
+    },
+    {
+      "Batch Sum": 185.41796875,
+      "accuracy": 0.703125,
+      "epoch": 0.1275,
+      "step": 51
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.1001558303833,
+      "learning_rate": 4.578947368421053e-06,
+      "loss": 0.5496,
+      "step": 52
+    },
+    {
+      "Batch Sum": 199.91015625,
+      "accuracy": 0.765625,
+      "epoch": 0.13,
+      "step": 52
+    },
+    {
+      "Batch Sum": 202.24609375,
+      "accuracy": 0.734375,
+      "epoch": 0.13,
+      "step": 52
+    },
+    {
+      "epoch": 0.1325,
+      "grad_norm": 11.745606422424316,
+      "learning_rate": 4.565789473684211e-06,
+      "loss": 0.4841,
+      "step": 53
+    },
+    {
+      "Batch Sum": 197.15625,
+      "accuracy": 0.765625,
+      "epoch": 0.1325,
+      "step": 53
+    },
+    {
+      "Batch Sum": 200.75,
+      "accuracy": 0.71875,
+      "epoch": 0.1325,
+      "step": 53
+    },
+    {
+      "epoch": 0.135,
+      "grad_norm": 12.604297637939453,
+      "learning_rate": 4.552631578947369e-06,
+      "loss": 0.4502,
+      "step": 54
+    },
+    {
+      "Batch Sum": 214.22265625,
+      "accuracy": 0.6875,
+      "epoch": 0.135,
+      "step": 54
+    },
+    {
+      "Batch Sum": 207.84765625,
+      "accuracy": 0.78125,
+      "epoch": 0.135,
+      "step": 54
+    },
+    {
+      "epoch": 0.1375,
+      "grad_norm": 14.610075950622559,
+      "learning_rate": 4.539473684210527e-06,
+      "loss": 0.5449,
+      "step": 55
+    },
+    {
+      "Batch Sum": 221.5546875,
+      "accuracy": 0.875,
+      "epoch": 0.1375,
+      "step": 55
+    },
+    {
+      "Batch Sum": 236.11328125,
+      "accuracy": 0.765625,
+      "epoch": 0.1375,
+      "step": 55
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.892423629760742,
+      "learning_rate": 4.526315789473685e-06,
+      "loss": 0.4002,
+      "step": 56
+    },
+    {
+      "Batch Sum": 246.265625,
+      "accuracy": 0.734375,
+      "epoch": 0.14,
+      "step": 56
+    },
+    {
+      "Batch Sum": 233.033203125,
+      "accuracy": 0.828125,
+      "epoch": 0.14,
+      "step": 56
+    },
+    {
+      "epoch": 0.1425,
+      "grad_norm": 13.1007719039917,
+      "learning_rate": 4.513157894736843e-06,
+      "loss": 0.4596,
+      "step": 57
+    },
+    {
+      "Batch Sum": 239.265625,
+      "accuracy": 0.78125,
+      "epoch": 0.1425,
+      "step": 57
+    },
+    {
+      "Batch Sum": 251.70703125,
+      "accuracy": 0.6875,
+      "epoch": 0.1425,
+      "step": 57
+    },
+    {
+      "epoch": 0.145,
+      "grad_norm": 14.813791275024414,
+      "learning_rate": 4.5e-06,
+      "loss": 0.5175,
+      "step": 58
+    },
+    {
+      "Batch Sum": 236.10546875,
+      "accuracy": 0.765625,
+      "epoch": 0.145,
+      "step": 58
+    },
+    {
+      "Batch Sum": 249.8046875,
+      "accuracy": 0.734375,
+      "epoch": 0.145,
+      "step": 58
+    },
+    {
+      "epoch": 0.1475,
+      "grad_norm": 14.486608505249023,
+      "learning_rate": 4.4868421052631584e-06,
+      "loss": 0.5035,
+      "step": 59
+    },
+    {
+      "Batch Sum": 241.12109375,
+      "accuracy": 0.78125,
+      "epoch": 0.1475,
+      "step": 59
+    },
+    {
+      "Batch Sum": 229.68359375,
+      "accuracy": 0.796875,
+      "epoch": 0.1475,
+      "step": 59
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.024293899536133,
+      "learning_rate": 4.473684210526316e-06,
+      "loss": 0.4143,
+      "step": 60
+    },
+    {
+      "Batch Sum": 251.0,
+      "accuracy": 0.796875,
+      "epoch": 0.15,
+      "step": 60
+    },
+    {
+      "Batch Sum": 245.59765625,
+      "accuracy": 0.78125,
+      "epoch": 0.15,
+      "step": 60
+    },
+    {
+      "epoch": 0.1525,
+      "grad_norm": 11.371909141540527,
+      "learning_rate": 4.460526315789474e-06,
+      "loss": 0.3862,
+      "step": 61
+    },
+    {
+      "Batch Sum": 239.98828125,
+      "accuracy": 0.875,
+      "epoch": 0.1525,
+      "step": 61
+    },
+    {
+      "Batch Sum": 243.82421875,
+      "accuracy": 0.78125,
+      "epoch": 0.1525,
+      "step": 61
+    },
+    {
+      "epoch": 0.155,
+      "grad_norm": 11.827494621276855,
+      "learning_rate": 4.447368421052632e-06,
+      "loss": 0.3865,
+      "step": 62
+    },
+    {
+      "Batch Sum": 230.6640625,
+      "accuracy": 0.734375,
+      "epoch": 0.155,
+      "step": 62
+    },
+    {
+      "Batch Sum": 221.58984375,
+      "accuracy": 0.796875,
+      "epoch": 0.155,
+      "step": 62
+    },
+    {
+      "epoch": 0.1575,
+      "grad_norm": 11.711631774902344,
+      "learning_rate": 4.43421052631579e-06,
+      "loss": 0.4655,
+      "step": 63
+    },
+    {
+      "Batch Sum": 232.0703125,
+      "accuracy": 0.75,
+      "epoch": 0.1575,
+      "step": 63
+    },
+    {
+      "Batch Sum": 224.037109375,
+      "accuracy": 0.8125,
+      "epoch": 0.1575,
+      "step": 63
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.687829971313477,
+      "learning_rate": 4.4210526315789476e-06,
+      "loss": 0.4065,
+      "step": 64
+    },
+    {
+      "Batch Sum": 232.53515625,
+      "accuracy": 0.671875,
+      "epoch": 0.16,
+      "step": 64
+    },
+    {
+      "Batch Sum": 225.68359375,
+      "accuracy": 0.734375,
+      "epoch": 0.16,
+      "step": 64
+    },
+    {
+      "epoch": 0.1625,
+      "grad_norm": 13.95382022857666,
+      "learning_rate": 4.407894736842105e-06,
+      "loss": 0.5705,
+      "step": 65
+    },
+    {
+      "Batch Sum": 237.1796875,
+      "accuracy": 0.859375,
+      "epoch": 0.1625,
+      "step": 65
+    },
+    {
+      "Batch Sum": 234.703125,
+      "accuracy": 0.84375,
+      "epoch": 0.1625,
+      "step": 65
+    },
+    {
+      "epoch": 0.165,
+      "grad_norm": 10.754307746887207,
+      "learning_rate": 4.394736842105263e-06,
+      "loss": 0.3935,
+      "step": 66
+    },
+    {
+      "Batch Sum": 216.4921875,
+      "accuracy": 0.8125,
+      "epoch": 0.165,
+      "step": 66
+    },
+    {
+      "Batch Sum": 238.1640625,
+      "accuracy": 0.828125,
+      "epoch": 0.165,
+      "step": 66
+    },
+    {
+      "epoch": 0.1675,
+      "grad_norm": 11.184181213378906,
+      "learning_rate": 4.381578947368421e-06,
+      "loss": 0.3747,
+      "step": 67
+    },
+    {
+      "Batch Sum": 211.755859375,
+      "accuracy": 0.828125,
+      "epoch": 0.1675,
+      "step": 67
+    },
+    {
+      "Batch Sum": 225.146484375,
+      "accuracy": 0.84375,
+      "epoch": 0.1675,
+      "step": 67
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.842012405395508,
+      "learning_rate": 4.368421052631579e-06,
+      "loss": 0.3372,
+      "step": 68
+    },
+    {
+      "Batch Sum": 227.8291015625,
+      "accuracy": 0.796875,
+      "epoch": 0.17,
+      "step": 68
+    },
+    {
+      "Batch Sum": 214.9453125,
+      "accuracy": 0.78125,
+      "epoch": 0.17,
+      "step": 68
+    },
+    {
+      "epoch": 0.1725,
+      "grad_norm": 12.21465015411377,
+      "learning_rate": 4.3552631578947375e-06,
+      "loss": 0.4673,
+      "step": 69
+    },
+    {
+      "Batch Sum": 230.5546875,
+      "accuracy": 0.859375,
+      "epoch": 0.1725,
+      "step": 69
+    },
+    {
+      "Batch Sum": 229.95703125,
+      "accuracy": 0.796875,
+      "epoch": 0.1725,
+      "step": 69
+    },
+    {
+      "epoch": 0.175,
+      "grad_norm": 12.969797134399414,
+      "learning_rate": 4.342105263157895e-06,
+      "loss": 0.4088,
+      "step": 70
+    },
+    {
+      "Batch Sum": 224.419921875,
+      "accuracy": 0.8125,
+      "epoch": 0.175,
+      "step": 70
+    },
+    {
+      "Batch Sum": 221.15234375,
+      "accuracy": 0.859375,
+      "epoch": 0.175,
+      "step": 70
+    },
+    {
+      "epoch": 0.1775,
+      "grad_norm": 12.190828323364258,
+      "learning_rate": 4.328947368421053e-06,
+      "loss": 0.4531,
+      "step": 71
+    },
+    {
+      "Batch Sum": 212.4345703125,
+      "accuracy": 0.78125,
+      "epoch": 0.1775,
+      "step": 71
+    },
+    {
+      "Batch Sum": 209.00390625,
+      "accuracy": 0.6875,
+      "epoch": 0.1775,
+      "step": 71
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.806035041809082,
+      "learning_rate": 4.315789473684211e-06,
+      "loss": 0.4743,
+      "step": 72
+    },
+    {
+      "Batch Sum": 203.03515625,
+      "accuracy": 0.84375,
+      "epoch": 0.18,
+      "step": 72
+    },
+    {
+      "Batch Sum": 223.71875,
+      "accuracy": 0.84375,
+      "epoch": 0.18,
+      "step": 72
+    },
+    {
+      "epoch": 0.1825,
+      "grad_norm": 11.65701961517334,
+      "learning_rate": 4.302631578947369e-06,
+      "loss": 0.3151,
+      "step": 73
+    },
+    {
+      "Batch Sum": 206.0234375,
+      "accuracy": 0.75,
+      "epoch": 0.1825,
+      "step": 73
+    },
+    {
+      "Batch Sum": 213.287109375,
+      "accuracy": 0.8125,
+      "epoch": 0.1825,
+      "step": 73
+    },
+    {
+      "epoch": 0.185,
+      "grad_norm": 14.198211669921875,
+      "learning_rate": 4.289473684210527e-06,
+      "loss": 0.4365,
+      "step": 74
+    },
+    {
+      "Batch Sum": 209.15234375,
+      "accuracy": 0.8125,
+      "epoch": 0.185,
+      "step": 74
+    },
+    {
+      "Batch Sum": 219.087890625,
+      "accuracy": 0.78125,
+      "epoch": 0.185,
+      "step": 74
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 13.063102722167969,
+      "learning_rate": 4.276315789473684e-06,
+      "loss": 0.4194,
+      "step": 75
+    },
+    {
+      "Batch Sum": 206.705078125,
+      "accuracy": 0.828125,
+      "epoch": 0.1875,
+      "step": 75
+    },
+    {
+      "Batch Sum": 208.162109375,
+      "accuracy": 0.828125,
+      "epoch": 0.1875,
+      "step": 75
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.24936580657959,
+      "learning_rate": 4.2631578947368425e-06,
+      "loss": 0.4291,
+      "step": 76
+    },
+    {
+      "Batch Sum": 198.51953125,
+      "accuracy": 0.703125,
+      "epoch": 0.19,
+      "step": 76
+    },
+    {
+      "Batch Sum": 197.17578125,
+      "accuracy": 0.78125,
+      "epoch": 0.19,
+      "step": 76
+    },
+    {
+      "epoch": 0.1925,
+      "grad_norm": 21.39528465270996,
+      "learning_rate": 4.25e-06,
+      "loss": 0.4907,
+      "step": 77
+    },
+    {
+      "Batch Sum": 189.150390625,
+      "accuracy": 0.75,
+      "epoch": 0.1925,
+      "step": 77
+    },
+    {
+      "Batch Sum": 189.181640625,
+      "accuracy": 0.8125,
+      "epoch": 0.1925,
+      "step": 77
+    },
+    {
+      "epoch": 0.195,
+      "grad_norm": 12.943084716796875,
+      "learning_rate": 4.236842105263158e-06,
+      "loss": 0.4025,
+      "step": 78
+    },
+    {
+      "Batch Sum": 199.0009765625,
+      "accuracy": 0.75,
+      "epoch": 0.195,
+      "step": 78
+    },
+    {
+      "Batch Sum": 197.556640625,
+      "accuracy": 0.71875,
+      "epoch": 0.195,
+      "step": 78
+    },
+    {
+      "epoch": 0.1975,
+      "grad_norm": 14.870590209960938,
+      "learning_rate": 4.223684210526316e-06,
+      "loss": 0.4874,
+      "step": 79
+    },
+    {
+      "Batch Sum": 187.6875,
+      "accuracy": 0.875,
+      "epoch": 0.1975,
+      "step": 79
+    },
+    {
+      "Batch Sum": 191.8916015625,
+      "accuracy": 0.765625,
+      "epoch": 0.1975,
+      "step": 79
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.593913078308105,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 0.3468,
+      "step": 80
+    },
+    {
+      "Batch Sum": 173.8515625,
+      "accuracy": 0.703125,
+      "epoch": 0.2,
+      "step": 80
+    },
+    {
+      "Batch Sum": 185.919921875,
+      "accuracy": 0.8125,
+      "epoch": 0.2,
+      "step": 80
+    },
+    {
+      "epoch": 0.2025,
+      "grad_norm": 11.869901657104492,
+      "learning_rate": 4.197368421052632e-06,
+      "loss": 0.4181,
+      "step": 81
+    },
+    {
+      "Batch Sum": 173.52197265625,
+      "accuracy": 0.828125,
+      "epoch": 0.2025,
+      "step": 81
+    },
+    {
+      "Batch Sum": 164.21044921875,
+      "accuracy": 0.84375,
+      "epoch": 0.2025,
+      "step": 81
+    },
+    {
+      "epoch": 0.205,
+      "grad_norm": 11.220733642578125,
+      "learning_rate": 4.18421052631579e-06,
+      "loss": 0.385,
+      "step": 82
+    },
+    {
+      "Batch Sum": 158.6143798828125,
+      "accuracy": 0.90625,
+      "epoch": 0.205,
+      "step": 82
+    },
+    {
+      "Batch Sum": 185.5107421875,
+      "accuracy": 0.796875,
+      "epoch": 0.205,
+      "step": 82
+    },
+    {
+      "epoch": 0.2075,
+      "grad_norm": 11.525442123413086,
+      "learning_rate": 4.171052631578948e-06,
+      "loss": 0.3701,
+      "step": 83
+    },
+    {
+      "Batch Sum": 162.90478515625,
+      "accuracy": 0.734375,
+      "epoch": 0.2075,
+      "step": 83
+    },
+    {
+      "Batch Sum": 163.460693359375,
+      "accuracy": 0.78125,
+      "epoch": 0.2075,
+      "step": 83
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.735023498535156,
+      "learning_rate": 4.157894736842106e-06,
+      "loss": 0.5058,
+      "step": 84
+    },
+    {
+      "Batch Sum": 172.984375,
+      "accuracy": 0.71875,
+      "epoch": 0.21,
+      "step": 84
+    },
+    {
+      "Batch Sum": 151.54931640625,
+      "accuracy": 0.84375,
+      "epoch": 0.21,
+      "step": 84
+    },
+    {
+      "epoch": 0.2125,
+      "grad_norm": 12.669657707214355,
+      "learning_rate": 4.144736842105263e-06,
+      "loss": 0.4216,
+      "step": 85
+    },
+    {
+      "Batch Sum": 156.759765625,
+      "accuracy": 0.78125,
+      "epoch": 0.2125,
+      "step": 85
+    },
+    {
+      "Batch Sum": 157.17529296875,
+      "accuracy": 0.765625,
+      "epoch": 0.2125,
+      "step": 85
+    },
+    {
+      "epoch": 0.215,
+      "grad_norm": 12.282760620117188,
+      "learning_rate": 4.1315789473684216e-06,
+      "loss": 0.4175,
+      "step": 86
+    },
+    {
+      "Batch Sum": 153.5771484375,
+      "accuracy": 0.78125,
+      "epoch": 0.215,
+      "step": 86
+    },
+    {
+      "Batch Sum": 147.1307373046875,
+      "accuracy": 0.8125,
+      "epoch": 0.215,
+      "step": 86
+    },
+    {
+      "epoch": 0.2175,
+      "grad_norm": 13.767319679260254,
+      "learning_rate": 4.118421052631579e-06,
+      "loss": 0.4217,
+      "step": 87
+    },
+    {
+      "Batch Sum": 151.193115234375,
+      "accuracy": 0.734375,
+      "epoch": 0.2175,
+      "step": 87
+    },
+    {
+      "Batch Sum": 143.0897216796875,
+      "accuracy": 0.8125,
+      "epoch": 0.2175,
+      "step": 87
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.140963554382324,
+      "learning_rate": 4.105263157894737e-06,
+      "loss": 0.4242,
+      "step": 88
+    },
+    {
+      "Batch Sum": 145.87548828125,
+      "accuracy": 0.765625,
+      "epoch": 0.22,
+      "step": 88
+    },
+    {
+      "Batch Sum": 150.4423828125,
+      "accuracy": 0.828125,
+      "epoch": 0.22,
+      "step": 88
+    },
+    {
+      "epoch": 0.2225,
+      "grad_norm": 10.421849250793457,
+      "learning_rate": 4.092105263157895e-06,
+      "loss": 0.4024,
+      "step": 89
+    },
+    {
+      "Batch Sum": 164.34619140625,
+      "accuracy": 0.84375,
+      "epoch": 0.2225,
+      "step": 89
+    },
+    {
+      "Batch Sum": 151.582275390625,
+      "accuracy": 0.84375,
+      "epoch": 0.2225,
+      "step": 89
+    },
+    {
+      "epoch": 0.225,
+      "grad_norm": 11.950529098510742,
+      "learning_rate": 4.078947368421053e-06,
+      "loss": 0.4162,
+      "step": 90
+    },
+    {
+      "Batch Sum": 177.598388671875,
+      "accuracy": 0.78125,
+      "epoch": 0.225,
+      "step": 90
+    },
+    {
+      "Batch Sum": 168.056884765625,
+      "accuracy": 0.75,
+      "epoch": 0.225,
+      "step": 90
+    },
+    {
+      "epoch": 0.2275,
+      "grad_norm": 13.336840629577637,
+      "learning_rate": 4.065789473684211e-06,
+      "loss": 0.4356,
+      "step": 91
+    },
+    {
+      "Batch Sum": 185.6875,
+      "accuracy": 0.828125,
+      "epoch": 0.2275,
+      "step": 91
+    },
+    {
+      "Batch Sum": 156.2733154296875,
+      "accuracy": 0.859375,
+      "epoch": 0.2275,
+      "step": 91
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.884383201599121,
+      "learning_rate": 4.052631578947368e-06,
+      "loss": 0.3845,
+      "step": 92
+    },
+    {
+      "Batch Sum": 179.4698486328125,
+      "accuracy": 0.90625,
+      "epoch": 0.23,
+      "step": 92
+    },
+    {
+      "Batch Sum": 171.6319580078125,
+      "accuracy": 0.796875,
+      "epoch": 0.23,
+      "step": 92
+    },
+    {
+      "epoch": 0.2325,
+      "grad_norm": 11.024062156677246,
+      "learning_rate": 4.0394736842105265e-06,
+      "loss": 0.3715,
+      "step": 93
+    },
+    {
+      "Batch Sum": 170.8381805419922,
+      "accuracy": 0.75,
+      "epoch": 0.2325,
+      "step": 93
+    },
+    {
+      "Batch Sum": 173.384765625,
+      "accuracy": 0.8125,
+      "epoch": 0.2325,
+      "step": 93
+    },
+    {
+      "epoch": 0.235,
+      "grad_norm": 10.82882022857666,
+      "learning_rate": 4.026315789473684e-06,
+      "loss": 0.4289,
+      "step": 94
+    },
+    {
+      "Batch Sum": 164.244140625,
+      "accuracy": 0.75,
+      "epoch": 0.235,
+      "step": 94
+    },
+    {
+      "Batch Sum": 180.776123046875,
+      "accuracy": 0.796875,
+      "epoch": 0.235,
+      "step": 94
+    },
+    {
+      "epoch": 0.2375,
+      "grad_norm": 14.97029972076416,
+      "learning_rate": 4.013157894736842e-06,
+      "loss": 0.4478,
+      "step": 95
+    },
+    {
+      "Batch Sum": 174.9443359375,
+      "accuracy": 0.859375,
+      "epoch": 0.2375,
+      "step": 95
+    },
+    {
+      "Batch Sum": 168.7977294921875,
+      "accuracy": 0.84375,
+      "epoch": 0.2375,
+      "step": 95
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.570242881774902,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.3363,
+      "step": 96
+    },
+    {
+      "Batch Sum": 162.29507446289062,
+      "accuracy": 0.75,
+      "epoch": 0.24,
+      "step": 96
+    },
+    {
+      "Batch Sum": 167.9951171875,
+      "accuracy": 0.734375,
+      "epoch": 0.24,
+      "step": 96
+    },
+    {
+      "epoch": 0.2425,
+      "grad_norm": 17.29254913330078,
+      "learning_rate": 3.986842105263158e-06,
+      "loss": 0.5527,
+      "step": 97
+    },
+    {
+      "Batch Sum": 154.8450927734375,
+      "accuracy": 0.859375,
+      "epoch": 0.2425,
+      "step": 97
+    },
+    {
+      "Batch Sum": 164.18359375,
+      "accuracy": 0.8125,
+      "epoch": 0.2425,
+      "step": 97
+    },
+    {
+      "epoch": 0.245,
+      "grad_norm": 10.639801025390625,
+      "learning_rate": 3.9736842105263165e-06,
+      "loss": 0.3685,
+      "step": 98
+    },
+    {
+      "Batch Sum": 161.786865234375,
+      "accuracy": 0.765625,
+      "epoch": 0.245,
+      "step": 98
+    },
+    {
+      "Batch Sum": 167.194091796875,
+      "accuracy": 0.796875,
+      "epoch": 0.245,
+      "step": 98
+    },
+    {
+      "epoch": 0.2475,
+      "grad_norm": 11.962891578674316,
+      "learning_rate": 3.960526315789474e-06,
+      "loss": 0.4209,
+      "step": 99
+    },
+    {
+      "Batch Sum": 156.116943359375,
+      "accuracy": 0.796875,
+      "epoch": 0.2475,
+      "step": 99
+    },
+    {
+      "Batch Sum": 158.47119140625,
+      "accuracy": 0.75,
+      "epoch": 0.2475,
+      "step": 99
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.670126914978027,
+      "learning_rate": 3.947368421052632e-06,
+      "loss": 0.4568,
+      "step": 100
+    },
+    {
+      "Batch Sum": 151.5293731689453,
+      "accuracy": 0.796875,
+      "epoch": 0.25,
+      "step": 100
+    },
+    {
+      "Batch Sum": 139.802490234375,
+      "accuracy": 0.75,
+      "epoch": 0.25,
+      "step": 100
+    },
+    {
+      "epoch": 0.2525,
+      "grad_norm": 10.257286071777344,
+      "learning_rate": 3.93421052631579e-06,
+      "loss": 0.4065,
+      "step": 101
+    },
+    {
+      "Batch Sum": 145.30331420898438,
+      "accuracy": 0.734375,
+      "epoch": 0.2525,
+      "step": 101
+    },
+    {
+      "Batch Sum": 142.9066162109375,
+      "accuracy": 0.84375,
+      "epoch": 0.2525,
+      "step": 101
+    },
+    {
+      "epoch": 0.255,
+      "grad_norm": 8.931941032409668,
+      "learning_rate": 3.921052631578947e-06,
+      "loss": 0.3486,
+      "step": 102
+    },
+    {
+      "Batch Sum": 129.5537109375,
+      "accuracy": 0.765625,
+      "epoch": 0.255,
+      "step": 102
+    },
+    {
+      "Batch Sum": 139.361328125,
+      "accuracy": 0.84375,
+      "epoch": 0.255,
+      "step": 102
+    },
+    {
+      "epoch": 0.2575,
+      "grad_norm": 11.819692611694336,
+      "learning_rate": 3.907894736842106e-06,
+      "loss": 0.4768,
+      "step": 103
+    },
+    {
+      "Batch Sum": 136.63330078125,
+      "accuracy": 0.859375,
+      "epoch": 0.2575,
+      "step": 103
+    },
+    {
+      "Batch Sum": 126.9212646484375,
+      "accuracy": 0.859375,
+      "epoch": 0.2575,
+      "step": 103
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.561305046081543,
+      "learning_rate": 3.894736842105263e-06,
+      "loss": 0.3266,
+      "step": 104
+    },
+    {
+      "Batch Sum": 119.14749145507812,
+      "accuracy": 0.75,
+      "epoch": 0.26,
+      "step": 104
+    },
+    {
+      "Batch Sum": 115.628662109375,
+      "accuracy": 0.8125,
+      "epoch": 0.26,
+      "step": 104
+    },
+    {
+      "epoch": 0.2625,
+      "grad_norm": 28.31745719909668,
+      "learning_rate": 3.8815789473684214e-06,
+      "loss": 0.4407,
+      "step": 105
+    },
+    {
+      "Batch Sum": 127.26849365234375,
+      "accuracy": 0.796875,
+      "epoch": 0.2625,
+      "step": 105
+    },
+    {
+      "Batch Sum": 140.0218505859375,
+      "accuracy": 0.828125,
+      "epoch": 0.2625,
+      "step": 105
+    },
+    {
+      "epoch": 0.265,
+      "grad_norm": 12.298383712768555,
+      "learning_rate": 3.868421052631579e-06,
+      "loss": 0.4529,
+      "step": 106
+    },
+    {
+      "Batch Sum": 136.11062622070312,
+      "accuracy": 0.828125,
+      "epoch": 0.265,
+      "step": 106
+    },
+    {
+      "Batch Sum": 139.4700927734375,
+      "accuracy": 0.765625,
+      "epoch": 0.265,
+      "step": 106
+    },
+    {
+      "epoch": 0.2675,
+      "grad_norm": 11.664482116699219,
+      "learning_rate": 3.855263157894737e-06,
+      "loss": 0.4132,
+      "step": 107
+    },
+    {
+      "Batch Sum": 139.60125732421875,
+      "accuracy": 0.734375,
+      "epoch": 0.2675,
+      "step": 107
+    },
+    {
+      "Batch Sum": 137.16650390625,
+      "accuracy": 0.796875,
+      "epoch": 0.2675,
+      "step": 107
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.46422290802002,
+      "learning_rate": 3.842105263157895e-06,
+      "loss": 0.3772,
+      "step": 108
+    },
+    {
+      "Batch Sum": 130.9759521484375,
+      "accuracy": 0.71875,
+      "epoch": 0.27,
+      "step": 108
+    },
+    {
+      "Batch Sum": 132.84173583984375,
+      "accuracy": 0.78125,
+      "epoch": 0.27,
+      "step": 108
+    },
+    {
+      "epoch": 0.2725,
+      "grad_norm": 11.786149024963379,
+      "learning_rate": 3.828947368421053e-06,
+      "loss": 0.4741,
+      "step": 109
+    },
+    {
+      "Batch Sum": 131.52923583984375,
+      "accuracy": 0.703125,
+      "epoch": 0.2725,
+      "step": 109
+    },
+    {
+      "Batch Sum": 133.90191650390625,
+      "accuracy": 0.78125,
+      "epoch": 0.2725,
+      "step": 109
+    },
+    {
+      "epoch": 0.275,
+      "grad_norm": 13.17595100402832,
+      "learning_rate": 3.815789473684211e-06,
+      "loss": 0.521,
+      "step": 110
+    },
+    {
+      "Batch Sum": 108.73193359375,
+      "accuracy": 0.828125,
+      "epoch": 0.275,
+      "step": 110
+    },
+    {
+      "Batch Sum": 112.35197448730469,
+      "accuracy": 0.78125,
+      "epoch": 0.275,
+      "step": 110
+    },
+    {
+      "epoch": 0.2775,
+      "grad_norm": 10.771636009216309,
+      "learning_rate": 3.802631578947369e-06,
+      "loss": 0.3971,
+      "step": 111
+    },
+    {
+      "Batch Sum": 121.26911926269531,
+      "accuracy": 0.828125,
+      "epoch": 0.2775,
+      "step": 111
+    },
+    {
+      "Batch Sum": 126.85107421875,
+      "accuracy": 0.90625,
+      "epoch": 0.2775,
+      "step": 111
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.355616569519043,
+      "learning_rate": 3.789473684210527e-06,
+      "loss": 0.3219,
+      "step": 112
+    },
+    {
+      "Batch Sum": 112.09989929199219,
+      "accuracy": 0.796875,
+      "epoch": 0.28,
+      "step": 112
+    },
+    {
+      "Batch Sum": 118.8616943359375,
+      "accuracy": 0.75,
+      "epoch": 0.28,
+      "step": 112
+    },
+    {
+      "epoch": 0.2825,
+      "grad_norm": 9.97263240814209,
+      "learning_rate": 3.7763157894736847e-06,
+      "loss": 0.4534,
+      "step": 113
+    },
+    {
+      "Batch Sum": 112.02926635742188,
+      "accuracy": 0.84375,
+      "epoch": 0.2825,
+      "step": 113
+    },
+    {
+      "Batch Sum": 128.11767578125,
+      "accuracy": 0.90625,
+      "epoch": 0.2825,
+      "step": 113
+    },
+    {
+      "epoch": 0.285,
+      "grad_norm": 8.701355934143066,
+      "learning_rate": 3.7631578947368426e-06,
+      "loss": 0.3521,
+      "step": 114
+    },
+    {
+      "Batch Sum": 114.733154296875,
+      "accuracy": 0.828125,
+      "epoch": 0.285,
+      "step": 114
+    },
+    {
+      "Batch Sum": 120.71221923828125,
+      "accuracy": 0.875,
+      "epoch": 0.285,
+      "step": 114
+    },
+    {
+      "epoch": 0.2875,
+      "grad_norm": 10.165462493896484,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.3602,
+      "step": 115
+    },
+    {
+      "Batch Sum": 130.4957275390625,
+      "accuracy": 0.875,
+      "epoch": 0.2875,
+      "step": 115
+    },
+    {
+      "Batch Sum": 129.851318359375,
+      "accuracy": 0.828125,
+      "epoch": 0.2875,
+      "step": 115
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.85599422454834,
+      "learning_rate": 3.736842105263158e-06,
+      "loss": 0.3406,
+      "step": 116
+    },
+    {
+      "Batch Sum": 153.54833984375,
+      "accuracy": 0.828125,
+      "epoch": 0.29,
+      "step": 116
+    },
+    {
+      "Batch Sum": 117.29022216796875,
+      "accuracy": 0.734375,
+      "epoch": 0.29,
+      "step": 116
+    },
+    {
+      "epoch": 0.2925,
+      "grad_norm": 11.120735168457031,
+      "learning_rate": 3.723684210526316e-06,
+      "loss": 0.4414,
+      "step": 117
+    },
+    {
+      "Batch Sum": 133.2943115234375,
+      "accuracy": 0.8125,
+      "epoch": 0.2925,
+      "step": 117
+    },
+    {
+      "Batch Sum": 154.751708984375,
+      "accuracy": 0.765625,
+      "epoch": 0.2925,
+      "step": 117
+    },
+    {
+      "epoch": 0.295,
+      "grad_norm": 9.896623611450195,
+      "learning_rate": 3.710526315789474e-06,
+      "loss": 0.4025,
+      "step": 118
+    },
+    {
+      "Batch Sum": 146.9268798828125,
+      "accuracy": 0.796875,
+      "epoch": 0.295,
+      "step": 118
+    },
+    {
+      "Batch Sum": 157.404541015625,
+      "accuracy": 0.84375,
+      "epoch": 0.295,
+      "step": 118
+    },
+    {
+      "epoch": 0.2975,
+      "grad_norm": 12.310927391052246,
+      "learning_rate": 3.6973684210526317e-06,
+      "loss": 0.3888,
+      "step": 119
+    },
+    {
+      "Batch Sum": 177.720458984375,
+      "accuracy": 0.734375,
+      "epoch": 0.2975,
+      "step": 119
+    },
+    {
+      "Batch Sum": 154.5662841796875,
+      "accuracy": 0.828125,
+      "epoch": 0.2975,
+      "step": 119
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.395303726196289,
+      "learning_rate": 3.6842105263157896e-06,
+      "loss": 0.4263,
+      "step": 120
+    },
+    {
+      "Batch Sum": 165.2161865234375,
+      "accuracy": 0.75,
+      "epoch": 0.3,
+      "step": 120
+    },
+    {
+      "Batch Sum": 154.25457763671875,
+      "accuracy": 0.765625,
+      "epoch": 0.3,
+      "step": 120
+    },
+    {
+      "epoch": 0.3025,
+      "grad_norm": 18.222015380859375,
+      "learning_rate": 3.6710526315789476e-06,
+      "loss": 0.5583,
+      "step": 121
+    },
+    {
+      "Batch Sum": 156.56472778320312,
+      "accuracy": 0.734375,
+      "epoch": 0.3025,
+      "step": 121
+    },
+    {
+      "Batch Sum": 154.8548583984375,
+      "accuracy": 0.8125,
+      "epoch": 0.3025,
+      "step": 121
+    },
+    {
+      "epoch": 0.305,
+      "grad_norm": 12.073637008666992,
+      "learning_rate": 3.657894736842106e-06,
+      "loss": 0.4433,
+      "step": 122
+    },
+    {
+      "Batch Sum": 191.345703125,
+      "accuracy": 0.8125,
+      "epoch": 0.305,
+      "step": 122
+    },
+    {
+      "Batch Sum": 194.02685546875,
+      "accuracy": 0.859375,
+      "epoch": 0.305,
+      "step": 122
+    },
+    {
+      "epoch": 0.3075,
+      "grad_norm": 11.347955703735352,
+      "learning_rate": 3.644736842105264e-06,
+      "loss": 0.3678,
+      "step": 123
+    },
+    {
+      "Batch Sum": 154.679443359375,
+      "accuracy": 0.859375,
+      "epoch": 0.3075,
+      "step": 123
+    },
+    {
+      "Batch Sum": 151.9825439453125,
+      "accuracy": 0.890625,
+      "epoch": 0.3075,
+      "step": 123
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.588033676147461,
+      "learning_rate": 3.6315789473684217e-06,
+      "loss": 0.2785,
+      "step": 124
+    },
+    {
+      "Batch Sum": 163.2354278564453,
+      "accuracy": 0.796875,
+      "epoch": 0.31,
+      "step": 124
+    },
+    {
+      "Batch Sum": 176.15380859375,
+      "accuracy": 0.765625,
+      "epoch": 0.31,
+      "step": 124
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 13.149937629699707,
+      "learning_rate": 3.618421052631579e-06,
+      "loss": 0.4441,
+      "step": 125
+    },
+    {
+      "Batch Sum": 168.91845703125,
+      "accuracy": 0.765625,
+      "epoch": 0.3125,
+      "step": 125
+    },
+    {
+      "Batch Sum": 160.57781982421875,
+      "accuracy": 0.875,
+      "epoch": 0.3125,
+      "step": 125
+    },
+    {
+      "epoch": 0.315,
+      "grad_norm": 11.270033836364746,
+      "learning_rate": 3.605263157894737e-06,
+      "loss": 0.3817,
+      "step": 126
+    },
+    {
+      "Batch Sum": 154.63726806640625,
+      "accuracy": 0.796875,
+      "epoch": 0.315,
+      "step": 126
+    },
+    {
+      "Batch Sum": 161.947265625,
+      "accuracy": 0.8125,
+      "epoch": 0.315,
+      "step": 126
+    },
+    {
+      "epoch": 0.3175,
+      "grad_norm": 11.978543281555176,
+      "learning_rate": 3.592105263157895e-06,
+      "loss": 0.4194,
+      "step": 127
+    },
+    {
+      "Batch Sum": 158.22998046875,
+      "accuracy": 0.8125,
+      "epoch": 0.3175,
+      "step": 127
+    },
+    {
+      "Batch Sum": 155.70001220703125,
+      "accuracy": 0.703125,
+      "epoch": 0.3175,
+      "step": 127
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.93179702758789,
+      "learning_rate": 3.578947368421053e-06,
+      "loss": 0.3997,
+      "step": 128
+    },
+    {
+      "Batch Sum": 156.26416015625,
+      "accuracy": 0.796875,
+      "epoch": 0.32,
+      "step": 128
+    },
+    {
+      "Batch Sum": 147.02413940429688,
+      "accuracy": 0.71875,
+      "epoch": 0.32,
+      "step": 128
+    },
+    {
+      "epoch": 0.3225,
+      "grad_norm": 11.40749740600586,
+      "learning_rate": 3.565789473684211e-06,
+      "loss": 0.4216,
+      "step": 129
+    },
+    {
+      "Batch Sum": 152.162109375,
+      "accuracy": 0.828125,
+      "epoch": 0.3225,
+      "step": 129
+    },
+    {
+      "Batch Sum": 156.4219970703125,
+      "accuracy": 0.828125,
+      "epoch": 0.3225,
+      "step": 129
+    },
+    {
+      "epoch": 0.325,
+      "grad_norm": 10.318909645080566,
+      "learning_rate": 3.5526315789473687e-06,
+      "loss": 0.3745,
+      "step": 130
+    },
+    {
+      "Batch Sum": 157.092529296875,
+      "accuracy": 0.75,
+      "epoch": 0.325,
+      "step": 130
+    },
+    {
+      "Batch Sum": 152.626953125,
+      "accuracy": 0.6875,
+      "epoch": 0.325,
+      "step": 130
+    },
+    {
+      "epoch": 0.3275,
+      "grad_norm": 14.140223503112793,
+      "learning_rate": 3.5394736842105266e-06,
+      "loss": 0.5467,
+      "step": 131
+    },
+    {
+      "Batch Sum": 148.7054443359375,
+      "accuracy": 0.78125,
+      "epoch": 0.3275,
+      "step": 131
+    },
+    {
+      "Batch Sum": 156.0888671875,
+      "accuracy": 0.796875,
+      "epoch": 0.3275,
+      "step": 131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.437580108642578,
+      "learning_rate": 3.5263157894736846e-06,
+      "loss": 0.426,
+      "step": 132
+    },
+    {
+      "Batch Sum": 143.746337890625,
+      "accuracy": 0.875,
+      "epoch": 0.33,
+      "step": 132
+    },
+    {
+      "Batch Sum": 159.1666259765625,
+      "accuracy": 0.890625,
+      "epoch": 0.33,
+      "step": 132
+    },
+    {
+      "epoch": 0.3325,
+      "grad_norm": 9.603164672851562,
+      "learning_rate": 3.513157894736842e-06,
+      "loss": 0.3198,
+      "step": 133
+    },
+    {
+      "Batch Sum": 142.6644287109375,
+      "accuracy": 0.765625,
+      "epoch": 0.3325,
+      "step": 133
+    },
+    {
+      "Batch Sum": 138.824951171875,
+      "accuracy": 0.734375,
+      "epoch": 0.3325,
+      "step": 133
+    },
+    {
+      "epoch": 0.335,
+      "grad_norm": 10.18635368347168,
+      "learning_rate": 3.5e-06,
+      "loss": 0.4675,
+      "step": 134
+    },
+    {
+      "Batch Sum": 145.2435302734375,
+      "accuracy": 0.828125,
+      "epoch": 0.335,
+      "step": 134
+    },
+    {
+      "Batch Sum": 133.55718994140625,
+      "accuracy": 0.84375,
+      "epoch": 0.335,
+      "step": 134
+    },
+    {
+      "epoch": 0.3375,
+      "grad_norm": 8.98958683013916,
+      "learning_rate": 3.486842105263158e-06,
+      "loss": 0.394,
+      "step": 135
+    },
+    {
+      "Batch Sum": 144.40631103515625,
+      "accuracy": 0.828125,
+      "epoch": 0.3375,
+      "step": 135
+    },
+    {
+      "Batch Sum": 155.13470458984375,
+      "accuracy": 0.828125,
+      "epoch": 0.3375,
+      "step": 135
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.116023063659668,
+      "learning_rate": 3.473684210526316e-06,
+      "loss": 0.3919,
+      "step": 136
+    },
+    {
+      "Batch Sum": 157.0986328125,
+      "accuracy": 0.828125,
+      "epoch": 0.34,
+      "step": 136
+    },
+    {
+      "Batch Sum": 151.594970703125,
+      "accuracy": 0.84375,
+      "epoch": 0.34,
+      "step": 136
+    },
+    {
+      "epoch": 0.3425,
+      "grad_norm": 9.4917573928833,
+      "learning_rate": 3.460526315789474e-06,
+      "loss": 0.3728,
+      "step": 137
+    },
+    {
+      "Batch Sum": 144.12567138671875,
+      "accuracy": 0.828125,
+      "epoch": 0.3425,
+      "step": 137
+    },
+    {
+      "Batch Sum": 158.22021484375,
+      "accuracy": 0.828125,
+      "epoch": 0.3425,
+      "step": 137
+    },
+    {
+      "epoch": 0.345,
+      "grad_norm": 9.91769027709961,
+      "learning_rate": 3.447368421052632e-06,
+      "loss": 0.3612,
+      "step": 138
+    },
+    {
+      "Batch Sum": 163.3560791015625,
+      "accuracy": 0.765625,
+      "epoch": 0.345,
+      "step": 138
+    },
+    {
+      "Batch Sum": 165.13037109375,
+      "accuracy": 0.8125,
+      "epoch": 0.345,
+      "step": 138
+    },
+    {
+      "epoch": 0.3475,
+      "grad_norm": 11.594426155090332,
+      "learning_rate": 3.43421052631579e-06,
+      "loss": 0.4229,
+      "step": 139
+    },
+    {
+      "Batch Sum": 176.0669708251953,
+      "accuracy": 0.75,
+      "epoch": 0.3475,
+      "step": 139
+    },
+    {
+      "Batch Sum": 170.3831787109375,
+      "accuracy": 0.8125,
+      "epoch": 0.3475,
+      "step": 139
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.000874519348145,
+      "learning_rate": 3.421052631578948e-06,
+      "loss": 0.448,
+      "step": 140
+    },
+    {
+      "Batch Sum": 173.9276123046875,
+      "accuracy": 0.8125,
+      "epoch": 0.35,
+      "step": 140
+    },
+    {
+      "Batch Sum": 171.8037109375,
+      "accuracy": 0.78125,
+      "epoch": 0.35,
+      "step": 140
+    },
+    {
+      "epoch": 0.3525,
+      "grad_norm": 13.811454772949219,
+      "learning_rate": 3.4078947368421057e-06,
+      "loss": 0.4739,
+      "step": 141
+    },
+    {
+      "Batch Sum": 167.4185791015625,
+      "accuracy": 0.796875,
+      "epoch": 0.3525,
+      "step": 141
+    },
+    {
+      "Batch Sum": 169.88824462890625,
+      "accuracy": 0.734375,
+      "epoch": 0.3525,
+      "step": 141
+    },
+    {
+      "epoch": 0.355,
+      "grad_norm": 12.386519432067871,
+      "learning_rate": 3.3947368421052636e-06,
+      "loss": 0.4468,
+      "step": 142
+    },
+    {
+      "Batch Sum": 172.253173828125,
+      "accuracy": 0.875,
+      "epoch": 0.355,
+      "step": 142
+    },
+    {
+      "Batch Sum": 179.5269775390625,
+      "accuracy": 0.765625,
+      "epoch": 0.355,
+      "step": 142
+    },
+    {
+      "epoch": 0.3575,
+      "grad_norm": 10.997251510620117,
+      "learning_rate": 3.381578947368421e-06,
+      "loss": 0.3703,
+      "step": 143
+    },
+    {
+      "Batch Sum": 177.33450317382812,
+      "accuracy": 0.859375,
+      "epoch": 0.3575,
+      "step": 143
+    },
+    {
+      "Batch Sum": 168.30322265625,
+      "accuracy": 0.796875,
+      "epoch": 0.3575,
+      "step": 143
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.775482177734375,
+      "learning_rate": 3.368421052631579e-06,
+      "loss": 0.3419,
+      "step": 144
+    },
+    {
+      "Batch Sum": 177.181396484375,
+      "accuracy": 0.78125,
+      "epoch": 0.36,
+      "step": 144
+    },
+    {
+      "Batch Sum": 179.53369140625,
+      "accuracy": 0.8125,
+      "epoch": 0.36,
+      "step": 144
+    },
+    {
+      "epoch": 0.3625,
+      "grad_norm": 11.77710247039795,
+      "learning_rate": 3.355263157894737e-06,
+      "loss": 0.442,
+      "step": 145
+    },
+    {
+      "Batch Sum": 163.1304931640625,
+      "accuracy": 0.84375,
+      "epoch": 0.3625,
+      "step": 145
+    },
+    {
+      "Batch Sum": 173.8359375,
+      "accuracy": 0.78125,
+      "epoch": 0.3625,
+      "step": 145
+    },
+    {
+      "epoch": 0.365,
+      "grad_norm": 10.837456703186035,
+      "learning_rate": 3.342105263157895e-06,
+      "loss": 0.3772,
+      "step": 146
+    },
+    {
+      "Batch Sum": 168.785888671875,
+      "accuracy": 0.765625,
+      "epoch": 0.365,
+      "step": 146
+    },
+    {
+      "Batch Sum": 162.473388671875,
+      "accuracy": 0.796875,
+      "epoch": 0.365,
+      "step": 146
+    },
+    {
+      "epoch": 0.3675,
+      "grad_norm": 12.898738861083984,
+      "learning_rate": 3.3289473684210528e-06,
+      "loss": 0.4284,
+      "step": 147
+    },
+    {
+      "Batch Sum": 155.8890380859375,
+      "accuracy": 0.8125,
+      "epoch": 0.3675,
+      "step": 147
+    },
+    {
+      "Batch Sum": 163.2218017578125,
+      "accuracy": 0.84375,
+      "epoch": 0.3675,
+      "step": 147
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.30620288848877,
+      "learning_rate": 3.3157894736842107e-06,
+      "loss": 0.3338,
+      "step": 148
+    },
+    {
+      "Batch Sum": 179.05615234375,
+      "accuracy": 0.84375,
+      "epoch": 0.37,
+      "step": 148
+    },
+    {
+      "Batch Sum": 165.72119140625,
+      "accuracy": 0.8125,
+      "epoch": 0.37,
+      "step": 148
+    },
+    {
+      "epoch": 0.3725,
+      "grad_norm": 10.406338691711426,
+      "learning_rate": 3.302631578947369e-06,
+      "loss": 0.3486,
+      "step": 149
+    },
+    {
+      "Batch Sum": 153.40414428710938,
+      "accuracy": 0.859375,
+      "epoch": 0.3725,
+      "step": 149
+    },
+    {
+      "Batch Sum": 157.75411987304688,
+      "accuracy": 0.796875,
+      "epoch": 0.3725,
+      "step": 149
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 11.733957290649414,
+      "learning_rate": 3.289473684210527e-06,
+      "loss": 0.36,
+      "step": 150
+    },
+    {
+      "Batch Sum": 179.1650390625,
+      "accuracy": 0.875,
+      "epoch": 0.375,
+      "step": 150
+    },
+    {
+      "Batch Sum": 178.9234619140625,
+      "accuracy": 0.875,
+      "epoch": 0.375,
+      "step": 150
+    },
+    {
+      "epoch": 0.3775,
+      "grad_norm": 9.269570350646973,
+      "learning_rate": 3.276315789473685e-06,
+      "loss": 0.2708,
+      "step": 151
+    },
+    {
+      "Batch Sum": 179.40478515625,
+      "accuracy": 0.84375,
+      "epoch": 0.3775,
+      "step": 151
+    },
+    {
+      "Batch Sum": 162.3435516357422,
+      "accuracy": 0.8125,
+      "epoch": 0.3775,
+      "step": 151
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.036487579345703,
+      "learning_rate": 3.2631578947368423e-06,
+      "loss": 0.3789,
+      "step": 152
+    },
+    {
+      "Batch Sum": 154.46905517578125,
+      "accuracy": 0.84375,
+      "epoch": 0.38,
+      "step": 152
+    },
+    {
+      "Batch Sum": 176.2408447265625,
+      "accuracy": 0.8125,
+      "epoch": 0.38,
+      "step": 152
+    },
+    {
+      "epoch": 0.3825,
+      "grad_norm": 10.829084396362305,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.3816,
+      "step": 153
+    },
+    {
+      "Batch Sum": 171.27947998046875,
+      "accuracy": 0.75,
+      "epoch": 0.3825,
+      "step": 153
+    },
+    {
+      "Batch Sum": 181.0279541015625,
+      "accuracy": 0.859375,
+      "epoch": 0.3825,
+      "step": 153
+    },
+    {
+      "epoch": 0.385,
+      "grad_norm": 13.027417182922363,
+      "learning_rate": 3.236842105263158e-06,
+      "loss": 0.4191,
+      "step": 154
+    },
+    {
+      "Batch Sum": 173.9217529296875,
+      "accuracy": 0.8125,
+      "epoch": 0.385,
+      "step": 154
+    },
+    {
+      "Batch Sum": 170.60922241210938,
+      "accuracy": 0.765625,
+      "epoch": 0.385,
+      "step": 154
+    },
+    {
+      "epoch": 0.3875,
+      "grad_norm": 15.362863540649414,
+      "learning_rate": 3.223684210526316e-06,
+      "loss": 0.4455,
+      "step": 155
+    },
+    {
+      "Batch Sum": 185.8359375,
+      "accuracy": 0.875,
+      "epoch": 0.3875,
+      "step": 155
+    },
+    {
+      "Batch Sum": 175.69873046875,
+      "accuracy": 0.859375,
+      "epoch": 0.3875,
+      "step": 155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.39478874206543,
+      "learning_rate": 3.210526315789474e-06,
+      "loss": 0.3245,
+      "step": 156
+    },
+    {
+      "Batch Sum": 187.887451171875,
+      "accuracy": 0.796875,
+      "epoch": 0.39,
+      "step": 156
+    },
+    {
+      "Batch Sum": 176.8115234375,
+      "accuracy": 0.8125,
+      "epoch": 0.39,
+      "step": 156
+    },
+    {
+      "epoch": 0.3925,
+      "grad_norm": 12.699715614318848,
+      "learning_rate": 3.197368421052632e-06,
+      "loss": 0.3727,
+      "step": 157
+    },
+    {
+      "Batch Sum": 174.041015625,
+      "accuracy": 0.875,
+      "epoch": 0.3925,
+      "step": 157
+    },
+    {
+      "Batch Sum": 175.4111328125,
+      "accuracy": 0.734375,
+      "epoch": 0.3925,
+      "step": 157
+    },
+    {
+      "epoch": 0.395,
+      "grad_norm": 13.248059272766113,
+      "learning_rate": 3.1842105263157898e-06,
+      "loss": 0.434,
+      "step": 158
+    },
+    {
+      "Batch Sum": 163.93408203125,
+      "accuracy": 0.796875,
+      "epoch": 0.395,
+      "step": 158
+    },
+    {
+      "Batch Sum": 172.42626953125,
+      "accuracy": 0.84375,
+      "epoch": 0.395,
+      "step": 158
+    },
+    {
+      "epoch": 0.3975,
+      "grad_norm": 12.247123718261719,
+      "learning_rate": 3.1710526315789477e-06,
+      "loss": 0.3552,
+      "step": 159
+    },
+    {
+      "Batch Sum": 159.5538330078125,
+      "accuracy": 0.796875,
+      "epoch": 0.3975,
+      "step": 159
+    },
+    {
+      "Batch Sum": 172.398193359375,
+      "accuracy": 0.84375,
+      "epoch": 0.3975,
+      "step": 159
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.754870414733887,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 0.4638,
+      "step": 160
+    },
+    {
+      "Batch Sum": 179.1923828125,
+      "accuracy": 0.859375,
+      "epoch": 0.4,
+      "step": 160
+    },
+    {
+      "Batch Sum": 161.6951904296875,
+      "accuracy": 0.8125,
+      "epoch": 0.4,
+      "step": 160
+    },
+    {
+      "epoch": 0.4025,
+      "grad_norm": 9.910033226013184,
+      "learning_rate": 3.144736842105263e-06,
+      "loss": 0.3066,
+      "step": 161
+    },
+    {
+      "Batch Sum": 169.45782470703125,
+      "accuracy": 0.859375,
+      "epoch": 0.4025,
+      "step": 161
+    },
+    {
+      "Batch Sum": 157.14697265625,
+      "accuracy": 0.890625,
+      "epoch": 0.4025,
+      "step": 161
+    },
+    {
+      "epoch": 0.405,
+      "grad_norm": 9.645856857299805,
+      "learning_rate": 3.131578947368421e-06,
+      "loss": 0.2836,
+      "step": 162
+    },
+    {
+      "Batch Sum": 153.429443359375,
+      "accuracy": 0.75,
+      "epoch": 0.405,
+      "step": 162
+    },
+    {
+      "Batch Sum": 159.56504821777344,
+      "accuracy": 0.859375,
+      "epoch": 0.405,
+      "step": 162
+    },
+    {
+      "epoch": 0.4075,
+      "grad_norm": 11.657127380371094,
+      "learning_rate": 3.1184210526315793e-06,
+      "loss": 0.4141,
+      "step": 163
+    },
+    {
+      "Batch Sum": 157.7459716796875,
+      "accuracy": 0.734375,
+      "epoch": 0.4075,
+      "step": 163
+    },
+    {
+      "Batch Sum": 163.69723510742188,
+      "accuracy": 0.796875,
+      "epoch": 0.4075,
+      "step": 163
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.670110702514648,
+      "learning_rate": 3.1052631578947372e-06,
+      "loss": 0.4772,
+      "step": 164
+    },
+    {
+      "Batch Sum": 157.6031494140625,
+      "accuracy": 0.890625,
+      "epoch": 0.41,
+      "step": 164
+    },
+    {
+      "Batch Sum": 172.3399658203125,
+      "accuracy": 0.828125,
+      "epoch": 0.41,
+      "step": 164
+    },
+    {
+      "epoch": 0.4125,
+      "grad_norm": 9.068910598754883,
+      "learning_rate": 3.092105263157895e-06,
+      "loss": 0.2752,
+      "step": 165
+    },
+    {
+      "Batch Sum": 160.44140625,
+      "accuracy": 0.859375,
+      "epoch": 0.4125,
+      "step": 165
+    },
+    {
+      "Batch Sum": 169.0107421875,
+      "accuracy": 0.71875,
+      "epoch": 0.4125,
+      "step": 165
+    },
+    {
+      "epoch": 0.415,
+      "grad_norm": 12.532942771911621,
+      "learning_rate": 3.078947368421053e-06,
+      "loss": 0.4162,
+      "step": 166
+    },
+    {
+      "Batch Sum": 150.088623046875,
+      "accuracy": 0.90625,
+      "epoch": 0.415,
+      "step": 166
+    },
+    {
+      "Batch Sum": 163.49560546875,
+      "accuracy": 0.765625,
+      "epoch": 0.415,
+      "step": 166
+    },
+    {
+      "epoch": 0.4175,
+      "grad_norm": 11.255102157592773,
+      "learning_rate": 3.065789473684211e-06,
+      "loss": 0.3426,
+      "step": 167
+    },
+    {
+      "Batch Sum": 158.72216796875,
+      "accuracy": 0.78125,
+      "epoch": 0.4175,
+      "step": 167
+    },
+    {
+      "Batch Sum": 157.4320068359375,
+      "accuracy": 0.75,
+      "epoch": 0.4175,
+      "step": 167
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.509016036987305,
+      "learning_rate": 3.052631578947369e-06,
+      "loss": 0.3974,
+      "step": 168
+    },
+    {
+      "Batch Sum": 170.9871826171875,
+      "accuracy": 0.859375,
+      "epoch": 0.42,
+      "step": 168
+    },
+    {
+      "Batch Sum": 151.966796875,
+      "accuracy": 0.859375,
+      "epoch": 0.42,
+      "step": 168
+    },
+    {
+      "epoch": 0.4225,
+      "grad_norm": 11.246017456054688,
+      "learning_rate": 3.0394736842105268e-06,
+      "loss": 0.346,
+      "step": 169
+    },
+    {
+      "Batch Sum": 141.69485473632812,
+      "accuracy": 0.8125,
+      "epoch": 0.4225,
+      "step": 169
+    },
+    {
+      "Batch Sum": 177.1005859375,
+      "accuracy": 0.84375,
+      "epoch": 0.4225,
+      "step": 169
+    },
+    {
+      "epoch": 0.425,
+      "grad_norm": 12.605325698852539,
+      "learning_rate": 3.0263157894736843e-06,
+      "loss": 0.3574,
+      "step": 170
+    },
+    {
+      "Batch Sum": 158.830078125,
+      "accuracy": 0.796875,
+      "epoch": 0.425,
+      "step": 170
+    },
+    {
+      "Batch Sum": 171.9595947265625,
+      "accuracy": 0.78125,
+      "epoch": 0.425,
+      "step": 170
+    },
+    {
+      "epoch": 0.4275,
+      "grad_norm": 11.472408294677734,
+      "learning_rate": 3.013157894736842e-06,
+      "loss": 0.3778,
+      "step": 171
+    },
+    {
+      "Batch Sum": 164.240966796875,
+      "accuracy": 0.75,
+      "epoch": 0.4275,
+      "step": 171
+    },
+    {
+      "Batch Sum": 187.87646484375,
+      "accuracy": 0.84375,
+      "epoch": 0.4275,
+      "step": 171
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 12.420621871948242,
+      "learning_rate": 3e-06,
+      "loss": 0.3898,
+      "step": 172
+    },
+    {
+      "Batch Sum": 168.580810546875,
+      "accuracy": 0.796875,
+      "epoch": 0.43,
+      "step": 172
+    },
+    {
+      "Batch Sum": 183.23199462890625,
+      "accuracy": 0.84375,
+      "epoch": 0.43,
+      "step": 172
+    },
+    {
+      "epoch": 0.4325,
+      "grad_norm": 11.917014122009277,
+      "learning_rate": 2.986842105263158e-06,
+      "loss": 0.3652,
+      "step": 173
+    },
+    {
+      "Batch Sum": 180.752685546875,
+      "accuracy": 0.8125,
+      "epoch": 0.4325,
+      "step": 173
+    },
+    {
+      "Batch Sum": 174.8634033203125,
+      "accuracy": 0.78125,
+      "epoch": 0.4325,
+      "step": 173
+    },
+    {
+      "epoch": 0.435,
+      "grad_norm": 13.711872100830078,
+      "learning_rate": 2.973684210526316e-06,
+      "loss": 0.4049,
+      "step": 174
+    },
+    {
+      "Batch Sum": 187.95513916015625,
+      "accuracy": 0.859375,
+      "epoch": 0.435,
+      "step": 174
+    },
+    {
+      "Batch Sum": 175.5262451171875,
+      "accuracy": 0.90625,
+      "epoch": 0.435,
+      "step": 174
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 13.002115249633789,
+      "learning_rate": 2.960526315789474e-06,
+      "loss": 0.3187,
+      "step": 175
+    },
+    {
+      "Batch Sum": 193.835693359375,
+      "accuracy": 0.890625,
+      "epoch": 0.4375,
+      "step": 175
+    },
+    {
+      "Batch Sum": 178.1373291015625,
+      "accuracy": 0.875,
+      "epoch": 0.4375,
+      "step": 175
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.474250793457031,
+      "learning_rate": 2.9473684210526317e-06,
+      "loss": 0.2761,
+      "step": 176
+    },
+    {
+      "Batch Sum": 204.98779296875,
+      "accuracy": 0.78125,
+      "epoch": 0.44,
+      "step": 176
+    },
+    {
+      "Batch Sum": 176.36172485351562,
+      "accuracy": 0.84375,
+      "epoch": 0.44,
+      "step": 176
+    },
+    {
+      "epoch": 0.4425,
+      "grad_norm": 13.780770301818848,
+      "learning_rate": 2.93421052631579e-06,
+      "loss": 0.4051,
+      "step": 177
+    },
+    {
+      "Batch Sum": 182.67578125,
+      "accuracy": 0.890625,
+      "epoch": 0.4425,
+      "step": 177
+    },
+    {
+      "Batch Sum": 186.214111328125,
+      "accuracy": 0.859375,
+      "epoch": 0.4425,
+      "step": 177
+    },
+    {
+      "epoch": 0.445,
+      "grad_norm": 11.39830207824707,
+      "learning_rate": 2.921052631578948e-06,
+      "loss": 0.3473,
+      "step": 178
+    },
+    {
+      "Batch Sum": 177.402587890625,
+      "accuracy": 0.84375,
+      "epoch": 0.445,
+      "step": 178
+    },
+    {
+      "Batch Sum": 189.29144287109375,
+      "accuracy": 0.828125,
+      "epoch": 0.445,
+      "step": 178
+    },
+    {
+      "epoch": 0.4475,
+      "grad_norm": 13.348224639892578,
+      "learning_rate": 2.907894736842106e-06,
+      "loss": 0.3464,
+      "step": 179
+    },
+    {
+      "Batch Sum": 184.62939453125,
+      "accuracy": 0.859375,
+      "epoch": 0.4475,
+      "step": 179
+    },
+    {
+      "Batch Sum": 171.792236328125,
+      "accuracy": 0.78125,
+      "epoch": 0.4475,
+      "step": 179
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.45236587524414,
+      "learning_rate": 2.8947368421052634e-06,
+      "loss": 0.3707,
+      "step": 180
+    },
+    {
+      "Batch Sum": 183.537353515625,
+      "accuracy": 0.828125,
+      "epoch": 0.45,
+      "step": 180
+    },
+    {
+      "Batch Sum": 184.44140625,
+      "accuracy": 0.875,
+      "epoch": 0.45,
+      "step": 180
+    },
+    {
+      "epoch": 0.4525,
+      "grad_norm": 11.033170700073242,
+      "learning_rate": 2.8815789473684213e-06,
+      "loss": 0.3141,
+      "step": 181
+    },
+    {
+      "Batch Sum": 157.82540893554688,
+      "accuracy": 0.796875,
+      "epoch": 0.4525,
+      "step": 181
+    },
+    {
+      "Batch Sum": 178.438720703125,
+      "accuracy": 0.796875,
+      "epoch": 0.4525,
+      "step": 181
+    },
+    {
+      "epoch": 0.455,
+      "grad_norm": 13.729015350341797,
+      "learning_rate": 2.868421052631579e-06,
+      "loss": 0.4089,
+      "step": 182
+    },
+    {
+      "Batch Sum": 178.9556884765625,
+      "accuracy": 0.734375,
+      "epoch": 0.455,
+      "step": 182
+    },
+    {
+      "Batch Sum": 181.88690185546875,
+      "accuracy": 0.84375,
+      "epoch": 0.455,
+      "step": 182
+    },
+    {
+      "epoch": 0.4575,
+      "grad_norm": 16.668292999267578,
+      "learning_rate": 2.855263157894737e-06,
+      "loss": 0.5041,
+      "step": 183
+    },
+    {
+      "Batch Sum": 170.711181640625,
+      "accuracy": 0.796875,
+      "epoch": 0.4575,
+      "step": 183
+    },
+    {
+      "Batch Sum": 169.305908203125,
+      "accuracy": 0.796875,
+      "epoch": 0.4575,
+      "step": 183
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.935322761535645,
+      "learning_rate": 2.842105263157895e-06,
+      "loss": 0.4036,
+      "step": 184
+    },
+    {
+      "Batch Sum": 159.673828125,
+      "accuracy": 0.796875,
+      "epoch": 0.46,
+      "step": 184
+    },
+    {
+      "Batch Sum": 148.835693359375,
+      "accuracy": 0.828125,
+      "epoch": 0.46,
+      "step": 184
+    },
+    {
+      "epoch": 0.4625,
+      "grad_norm": 12.534523010253906,
+      "learning_rate": 2.828947368421053e-06,
+      "loss": 0.3995,
+      "step": 185
+    },
+    {
+      "Batch Sum": 137.3509521484375,
+      "accuracy": 0.859375,
+      "epoch": 0.4625,
+      "step": 185
+    },
+    {
+      "Batch Sum": 137.57501220703125,
+      "accuracy": 0.875,
+      "epoch": 0.4625,
+      "step": 185
+    },
+    {
+      "epoch": 0.465,
+      "grad_norm": 10.489812850952148,
+      "learning_rate": 2.815789473684211e-06,
+      "loss": 0.3355,
+      "step": 186
+    },
+    {
+      "Batch Sum": 149.32647705078125,
+      "accuracy": 0.84375,
+      "epoch": 0.465,
+      "step": 186
+    },
+    {
+      "Batch Sum": 137.6878662109375,
+      "accuracy": 0.890625,
+      "epoch": 0.465,
+      "step": 186
+    },
+    {
+      "epoch": 0.4675,
+      "grad_norm": 10.799548149108887,
+      "learning_rate": 2.8026315789473683e-06,
+      "loss": 0.3154,
+      "step": 187
+    },
+    {
+      "Batch Sum": 140.84820556640625,
+      "accuracy": 0.84375,
+      "epoch": 0.4675,
+      "step": 187
+    },
+    {
+      "Batch Sum": 137.59814453125,
+      "accuracy": 0.796875,
+      "epoch": 0.4675,
+      "step": 187
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.771398544311523,
+      "learning_rate": 2.789473684210526e-06,
+      "loss": 0.3794,
+      "step": 188
+    },
+    {
+      "Batch Sum": 120.0439453125,
+      "accuracy": 0.828125,
+      "epoch": 0.47,
+      "step": 188
+    },
+    {
+      "Batch Sum": 132.53411865234375,
+      "accuracy": 0.890625,
+      "epoch": 0.47,
+      "step": 188
+    },
+    {
+      "epoch": 0.4725,
+      "grad_norm": 8.793214797973633,
+      "learning_rate": 2.776315789473684e-06,
+      "loss": 0.3329,
+      "step": 189
+    },
+    {
+      "Batch Sum": 148.12432861328125,
+      "accuracy": 0.796875,
+      "epoch": 0.4725,
+      "step": 189
+    },
+    {
+      "Batch Sum": 123.746826171875,
+      "accuracy": 0.765625,
+      "epoch": 0.4725,
+      "step": 189
+    },
+    {
+      "epoch": 0.475,
+      "grad_norm": 10.795166969299316,
+      "learning_rate": 2.7631578947368424e-06,
+      "loss": 0.3926,
+      "step": 190
+    },
+    {
+      "Batch Sum": 144.51519775390625,
+      "accuracy": 0.796875,
+      "epoch": 0.475,
+      "step": 190
+    },
+    {
+      "Batch Sum": 156.36366271972656,
+      "accuracy": 0.84375,
+      "epoch": 0.475,
+      "step": 190
+    },
+    {
+      "epoch": 0.4775,
+      "grad_norm": 11.242524147033691,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.4088,
+      "step": 191
+    },
+    {
+      "Batch Sum": 132.588134765625,
+      "accuracy": 0.734375,
+      "epoch": 0.4775,
+      "step": 191
+    },
+    {
+      "Batch Sum": 132.279296875,
+      "accuracy": 0.765625,
+      "epoch": 0.4775,
+      "step": 191
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.10943603515625,
+      "learning_rate": 2.7368421052631583e-06,
+      "loss": 0.4502,
+      "step": 192
+    },
+    {
+      "Batch Sum": 127.26214599609375,
+      "accuracy": 0.765625,
+      "epoch": 0.48,
+      "step": 192
+    },
+    {
+      "Batch Sum": 120.28961181640625,
+      "accuracy": 0.734375,
+      "epoch": 0.48,
+      "step": 192
+    },
+    {
+      "epoch": 0.4825,
+      "grad_norm": 10.466278076171875,
+      "learning_rate": 2.723684210526316e-06,
+      "loss": 0.4388,
+      "step": 193
+    },
+    {
+      "Batch Sum": 137.3543701171875,
+      "accuracy": 0.875,
+      "epoch": 0.4825,
+      "step": 193
+    },
+    {
+      "Batch Sum": 126.865234375,
+      "accuracy": 0.875,
+      "epoch": 0.4825,
+      "step": 193
+    },
+    {
+      "epoch": 0.485,
+      "grad_norm": 9.617724418640137,
+      "learning_rate": 2.710526315789474e-06,
+      "loss": 0.3097,
+      "step": 194
+    },
+    {
+      "Batch Sum": 141.0351104736328,
+      "accuracy": 0.859375,
+      "epoch": 0.485,
+      "step": 194
+    },
+    {
+      "Batch Sum": 115.21342468261719,
+      "accuracy": 0.765625,
+      "epoch": 0.485,
+      "step": 194
+    },
+    {
+      "epoch": 0.4875,
+      "grad_norm": 9.829240798950195,
+      "learning_rate": 2.697368421052632e-06,
+      "loss": 0.4003,
+      "step": 195
+    },
+    {
+      "Batch Sum": 140.189697265625,
+      "accuracy": 0.875,
+      "epoch": 0.4875,
+      "step": 195
+    },
+    {
+      "Batch Sum": 147.9466552734375,
+      "accuracy": 0.84375,
+      "epoch": 0.4875,
+      "step": 195
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.37118911743164,
+      "learning_rate": 2.68421052631579e-06,
+      "loss": 0.3677,
+      "step": 196
+    },
+    {
+      "Batch Sum": 148.1060791015625,
+      "accuracy": 0.71875,
+      "epoch": 0.49,
+      "step": 196
+    },
+    {
+      "Batch Sum": 131.451171875,
+      "accuracy": 0.84375,
+      "epoch": 0.49,
+      "step": 196
+    },
+    {
+      "epoch": 0.4925,
+      "grad_norm": 11.06047534942627,
+      "learning_rate": 2.6710526315789474e-06,
+      "loss": 0.4126,
+      "step": 197
+    },
+    {
+      "Batch Sum": 144.2012939453125,
+      "accuracy": 0.828125,
+      "epoch": 0.4925,
+      "step": 197
+    },
+    {
+      "Batch Sum": 141.30419921875,
+      "accuracy": 0.8125,
+      "epoch": 0.4925,
+      "step": 197
+    },
+    {
+      "epoch": 0.495,
+      "grad_norm": 9.70101261138916,
+      "learning_rate": 2.6578947368421053e-06,
+      "loss": 0.3397,
+      "step": 198
+    },
+    {
+      "Batch Sum": 149.09429931640625,
+      "accuracy": 0.828125,
+      "epoch": 0.495,
+      "step": 198
+    },
+    {
+      "Batch Sum": 141.2196044921875,
+      "accuracy": 0.828125,
+      "epoch": 0.495,
+      "step": 198
+    },
+    {
+      "epoch": 0.4975,
+      "grad_norm": 9.994836807250977,
+      "learning_rate": 2.644736842105263e-06,
+      "loss": 0.3922,
+      "step": 199
+    },
+    {
+      "Batch Sum": 150.16015625,
+      "accuracy": 0.78125,
+      "epoch": 0.4975,
+      "step": 199
+    },
+    {
+      "Batch Sum": 149.1392822265625,
+      "accuracy": 0.75,
+      "epoch": 0.4975,
+      "step": 199
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.291372299194336,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 0.4164,
+      "step": 200
+    },
+    {
+      "Batch Sum": 147.03662109375,
+      "accuracy": 0.796875,
+      "epoch": 0.5,
+      "step": 200
+    },
+    {
+      "Batch Sum": 140.141357421875,
+      "accuracy": 0.78125,
+      "epoch": 0.5,
+      "step": 200
+    },
+    {
+      "epoch": 0.5025,
+      "grad_norm": 11.896326065063477,
+      "learning_rate": 2.618421052631579e-06,
+      "loss": 0.4281,
+      "step": 201
+    },
+    {
+      "Batch Sum": 155.37576293945312,
+      "accuracy": 0.84375,
+      "epoch": 0.5025,
+      "step": 201
+    },
+    {
+      "Batch Sum": 145.00149536132812,
+      "accuracy": 0.921875,
+      "epoch": 0.5025,
+      "step": 201
+    },
+    {
+      "epoch": 0.505,
+      "grad_norm": 9.94739055633545,
+      "learning_rate": 2.605263157894737e-06,
+      "loss": 0.3113,
+      "step": 202
+    },
+    {
+      "Batch Sum": 158.8394775390625,
+      "accuracy": 0.71875,
+      "epoch": 0.505,
+      "step": 202
+    },
+    {
+      "Batch Sum": 144.2637939453125,
+      "accuracy": 0.734375,
+      "epoch": 0.505,
+      "step": 202
+    },
+    {
+      "epoch": 0.5075,
+      "grad_norm": 11.568675994873047,
+      "learning_rate": 2.592105263157895e-06,
+      "loss": 0.4255,
+      "step": 203
+    },
+    {
+      "Batch Sum": 150.56719970703125,
+      "accuracy": 0.828125,
+      "epoch": 0.5075,
+      "step": 203
+    },
+    {
+      "Batch Sum": 152.44580078125,
+      "accuracy": 0.796875,
+      "epoch": 0.5075,
+      "step": 203
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.738829612731934,
+      "learning_rate": 2.578947368421053e-06,
+      "loss": 0.4118,
+      "step": 204
+    },
+    {
+      "Batch Sum": 176.9063720703125,
+      "accuracy": 0.890625,
+      "epoch": 0.51,
+      "step": 204
+    },
+    {
+      "Batch Sum": 152.7841796875,
+      "accuracy": 0.859375,
+      "epoch": 0.51,
+      "step": 204
+    },
+    {
+      "epoch": 0.5125,
+      "grad_norm": 9.963852882385254,
+      "learning_rate": 2.565789473684211e-06,
+      "loss": 0.3268,
+      "step": 205
+    },
+    {
+      "Batch Sum": 135.9202880859375,
+      "accuracy": 0.875,
+      "epoch": 0.5125,
+      "step": 205
+    },
+    {
+      "Batch Sum": 169.02813720703125,
+      "accuracy": 0.796875,
+      "epoch": 0.5125,
+      "step": 205
+    },
+    {
+      "epoch": 0.515,
+      "grad_norm": 10.456665992736816,
+      "learning_rate": 2.552631578947369e-06,
+      "loss": 0.3363,
+      "step": 206
+    },
+    {
+      "Batch Sum": 150.61962890625,
+      "accuracy": 0.828125,
+      "epoch": 0.515,
+      "step": 206
+    },
+    {
+      "Batch Sum": 168.18505859375,
+      "accuracy": 0.796875,
+      "epoch": 0.515,
+      "step": 206
+    },
+    {
+      "epoch": 0.5175,
+      "grad_norm": 10.877069473266602,
+      "learning_rate": 2.5394736842105265e-06,
+      "loss": 0.4049,
+      "step": 207
+    },
+    {
+      "Batch Sum": 152.3681640625,
+      "accuracy": 0.75,
+      "epoch": 0.5175,
+      "step": 207
+    },
+    {
+      "Batch Sum": 169.8326416015625,
+      "accuracy": 0.859375,
+      "epoch": 0.5175,
+      "step": 207
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.995158195495605,
+      "learning_rate": 2.5263157894736844e-06,
+      "loss": 0.4583,
+      "step": 208
+    },
+    {
+      "Batch Sum": 179.8681640625,
+      "accuracy": 0.828125,
+      "epoch": 0.52,
+      "step": 208
+    },
+    {
+      "Batch Sum": 166.66302490234375,
+      "accuracy": 0.8125,
+      "epoch": 0.52,
+      "step": 208
+    },
+    {
+      "epoch": 0.5225,
+      "grad_norm": 11.114405632019043,
+      "learning_rate": 2.5131578947368423e-06,
+      "loss": 0.4107,
+      "step": 209
+    },
+    {
+      "Batch Sum": 167.494384765625,
+      "accuracy": 0.890625,
+      "epoch": 0.5225,
+      "step": 209
+    },
+    {
+      "Batch Sum": 166.34371948242188,
+      "accuracy": 0.8125,
+      "epoch": 0.5225,
+      "step": 209
+    },
+    {
+      "epoch": 0.525,
+      "grad_norm": 10.482272148132324,
+      "learning_rate": 2.5e-06,
+      "loss": 0.3186,
+      "step": 210
+    },
+    {
+      "Batch Sum": 170.8397216796875,
+      "accuracy": 0.796875,
+      "epoch": 0.525,
+      "step": 210
+    },
+    {
+      "Batch Sum": 156.501953125,
+      "accuracy": 0.765625,
+      "epoch": 0.525,
+      "step": 210
+    },
+    {
+      "epoch": 0.5275,
+      "grad_norm": 11.457318305969238,
+      "learning_rate": 2.486842105263158e-06,
+      "loss": 0.4198,
+      "step": 211
+    },
+    {
+      "Batch Sum": 166.90213012695312,
+      "accuracy": 0.796875,
+      "epoch": 0.5275,
+      "step": 211
+    },
+    {
+      "Batch Sum": 171.12567138671875,
+      "accuracy": 0.75,
+      "epoch": 0.5275,
+      "step": 211
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.193228721618652,
+      "learning_rate": 2.473684210526316e-06,
+      "loss": 0.4258,
+      "step": 212
+    },
+    {
+      "Batch Sum": 152.46875,
+      "accuracy": 0.828125,
+      "epoch": 0.53,
+      "step": 212
+    },
+    {
+      "Batch Sum": 161.25390625,
+      "accuracy": 0.8125,
+      "epoch": 0.53,
+      "step": 212
+    },
+    {
+      "epoch": 0.5325,
+      "grad_norm": 10.542277336120605,
+      "learning_rate": 2.460526315789474e-06,
+      "loss": 0.3789,
+      "step": 213
+    },
+    {
+      "Batch Sum": 171.485595703125,
+      "accuracy": 0.828125,
+      "epoch": 0.5325,
+      "step": 213
+    },
+    {
+      "Batch Sum": 176.388671875,
+      "accuracy": 0.796875,
+      "epoch": 0.5325,
+      "step": 213
+    },
+    {
+      "epoch": 0.535,
+      "grad_norm": 9.065043449401855,
+      "learning_rate": 2.447368421052632e-06,
+      "loss": 0.3522,
+      "step": 214
+    },
+    {
+      "Batch Sum": 159.828369140625,
+      "accuracy": 0.765625,
+      "epoch": 0.535,
+      "step": 214
+    },
+    {
+      "Batch Sum": 153.4268798828125,
+      "accuracy": 0.875,
+      "epoch": 0.535,
+      "step": 214
+    },
+    {
+      "epoch": 0.5375,
+      "grad_norm": 10.286127090454102,
+      "learning_rate": 2.4342105263157898e-06,
+      "loss": 0.3434,
+      "step": 215
+    },
+    {
+      "Batch Sum": 148.36474609375,
+      "accuracy": 0.8125,
+      "epoch": 0.5375,
+      "step": 215
+    },
+    {
+      "Batch Sum": 160.15513610839844,
+      "accuracy": 0.828125,
+      "epoch": 0.5375,
+      "step": 215
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.205731391906738,
+      "learning_rate": 2.4210526315789477e-06,
+      "loss": 0.3791,
+      "step": 216
+    },
+    {
+      "Batch Sum": 170.2554931640625,
+      "accuracy": 0.90625,
+      "epoch": 0.54,
+      "step": 216
+    },
+    {
+      "Batch Sum": 176.7681884765625,
+      "accuracy": 0.84375,
+      "epoch": 0.54,
+      "step": 216
+    },
+    {
+      "epoch": 0.5425,
+      "grad_norm": 8.758590698242188,
+      "learning_rate": 2.4078947368421056e-06,
+      "loss": 0.2778,
+      "step": 217
+    },
+    {
+      "Batch Sum": 153.43190002441406,
+      "accuracy": 0.890625,
+      "epoch": 0.5425,
+      "step": 217
+    },
+    {
+      "Batch Sum": 154.661865234375,
+      "accuracy": 0.859375,
+      "epoch": 0.5425,
+      "step": 217
+    },
+    {
+      "epoch": 0.545,
+      "grad_norm": 10.250535011291504,
+      "learning_rate": 2.3947368421052635e-06,
+      "loss": 0.2675,
+      "step": 218
+    },
+    {
+      "Batch Sum": 183.23333740234375,
+      "accuracy": 0.828125,
+      "epoch": 0.545,
+      "step": 218
+    },
+    {
+      "Batch Sum": 190.609619140625,
+      "accuracy": 0.90625,
+      "epoch": 0.545,
+      "step": 218
+    },
+    {
+      "epoch": 0.5475,
+      "grad_norm": 11.179657936096191,
+      "learning_rate": 2.381578947368421e-06,
+      "loss": 0.3367,
+      "step": 219
+    },
+    {
+      "Batch Sum": 172.64797973632812,
+      "accuracy": 0.71875,
+      "epoch": 0.5475,
+      "step": 219
+    },
+    {
+      "Batch Sum": 178.2139892578125,
+      "accuracy": 0.84375,
+      "epoch": 0.5475,
+      "step": 219
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.280330657958984,
+      "learning_rate": 2.368421052631579e-06,
+      "loss": 0.426,
+      "step": 220
+    },
+    {
+      "Batch Sum": 161.825439453125,
+      "accuracy": 0.796875,
+      "epoch": 0.55,
+      "step": 220
+    },
+    {
+      "Batch Sum": 177.78564453125,
+      "accuracy": 0.859375,
+      "epoch": 0.55,
+      "step": 220
+    },
+    {
+      "epoch": 0.5525,
+      "grad_norm": 12.05972671508789,
+      "learning_rate": 2.355263157894737e-06,
+      "loss": 0.3653,
+      "step": 221
+    },
+    {
+      "Batch Sum": 211.14947509765625,
+      "accuracy": 0.875,
+      "epoch": 0.5525,
+      "step": 221
+    },
+    {
+      "Batch Sum": 187.81695556640625,
+      "accuracy": 0.859375,
+      "epoch": 0.5525,
+      "step": 221
+    },
+    {
+      "epoch": 0.555,
+      "grad_norm": 11.996454238891602,
+      "learning_rate": 2.342105263157895e-06,
+      "loss": 0.317,
+      "step": 222
+    },
+    {
+      "Batch Sum": 199.1517333984375,
+      "accuracy": 0.84375,
+      "epoch": 0.555,
+      "step": 222
+    },
+    {
+      "Batch Sum": 213.52239990234375,
+      "accuracy": 0.84375,
+      "epoch": 0.555,
+      "step": 222
+    },
+    {
+      "epoch": 0.5575,
+      "grad_norm": 12.501474380493164,
+      "learning_rate": 2.328947368421053e-06,
+      "loss": 0.3199,
+      "step": 223
+    },
+    {
+      "Batch Sum": 221.9879150390625,
+      "accuracy": 0.90625,
+      "epoch": 0.5575,
+      "step": 223
+    },
+    {
+      "Batch Sum": 210.73193359375,
+      "accuracy": 0.859375,
+      "epoch": 0.5575,
+      "step": 223
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.66073989868164,
+      "learning_rate": 2.3157894736842105e-06,
+      "loss": 0.2951,
+      "step": 224
+    },
+    {
+      "Batch Sum": 209.3916015625,
+      "accuracy": 0.8125,
+      "epoch": 0.56,
+      "step": 224
+    },
+    {
+      "Batch Sum": 204.224609375,
+      "accuracy": 0.875,
+      "epoch": 0.56,
+      "step": 224
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 14.799392700195312,
+      "learning_rate": 2.3026315789473684e-06,
+      "loss": 0.4081,
+      "step": 225
+    },
+    {
+      "Batch Sum": 189.40283203125,
+      "accuracy": 0.859375,
+      "epoch": 0.5625,
+      "step": 225
+    },
+    {
+      "Batch Sum": 192.8564453125,
+      "accuracy": 0.84375,
+      "epoch": 0.5625,
+      "step": 225
+    },
+    {
+      "epoch": 0.565,
+      "grad_norm": 11.401546478271484,
+      "learning_rate": 2.2894736842105263e-06,
+      "loss": 0.302,
+      "step": 226
+    },
+    {
+      "Batch Sum": 196.44488525390625,
+      "accuracy": 0.84375,
+      "epoch": 0.565,
+      "step": 226
+    },
+    {
+      "Batch Sum": 202.50390625,
+      "accuracy": 0.859375,
+      "epoch": 0.565,
+      "step": 226
+    },
+    {
+      "epoch": 0.5675,
+      "grad_norm": 11.403398513793945,
+      "learning_rate": 2.2763157894736847e-06,
+      "loss": 0.3343,
+      "step": 227
+    },
+    {
+      "Batch Sum": 201.01004028320312,
+      "accuracy": 0.75,
+      "epoch": 0.5675,
+      "step": 227
+    },
+    {
+      "Batch Sum": 214.990478515625,
+      "accuracy": 0.84375,
+      "epoch": 0.5675,
+      "step": 227
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 14.557242393493652,
+      "learning_rate": 2.2631578947368426e-06,
+      "loss": 0.4123,
+      "step": 228
+    },
+    {
+      "Batch Sum": 202.841064453125,
+      "accuracy": 0.859375,
+      "epoch": 0.57,
+      "step": 228
+    },
+    {
+      "Batch Sum": 200.405029296875,
+      "accuracy": 0.859375,
+      "epoch": 0.57,
+      "step": 228
+    },
+    {
+      "epoch": 0.5725,
+      "grad_norm": 12.387444496154785,
+      "learning_rate": 2.25e-06,
+      "loss": 0.3513,
+      "step": 229
+    },
+    {
+      "Batch Sum": 181.8583984375,
+      "accuracy": 0.859375,
+      "epoch": 0.5725,
+      "step": 229
+    },
+    {
+      "Batch Sum": 190.2799072265625,
+      "accuracy": 0.84375,
+      "epoch": 0.5725,
+      "step": 229
+    },
+    {
+      "epoch": 0.575,
+      "grad_norm": 11.778051376342773,
+      "learning_rate": 2.236842105263158e-06,
+      "loss": 0.3423,
+      "step": 230
+    },
+    {
+      "Batch Sum": 184.83056640625,
+      "accuracy": 0.84375,
+      "epoch": 0.575,
+      "step": 230
+    },
+    {
+      "Batch Sum": 205.462890625,
+      "accuracy": 0.875,
+      "epoch": 0.575,
+      "step": 230
+    },
+    {
+      "epoch": 0.5775,
+      "grad_norm": 10.590459823608398,
+      "learning_rate": 2.223684210526316e-06,
+      "loss": 0.2898,
+      "step": 231
+    },
+    {
+      "Batch Sum": 215.6883544921875,
+      "accuracy": 0.875,
+      "epoch": 0.5775,
+      "step": 231
+    },
+    {
+      "Batch Sum": 196.66650390625,
+      "accuracy": 0.78125,
+      "epoch": 0.5775,
+      "step": 231
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 15.713869094848633,
+      "learning_rate": 2.2105263157894738e-06,
+      "loss": 0.4,
+      "step": 232
+    },
+    {
+      "Batch Sum": 193.6600341796875,
+      "accuracy": 0.859375,
+      "epoch": 0.58,
+      "step": 232
+    },
+    {
+      "Batch Sum": 207.85665893554688,
+      "accuracy": 0.921875,
+      "epoch": 0.58,
+      "step": 232
+    },
+    {
+      "epoch": 0.5825,
+      "grad_norm": 10.010421752929688,
+      "learning_rate": 2.1973684210526317e-06,
+      "loss": 0.2885,
+      "step": 233
+    },
+    {
+      "Batch Sum": 171.9423828125,
+      "accuracy": 0.828125,
+      "epoch": 0.5825,
+      "step": 233
+    },
+    {
+      "Batch Sum": 174.4212646484375,
+      "accuracy": 0.875,
+      "epoch": 0.5825,
+      "step": 233
+    },
+    {
+      "epoch": 0.585,
+      "grad_norm": 10.00387954711914,
+      "learning_rate": 2.1842105263157896e-06,
+      "loss": 0.2916,
+      "step": 234
+    },
+    {
+      "Batch Sum": 187.0257568359375,
+      "accuracy": 0.828125,
+      "epoch": 0.585,
+      "step": 234
+    },
+    {
+      "Batch Sum": 201.4970703125,
+      "accuracy": 0.828125,
+      "epoch": 0.585,
+      "step": 234
+    },
+    {
+      "epoch": 0.5875,
+      "grad_norm": 10.736607551574707,
+      "learning_rate": 2.1710526315789475e-06,
+      "loss": 0.2984,
+      "step": 235
+    },
+    {
+      "Batch Sum": 203.65283203125,
+      "accuracy": 0.765625,
+      "epoch": 0.5875,
+      "step": 235
+    },
+    {
+      "Batch Sum": 171.0863037109375,
+      "accuracy": 0.75,
+      "epoch": 0.5875,
+      "step": 235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.4028902053833,
+      "learning_rate": 2.1578947368421054e-06,
+      "loss": 0.4165,
+      "step": 236
+    },
+    {
+      "Batch Sum": 188.96435546875,
+      "accuracy": 0.828125,
+      "epoch": 0.59,
+      "step": 236
+    },
+    {
+      "Batch Sum": 184.859130859375,
+      "accuracy": 0.828125,
+      "epoch": 0.59,
+      "step": 236
+    },
+    {
+      "epoch": 0.5925,
+      "grad_norm": 13.162971496582031,
+      "learning_rate": 2.1447368421052633e-06,
+      "loss": 0.39,
+      "step": 237
+    },
+    {
+      "Batch Sum": 182.19122314453125,
+      "accuracy": 0.796875,
+      "epoch": 0.5925,
+      "step": 237
+    },
+    {
+      "Batch Sum": 208.36264038085938,
+      "accuracy": 0.8125,
+      "epoch": 0.5925,
+      "step": 237
+    },
+    {
+      "epoch": 0.595,
+      "grad_norm": 12.986008644104004,
+      "learning_rate": 2.1315789473684212e-06,
+      "loss": 0.4015,
+      "step": 238
+    },
+    {
+      "Batch Sum": 183.390869140625,
+      "accuracy": 0.828125,
+      "epoch": 0.595,
+      "step": 238
+    },
+    {
+      "Batch Sum": 203.293701171875,
+      "accuracy": 0.84375,
+      "epoch": 0.595,
+      "step": 238
+    },
+    {
+      "epoch": 0.5975,
+      "grad_norm": 14.014009475708008,
+      "learning_rate": 2.118421052631579e-06,
+      "loss": 0.4549,
+      "step": 239
+    },
+    {
+      "Batch Sum": 176.718017578125,
+      "accuracy": 0.78125,
+      "epoch": 0.5975,
+      "step": 239
+    },
+    {
+      "Batch Sum": 168.53076171875,
+      "accuracy": 0.765625,
+      "epoch": 0.5975,
+      "step": 239
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.759445190429688,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 0.4158,
+      "step": 240
+    },
+    {
+      "Batch Sum": 169.9306640625,
+      "accuracy": 0.796875,
+      "epoch": 0.6,
+      "step": 240
+    },
+    {
+      "Batch Sum": 161.29913330078125,
+      "accuracy": 0.875,
+      "epoch": 0.6,
+      "step": 240
+    },
+    {
+      "epoch": 0.6025,
+      "grad_norm": 10.13369369506836,
+      "learning_rate": 2.092105263157895e-06,
+      "loss": 0.3565,
+      "step": 241
+    },
+    {
+      "Batch Sum": 169.426513671875,
+      "accuracy": 0.84375,
+      "epoch": 0.6025,
+      "step": 241
+    },
+    {
+      "Batch Sum": 177.166259765625,
+      "accuracy": 0.828125,
+      "epoch": 0.6025,
+      "step": 241
+    },
+    {
+      "epoch": 0.605,
+      "grad_norm": 12.44567584991455,
+      "learning_rate": 2.078947368421053e-06,
+      "loss": 0.3771,
+      "step": 242
+    },
+    {
+      "Batch Sum": 166.62744140625,
+      "accuracy": 0.84375,
+      "epoch": 0.605,
+      "step": 242
+    },
+    {
+      "Batch Sum": 178.87554931640625,
+      "accuracy": 0.875,
+      "epoch": 0.605,
+      "step": 242
+    },
+    {
+      "epoch": 0.6075,
+      "grad_norm": 11.519309043884277,
+      "learning_rate": 2.0657894736842108e-06,
+      "loss": 0.2864,
+      "step": 243
+    },
+    {
+      "Batch Sum": 157.76678466796875,
+      "accuracy": 0.765625,
+      "epoch": 0.6075,
+      "step": 243
+    },
+    {
+      "Batch Sum": 164.535888671875,
+      "accuracy": 0.890625,
+      "epoch": 0.6075,
+      "step": 243
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.705724716186523,
+      "learning_rate": 2.0526315789473687e-06,
+      "loss": 0.4141,
+      "step": 244
+    },
+    {
+      "Batch Sum": 175.552490234375,
+      "accuracy": 0.796875,
+      "epoch": 0.61,
+      "step": 244
+    },
+    {
+      "Batch Sum": 163.62815856933594,
+      "accuracy": 0.90625,
+      "epoch": 0.61,
+      "step": 244
+    },
+    {
+      "epoch": 0.6125,
+      "grad_norm": 11.879432678222656,
+      "learning_rate": 2.0394736842105266e-06,
+      "loss": 0.3428,
+      "step": 245
+    },
+    {
+      "Batch Sum": 159.4302978515625,
+      "accuracy": 0.828125,
+      "epoch": 0.6125,
+      "step": 245
+    },
+    {
+      "Batch Sum": 159.87109375,
+      "accuracy": 0.765625,
+      "epoch": 0.6125,
+      "step": 245
+    },
+    {
+      "epoch": 0.615,
+      "grad_norm": 13.999123573303223,
+      "learning_rate": 2.026315789473684e-06,
+      "loss": 0.4236,
+      "step": 246
+    },
+    {
+      "Batch Sum": 180.75390625,
+      "accuracy": 0.796875,
+      "epoch": 0.615,
+      "step": 246
+    },
+    {
+      "Batch Sum": 159.6279296875,
+      "accuracy": 0.859375,
+      "epoch": 0.615,
+      "step": 246
+    },
+    {
+      "epoch": 0.6175,
+      "grad_norm": 11.608491897583008,
+      "learning_rate": 2.013157894736842e-06,
+      "loss": 0.3972,
+      "step": 247
+    },
+    {
+      "Batch Sum": 164.1768798828125,
+      "accuracy": 0.734375,
+      "epoch": 0.6175,
+      "step": 247
+    },
+    {
+      "Batch Sum": 155.8487548828125,
+      "accuracy": 0.859375,
+      "epoch": 0.6175,
+      "step": 247
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.695984840393066,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.3948,
+      "step": 248
+    },
+    {
+      "Batch Sum": 152.81097412109375,
+      "accuracy": 0.875,
+      "epoch": 0.62,
+      "step": 248
+    },
+    {
+      "Batch Sum": 144.8948974609375,
+      "accuracy": 0.84375,
+      "epoch": 0.62,
+      "step": 248
+    },
+    {
+      "epoch": 0.6225,
+      "grad_norm": 9.89355754852295,
+      "learning_rate": 1.9868421052631582e-06,
+      "loss": 0.2686,
+      "step": 249
+    },
+    {
+      "Batch Sum": 167.4403076171875,
+      "accuracy": 0.828125,
+      "epoch": 0.6225,
+      "step": 249
+    },
+    {
+      "Batch Sum": 148.5291748046875,
+      "accuracy": 0.84375,
+      "epoch": 0.6225,
+      "step": 249
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 10.045268058776855,
+      "learning_rate": 1.973684210526316e-06,
+      "loss": 0.3181,
+      "step": 250
+    },
+    {
+      "Batch Sum": 175.96142578125,
+      "accuracy": 0.890625,
+      "epoch": 0.625,
+      "step": 250
+    },
+    {
+      "Batch Sum": 170.6485595703125,
+      "accuracy": 0.875,
+      "epoch": 0.625,
+      "step": 250
+    },
+    {
+      "epoch": 0.6275,
+      "grad_norm": 10.1083402633667,
+      "learning_rate": 1.9605263157894736e-06,
+      "loss": 0.2626,
+      "step": 251
+    },
+    {
+      "Batch Sum": 161.82470703125,
+      "accuracy": 0.84375,
+      "epoch": 0.6275,
+      "step": 251
+    },
+    {
+      "Batch Sum": 149.327392578125,
+      "accuracy": 0.78125,
+      "epoch": 0.6275,
+      "step": 251
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.48316764831543,
+      "learning_rate": 1.9473684210526315e-06,
+      "loss": 0.3869,
+      "step": 252
+    },
+    {
+      "Batch Sum": 184.88150024414062,
+      "accuracy": 0.875,
+      "epoch": 0.63,
+      "step": 252
+    },
+    {
+      "Batch Sum": 180.953125,
+      "accuracy": 0.890625,
+      "epoch": 0.63,
+      "step": 252
+    },
+    {
+      "epoch": 0.6325,
+      "grad_norm": 10.67346477508545,
+      "learning_rate": 1.9342105263157895e-06,
+      "loss": 0.3168,
+      "step": 253
+    },
+    {
+      "Batch Sum": 182.32440185546875,
+      "accuracy": 0.859375,
+      "epoch": 0.6325,
+      "step": 253
+    },
+    {
+      "Batch Sum": 175.0447998046875,
+      "accuracy": 0.8125,
+      "epoch": 0.6325,
+      "step": 253
+    },
+    {
+      "epoch": 0.635,
+      "grad_norm": 15.229326248168945,
+      "learning_rate": 1.9210526315789474e-06,
+      "loss": 0.3236,
+      "step": 254
+    },
+    {
+      "Batch Sum": 177.50146484375,
+      "accuracy": 0.71875,
+      "epoch": 0.635,
+      "step": 254
+    },
+    {
+      "Batch Sum": 173.072509765625,
+      "accuracy": 0.84375,
+      "epoch": 0.635,
+      "step": 254
+    },
+    {
+      "epoch": 0.6375,
+      "grad_norm": 15.01415729522705,
+      "learning_rate": 1.9078947368421057e-06,
+      "loss": 0.4247,
+      "step": 255
+    },
+    {
+      "Batch Sum": 179.178466796875,
+      "accuracy": 0.765625,
+      "epoch": 0.6375,
+      "step": 255
+    },
+    {
+      "Batch Sum": 191.432373046875,
+      "accuracy": 0.8125,
+      "epoch": 0.6375,
+      "step": 255
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.656612396240234,
+      "learning_rate": 1.8947368421052634e-06,
+      "loss": 0.3423,
+      "step": 256
+    },
+    {
+      "Batch Sum": 174.36767578125,
+      "accuracy": 0.859375,
+      "epoch": 0.64,
+      "step": 256
+    },
+    {
+      "Batch Sum": 200.411376953125,
+      "accuracy": 0.78125,
+      "epoch": 0.64,
+      "step": 256
+    },
+    {
+      "epoch": 0.6425,
+      "grad_norm": 12.723453521728516,
+      "learning_rate": 1.8815789473684213e-06,
+      "loss": 0.3384,
+      "step": 257
+    },
+    {
+      "Batch Sum": 174.244140625,
+      "accuracy": 0.828125,
+      "epoch": 0.6425,
+      "step": 257
+    },
+    {
+      "Batch Sum": 187.5958251953125,
+      "accuracy": 0.765625,
+      "epoch": 0.6425,
+      "step": 257
+    },
+    {
+      "epoch": 0.645,
+      "grad_norm": 13.295060157775879,
+      "learning_rate": 1.868421052631579e-06,
+      "loss": 0.3964,
+      "step": 258
+    },
+    {
+      "Batch Sum": 172.2286376953125,
+      "accuracy": 0.796875,
+      "epoch": 0.645,
+      "step": 258
+    },
+    {
+      "Batch Sum": 178.4345703125,
+      "accuracy": 0.890625,
+      "epoch": 0.645,
+      "step": 258
+    },
+    {
+      "epoch": 0.6475,
+      "grad_norm": 13.88492488861084,
+      "learning_rate": 1.855263157894737e-06,
+      "loss": 0.3701,
+      "step": 259
+    },
+    {
+      "Batch Sum": 168.85760498046875,
+      "accuracy": 0.84375,
+      "epoch": 0.6475,
+      "step": 259
+    },
+    {
+      "Batch Sum": 174.300048828125,
+      "accuracy": 0.921875,
+      "epoch": 0.6475,
+      "step": 259
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.077507019042969,
+      "learning_rate": 1.8421052631578948e-06,
+      "loss": 0.3483,
+      "step": 260
+    },
+    {
+      "Batch Sum": 188.45849609375,
+      "accuracy": 0.75,
+      "epoch": 0.65,
+      "step": 260
+    },
+    {
+      "Batch Sum": 176.146484375,
+      "accuracy": 0.859375,
+      "epoch": 0.65,
+      "step": 260
+    },
+    {
+      "epoch": 0.6525,
+      "grad_norm": 16.151479721069336,
+      "learning_rate": 1.828947368421053e-06,
+      "loss": 0.4429,
+      "step": 261
+    },
+    {
+      "Batch Sum": 158.18531799316406,
+      "accuracy": 0.78125,
+      "epoch": 0.6525,
+      "step": 261
+    },
+    {
+      "Batch Sum": 176.09814453125,
+      "accuracy": 0.765625,
+      "epoch": 0.6525,
+      "step": 261
+    },
+    {
+      "epoch": 0.655,
+      "grad_norm": 13.163310050964355,
+      "learning_rate": 1.8157894736842109e-06,
+      "loss": 0.41,
+      "step": 262
+    },
+    {
+      "Batch Sum": 175.75927734375,
+      "accuracy": 0.859375,
+      "epoch": 0.655,
+      "step": 262
+    },
+    {
+      "Batch Sum": 163.3080596923828,
+      "accuracy": 0.890625,
+      "epoch": 0.655,
+      "step": 262
+    },
+    {
+      "epoch": 0.6575,
+      "grad_norm": 11.67258071899414,
+      "learning_rate": 1.8026315789473685e-06,
+      "loss": 0.363,
+      "step": 263
+    },
+    {
+      "Batch Sum": 166.0087890625,
+      "accuracy": 0.78125,
+      "epoch": 0.6575,
+      "step": 263
+    },
+    {
+      "Batch Sum": 189.489501953125,
+      "accuracy": 0.828125,
+      "epoch": 0.6575,
+      "step": 263
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.687853813171387,
+      "learning_rate": 1.7894736842105265e-06,
+      "loss": 0.3771,
+      "step": 264
+    },
+    {
+      "Batch Sum": 159.7763671875,
+      "accuracy": 0.75,
+      "epoch": 0.66,
+      "step": 264
+    },
+    {
+      "Batch Sum": 180.8643798828125,
+      "accuracy": 0.859375,
+      "epoch": 0.66,
+      "step": 264
+    },
+    {
+      "epoch": 0.6625,
+      "grad_norm": 11.59654426574707,
+      "learning_rate": 1.7763157894736844e-06,
+      "loss": 0.4142,
+      "step": 265
+    },
+    {
+      "Batch Sum": 165.2060546875,
+      "accuracy": 0.8125,
+      "epoch": 0.6625,
+      "step": 265
+    },
+    {
+      "Batch Sum": 160.62353515625,
+      "accuracy": 0.8125,
+      "epoch": 0.6625,
+      "step": 265
+    },
+    {
+      "epoch": 0.665,
+      "grad_norm": 12.089131355285645,
+      "learning_rate": 1.7631578947368423e-06,
+      "loss": 0.408,
+      "step": 266
+    },
+    {
+      "Batch Sum": 152.2724609375,
+      "accuracy": 0.796875,
+      "epoch": 0.665,
+      "step": 266
+    },
+    {
+      "Batch Sum": 146.29530334472656,
+      "accuracy": 0.75,
+      "epoch": 0.665,
+      "step": 266
+    },
+    {
+      "epoch": 0.6675,
+      "grad_norm": 11.259248733520508,
+      "learning_rate": 1.75e-06,
+      "loss": 0.3879,
+      "step": 267
+    },
+    {
+      "Batch Sum": 152.69775390625,
+      "accuracy": 0.75,
+      "epoch": 0.6675,
+      "step": 267
+    },
+    {
+      "Batch Sum": 167.01861572265625,
+      "accuracy": 0.8125,
+      "epoch": 0.6675,
+      "step": 267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.48369312286377,
+      "learning_rate": 1.736842105263158e-06,
+      "loss": 0.4232,
+      "step": 268
+    },
+    {
+      "Batch Sum": 158.3505859375,
+      "accuracy": 0.8125,
+      "epoch": 0.67,
+      "step": 268
+    },
+    {
+      "Batch Sum": 139.9046630859375,
+      "accuracy": 0.8125,
+      "epoch": 0.67,
+      "step": 268
+    },
+    {
+      "epoch": 0.6725,
+      "grad_norm": 11.154230117797852,
+      "learning_rate": 1.723684210526316e-06,
+      "loss": 0.3859,
+      "step": 269
+    },
+    {
+      "Batch Sum": 158.55126953125,
+      "accuracy": 0.703125,
+      "epoch": 0.6725,
+      "step": 269
+    },
+    {
+      "Batch Sum": 147.5528564453125,
+      "accuracy": 0.84375,
+      "epoch": 0.6725,
+      "step": 269
+    },
+    {
+      "epoch": 0.675,
+      "grad_norm": 11.144701957702637,
+      "learning_rate": 1.710526315789474e-06,
+      "loss": 0.424,
+      "step": 270
+    },
+    {
+      "Batch Sum": 138.054931640625,
+      "accuracy": 0.734375,
+      "epoch": 0.675,
+      "step": 270
+    },
+    {
+      "Batch Sum": 124.85104370117188,
+      "accuracy": 0.78125,
+      "epoch": 0.675,
+      "step": 270
+    },
+    {
+      "epoch": 0.6775,
+      "grad_norm": 10.864275932312012,
+      "learning_rate": 1.6973684210526318e-06,
+      "loss": 0.4427,
+      "step": 271
+    },
+    {
+      "Batch Sum": 138.0224609375,
+      "accuracy": 0.890625,
+      "epoch": 0.6775,
+      "step": 271
+    },
+    {
+      "Batch Sum": 151.0075225830078,
+      "accuracy": 0.75,
+      "epoch": 0.6775,
+      "step": 271
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.548851013183594,
+      "learning_rate": 1.6842105263157895e-06,
+      "loss": 0.3706,
+      "step": 272
+    },
+    {
+      "Batch Sum": 148.39080810546875,
+      "accuracy": 0.765625,
+      "epoch": 0.68,
+      "step": 272
+    },
+    {
+      "Batch Sum": 135.73193359375,
+      "accuracy": 0.84375,
+      "epoch": 0.68,
+      "step": 272
+    },
+    {
+      "epoch": 0.6825,
+      "grad_norm": 10.340380668640137,
+      "learning_rate": 1.6710526315789474e-06,
+      "loss": 0.3761,
+      "step": 273
+    },
+    {
+      "Batch Sum": 133.164794921875,
+      "accuracy": 0.71875,
+      "epoch": 0.6825,
+      "step": 273
+    },
+    {
+      "Batch Sum": 129.74310302734375,
+      "accuracy": 0.828125,
+      "epoch": 0.6825,
+      "step": 273
+    },
+    {
+      "epoch": 0.685,
+      "grad_norm": 9.391560554504395,
+      "learning_rate": 1.6578947368421053e-06,
+      "loss": 0.393,
+      "step": 274
+    },
+    {
+      "Batch Sum": 138.79864501953125,
+      "accuracy": 0.859375,
+      "epoch": 0.685,
+      "step": 274
+    },
+    {
+      "Batch Sum": 121.115966796875,
+      "accuracy": 0.703125,
+      "epoch": 0.685,
+      "step": 274
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 10.134588241577148,
+      "learning_rate": 1.6447368421052635e-06,
+      "loss": 0.4303,
+      "step": 275
+    },
+    {
+      "Batch Sum": 130.93405151367188,
+      "accuracy": 0.875,
+      "epoch": 0.6875,
+      "step": 275
+    },
+    {
+      "Batch Sum": 145.55398559570312,
+      "accuracy": 0.859375,
+      "epoch": 0.6875,
+      "step": 275
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.827194213867188,
+      "learning_rate": 1.6315789473684212e-06,
+      "loss": 0.3539,
+      "step": 276
+    },
+    {
+      "Batch Sum": 152.80908203125,
+      "accuracy": 0.75,
+      "epoch": 0.69,
+      "step": 276
+    },
+    {
+      "Batch Sum": 144.02023315429688,
+      "accuracy": 0.765625,
+      "epoch": 0.69,
+      "step": 276
+    },
+    {
+      "epoch": 0.6925,
+      "grad_norm": 11.620319366455078,
+      "learning_rate": 1.618421052631579e-06,
+      "loss": 0.448,
+      "step": 277
+    },
+    {
+      "Batch Sum": 139.06533813476562,
+      "accuracy": 0.84375,
+      "epoch": 0.6925,
+      "step": 277
+    },
+    {
+      "Batch Sum": 153.028076171875,
+      "accuracy": 0.875,
+      "epoch": 0.6925,
+      "step": 277
+    },
+    {
+      "epoch": 0.695,
+      "grad_norm": 9.624265670776367,
+      "learning_rate": 1.605263157894737e-06,
+      "loss": 0.3399,
+      "step": 278
+    },
+    {
+      "Batch Sum": 153.900634765625,
+      "accuracy": 0.75,
+      "epoch": 0.695,
+      "step": 278
+    },
+    {
+      "Batch Sum": 158.28204345703125,
+      "accuracy": 0.84375,
+      "epoch": 0.695,
+      "step": 278
+    },
+    {
+      "epoch": 0.6975,
+      "grad_norm": 10.872698783874512,
+      "learning_rate": 1.5921052631578949e-06,
+      "loss": 0.3803,
+      "step": 279
+    },
+    {
+      "Batch Sum": 156.890869140625,
+      "accuracy": 0.828125,
+      "epoch": 0.6975,
+      "step": 279
+    },
+    {
+      "Batch Sum": 143.0692138671875,
+      "accuracy": 0.828125,
+      "epoch": 0.6975,
+      "step": 279
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.971992492675781,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 0.3746,
+      "step": 280
+    },
+    {
+      "Batch Sum": 165.6650390625,
+      "accuracy": 0.796875,
+      "epoch": 0.7,
+      "step": 280
+    },
+    {
+      "Batch Sum": 170.763427734375,
+      "accuracy": 0.828125,
+      "epoch": 0.7,
+      "step": 280
+    },
+    {
+      "epoch": 0.7025,
+      "grad_norm": 9.986807823181152,
+      "learning_rate": 1.5657894736842105e-06,
+      "loss": 0.3516,
+      "step": 281
+    },
+    {
+      "Batch Sum": 169.47926330566406,
+      "accuracy": 0.78125,
+      "epoch": 0.7025,
+      "step": 281
+    },
+    {
+      "Batch Sum": 172.787353515625,
+      "accuracy": 0.875,
+      "epoch": 0.7025,
+      "step": 281
+    },
+    {
+      "epoch": 0.705,
+      "grad_norm": 11.13526725769043,
+      "learning_rate": 1.5526315789473686e-06,
+      "loss": 0.3736,
+      "step": 282
+    },
+    {
+      "Batch Sum": 165.2123260498047,
+      "accuracy": 0.796875,
+      "epoch": 0.705,
+      "step": 282
+    },
+    {
+      "Batch Sum": 179.7053680419922,
+      "accuracy": 0.734375,
+      "epoch": 0.705,
+      "step": 282
+    },
+    {
+      "epoch": 0.7075,
+      "grad_norm": 11.187799453735352,
+      "learning_rate": 1.5394736842105265e-06,
+      "loss": 0.3895,
+      "step": 283
+    },
+    {
+      "Batch Sum": 163.5240478515625,
+      "accuracy": 0.828125,
+      "epoch": 0.7075,
+      "step": 283
+    },
+    {
+      "Batch Sum": 167.626953125,
+      "accuracy": 0.890625,
+      "epoch": 0.7075,
+      "step": 283
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.662762641906738,
+      "learning_rate": 1.5263157894736844e-06,
+      "loss": 0.317,
+      "step": 284
+    },
+    {
+      "Batch Sum": 175.9295654296875,
+      "accuracy": 0.765625,
+      "epoch": 0.71,
+      "step": 284
+    },
+    {
+      "Batch Sum": 170.70166015625,
+      "accuracy": 0.796875,
+      "epoch": 0.71,
+      "step": 284
+    },
+    {
+      "epoch": 0.7125,
+      "grad_norm": 12.166522979736328,
+      "learning_rate": 1.5131578947368421e-06,
+      "loss": 0.4163,
+      "step": 285
+    },
+    {
+      "Batch Sum": 192.76531982421875,
+      "accuracy": 0.828125,
+      "epoch": 0.7125,
+      "step": 285
+    },
+    {
+      "Batch Sum": 178.5068359375,
+      "accuracy": 0.796875,
+      "epoch": 0.7125,
+      "step": 285
+    },
+    {
+      "epoch": 0.715,
+      "grad_norm": 11.759383201599121,
+      "learning_rate": 1.5e-06,
+      "loss": 0.3661,
+      "step": 286
+    },
+    {
+      "Batch Sum": 176.7509765625,
+      "accuracy": 0.84375,
+      "epoch": 0.715,
+      "step": 286
+    },
+    {
+      "Batch Sum": 181.384521484375,
+      "accuracy": 0.875,
+      "epoch": 0.715,
+      "step": 286
+    },
+    {
+      "epoch": 0.7175,
+      "grad_norm": 11.38322925567627,
+      "learning_rate": 1.486842105263158e-06,
+      "loss": 0.3306,
+      "step": 287
+    },
+    {
+      "Batch Sum": 178.4794921875,
+      "accuracy": 0.84375,
+      "epoch": 0.7175,
+      "step": 287
+    },
+    {
+      "Batch Sum": 184.6533203125,
+      "accuracy": 0.78125,
+      "epoch": 0.7175,
+      "step": 287
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.384878158569336,
+      "learning_rate": 1.4736842105263159e-06,
+      "loss": 0.3039,
+      "step": 288
+    },
+    {
+      "Batch Sum": 172.48699951171875,
+      "accuracy": 0.8125,
+      "epoch": 0.72,
+      "step": 288
+    },
+    {
+      "Batch Sum": 183.55014038085938,
+      "accuracy": 0.78125,
+      "epoch": 0.72,
+      "step": 288
+    },
+    {
+      "epoch": 0.7225,
+      "grad_norm": 12.995137214660645,
+      "learning_rate": 1.460526315789474e-06,
+      "loss": 0.3493,
+      "step": 289
+    },
+    {
+      "Batch Sum": 182.96871948242188,
+      "accuracy": 0.84375,
+      "epoch": 0.7225,
+      "step": 289
+    },
+    {
+      "Batch Sum": 203.369140625,
+      "accuracy": 0.859375,
+      "epoch": 0.7225,
+      "step": 289
+    },
+    {
+      "epoch": 0.725,
+      "grad_norm": 10.916609764099121,
+      "learning_rate": 1.4473684210526317e-06,
+      "loss": 0.2863,
+      "step": 290
+    },
+    {
+      "Batch Sum": 178.37255859375,
+      "accuracy": 0.828125,
+      "epoch": 0.725,
+      "step": 290
+    },
+    {
+      "Batch Sum": 163.4500732421875,
+      "accuracy": 0.859375,
+      "epoch": 0.725,
+      "step": 290
+    },
+    {
+      "epoch": 0.7275,
+      "grad_norm": 12.26250171661377,
+      "learning_rate": 1.4342105263157896e-06,
+      "loss": 0.3507,
+      "step": 291
+    },
+    {
+      "Batch Sum": 204.6234130859375,
+      "accuracy": 0.796875,
+      "epoch": 0.7275,
+      "step": 291
+    },
+    {
+      "Batch Sum": 185.291015625,
+      "accuracy": 0.875,
+      "epoch": 0.7275,
+      "step": 291
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.206491470336914,
+      "learning_rate": 1.4210526315789475e-06,
+      "loss": 0.3976,
+      "step": 292
+    },
+    {
+      "Batch Sum": 181.6513671875,
+      "accuracy": 0.890625,
+      "epoch": 0.73,
+      "step": 292
+    },
+    {
+      "Batch Sum": 223.3564453125,
+      "accuracy": 0.8125,
+      "epoch": 0.73,
+      "step": 292
+    },
+    {
+      "epoch": 0.7325,
+      "grad_norm": 10.997773170471191,
+      "learning_rate": 1.4078947368421054e-06,
+      "loss": 0.3032,
+      "step": 293
+    },
+    {
+      "Batch Sum": 182.54302978515625,
+      "accuracy": 0.8125,
+      "epoch": 0.7325,
+      "step": 293
+    },
+    {
+      "Batch Sum": 199.45068359375,
+      "accuracy": 0.8125,
+      "epoch": 0.7325,
+      "step": 293
+    },
+    {
+      "epoch": 0.735,
+      "grad_norm": 14.625078201293945,
+      "learning_rate": 1.394736842105263e-06,
+      "loss": 0.416,
+      "step": 294
+    },
+    {
+      "Batch Sum": 204.288818359375,
+      "accuracy": 0.84375,
+      "epoch": 0.735,
+      "step": 294
+    },
+    {
+      "Batch Sum": 185.4790496826172,
+      "accuracy": 0.828125,
+      "epoch": 0.735,
+      "step": 294
+    },
+    {
+      "epoch": 0.7375,
+      "grad_norm": 11.689982414245605,
+      "learning_rate": 1.3815789473684212e-06,
+      "loss": 0.3305,
+      "step": 295
+    },
+    {
+      "Batch Sum": 176.4525146484375,
+      "accuracy": 0.90625,
+      "epoch": 0.7375,
+      "step": 295
+    },
+    {
+      "Batch Sum": 186.1865234375,
+      "accuracy": 0.890625,
+      "epoch": 0.7375,
+      "step": 295
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.708041191101074,
+      "learning_rate": 1.3684210526315791e-06,
+      "loss": 0.2496,
+      "step": 296
+    },
+    {
+      "Batch Sum": 186.8370361328125,
+      "accuracy": 0.78125,
+      "epoch": 0.74,
+      "step": 296
+    },
+    {
+      "Batch Sum": 196.197998046875,
+      "accuracy": 0.828125,
+      "epoch": 0.74,
+      "step": 296
+    },
+    {
+      "epoch": 0.7425,
+      "grad_norm": 13.83351993560791,
+      "learning_rate": 1.355263157894737e-06,
+      "loss": 0.4217,
+      "step": 297
+    },
+    {
+      "Batch Sum": 187.9921875,
+      "accuracy": 0.78125,
+      "epoch": 0.7425,
+      "step": 297
+    },
+    {
+      "Batch Sum": 194.74755859375,
+      "accuracy": 0.84375,
+      "epoch": 0.7425,
+      "step": 297
+    },
+    {
+      "epoch": 0.745,
+      "grad_norm": 15.314090728759766,
+      "learning_rate": 1.342105263157895e-06,
+      "loss": 0.4151,
+      "step": 298
+    },
+    {
+      "Batch Sum": 221.93280029296875,
+      "accuracy": 0.875,
+      "epoch": 0.745,
+      "step": 298
+    },
+    {
+      "Batch Sum": 197.100341796875,
+      "accuracy": 0.828125,
+      "epoch": 0.745,
+      "step": 298
+    },
+    {
+      "epoch": 0.7475,
+      "grad_norm": 11.85640811920166,
+      "learning_rate": 1.3289473684210526e-06,
+      "loss": 0.2987,
+      "step": 299
+    },
+    {
+      "Batch Sum": 200.4658203125,
+      "accuracy": 0.71875,
+      "epoch": 0.7475,
+      "step": 299
+    },
+    {
+      "Batch Sum": 192.3809814453125,
+      "accuracy": 0.796875,
+      "epoch": 0.7475,
+      "step": 299
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 16.492263793945312,
+      "learning_rate": 1.3157894736842106e-06,
+      "loss": 0.453,
+      "step": 300
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 400,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}