diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,34883 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9998192662208565,
+  "eval_steps": 500,
+  "global_step": 24897,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.06317149847745895,
+      "learning_rate": 7.499999999999999e-07,
+      "loss": 1.0918,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.06490806490182877,
+      "learning_rate": 1.4999999999999998e-06,
+      "loss": 1.1945,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.06881864368915558,
+      "learning_rate": 2.2499999999999996e-06,
+      "loss": 1.1409,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.07844685763120651,
+      "learning_rate": 2.9999999999999997e-06,
+      "loss": 1.2207,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.08064830303192139,
+      "learning_rate": 3.7499999999999997e-06,
+      "loss": 1.1945,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.07337764650583267,
+      "learning_rate": 4.499999999999999e-06,
+      "loss": 1.1097,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.07508231699466705,
+      "learning_rate": 5.25e-06,
+      "loss": 1.1413,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0660620704293251,
+      "learning_rate": 5.999999999999999e-06,
+      "loss": 1.0751,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0696869045495987,
+      "learning_rate": 6.749999999999999e-06,
+      "loss": 1.1048,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0708666518330574,
+      "learning_rate": 7.499999999999999e-06,
+      "loss": 1.1369,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.07739271968603134,
+      "learning_rate": 8.249999999999999e-06,
+      "loss": 1.1332,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0700831338763237,
+      "learning_rate": 8.999999999999999e-06,
+      "loss": 1.2094,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.07708919048309326,
+      "learning_rate": 9.75e-06,
+      "loss": 1.1344,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.07628925144672394,
+      "learning_rate": 1.05e-05,
+      "loss": 1.0817,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.08013926446437836,
+      "learning_rate": 1.1249999999999999e-05,
+      "loss": 1.0898,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.09165573865175247,
+      "learning_rate": 1.1999999999999999e-05,
+      "loss": 1.1173,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.08685644716024399,
+      "learning_rate": 1.275e-05,
+      "loss": 1.1268,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.08818597346544266,
+      "learning_rate": 1.3499999999999998e-05,
+      "loss": 1.1928,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0848746970295906,
+      "learning_rate": 1.4249999999999999e-05,
+      "loss": 1.1179,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.09676993638277054,
+      "learning_rate": 1.4999999999999999e-05,
+      "loss": 1.1478,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.09682285040616989,
+      "learning_rate": 1.5749999999999997e-05,
+      "loss": 1.0813,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.09817025810480118,
+      "learning_rate": 1.6499999999999998e-05,
+      "loss": 1.096,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10309538245201111,
+      "learning_rate": 1.725e-05,
+      "loss": 1.1418,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10393266379833221,
+      "learning_rate": 1.7999999999999997e-05,
+      "loss": 1.1351,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.10402549803256989,
+      "learning_rate": 1.875e-05,
+      "loss": 1.0959,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.10404133796691895,
+      "learning_rate": 1.95e-05,
+      "loss": 1.1869,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12272530794143677,
+      "learning_rate": 2.025e-05,
+      "loss": 1.1436,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1208878755569458,
+      "learning_rate": 2.1e-05,
+      "loss": 1.1444,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12307845801115036,
+      "learning_rate": 2.1749999999999997e-05,
+      "loss": 1.0762,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11315160989761353,
+      "learning_rate": 2.2499999999999998e-05,
+      "loss": 1.0269,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.10878825187683105,
+      "learning_rate": 2.325e-05,
+      "loss": 1.1535,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11939235776662827,
+      "learning_rate": 2.3999999999999997e-05,
+      "loss": 1.1705,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11615506559610367,
+      "learning_rate": 2.475e-05,
+      "loss": 1.1406,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11318700015544891,
+      "learning_rate": 2.55e-05,
+      "loss": 1.0874,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1265944391489029,
+      "learning_rate": 2.6249999999999998e-05,
+      "loss": 1.1317,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.153924360871315,
+      "learning_rate": 2.6999999999999996e-05,
+      "loss": 1.102,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1322435736656189,
+      "learning_rate": 2.7749999999999997e-05,
+      "loss": 1.0387,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1285255402326584,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 1.1082,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.13724276423454285,
+      "learning_rate": 2.925e-05,
+      "loss": 1.082,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.13925635814666748,
+      "learning_rate": 2.9999999999999997e-05,
+      "loss": 1.2066,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.16304455697536469,
+      "learning_rate": 3.0749999999999995e-05,
+      "loss": 1.1404,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1349136382341385,
+      "learning_rate": 3.149999999999999e-05,
+      "loss": 1.0513,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.150814950466156,
+      "learning_rate": 3.225e-05,
+      "loss": 1.0785,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13001343607902527,
+      "learning_rate": 3.2999999999999996e-05,
+      "loss": 1.121,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1322905719280243,
+      "learning_rate": 3.375e-05,
+      "loss": 1.1,
+      "step": 225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1382080763578415,
+      "learning_rate": 3.45e-05,
+      "loss": 1.1415,
+      "step": 230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.14844882488250732,
+      "learning_rate": 3.5249999999999996e-05,
+      "loss": 1.1337,
+      "step": 235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13436168432235718,
+      "learning_rate": 3.5999999999999994e-05,
+      "loss": 1.1736,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1420661062002182,
+      "learning_rate": 3.675e-05,
+      "loss": 1.0822,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.14159414172172546,
+      "learning_rate": 3.75e-05,
+      "loss": 1.0864,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.15307262539863586,
+      "learning_rate": 3.8249999999999995e-05,
+      "loss": 1.114,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1432732194662094,
+      "learning_rate": 3.9e-05,
+      "loss": 1.1041,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.14971235394477844,
+      "learning_rate": 3.975e-05,
+      "loss": 1.1015,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13618604838848114,
+      "learning_rate": 4.05e-05,
+      "loss": 1.1041,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.15167336165905,
+      "learning_rate": 4.125e-05,
+      "loss": 1.1905,
+      "step": 275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17375151813030243,
+      "learning_rate": 4.2e-05,
+      "loss": 1.0554,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.15129323303699493,
+      "learning_rate": 4.2749999999999996e-05,
+      "loss": 1.1595,
+      "step": 285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13002634048461914,
+      "learning_rate": 4.3499999999999993e-05,
+      "loss": 1.0993,
+      "step": 290
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15022237598896027,
+      "learning_rate": 4.424999999999999e-05,
+      "loss": 1.1154,
+      "step": 295
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15329715609550476,
+      "learning_rate": 4.4999999999999996e-05,
+      "loss": 1.1181,
+      "step": 300
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14781033992767334,
+      "learning_rate": 4.5749999999999994e-05,
+      "loss": 1.1708,
+      "step": 305
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13594377040863037,
+      "learning_rate": 4.65e-05,
+      "loss": 1.0517,
+      "step": 310
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14029279351234436,
+      "learning_rate": 4.7249999999999997e-05,
+      "loss": 1.0883,
+      "step": 315
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14294695854187012,
+      "learning_rate": 4.7999999999999994e-05,
+      "loss": 1.166,
+      "step": 320
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1437993049621582,
+      "learning_rate": 4.875e-05,
+      "loss": 1.0838,
+      "step": 325
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13464024662971497,
+      "learning_rate": 4.95e-05,
+      "loss": 1.076,
+      "step": 330
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15120777487754822,
+      "learning_rate": 5.025e-05,
+      "loss": 1.1465,
+      "step": 335
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13077160716056824,
+      "learning_rate": 5.1e-05,
+      "loss": 0.9944,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14656583964824677,
+      "learning_rate": 5.174999999999999e-05,
+      "loss": 1.142,
+      "step": 345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14403922855854034,
+      "learning_rate": 5.2499999999999995e-05,
+      "loss": 1.1545,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14558084309101105,
+      "learning_rate": 5.324999999999999e-05,
+      "loss": 1.1135,
+      "step": 355
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13940554857254028,
+      "learning_rate": 5.399999999999999e-05,
+      "loss": 1.1143,
+      "step": 360
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.12889456748962402,
+      "learning_rate": 5.4749999999999996e-05,
+      "loss": 1.0995,
+      "step": 365
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13410042226314545,
+      "learning_rate": 5.5499999999999994e-05,
+      "loss": 1.0377,
+      "step": 370
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1537715345621109,
+      "learning_rate": 5.625e-05,
+      "loss": 1.1654,
+      "step": 375
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14372758567333221,
+      "learning_rate": 5.6999999999999996e-05,
+      "loss": 1.0393,
+      "step": 380
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14559227228164673,
+      "learning_rate": 5.7749999999999994e-05,
+      "loss": 1.1712,
+      "step": 385
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14143848419189453,
+      "learning_rate": 5.85e-05,
+      "loss": 1.0294,
+      "step": 390
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13969220221042633,
+      "learning_rate": 5.925e-05,
+      "loss": 1.1678,
+      "step": 395
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14875860512256622,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 1.0829,
+      "step": 400
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13244011998176575,
+      "learning_rate": 6.075e-05,
+      "loss": 1.173,
+      "step": 405
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14143739640712738,
+      "learning_rate": 6.149999999999999e-05,
+      "loss": 1.0446,
+      "step": 410
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13888464868068695,
+      "learning_rate": 6.225e-05,
+      "loss": 1.1222,
+      "step": 415
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1454809457063675,
+      "learning_rate": 6.299999999999999e-05,
+      "loss": 1.0266,
+      "step": 420
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1294557899236679,
+      "learning_rate": 6.374999999999999e-05,
+      "loss": 1.0773,
+      "step": 425
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13563519716262817,
+      "learning_rate": 6.45e-05,
+      "loss": 1.0808,
+      "step": 430
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1332767903804779,
+      "learning_rate": 6.525e-05,
+      "loss": 1.1356,
+      "step": 435
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14194917678833008,
+      "learning_rate": 6.599999999999999e-05,
+      "loss": 1.0592,
+      "step": 440
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1359052062034607,
+      "learning_rate": 6.675e-05,
+      "loss": 1.1103,
+      "step": 445
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.12448085099458694,
+      "learning_rate": 6.75e-05,
+      "loss": 1.0912,
+      "step": 450
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14592844247817993,
+      "learning_rate": 6.824999999999999e-05,
+      "loss": 1.0549,
+      "step": 455
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14550206065177917,
+      "learning_rate": 6.9e-05,
+      "loss": 1.0997,
+      "step": 460
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13571631908416748,
+      "learning_rate": 6.975e-05,
+      "loss": 1.1559,
+      "step": 465
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12359726428985596,
+      "learning_rate": 7.049999999999999e-05,
+      "loss": 1.1352,
+      "step": 470
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1370527297258377,
+      "learning_rate": 7.125e-05,
+      "loss": 1.1114,
+      "step": 475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1392049938440323,
+      "learning_rate": 7.199999999999999e-05,
+      "loss": 1.1121,
+      "step": 480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13348931074142456,
+      "learning_rate": 7.274999999999999e-05,
+      "loss": 1.1471,
+      "step": 485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1287151575088501,
+      "learning_rate": 7.35e-05,
+      "loss": 1.1812,
+      "step": 490
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13077159225940704,
+      "learning_rate": 7.424999999999999e-05,
+      "loss": 1.0183,
+      "step": 495
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13283278048038483,
+      "learning_rate": 7.5e-05,
+      "loss": 1.1351,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14058330655097961,
+      "learning_rate": 7.575e-05,
+      "loss": 1.0175,
+      "step": 505
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1260114163160324,
+      "learning_rate": 7.649999999999999e-05,
+      "loss": 1.0918,
+      "step": 510
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12077819555997849,
+      "learning_rate": 7.725e-05,
+      "loss": 1.2125,
+      "step": 515
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13014502823352814,
+      "learning_rate": 7.8e-05,
+      "loss": 1.0959,
+      "step": 520
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13080598413944244,
+      "learning_rate": 7.874999999999999e-05,
+      "loss": 1.1083,
+      "step": 525
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12461178004741669,
+      "learning_rate": 7.95e-05,
+      "loss": 1.0968,
+      "step": 530
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13079579174518585,
+      "learning_rate": 8.025e-05,
+      "loss": 1.1039,
+      "step": 535
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12605786323547363,
+      "learning_rate": 8.1e-05,
+      "loss": 1.0558,
+      "step": 540
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.13258296251296997,
+      "learning_rate": 8.175e-05,
+      "loss": 1.0635,
+      "step": 545
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1320507675409317,
+      "learning_rate": 8.25e-05,
+      "loss": 1.042,
+      "step": 550
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11775451898574829,
+      "learning_rate": 8.325e-05,
+      "loss": 1.1299,
+      "step": 555
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12879742681980133,
+      "learning_rate": 8.4e-05,
+      "loss": 1.0587,
+      "step": 560
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1299828737974167,
+      "learning_rate": 8.474999999999999e-05,
+      "loss": 1.0773,
+      "step": 565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1389782577753067,
+      "learning_rate": 8.549999999999999e-05,
+      "loss": 1.066,
+      "step": 570
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12050167471170425,
+      "learning_rate": 8.624999999999998e-05,
+      "loss": 1.0786,
+      "step": 575
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14948152005672455,
+      "learning_rate": 8.699999999999999e-05,
+      "loss": 1.1208,
+      "step": 580
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.13367661833763123,
+      "learning_rate": 8.774999999999999e-05,
+      "loss": 1.0754,
+      "step": 585
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12811419367790222,
+      "learning_rate": 8.849999999999998e-05,
+      "loss": 1.1275,
+      "step": 590
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12895521521568298,
+      "learning_rate": 8.924999999999999e-05,
+      "loss": 1.112,
+      "step": 595
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1275772899389267,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 1.1171,
+      "step": 600
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12136266380548477,
+      "learning_rate": 9.074999999999998e-05,
+      "loss": 1.119,
+      "step": 605
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12577873468399048,
+      "learning_rate": 9.149999999999999e-05,
+      "loss": 1.059,
+      "step": 610
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.11611288040876389,
+      "learning_rate": 9.224999999999999e-05,
+      "loss": 1.093,
+      "step": 615
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1267736703157425,
+      "learning_rate": 9.3e-05,
+      "loss": 1.1074,
+      "step": 620
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12328934669494629,
+      "learning_rate": 9.374999999999999e-05,
+      "loss": 1.0373,
+      "step": 625
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.13137248158454895,
+      "learning_rate": 9.449999999999999e-05,
+      "loss": 1.0127,
+      "step": 630
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.13046279549598694,
+      "learning_rate": 9.525e-05,
+      "loss": 1.0549,
+      "step": 635
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11722920835018158,
+      "learning_rate": 9.599999999999999e-05,
+      "loss": 1.0051,
+      "step": 640
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.14764057099819183,
+      "learning_rate": 9.675e-05,
+      "loss": 1.1092,
+      "step": 645
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12383400648832321,
+      "learning_rate": 9.75e-05,
+      "loss": 1.1019,
+      "step": 650
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.13521508872509003,
+      "learning_rate": 9.824999999999999e-05,
+      "loss": 1.0963,
+      "step": 655
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12719859182834625,
+      "learning_rate": 9.9e-05,
+      "loss": 1.1604,
+      "step": 660
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.14219126105308533,
+      "learning_rate": 9.975e-05,
+      "loss": 1.0636,
+      "step": 665
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1256975382566452,
+      "learning_rate": 0.0001005,
+      "loss": 1.1046,
+      "step": 670
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12861751019954681,
+      "learning_rate": 0.00010125,
+      "loss": 1.0016,
+      "step": 675
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1278897374868393,
+      "learning_rate": 0.000102,
+      "loss": 1.108,
+      "step": 680
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11729888617992401,
+      "learning_rate": 0.00010275,
+      "loss": 1.1411,
+      "step": 685
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12147238105535507,
+      "learning_rate": 0.00010349999999999998,
+      "loss": 1.094,
+      "step": 690
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12336426228284836,
+      "learning_rate": 0.00010424999999999999,
+      "loss": 1.0397,
+      "step": 695
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12023478001356125,
+      "learning_rate": 0.00010499999999999999,
+      "loss": 1.0665,
+      "step": 700
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.11755697429180145,
+      "learning_rate": 0.00010574999999999998,
+      "loss": 1.061,
+      "step": 705
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11740587651729584,
+      "learning_rate": 0.00010649999999999999,
+      "loss": 1.0637,
+      "step": 710
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.13212309777736664,
+      "learning_rate": 0.00010724999999999999,
+      "loss": 1.0128,
+      "step": 715
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.13335004448890686,
+      "learning_rate": 0.00010799999999999998,
+      "loss": 1.064,
+      "step": 720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12249240279197693,
+      "learning_rate": 0.00010874999999999999,
+      "loss": 1.0939,
+      "step": 725
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.13641893863677979,
+      "learning_rate": 0.00010949999999999999,
+      "loss": 1.1127,
+      "step": 730
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.127156600356102,
+      "learning_rate": 0.00011024999999999998,
+      "loss": 1.0527,
+      "step": 735
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1096034049987793,
+      "learning_rate": 0.00011099999999999999,
+      "loss": 1.1304,
+      "step": 740
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12487472593784332,
+      "learning_rate": 0.00011174999999999999,
+      "loss": 1.0862,
+      "step": 745
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12395402044057846,
+      "learning_rate": 0.0001125,
+      "loss": 1.0641,
+      "step": 750
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12146257609128952,
+      "learning_rate": 0.00011324999999999999,
+      "loss": 1.1303,
+      "step": 755
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12403228133916855,
+      "learning_rate": 0.00011399999999999999,
+      "loss": 1.0214,
+      "step": 760
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11013570427894592,
+      "learning_rate": 0.00011475,
+      "loss": 1.0696,
+      "step": 765
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11409767717123032,
+      "learning_rate": 0.00011549999999999999,
+      "loss": 1.2188,
+      "step": 770
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.12632502615451813,
+      "learning_rate": 0.00011624999999999999,
+      "loss": 1.113,
+      "step": 775
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.11955476552248001,
+      "learning_rate": 0.000117,
+      "loss": 1.1429,
+      "step": 780
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.13300061225891113,
+      "learning_rate": 0.00011774999999999999,
+      "loss": 1.1171,
+      "step": 785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12019680440425873,
+      "learning_rate": 0.0001185,
+      "loss": 1.0675,
+      "step": 790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12077269703149796,
+      "learning_rate": 0.00011925,
+      "loss": 1.0694,
+      "step": 795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12203273177146912,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 1.0378,
+      "step": 800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1268533170223236,
+      "learning_rate": 0.00012075,
+      "loss": 1.1294,
+      "step": 805
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1281730681657791,
+      "learning_rate": 0.0001215,
+      "loss": 1.024,
+      "step": 810
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11289986968040466,
+      "learning_rate": 0.00012225,
+      "loss": 1.0776,
+      "step": 815
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11376605927944183,
+      "learning_rate": 0.00012299999999999998,
+      "loss": 1.107,
+      "step": 820
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1271873265504837,
+      "learning_rate": 0.00012374999999999997,
+      "loss": 1.1817,
+      "step": 825
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10974998027086258,
+      "learning_rate": 0.0001245,
+      "loss": 1.0228,
+      "step": 830
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.13189394772052765,
+      "learning_rate": 0.00012524999999999998,
+      "loss": 1.1298,
+      "step": 835
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10656459629535675,
+      "learning_rate": 0.00012599999999999997,
+      "loss": 1.0833,
+      "step": 840
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.14083336293697357,
+      "learning_rate": 0.00012675,
+      "loss": 1.0027,
+      "step": 845
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1123281791806221,
+      "learning_rate": 0.00012749999999999998,
+      "loss": 1.0953,
+      "step": 850
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12783868610858917,
+      "learning_rate": 0.00012824999999999997,
+      "loss": 1.0793,
+      "step": 855
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12001547962427139,
+      "learning_rate": 0.000129,
+      "loss": 1.0685,
+      "step": 860
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.11165610700845718,
+      "learning_rate": 0.00012974999999999998,
+      "loss": 1.111,
+      "step": 865
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.12007506936788559,
+      "learning_rate": 0.0001305,
+      "loss": 1.0681,
+      "step": 870
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.12255184352397919,
+      "learning_rate": 0.00013125,
+      "loss": 1.1185,
+      "step": 875
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.12230581790208817,
+      "learning_rate": 0.00013199999999999998,
+      "loss": 1.0196,
+      "step": 880
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11624744534492493,
+      "learning_rate": 0.00013275,
+      "loss": 1.0478,
+      "step": 885
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11575216054916382,
+      "learning_rate": 0.0001335,
+      "loss": 1.0387,
+      "step": 890
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1169903427362442,
+      "learning_rate": 0.00013424999999999998,
+      "loss": 1.1127,
+      "step": 895
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11636984348297119,
+      "learning_rate": 0.000135,
+      "loss": 1.0555,
+      "step": 900
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.12122248113155365,
+      "learning_rate": 0.00013575,
+      "loss": 1.1497,
+      "step": 905
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1243358924984932,
+      "learning_rate": 0.00013649999999999998,
+      "loss": 1.0675,
+      "step": 910
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11892006546258926,
+      "learning_rate": 0.00013725,
+      "loss": 1.1499,
+      "step": 915
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.12357810884714127,
+      "learning_rate": 0.000138,
+      "loss": 1.0508,
+      "step": 920
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11074083298444748,
+      "learning_rate": 0.00013874999999999998,
+      "loss": 0.9855,
+      "step": 925
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1215604916214943,
+      "learning_rate": 0.0001395,
+      "loss": 1.0723,
+      "step": 930
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.11568805575370789,
+      "learning_rate": 0.00014025,
+      "loss": 1.141,
+      "step": 935
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1251344233751297,
+      "learning_rate": 0.00014099999999999998,
+      "loss": 1.0817,
+      "step": 940
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.12408124655485153,
+      "learning_rate": 0.00014174999999999998,
+      "loss": 0.9859,
+      "step": 945
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.13382701575756073,
+      "learning_rate": 0.0001425,
+      "loss": 1.0902,
+      "step": 950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.12058982998132706,
+      "learning_rate": 0.00014324999999999999,
+      "loss": 1.0817,
+      "step": 955
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.12119226902723312,
+      "learning_rate": 0.00014399999999999998,
+      "loss": 1.0502,
+      "step": 960
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.14532142877578735,
+      "learning_rate": 0.00014475,
+      "loss": 1.0244,
+      "step": 965
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.12690915167331696,
+      "learning_rate": 0.00014549999999999999,
+      "loss": 1.1313,
+      "step": 970
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11626715213060379,
+      "learning_rate": 0.00014624999999999998,
+      "loss": 1.088,
+      "step": 975
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11747872829437256,
+      "learning_rate": 0.000147,
+      "loss": 1.0855,
+      "step": 980
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.12771141529083252,
+      "learning_rate": 0.00014774999999999999,
+      "loss": 1.1304,
+      "step": 985
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1297111064195633,
+      "learning_rate": 0.00014849999999999998,
+      "loss": 1.0307,
+      "step": 990
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.12018551677465439,
+      "learning_rate": 0.00014925,
+      "loss": 1.0946,
+      "step": 995
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11594726890325546,
+      "learning_rate": 0.00015,
+      "loss": 1.0314,
+      "step": 1000
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.133875772356987,
+      "learning_rate": 0.00015074999999999998,
+      "loss": 1.0018,
+      "step": 1005
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1387271285057068,
+      "learning_rate": 0.0001515,
+      "loss": 1.0417,
+      "step": 1010
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11644894629716873,
+      "learning_rate": 0.00015224999999999996,
+      "loss": 1.045,
+      "step": 1015
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.13126513361930847,
+      "learning_rate": 0.00015299999999999998,
+      "loss": 1.0666,
+      "step": 1020
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.12716297805309296,
+      "learning_rate": 0.00015374999999999997,
+      "loss": 1.0578,
+      "step": 1025
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.11895552277565002,
+      "learning_rate": 0.0001545,
+      "loss": 1.1052,
+      "step": 1030
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.12736555933952332,
+      "learning_rate": 0.00015524999999999998,
+      "loss": 0.9881,
+      "step": 1035
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11497914791107178,
+      "learning_rate": 0.000156,
+      "loss": 1.0462,
+      "step": 1040
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.13537730276584625,
+      "learning_rate": 0.00015675,
+      "loss": 0.9891,
+      "step": 1045
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.12179003655910492,
+      "learning_rate": 0.00015749999999999998,
+      "loss": 1.0565,
+      "step": 1050
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.12696489691734314,
+      "learning_rate": 0.00015824999999999997,
+      "loss": 1.0551,
+      "step": 1055
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.11586569994688034,
+      "learning_rate": 0.000159,
+      "loss": 1.1354,
+      "step": 1060
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1284131109714508,
+      "learning_rate": 0.00015974999999999998,
+      "loss": 1.1317,
+      "step": 1065
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.13363447785377502,
+      "learning_rate": 0.0001605,
+      "loss": 1.0358,
+      "step": 1070
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.13264328241348267,
+      "learning_rate": 0.00016125,
+      "loss": 1.0852,
+      "step": 1075
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.12372339516878128,
+      "learning_rate": 0.000162,
+      "loss": 0.997,
+      "step": 1080
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.12511901557445526,
+      "learning_rate": 0.00016274999999999997,
+      "loss": 1.0837,
+      "step": 1085
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.12333539128303528,
+      "learning_rate": 0.0001635,
+      "loss": 1.0565,
+      "step": 1090
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1194729283452034,
+      "learning_rate": 0.00016424999999999998,
+      "loss": 1.0099,
+      "step": 1095
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.12635278701782227,
+      "learning_rate": 0.000165,
+      "loss": 1.039,
+      "step": 1100
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.124270498752594,
+      "learning_rate": 0.00016575,
+      "loss": 1.165,
+      "step": 1105
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.12829731404781342,
+      "learning_rate": 0.0001665,
+      "loss": 1.0653,
+      "step": 1110
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.13601495325565338,
+      "learning_rate": 0.00016724999999999997,
+      "loss": 1.0509,
+      "step": 1115
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1377023309469223,
+      "learning_rate": 0.000168,
+      "loss": 1.0376,
+      "step": 1120
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.12421699613332748,
+      "learning_rate": 0.00016874999999999998,
+      "loss": 1.0082,
+      "step": 1125
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.12038341909646988,
+      "learning_rate": 0.00016949999999999997,
+      "loss": 1.0865,
+      "step": 1130
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.11900181323289871,
+      "learning_rate": 0.00017025,
+      "loss": 1.1051,
+      "step": 1135
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.12548121809959412,
+      "learning_rate": 0.00017099999999999998,
+      "loss": 1.1436,
+      "step": 1140
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.12389590591192245,
+      "learning_rate": 0.00017175,
+      "loss": 0.9946,
+      "step": 1145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.13046157360076904,
+      "learning_rate": 0.00017249999999999996,
+      "loss": 1.0448,
+      "step": 1150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.13118329644203186,
+      "learning_rate": 0.00017324999999999998,
+      "loss": 1.0698,
+      "step": 1155
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.11619829386472702,
+      "learning_rate": 0.00017399999999999997,
+      "loss": 1.0619,
+      "step": 1160
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.12934699654579163,
+      "learning_rate": 0.00017475,
+      "loss": 1.0755,
+      "step": 1165
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.13581542670726776,
+      "learning_rate": 0.00017549999999999998,
+      "loss": 1.0456,
+      "step": 1170
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1390586644411087,
+      "learning_rate": 0.00017625,
+      "loss": 1.0822,
+      "step": 1175
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.13362278044223785,
+      "learning_rate": 0.00017699999999999997,
+      "loss": 1.0588,
+      "step": 1180
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1188979223370552,
+      "learning_rate": 0.00017774999999999998,
+      "loss": 1.0941,
+      "step": 1185
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.11467181891202927,
+      "learning_rate": 0.00017849999999999997,
+      "loss": 1.0485,
+      "step": 1190
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.13196997344493866,
+      "learning_rate": 0.00017925,
+      "loss": 1.1128,
+      "step": 1195
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.13005390763282776,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 1.0536,
+      "step": 1200
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.13372902572155,
+      "learning_rate": 0.00018075,
+      "loss": 1.1123,
+      "step": 1205
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.13549335300922394,
+      "learning_rate": 0.00018149999999999997,
+      "loss": 1.0998,
+      "step": 1210
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.13829483091831207,
+      "learning_rate": 0.00018224999999999998,
+      "loss": 1.0174,
+      "step": 1215
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1173296719789505,
+      "learning_rate": 0.00018299999999999998,
+      "loss": 1.0898,
+      "step": 1220
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.12226750701665878,
+      "learning_rate": 0.00018375,
+      "loss": 1.0516,
+      "step": 1225
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.12172769755125046,
+      "learning_rate": 0.00018449999999999999,
+      "loss": 1.0495,
+      "step": 1230
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.12832506000995636,
+      "learning_rate": 0.00018525,
+      "loss": 1.0816,
+      "step": 1235
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.12533822655677795,
+      "learning_rate": 0.000186,
+      "loss": 1.0122,
+      "step": 1240
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.13508310914039612,
+      "learning_rate": 0.00018675,
+      "loss": 1.0122,
+      "step": 1245
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.11473240703344345,
+      "learning_rate": 0.00018749999999999998,
+      "loss": 1.0813,
+      "step": 1250
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.12293410301208496,
+      "learning_rate": 0.00018824999999999997,
+      "loss": 1.03,
+      "step": 1255
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.14231853187084198,
+      "learning_rate": 0.00018899999999999999,
+      "loss": 1.0724,
+      "step": 1260
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1204795315861702,
+      "learning_rate": 0.00018974999999999998,
+      "loss": 1.0449,
+      "step": 1265
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.12521016597747803,
+      "learning_rate": 0.0001905,
+      "loss": 1.1008,
+      "step": 1270
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.12301227450370789,
+      "learning_rate": 0.00019124999999999996,
+      "loss": 1.1471,
+      "step": 1275
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1287146359682083,
+      "learning_rate": 0.00019199999999999998,
+      "loss": 1.0501,
+      "step": 1280
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.13978512585163116,
+      "learning_rate": 0.00019274999999999997,
+      "loss": 1.03,
+      "step": 1285
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.12224988639354706,
+      "learning_rate": 0.0001935,
+      "loss": 1.0996,
+      "step": 1290
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1454314887523651,
+      "learning_rate": 0.00019424999999999998,
+      "loss": 1.0529,
+      "step": 1295
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.12719061970710754,
+      "learning_rate": 0.000195,
+      "loss": 1.0984,
+      "step": 1300
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.13518300652503967,
+      "learning_rate": 0.00019574999999999996,
+      "loss": 1.0336,
+      "step": 1305
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1353674679994583,
+      "learning_rate": 0.00019649999999999998,
+      "loss": 1.1046,
+      "step": 1310
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.13263052701950073,
+      "learning_rate": 0.00019724999999999997,
+      "loss": 0.9654,
+      "step": 1315
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.13989952206611633,
+      "learning_rate": 0.000198,
+      "loss": 1.0994,
+      "step": 1320
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.13030202686786652,
+      "learning_rate": 0.00019874999999999998,
+      "loss": 1.1445,
+      "step": 1325
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.12201202660799026,
+      "learning_rate": 0.0001995,
+      "loss": 1.0605,
+      "step": 1330
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1385253220796585,
+      "learning_rate": 0.00020025,
+      "loss": 1.0405,
+      "step": 1335
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.14180108904838562,
+      "learning_rate": 0.000201,
+      "loss": 1.0612,
+      "step": 1340
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1332985907793045,
+      "learning_rate": 0.00020174999999999997,
+      "loss": 1.0653,
+      "step": 1345
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.14803795516490936,
+      "learning_rate": 0.0002025,
+      "loss": 1.1237,
+      "step": 1350
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.13238713145256042,
+      "learning_rate": 0.00020324999999999998,
+      "loss": 1.1257,
+      "step": 1355
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.12412701547145844,
+      "learning_rate": 0.000204,
+      "loss": 1.0947,
+      "step": 1360
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.13152962923049927,
+      "learning_rate": 0.00020475,
+      "loss": 1.1676,
+      "step": 1365
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13383346796035767,
+      "learning_rate": 0.0002055,
+      "loss": 1.0314,
+      "step": 1370
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.12599609792232513,
+      "learning_rate": 0.00020624999999999997,
+      "loss": 0.9767,
+      "step": 1375
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.12485210597515106,
+      "learning_rate": 0.00020699999999999996,
+      "loss": 1.1578,
+      "step": 1380
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15362361073493958,
+      "learning_rate": 0.00020774999999999998,
+      "loss": 1.0263,
+      "step": 1385
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13580173254013062,
+      "learning_rate": 0.00020849999999999997,
+      "loss": 1.0553,
+      "step": 1390
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13229776918888092,
+      "learning_rate": 0.00020925,
+      "loss": 1.0242,
+      "step": 1395
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14048846065998077,
+      "learning_rate": 0.00020999999999999998,
+      "loss": 1.0541,
+      "step": 1400
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1278860718011856,
+      "learning_rate": 0.00021074999999999997,
+      "loss": 1.0295,
+      "step": 1405
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.12558186054229736,
+      "learning_rate": 0.00021149999999999996,
+      "loss": 0.9833,
+      "step": 1410
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13044369220733643,
+      "learning_rate": 0.00021224999999999998,
+      "loss": 1.0244,
+      "step": 1415
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.12153918296098709,
+      "learning_rate": 0.00021299999999999997,
+      "loss": 1.0513,
+      "step": 1420
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1474679559469223,
+      "learning_rate": 0.00021375,
+      "loss": 1.0771,
+      "step": 1425
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13985954225063324,
+      "learning_rate": 0.00021449999999999998,
+      "loss": 1.0419,
+      "step": 1430
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13685761392116547,
+      "learning_rate": 0.00021525,
+      "loss": 1.045,
+      "step": 1435
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.13151884078979492,
+      "learning_rate": 0.00021599999999999996,
+      "loss": 1.0856,
+      "step": 1440
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.12494616955518723,
+      "learning_rate": 0.00021674999999999998,
+      "loss": 1.093,
+      "step": 1445
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1363985389471054,
+      "learning_rate": 0.00021749999999999997,
+      "loss": 1.0921,
+      "step": 1450
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13812515139579773,
+      "learning_rate": 0.00021825,
+      "loss": 1.0832,
+      "step": 1455
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1363314837217331,
+      "learning_rate": 0.00021899999999999998,
+      "loss": 1.0995,
+      "step": 1460
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.12492544949054718,
+      "learning_rate": 0.00021975,
+      "loss": 1.0806,
+      "step": 1465
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.12725725769996643,
+      "learning_rate": 0.00022049999999999997,
+      "loss": 1.0276,
+      "step": 1470
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.11777817457914352,
+      "learning_rate": 0.00022124999999999998,
+      "loss": 0.9419,
+      "step": 1475
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13982278108596802,
+      "learning_rate": 0.00022199999999999998,
+      "loss": 1.0673,
+      "step": 1480
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15291635692119598,
+      "learning_rate": 0.00022275,
+      "loss": 1.0753,
+      "step": 1485
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1316549926996231,
+      "learning_rate": 0.00022349999999999998,
+      "loss": 1.0793,
+      "step": 1490
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13658076524734497,
+      "learning_rate": 0.00022425,
+      "loss": 1.0742,
+      "step": 1495
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.12889617681503296,
+      "learning_rate": 0.000225,
+      "loss": 1.0272,
+      "step": 1500
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1431025117635727,
+      "learning_rate": 0.00022574999999999996,
+      "loss": 1.056,
+      "step": 1505
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.12882299721240997,
+      "learning_rate": 0.00022649999999999998,
+      "loss": 1.0965,
+      "step": 1510
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1326344758272171,
+      "learning_rate": 0.00022724999999999997,
+      "loss": 1.1601,
+      "step": 1515
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1312105506658554,
+      "learning_rate": 0.00022799999999999999,
+      "loss": 1.0286,
+      "step": 1520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.13419494032859802,
+      "learning_rate": 0.00022874999999999998,
+      "loss": 1.0998,
+      "step": 1525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.12598495185375214,
+      "learning_rate": 0.0002295,
+      "loss": 1.0099,
+      "step": 1530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.12234248965978622,
+      "learning_rate": 0.00023024999999999996,
+      "loss": 1.0512,
+      "step": 1535
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1486140936613083,
+      "learning_rate": 0.00023099999999999998,
+      "loss": 1.0563,
+      "step": 1540
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13529768586158752,
+      "learning_rate": 0.00023174999999999997,
+      "loss": 0.9952,
+      "step": 1545
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1367512345314026,
+      "learning_rate": 0.00023249999999999999,
+      "loss": 0.9924,
+      "step": 1550
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14160625636577606,
+      "learning_rate": 0.00023324999999999998,
+      "loss": 0.9568,
+      "step": 1555
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13413269817829132,
+      "learning_rate": 0.000234,
+      "loss": 1.0429,
+      "step": 1560
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13511444628238678,
+      "learning_rate": 0.00023474999999999996,
+      "loss": 1.0548,
+      "step": 1565
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13897758722305298,
+      "learning_rate": 0.00023549999999999998,
+      "loss": 1.0312,
+      "step": 1570
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13271476328372955,
+      "learning_rate": 0.00023624999999999997,
+      "loss": 1.1496,
+      "step": 1575
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.12695930898189545,
+      "learning_rate": 0.000237,
+      "loss": 1.0083,
+      "step": 1580
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1538165807723999,
+      "learning_rate": 0.00023774999999999998,
+      "loss": 1.1104,
+      "step": 1585
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13440971076488495,
+      "learning_rate": 0.0002385,
+      "loss": 1.0746,
+      "step": 1590
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15520578622817993,
+      "learning_rate": 0.00023925,
+      "loss": 1.0702,
+      "step": 1595
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1417379379272461,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 1.0313,
+      "step": 1600
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1348305493593216,
+      "learning_rate": 0.00024074999999999997,
+      "loss": 1.0489,
+      "step": 1605
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.13864970207214355,
+      "learning_rate": 0.0002415,
+      "loss": 1.0563,
+      "step": 1610
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.12686601281166077,
+      "learning_rate": 0.00024224999999999998,
+      "loss": 1.1186,
+      "step": 1615
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1524277776479721,
+      "learning_rate": 0.000243,
+      "loss": 1.0285,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13417251408100128,
+      "learning_rate": 0.00024375,
+      "loss": 1.0816,
+      "step": 1625
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13841432332992554,
+      "learning_rate": 0.0002445,
+      "loss": 0.9558,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14064748585224152,
+      "learning_rate": 0.00024524999999999997,
+      "loss": 1.0487,
+      "step": 1635
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15057429671287537,
+      "learning_rate": 0.00024599999999999996,
+      "loss": 1.0382,
+      "step": 1640
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.140291228890419,
+      "learning_rate": 0.00024675,
+      "loss": 1.0297,
+      "step": 1645
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1320800930261612,
+      "learning_rate": 0.00024749999999999994,
+      "loss": 1.1047,
+      "step": 1650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.13335539400577545,
+      "learning_rate": 0.00024825,
+      "loss": 1.0456,
+      "step": 1655
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14252524077892303,
+      "learning_rate": 0.000249,
+      "loss": 1.1067,
+      "step": 1660
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14384162425994873,
+      "learning_rate": 0.00024974999999999997,
+      "loss": 1.0627,
+      "step": 1665
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1295383721590042,
+      "learning_rate": 0.00025049999999999996,
+      "loss": 0.9831,
+      "step": 1670
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14580777287483215,
+      "learning_rate": 0.00025125,
+      "loss": 1.1085,
+      "step": 1675
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1554543673992157,
+      "learning_rate": 0.00025199999999999995,
+      "loss": 1.0539,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1461145579814911,
+      "learning_rate": 0.00025275,
+      "loss": 1.0761,
+      "step": 1685
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.12733665108680725,
+      "learning_rate": 0.0002535,
+      "loss": 1.0143,
+      "step": 1690
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16144566237926483,
+      "learning_rate": 0.00025425,
+      "loss": 0.9641,
+      "step": 1695
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.14675092697143555,
+      "learning_rate": 0.00025499999999999996,
+      "loss": 1.029,
+      "step": 1700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15714582800865173,
+      "learning_rate": 0.00025575,
+      "loss": 1.0251,
+      "step": 1705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1321059912443161,
+      "learning_rate": 0.00025649999999999995,
+      "loss": 1.0629,
+      "step": 1710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13915494084358215,
+      "learning_rate": 0.00025725,
+      "loss": 1.0581,
+      "step": 1715
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1294393241405487,
+      "learning_rate": 0.000258,
+      "loss": 0.939,
+      "step": 1720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14472445845603943,
+      "learning_rate": 0.00025875,
+      "loss": 1.0293,
+      "step": 1725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1563960462808609,
+      "learning_rate": 0.00025949999999999997,
+      "loss": 1.0984,
+      "step": 1730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13787269592285156,
+      "learning_rate": 0.00026025,
+      "loss": 1.0645,
+      "step": 1735
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.142653688788414,
+      "learning_rate": 0.000261,
+      "loss": 1.0271,
+      "step": 1740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16362793743610382,
+      "learning_rate": 0.00026175,
+      "loss": 0.9973,
+      "step": 1745
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13521093130111694,
+      "learning_rate": 0.0002625,
+      "loss": 1.0968,
+      "step": 1750
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1348579227924347,
+      "learning_rate": 0.00026325,
+      "loss": 1.084,
+      "step": 1755
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15295952558517456,
+      "learning_rate": 0.00026399999999999997,
+      "loss": 1.0843,
+      "step": 1760
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.12325224280357361,
+      "learning_rate": 0.00026474999999999996,
+      "loss": 1.0448,
+      "step": 1765
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13043299317359924,
+      "learning_rate": 0.0002655,
+      "loss": 1.0078,
+      "step": 1770
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13307566940784454,
+      "learning_rate": 0.00026624999999999994,
+      "loss": 1.0496,
+      "step": 1775
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1296854168176651,
+      "learning_rate": 0.000267,
+      "loss": 1.0945,
+      "step": 1780
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14310450851917267,
+      "learning_rate": 0.00026775,
+      "loss": 0.9254,
+      "step": 1785
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14738260209560394,
+      "learning_rate": 0.00026849999999999997,
+      "loss": 1.0589,
+      "step": 1790
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1367303431034088,
+      "learning_rate": 0.00026924999999999996,
+      "loss": 1.1256,
+      "step": 1795
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1518729031085968,
+      "learning_rate": 0.00027,
+      "loss": 1.1417,
+      "step": 1800
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14660024642944336,
+      "learning_rate": 0.00027074999999999994,
+      "loss": 1.0219,
+      "step": 1805
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14216068387031555,
+      "learning_rate": 0.0002715,
+      "loss": 1.0597,
+      "step": 1810
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1290411353111267,
+      "learning_rate": 0.00027225,
+      "loss": 1.0887,
+      "step": 1815
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14220093190670013,
+      "learning_rate": 0.00027299999999999997,
+      "loss": 1.0228,
+      "step": 1820
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.15115365386009216,
+      "learning_rate": 0.00027374999999999996,
+      "loss": 1.0877,
+      "step": 1825
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.12890294194221497,
+      "learning_rate": 0.0002745,
+      "loss": 1.1167,
+      "step": 1830
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14485861361026764,
+      "learning_rate": 0.00027525,
+      "loss": 1.0937,
+      "step": 1835
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1425914615392685,
+      "learning_rate": 0.000276,
+      "loss": 1.0565,
+      "step": 1840
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16691897809505463,
+      "learning_rate": 0.00027675,
+      "loss": 1.0881,
+      "step": 1845
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1356891691684723,
+      "learning_rate": 0.00027749999999999997,
+      "loss": 0.9696,
+      "step": 1850
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14420901238918304,
+      "learning_rate": 0.00027824999999999996,
+      "loss": 0.9847,
+      "step": 1855
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1378943920135498,
+      "learning_rate": 0.000279,
+      "loss": 0.9863,
+      "step": 1860
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.13317394256591797,
+      "learning_rate": 0.00027975,
+      "loss": 1.0724,
+      "step": 1865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14231087267398834,
+      "learning_rate": 0.0002805,
+      "loss": 1.0886,
+      "step": 1870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13940946757793427,
+      "learning_rate": 0.00028125,
+      "loss": 1.047,
+      "step": 1875
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13190115988254547,
+      "learning_rate": 0.00028199999999999997,
+      "loss": 1.0192,
+      "step": 1880
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16745632886886597,
+      "learning_rate": 0.00028274999999999996,
+      "loss": 0.9855,
+      "step": 1885
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1373772770166397,
+      "learning_rate": 0.00028349999999999995,
+      "loss": 1.0242,
+      "step": 1890
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13555435836315155,
+      "learning_rate": 0.00028425,
+      "loss": 1.075,
+      "step": 1895
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.145888552069664,
+      "learning_rate": 0.000285,
+      "loss": 1.0011,
+      "step": 1900
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14030475914478302,
+      "learning_rate": 0.00028575,
+      "loss": 1.0065,
+      "step": 1905
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.13973671197891235,
+      "learning_rate": 0.00028649999999999997,
+      "loss": 0.9631,
+      "step": 1910
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15829509496688843,
+      "learning_rate": 0.00028724999999999996,
+      "loss": 1.0565,
+      "step": 1915
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14472809433937073,
+      "learning_rate": 0.00028799999999999995,
+      "loss": 0.9725,
+      "step": 1920
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16352474689483643,
+      "learning_rate": 0.00028875,
+      "loss": 0.9782,
+      "step": 1925
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14885759353637695,
+      "learning_rate": 0.0002895,
+      "loss": 1.0521,
+      "step": 1930
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14806737005710602,
+      "learning_rate": 0.00029025,
+      "loss": 1.0596,
+      "step": 1935
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15192674100399017,
+      "learning_rate": 0.00029099999999999997,
+      "loss": 1.0307,
+      "step": 1940
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15374353528022766,
+      "learning_rate": 0.00029174999999999996,
+      "loss": 1.0676,
+      "step": 1945
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1534864455461502,
+      "learning_rate": 0.00029249999999999995,
+      "loss": 0.9044,
+      "step": 1950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14087004959583282,
+      "learning_rate": 0.00029325,
+      "loss": 1.0849,
+      "step": 1955
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13771598041057587,
+      "learning_rate": 0.000294,
+      "loss": 1.0698,
+      "step": 1960
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17788104712963104,
+      "learning_rate": 0.00029475,
+      "loss": 1.0747,
+      "step": 1965
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13944664597511292,
+      "learning_rate": 0.00029549999999999997,
+      "loss": 1.0289,
+      "step": 1970
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14483226835727692,
+      "learning_rate": 0.00029624999999999996,
+      "loss": 1.0552,
+      "step": 1975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16384923458099365,
+      "learning_rate": 0.00029699999999999996,
+      "loss": 1.103,
+      "step": 1980
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17012694478034973,
+      "learning_rate": 0.00029775,
+      "loss": 1.0814,
+      "step": 1985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1577819287776947,
+      "learning_rate": 0.0002985,
+      "loss": 0.9313,
+      "step": 1990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13799041509628296,
+      "learning_rate": 0.00029925,
+      "loss": 0.9898,
+      "step": 1995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15773248672485352,
+      "learning_rate": 0.0003,
+      "loss": 1.0522,
+      "step": 2000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1652466058731079,
+      "learning_rate": 0.0002999999647025093,
+      "loss": 1.1033,
+      "step": 2005
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14737606048583984,
+      "learning_rate": 0.0002999998588100539,
+      "loss": 1.024,
+      "step": 2010
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15708868205547333,
+      "learning_rate": 0.0002999996823226836,
+      "loss": 1.055,
+      "step": 2015
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13661877810955048,
+      "learning_rate": 0.00029999943524048147,
+      "loss": 1.0728,
+      "step": 2020
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.13963155448436737,
+      "learning_rate": 0.00029999911756356377,
+      "loss": 1.0962,
+      "step": 2025
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.14523880183696747,
+      "learning_rate": 0.00029999872929208,
+      "loss": 0.9992,
+      "step": 2030
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1487034410238266,
+      "learning_rate": 0.000299998270426213,
+      "loss": 0.9902,
+      "step": 2035
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14978134632110596,
+      "learning_rate": 0.0002999977409661786,
+      "loss": 1.0294,
+      "step": 2040
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15391825139522552,
+      "learning_rate": 0.00029999714091222604,
+      "loss": 0.9604,
+      "step": 2045
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15368473529815674,
+      "learning_rate": 0.0002999964702646377,
+      "loss": 0.9744,
+      "step": 2050
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15019111335277557,
+      "learning_rate": 0.00029999572902372925,
+      "loss": 0.9179,
+      "step": 2055
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.151919886469841,
+      "learning_rate": 0.00029999491718984945,
+      "loss": 1.0588,
+      "step": 2060
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14386044442653656,
+      "learning_rate": 0.00029999403476338053,
+      "loss": 0.9972,
+      "step": 2065
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1599544882774353,
+      "learning_rate": 0.0002999930817447377,
+      "loss": 1.01,
+      "step": 2070
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14225023984909058,
+      "learning_rate": 0.00029999205813436945,
+      "loss": 1.1015,
+      "step": 2075
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1291441023349762,
+      "learning_rate": 0.00029999096393275754,
+      "loss": 1.0002,
+      "step": 2080
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.140071839094162,
+      "learning_rate": 0.000299989799140417,
+      "loss": 1.1162,
+      "step": 2085
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1789652705192566,
+      "learning_rate": 0.00029998856375789594,
+      "loss": 0.9408,
+      "step": 2090
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13465484976768494,
+      "learning_rate": 0.00029998725778577584,
+      "loss": 1.0336,
+      "step": 2095
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.13943049311637878,
+      "learning_rate": 0.0002999858812246713,
+      "loss": 1.0623,
+      "step": 2100
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15316350758075714,
+      "learning_rate": 0.0002999844340752302,
+      "loss": 1.0793,
+      "step": 2105
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15929552912712097,
+      "learning_rate": 0.00029998291633813353,
+      "loss": 1.0003,
+      "step": 2110
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.14730267226696014,
+      "learning_rate": 0.00029998132801409565,
+      "loss": 1.0309,
+      "step": 2115
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1534448117017746,
+      "learning_rate": 0.00029997966910386413,
+      "loss": 1.0758,
+      "step": 2120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14620815217494965,
+      "learning_rate": 0.00029997793960821967,
+      "loss": 1.1534,
+      "step": 2125
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16130419075489044,
+      "learning_rate": 0.00029997613952797617,
+      "loss": 0.9605,
+      "step": 2130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13983474671840668,
+      "learning_rate": 0.00029997426886398094,
+      "loss": 1.0272,
+      "step": 2135
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13644209504127502,
+      "learning_rate": 0.00029997232761711423,
+      "loss": 1.0924,
+      "step": 2140
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15886028110980988,
+      "learning_rate": 0.0002999703157882897,
+      "loss": 1.1073,
+      "step": 2145
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.13936792314052582,
+      "learning_rate": 0.0002999682333784542,
+      "loss": 1.0108,
+      "step": 2150
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1483025848865509,
+      "learning_rate": 0.0002999660803885878,
+      "loss": 1.0614,
+      "step": 2155
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16978415846824646,
+      "learning_rate": 0.00029996385681970377,
+      "loss": 1.0051,
+      "step": 2160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1511315107345581,
+      "learning_rate": 0.0002999615626728486,
+      "loss": 0.9862,
+      "step": 2165
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16898401081562042,
+      "learning_rate": 0.00029995919794910186,
+      "loss": 1.0046,
+      "step": 2170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1466691642999649,
+      "learning_rate": 0.00029995676264957667,
+      "loss": 1.0005,
+      "step": 2175
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16119731962680817,
+      "learning_rate": 0.000299954256775419,
+      "loss": 0.9917,
+      "step": 2180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17459401488304138,
+      "learning_rate": 0.00029995168032780826,
+      "loss": 1.0336,
+      "step": 2185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14910408854484558,
+      "learning_rate": 0.0002999490333079571,
+      "loss": 0.9864,
+      "step": 2190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16221760213375092,
+      "learning_rate": 0.00029994631571711114,
+      "loss": 1.0274,
+      "step": 2195
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1858038604259491,
+      "learning_rate": 0.0002999435275565495,
+      "loss": 1.0693,
+      "step": 2200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1565389484167099,
+      "learning_rate": 0.00029994066882758425,
+      "loss": 0.958,
+      "step": 2205
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1515437662601471,
+      "learning_rate": 0.00029993773953156095,
+      "loss": 1.1011,
+      "step": 2210
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16269664466381073,
+      "learning_rate": 0.0002999347396698581,
+      "loss": 1.0161,
+      "step": 2215
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.18648435175418854,
+      "learning_rate": 0.00029993166924388755,
+      "loss": 0.9996,
+      "step": 2220
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15165026485919952,
+      "learning_rate": 0.00029992852825509443,
+      "loss": 1.0186,
+      "step": 2225
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15864279866218567,
+      "learning_rate": 0.00029992531670495695,
+      "loss": 1.1264,
+      "step": 2230
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.13871487975120544,
+      "learning_rate": 0.00029992203459498654,
+      "loss": 1.0645,
+      "step": 2235
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15242180228233337,
+      "learning_rate": 0.0002999186819267279,
+      "loss": 1.004,
+      "step": 2240
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16039399802684784,
+      "learning_rate": 0.0002999152587017589,
+      "loss": 1.0212,
+      "step": 2245
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1544654369354248,
+      "learning_rate": 0.0002999117649216906,
+      "loss": 1.0053,
+      "step": 2250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15874332189559937,
+      "learning_rate": 0.0002999082005881673,
+      "loss": 0.991,
+      "step": 2255
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1567724049091339,
+      "learning_rate": 0.0002999045657028666,
+      "loss": 0.9868,
+      "step": 2260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.168124258518219,
+      "learning_rate": 0.000299900860267499,
+      "loss": 0.9966,
+      "step": 2265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14472615718841553,
+      "learning_rate": 0.0002998970842838086,
+      "loss": 1.0268,
+      "step": 2270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1609342396259308,
+      "learning_rate": 0.0002998932377535723,
+      "loss": 1.1297,
+      "step": 2275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14865976572036743,
+      "learning_rate": 0.0002998893206786006,
+      "loss": 1.0403,
+      "step": 2280
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.162654310464859,
+      "learning_rate": 0.0002998853330607369,
+      "loss": 1.0732,
+      "step": 2285
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15120600163936615,
+      "learning_rate": 0.0002998812749018579,
+      "loss": 1.0225,
+      "step": 2290
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14963847398757935,
+      "learning_rate": 0.0002998771462038735,
+      "loss": 0.9712,
+      "step": 2295
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14909087121486664,
+      "learning_rate": 0.00029987294696872687,
+      "loss": 1.0717,
+      "step": 2300
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15834349393844604,
+      "learning_rate": 0.00029986867719839427,
+      "loss": 1.1024,
+      "step": 2305
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14275187253952026,
+      "learning_rate": 0.00029986433689488515,
+      "loss": 1.0141,
+      "step": 2310
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.148543119430542,
+      "learning_rate": 0.0002998599260602423,
+      "loss": 1.0205,
+      "step": 2315
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14403221011161804,
+      "learning_rate": 0.00029985544469654155,
+      "loss": 1.0232,
+      "step": 2320
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1535564363002777,
+      "learning_rate": 0.000299850892805892,
+      "loss": 0.9793,
+      "step": 2325
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.143857941031456,
+      "learning_rate": 0.00029984627039043583,
+      "loss": 0.9961,
+      "step": 2330
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17255234718322754,
+      "learning_rate": 0.0002998415774523486,
+      "loss": 0.9796,
+      "step": 2335
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15468095242977142,
+      "learning_rate": 0.00029983681399383896,
+      "loss": 1.0223,
+      "step": 2340
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14770975708961487,
+      "learning_rate": 0.00029983198001714873,
+      "loss": 1.0694,
+      "step": 2345
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14611324667930603,
+      "learning_rate": 0.00029982707552455293,
+      "loss": 1.0747,
+      "step": 2350
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.19976738095283508,
+      "learning_rate": 0.0002998221005183598,
+      "loss": 1.0519,
+      "step": 2355
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1997351050376892,
+      "learning_rate": 0.0002998170550009107,
+      "loss": 1.1093,
+      "step": 2360
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15991829335689545,
+      "learning_rate": 0.0002998119389745802,
+      "loss": 1.029,
+      "step": 2365
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1683609038591385,
+      "learning_rate": 0.0002998067524417762,
+      "loss": 0.8991,
+      "step": 2370
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1613713800907135,
+      "learning_rate": 0.00029980149540493955,
+      "loss": 1.0849,
+      "step": 2375
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15293876826763153,
+      "learning_rate": 0.0002997961678665444,
+      "loss": 1.0498,
+      "step": 2380
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15544572472572327,
+      "learning_rate": 0.0002997907698290981,
+      "loss": 0.9695,
+      "step": 2385
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16265787184238434,
+      "learning_rate": 0.0002997853012951411,
+      "loss": 0.9936,
+      "step": 2390
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.144724041223526,
+      "learning_rate": 0.00029977976226724706,
+      "loss": 1.0893,
+      "step": 2395
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1512562781572342,
+      "learning_rate": 0.00029977415274802294,
+      "loss": 1.0602,
+      "step": 2400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.135329931974411,
+      "learning_rate": 0.0002997684727401086,
+      "loss": 1.0371,
+      "step": 2405
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15818721055984497,
+      "learning_rate": 0.00029976272224617744,
+      "loss": 1.0003,
+      "step": 2410
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16431140899658203,
+      "learning_rate": 0.00029975690126893566,
+      "loss": 1.017,
+      "step": 2415
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1531086415052414,
+      "learning_rate": 0.00029975100981112284,
+      "loss": 1.0058,
+      "step": 2420
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15581297874450684,
+      "learning_rate": 0.0002997450478755118,
+      "loss": 0.9911,
+      "step": 2425
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.14527705311775208,
+      "learning_rate": 0.0002997390154649083,
+      "loss": 1.0665,
+      "step": 2430
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1660381704568863,
+      "learning_rate": 0.0002997329125821515,
+      "loss": 0.9779,
+      "step": 2435
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.18006062507629395,
+      "learning_rate": 0.0002997267392301135,
+      "loss": 1.0496,
+      "step": 2440
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16907352209091187,
+      "learning_rate": 0.00029972049541169974,
+      "loss": 0.9899,
+      "step": 2445
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15815675258636475,
+      "learning_rate": 0.00029971418112984883,
+      "loss": 1.061,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15669722855091095,
+      "learning_rate": 0.0002997077963875324,
+      "loss": 1.0424,
+      "step": 2455
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14894799888134003,
+      "learning_rate": 0.00029970134118775533,
+      "loss": 0.9602,
+      "step": 2460
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16584117710590363,
+      "learning_rate": 0.00029969481553355565,
+      "loss": 0.9836,
+      "step": 2465
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15635208785533905,
+      "learning_rate": 0.0002996882194280046,
+      "loss": 0.9357,
+      "step": 2470
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16133363544940948,
+      "learning_rate": 0.0002996815528742065,
+      "loss": 0.9737,
+      "step": 2475
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1643681675195694,
+      "learning_rate": 0.00029967481587529884,
+      "loss": 1.0039,
+      "step": 2480
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16857954859733582,
+      "learning_rate": 0.0002996680084344523,
+      "loss": 1.0575,
+      "step": 2485
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16123433411121368,
+      "learning_rate": 0.0002996611305548707,
+      "loss": 0.9418,
+      "step": 2490
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1680585891008377,
+      "learning_rate": 0.0002996541822397909,
+      "loss": 0.9875,
+      "step": 2495
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1484622061252594,
+      "learning_rate": 0.00029964716349248306,
+      "loss": 0.824,
+      "step": 2500
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14574463665485382,
+      "learning_rate": 0.0002996400743162505,
+      "loss": 1.0491,
+      "step": 2505
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15741370618343353,
+      "learning_rate": 0.0002996329147144296,
+      "loss": 0.9946,
+      "step": 2510
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15442806482315063,
+      "learning_rate": 0.0002996256846903898,
+      "loss": 0.9688,
+      "step": 2515
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.15500134229660034,
+      "learning_rate": 0.00029961838424753394,
+      "loss": 1.0,
+      "step": 2520
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13550515472888947,
+      "learning_rate": 0.0002996110133892978,
+      "loss": 0.9205,
+      "step": 2525
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14668314158916473,
+      "learning_rate": 0.00029960357211915024,
+      "loss": 0.9877,
+      "step": 2530
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1662401258945465,
+      "learning_rate": 0.0002995960604405935,
+      "loss": 1.0079,
+      "step": 2535
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.14927972853183746,
+      "learning_rate": 0.00029958847835716285,
+      "loss": 0.8561,
+      "step": 2540
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.14817696809768677,
+      "learning_rate": 0.0002995808258724265,
+      "loss": 0.9245,
+      "step": 2545
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1539267748594284,
+      "learning_rate": 0.00029957310298998614,
+      "loss": 1.0536,
+      "step": 2550
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16719764471054077,
+      "learning_rate": 0.00029956530971347634,
+      "loss": 0.9806,
+      "step": 2555
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1492961347103119,
+      "learning_rate": 0.0002995574460465648,
+      "loss": 0.9849,
+      "step": 2560
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16492590308189392,
+      "learning_rate": 0.00029954951199295257,
+      "loss": 1.0159,
+      "step": 2565
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16332361102104187,
+      "learning_rate": 0.0002995415075563736,
+      "loss": 1.1316,
+      "step": 2570
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.15671640634536743,
+      "learning_rate": 0.0002995334327405951,
+      "loss": 0.9887,
+      "step": 2575
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1556137055158615,
+      "learning_rate": 0.00029952528754941725,
+      "loss": 1.0926,
+      "step": 2580
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1703471690416336,
+      "learning_rate": 0.00029951707198667347,
+      "loss": 0.9838,
+      "step": 2585
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.15181928873062134,
+      "learning_rate": 0.0002995087860562304,
+      "loss": 0.9467,
+      "step": 2590
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16134661436080933,
+      "learning_rate": 0.0002995004297619875,
+      "loss": 1.0854,
+      "step": 2595
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17149412631988525,
+      "learning_rate": 0.0002994920031078776,
+      "loss": 1.0174,
+      "step": 2600
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.15041255950927734,
+      "learning_rate": 0.0002994835060978666,
+      "loss": 1.0738,
+      "step": 2605
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1523531824350357,
+      "learning_rate": 0.0002994749387359534,
+      "loss": 1.0006,
+      "step": 2610
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.15003371238708496,
+      "learning_rate": 0.0002994663010261701,
+      "loss": 1.0203,
+      "step": 2615
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1510554850101471,
+      "learning_rate": 0.000299457592972582,
+      "loss": 0.9488,
+      "step": 2620
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1823001652956009,
+      "learning_rate": 0.0002994488145792872,
+      "loss": 0.9754,
+      "step": 2625
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.15630380809307098,
+      "learning_rate": 0.00029943996585041736,
+      "loss": 1.1082,
+      "step": 2630
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1587289720773697,
+      "learning_rate": 0.0002994310467901367,
+      "loss": 1.0511,
+      "step": 2635
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1543097347021103,
+      "learning_rate": 0.00029942205740264306,
+      "loss": 1.0272,
+      "step": 2640
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.16899822652339935,
+      "learning_rate": 0.00029941299769216704,
+      "loss": 1.0257,
+      "step": 2645
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.14778625965118408,
+      "learning_rate": 0.00029940386766297246,
+      "loss": 0.965,
+      "step": 2650
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.15615618228912354,
+      "learning_rate": 0.00029939466731935616,
+      "loss": 1.0942,
+      "step": 2655
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.15470904111862183,
+      "learning_rate": 0.0002993853966656482,
+      "loss": 0.9682,
+      "step": 2660
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.16802319884300232,
+      "learning_rate": 0.0002993760557062117,
+      "loss": 1.1369,
+      "step": 2665
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.15009041130542755,
+      "learning_rate": 0.0002993666444454426,
+      "loss": 0.9354,
+      "step": 2670
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1728777289390564,
+      "learning_rate": 0.0002993571628877704,
+      "loss": 1.0449,
+      "step": 2675
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1533242017030716,
+      "learning_rate": 0.0002993476110376574,
+      "loss": 0.9436,
+      "step": 2680
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.15997876226902008,
+      "learning_rate": 0.0002993379888995989,
+      "loss": 1.0483,
+      "step": 2685
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.16369123756885529,
+      "learning_rate": 0.00029932829647812354,
+      "loss": 0.9826,
+      "step": 2690
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1492471545934677,
+      "learning_rate": 0.0002993185337777927,
+      "loss": 1.0283,
+      "step": 2695
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.16298407316207886,
+      "learning_rate": 0.00029930870080320125,
+      "loss": 0.9464,
+      "step": 2700
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.20998337864875793,
+      "learning_rate": 0.00029929879755897674,
+      "loss": 1.0735,
+      "step": 2705
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.15956445038318634,
+      "learning_rate": 0.0002992888240497801,
+      "loss": 1.037,
+      "step": 2710
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.15296678245067596,
+      "learning_rate": 0.0002992787802803051,
+      "loss": 1.0028,
+      "step": 2715
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1596369743347168,
+      "learning_rate": 0.0002992686662552787,
+      "loss": 1.0732,
+      "step": 2720
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.16030800342559814,
+      "learning_rate": 0.0002992584819794609,
+      "loss": 1.0009,
+      "step": 2725
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.16185402870178223,
+      "learning_rate": 0.00029924822745764485,
+      "loss": 0.9573,
+      "step": 2730
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.15265031158924103,
+      "learning_rate": 0.0002992379026946565,
+      "loss": 1.0878,
+      "step": 2735
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17246966063976288,
+      "learning_rate": 0.00029922750769535505,
+      "loss": 0.9797,
+      "step": 2740
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.16208067536354065,
+      "learning_rate": 0.00029921704246463284,
+      "loss": 1.0398,
+      "step": 2745
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.15843920409679413,
+      "learning_rate": 0.0002992065070074151,
+      "loss": 0.9905,
+      "step": 2750
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.15435267984867096,
+      "learning_rate": 0.0002991959013286602,
+      "loss": 1.0302,
+      "step": 2755
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.14609667658805847,
+      "learning_rate": 0.00029918522543335947,
+      "loss": 0.9949,
+      "step": 2760
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1628682017326355,
+      "learning_rate": 0.00029917447932653737,
+      "loss": 0.9568,
+      "step": 2765
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1572001874446869,
+      "learning_rate": 0.0002991636630132513,
+      "loss": 1.0415,
+      "step": 2770
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1506408005952835,
+      "learning_rate": 0.0002991527764985919,
+      "loss": 0.9368,
+      "step": 2775
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1670454889535904,
+      "learning_rate": 0.00029914181978768267,
+      "loss": 1.0478,
+      "step": 2780
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16997024416923523,
+      "learning_rate": 0.0002991307928856802,
+      "loss": 1.0219,
+      "step": 2785
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.15132923424243927,
+      "learning_rate": 0.00029911969579777414,
+      "loss": 1.0253,
+      "step": 2790
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1776067316532135,
+      "learning_rate": 0.00029910852852918713,
+      "loss": 1.0261,
+      "step": 2795
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16810853779315948,
+      "learning_rate": 0.0002990972910851748,
+      "loss": 1.1051,
+      "step": 2800
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16523437201976776,
+      "learning_rate": 0.0002990859834710259,
+      "loss": 1.0066,
+      "step": 2805
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.15905992686748505,
+      "learning_rate": 0.0002990746056920623,
+      "loss": 1.0197,
+      "step": 2810
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16057956218719482,
+      "learning_rate": 0.00029906315775363857,
+      "loss": 0.9492,
+      "step": 2815
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1586558222770691,
+      "learning_rate": 0.0002990516396611425,
+      "loss": 1.0266,
+      "step": 2820
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17754611372947693,
+      "learning_rate": 0.0002990400514199951,
+      "loss": 0.9497,
+      "step": 2825
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1540936827659607,
+      "learning_rate": 0.00029902839303564994,
+      "loss": 0.9285,
+      "step": 2830
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1709967404603958,
+      "learning_rate": 0.00029901666451359393,
+      "loss": 0.9819,
+      "step": 2835
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16606543958187103,
+      "learning_rate": 0.0002990048658593469,
+      "loss": 0.999,
+      "step": 2840
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16725273430347443,
+      "learning_rate": 0.0002989929970784618,
+      "loss": 1.0302,
+      "step": 2845
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16970454156398773,
+      "learning_rate": 0.0002989810581765243,
+      "loss": 1.0709,
+      "step": 2850
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.18702805042266846,
+      "learning_rate": 0.0002989690491591533,
+      "loss": 0.9305,
+      "step": 2855
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16489920020103455,
+      "learning_rate": 0.0002989569700320007,
+      "loss": 0.9604,
+      "step": 2860
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.21395571529865265,
+      "learning_rate": 0.0002989448208007513,
+      "loss": 0.9137,
+      "step": 2865
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.157196506857872,
+      "learning_rate": 0.00029893260147112287,
+      "loss": 1.0027,
+      "step": 2870
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16530407965183258,
+      "learning_rate": 0.0002989203120488663,
+      "loss": 1.0449,
+      "step": 2875
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15222971141338348,
+      "learning_rate": 0.0002989079525397654,
+      "loss": 0.985,
+      "step": 2880
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1723487377166748,
+      "learning_rate": 0.00029889552294963697,
+      "loss": 0.9842,
+      "step": 2885
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15952201187610626,
+      "learning_rate": 0.0002988830232843308,
+      "loss": 1.0918,
+      "step": 2890
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1670973002910614,
+      "learning_rate": 0.00029887045354972953,
+      "loss": 1.0285,
+      "step": 2895
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16919025778770447,
+      "learning_rate": 0.00029885781375174906,
+      "loss": 0.9887,
+      "step": 2900
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15993523597717285,
+      "learning_rate": 0.000298845103896338,
+      "loss": 1.0421,
+      "step": 2905
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15911990404129028,
+      "learning_rate": 0.00029883232398947806,
+      "loss": 1.0282,
+      "step": 2910
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15578506886959076,
+      "learning_rate": 0.0002988194740371839,
+      "loss": 1.0445,
+      "step": 2915
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1739857941865921,
+      "learning_rate": 0.0002988065540455031,
+      "loss": 1.1877,
+      "step": 2920
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15756095945835114,
+      "learning_rate": 0.0002987935640205162,
+      "loss": 0.997,
+      "step": 2925
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1599845141172409,
+      "learning_rate": 0.00029878050396833685,
+      "loss": 0.998,
+      "step": 2930
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15747900307178497,
+      "learning_rate": 0.0002987673738951115,
+      "loss": 0.9975,
+      "step": 2935
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17130185663700104,
+      "learning_rate": 0.00029875417380701954,
+      "loss": 1.0535,
+      "step": 2940
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17812122404575348,
+      "learning_rate": 0.0002987409037102734,
+      "loss": 0.9872,
+      "step": 2945
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16236154735088348,
+      "learning_rate": 0.0002987275636111185,
+      "loss": 1.0886,
+      "step": 2950
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16785535216331482,
+      "learning_rate": 0.0002987141535158331,
+      "loss": 0.9917,
+      "step": 2955
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1586502194404602,
+      "learning_rate": 0.0002987006734307283,
+      "loss": 1.0217,
+      "step": 2960
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.15011291205883026,
+      "learning_rate": 0.0002986871233621484,
+      "loss": 1.0619,
+      "step": 2965
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.14755620062351227,
+      "learning_rate": 0.0002986735033164706,
+      "loss": 1.0555,
+      "step": 2970
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1600685715675354,
+      "learning_rate": 0.0002986598133001048,
+      "loss": 0.9612,
+      "step": 2975
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.15799719095230103,
+      "learning_rate": 0.00029864605331949396,
+      "loss": 0.9002,
+      "step": 2980
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16441842913627625,
+      "learning_rate": 0.0002986322233811141,
+      "loss": 0.9336,
+      "step": 2985
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16389460861682892,
+      "learning_rate": 0.000298618323491474,
+      "loss": 0.8653,
+      "step": 2990
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16231980919837952,
+      "learning_rate": 0.00029860435365711537,
+      "loss": 1.0476,
+      "step": 2995
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.15994654595851898,
+      "learning_rate": 0.00029859031388461296,
+      "loss": 0.9567,
+      "step": 3000
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16480427980422974,
+      "learning_rate": 0.00029857620418057424,
+      "loss": 0.9938,
+      "step": 3005
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16214433312416077,
+      "learning_rate": 0.0002985620245516398,
+      "loss": 0.9345,
+      "step": 3010
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1897832155227661,
+      "learning_rate": 0.00029854777500448303,
+      "loss": 0.9927,
+      "step": 3015
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1605633646249771,
+      "learning_rate": 0.00029853345554581024,
+      "loss": 0.9714,
+      "step": 3020
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17080001533031464,
+      "learning_rate": 0.0002985190661823606,
+      "loss": 1.0147,
+      "step": 3025
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16237355768680573,
+      "learning_rate": 0.0002985046069209062,
+      "loss": 1.0514,
+      "step": 3030
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17489008605480194,
+      "learning_rate": 0.0002984900777682522,
+      "loss": 0.9702,
+      "step": 3035
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1539561152458191,
+      "learning_rate": 0.00029847547873123627,
+      "loss": 0.9965,
+      "step": 3040
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.15789541602134705,
+      "learning_rate": 0.0002984608098167295,
+      "loss": 1.0079,
+      "step": 3045
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17130644619464874,
+      "learning_rate": 0.0002984460710316353,
+      "loss": 1.006,
+      "step": 3050
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.19519822299480438,
+      "learning_rate": 0.0002984312623828903,
+      "loss": 0.9245,
+      "step": 3055
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16348280012607574,
+      "learning_rate": 0.000298416383877464,
+      "loss": 1.042,
+      "step": 3060
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.14846982061862946,
+      "learning_rate": 0.0002984014355223587,
+      "loss": 1.0291,
+      "step": 3065
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18363744020462036,
+      "learning_rate": 0.0002983864173246096,
+      "loss": 1.0021,
+      "step": 3070
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.15816280245780945,
+      "learning_rate": 0.00029837132929128474,
+      "loss": 0.9942,
+      "step": 3075
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1437516212463379,
+      "learning_rate": 0.00029835617142948503,
+      "loss": 1.0911,
+      "step": 3080
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16180740296840668,
+      "learning_rate": 0.0002983409437463443,
+      "loss": 1.0143,
+      "step": 3085
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1789834052324295,
+      "learning_rate": 0.0002983256462490292,
+      "loss": 1.0422,
+      "step": 3090
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17736496031284332,
+      "learning_rate": 0.00029831027894473925,
+      "loss": 1.028,
+      "step": 3095
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16583746671676636,
+      "learning_rate": 0.00029829484184070674,
+      "loss": 1.0083,
+      "step": 3100
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1849256455898285,
+      "learning_rate": 0.000298279334944197,
+      "loss": 1.0822,
+      "step": 3105
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16508488357067108,
+      "learning_rate": 0.000298263758262508,
+      "loss": 0.9659,
+      "step": 3110
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1658620983362198,
+      "learning_rate": 0.0002982481118029707,
+      "loss": 1.0032,
+      "step": 3115
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16383464634418488,
+      "learning_rate": 0.0002982323955729488,
+      "loss": 1.012,
+      "step": 3120
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16278916597366333,
+      "learning_rate": 0.0002982166095798389,
+      "loss": 1.0042,
+      "step": 3125
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16901999711990356,
+      "learning_rate": 0.0002982007538310704,
+      "loss": 0.9857,
+      "step": 3130
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.15740418434143066,
+      "learning_rate": 0.0002981848283341056,
+      "loss": 0.9621,
+      "step": 3135
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1794396936893463,
+      "learning_rate": 0.00029816883309643946,
+      "loss": 0.9642,
+      "step": 3140
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16382227838039398,
+      "learning_rate": 0.0002981527681255999,
+      "loss": 1.0157,
+      "step": 3145
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17343173921108246,
+      "learning_rate": 0.00029813663342914774,
+      "loss": 1.0122,
+      "step": 3150
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16225513815879822,
+      "learning_rate": 0.0002981204290146764,
+      "loss": 0.9324,
+      "step": 3155
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1837487518787384,
+      "learning_rate": 0.00029810415488981223,
+      "loss": 0.9521,
+      "step": 3160
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16996806859970093,
+      "learning_rate": 0.0002980878110622144,
+      "loss": 0.9372,
+      "step": 3165
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1729620099067688,
+      "learning_rate": 0.0002980713975395748,
+      "loss": 1.0272,
+      "step": 3170
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.15893127024173737,
+      "learning_rate": 0.0002980549143296182,
+      "loss": 1.0605,
+      "step": 3175
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.19714705646038055,
+      "learning_rate": 0.0002980383614401023,
+      "loss": 1.0335,
+      "step": 3180
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17942222952842712,
+      "learning_rate": 0.0002980217388788172,
+      "loss": 1.0134,
+      "step": 3185
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1684209406375885,
+      "learning_rate": 0.0002980050466535861,
+      "loss": 1.0344,
+      "step": 3190
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16573862731456757,
+      "learning_rate": 0.000297988284772265,
+      "loss": 1.0624,
+      "step": 3195
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16821695864200592,
+      "learning_rate": 0.00029797145324274256,
+      "loss": 0.9756,
+      "step": 3200
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.15828034281730652,
+      "learning_rate": 0.0002979545520729402,
+      "loss": 0.9553,
+      "step": 3205
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17686916887760162,
+      "learning_rate": 0.00029793758127081226,
+      "loss": 0.9576,
+      "step": 3210
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16397157311439514,
+      "learning_rate": 0.00029792054084434573,
+      "loss": 0.9603,
+      "step": 3215
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16657304763793945,
+      "learning_rate": 0.0002979034308015603,
+      "loss": 1.0867,
+      "step": 3220
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16404330730438232,
+      "learning_rate": 0.00029788625115050873,
+      "loss": 1.0548,
+      "step": 3225
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16127341985702515,
+      "learning_rate": 0.0002978690018992761,
+      "loss": 1.0754,
+      "step": 3230
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1616549789905548,
+      "learning_rate": 0.0002978516830559807,
+      "loss": 0.9912,
+      "step": 3235
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.15977180004119873,
+      "learning_rate": 0.0002978342946287732,
+      "loss": 0.9411,
+      "step": 3240
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.18043774366378784,
+      "learning_rate": 0.00029781683662583725,
+      "loss": 1.0057,
+      "step": 3245
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16147902607917786,
+      "learning_rate": 0.00029779930905538915,
+      "loss": 0.9489,
+      "step": 3250
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17931294441223145,
+      "learning_rate": 0.0002977817119256779,
+      "loss": 1.0077,
+      "step": 3255
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16193710267543793,
+      "learning_rate": 0.00029776404524498533,
+      "loss": 1.0576,
+      "step": 3260
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17987242341041565,
+      "learning_rate": 0.00029774630902162604,
+      "loss": 0.9772,
+      "step": 3265
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.15659894049167633,
+      "learning_rate": 0.0002977285032639472,
+      "loss": 0.9922,
+      "step": 3270
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17230859398841858,
+      "learning_rate": 0.0002977106279803288,
+      "loss": 1.0216,
+      "step": 3275
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16389912366867065,
+      "learning_rate": 0.00029769268317918354,
+      "loss": 0.9859,
+      "step": 3280
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1696440726518631,
+      "learning_rate": 0.00029767466886895685,
+      "loss": 1.0755,
+      "step": 3285
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.18958187103271484,
+      "learning_rate": 0.0002976565850581269,
+      "loss": 0.9506,
+      "step": 3290
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1598709523677826,
+      "learning_rate": 0.0002976384317552044,
+      "loss": 0.9734,
+      "step": 3295
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17023463547229767,
+      "learning_rate": 0.0002976202089687331,
+      "loss": 0.9537,
+      "step": 3300
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.19109511375427246,
+      "learning_rate": 0.0002976019167072891,
+      "loss": 0.9778,
+      "step": 3305
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1838565170764923,
+      "learning_rate": 0.00029758355497948145,
+      "loss": 1.0258,
+      "step": 3310
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16939514875411987,
+      "learning_rate": 0.0002975651237939517,
+      "loss": 1.0244,
+      "step": 3315
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1645985096693039,
+      "learning_rate": 0.0002975466231593742,
+      "loss": 0.9729,
+      "step": 3320
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17347556352615356,
+      "learning_rate": 0.0002975280530844559,
+      "loss": 1.0492,
+      "step": 3325
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.21827654540538788,
+      "learning_rate": 0.00029750941357793666,
+      "loss": 0.9775,
+      "step": 3330
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1544768363237381,
+      "learning_rate": 0.00029749070464858875,
+      "loss": 1.0207,
+      "step": 3335
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.15815982222557068,
+      "learning_rate": 0.00029747192630521715,
+      "loss": 1.0502,
+      "step": 3340
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17267605662345886,
+      "learning_rate": 0.0002974530785566597,
+      "loss": 0.9846,
+      "step": 3345
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1674361526966095,
+      "learning_rate": 0.00029743416141178667,
+      "loss": 0.9606,
+      "step": 3350
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16751405596733093,
+      "learning_rate": 0.00029741517487950116,
+      "loss": 1.0469,
+      "step": 3355
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.20405547320842743,
+      "learning_rate": 0.00029739611896873884,
+      "loss": 0.9714,
+      "step": 3360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.15737318992614746,
+      "learning_rate": 0.00029737699368846806,
+      "loss": 0.982,
+      "step": 3365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.16428694128990173,
+      "learning_rate": 0.0002973577990476899,
+      "loss": 1.0685,
+      "step": 3370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.16110824048519135,
+      "learning_rate": 0.0002973385350554378,
+      "loss": 0.947,
+      "step": 3375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1924237459897995,
+      "learning_rate": 0.00029731920172077815,
+      "loss": 0.9267,
+      "step": 3380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17750895023345947,
+      "learning_rate": 0.00029729979905280987,
+      "loss": 1.0132,
+      "step": 3385
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.16638635098934174,
+      "learning_rate": 0.0002972803270606645,
+      "loss": 0.9343,
+      "step": 3390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.16308121383190155,
+      "learning_rate": 0.00029726078575350613,
+      "loss": 1.0279,
+      "step": 3395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.16061437129974365,
+      "learning_rate": 0.00029724117514053164,
+      "loss": 0.9822,
+      "step": 3400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1641559600830078,
+      "learning_rate": 0.00029722149523097046,
+      "loss": 1.014,
+      "step": 3405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.16438022255897522,
+      "learning_rate": 0.0002972017460340845,
+      "loss": 1.0781,
+      "step": 3410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.16575871407985687,
+      "learning_rate": 0.0002971819275591684,
+      "loss": 1.0313,
+      "step": 3415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1603136956691742,
+      "learning_rate": 0.00029716203981554947,
+      "loss": 0.9883,
+      "step": 3420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1538233458995819,
+      "learning_rate": 0.0002971420828125875,
+      "loss": 1.0134,
+      "step": 3425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.15688014030456543,
+      "learning_rate": 0.0002971220565596749,
+      "loss": 0.9029,
+      "step": 3430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17302924394607544,
+      "learning_rate": 0.0002971019610662367,
+      "loss": 0.9459,
+      "step": 3435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17244279384613037,
+      "learning_rate": 0.00029708179634173055,
+      "loss": 0.9259,
+      "step": 3440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17056743800640106,
+      "learning_rate": 0.00029706156239564665,
+      "loss": 0.8916,
+      "step": 3445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.18607419729232788,
+      "learning_rate": 0.00029704125923750766,
+      "loss": 0.9067,
+      "step": 3450
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.18427631258964539,
+      "learning_rate": 0.000297020886876869,
+      "loss": 0.9818,
+      "step": 3455
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17508511245250702,
+      "learning_rate": 0.00029700044532331854,
+      "loss": 0.9721,
+      "step": 3460
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1710289865732193,
+      "learning_rate": 0.0002969799345864768,
+      "loss": 1.0841,
+      "step": 3465
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.16923823952674866,
+      "learning_rate": 0.00029695935467599676,
+      "loss": 1.036,
+      "step": 3470
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1833118349313736,
+      "learning_rate": 0.00029693870560156406,
+      "loss": 1.0369,
+      "step": 3475
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.158515065908432,
+      "learning_rate": 0.0002969179873728968,
+      "loss": 0.9474,
+      "step": 3480
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.18448443710803986,
+      "learning_rate": 0.0002968971999997458,
+      "loss": 1.036,
+      "step": 3485
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17607718706130981,
+      "learning_rate": 0.000296876343491894,
+      "loss": 1.1111,
+      "step": 3490
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.16396880149841309,
+      "learning_rate": 0.0002968554178591575,
+      "loss": 1.0609,
+      "step": 3495
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1581658273935318,
+      "learning_rate": 0.00029683442311138436,
+      "loss": 1.0826,
+      "step": 3500
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.16231244802474976,
+      "learning_rate": 0.00029681335925845544,
+      "loss": 0.9405,
+      "step": 3505
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17052994668483734,
+      "learning_rate": 0.0002967922263102842,
+      "loss": 1.057,
+      "step": 3510
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1752976030111313,
+      "learning_rate": 0.00029677102427681643,
+      "loss": 0.9852,
+      "step": 3515
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1679937094449997,
+      "learning_rate": 0.00029674975316803056,
+      "loss": 1.053,
+      "step": 3520
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17589174211025238,
+      "learning_rate": 0.00029672841299393734,
+      "loss": 1.0338,
+      "step": 3525
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17791646718978882,
+      "learning_rate": 0.00029670700376458034,
+      "loss": 0.9818,
+      "step": 3530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1609509289264679,
+      "learning_rate": 0.0002966855254900353,
+      "loss": 0.9841,
+      "step": 3535
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1620442420244217,
+      "learning_rate": 0.0002966639781804108,
+      "loss": 1.0825,
+      "step": 3540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18047846853733063,
+      "learning_rate": 0.00029664236184584757,
+      "loss": 1.0661,
+      "step": 3545
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.16510747373104095,
+      "learning_rate": 0.00029662067649651895,
+      "loss": 0.8942,
+      "step": 3550
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18629179894924164,
+      "learning_rate": 0.00029659892214263094,
+      "loss": 1.0014,
+      "step": 3555
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18440277874469757,
+      "learning_rate": 0.0002965770987944217,
+      "loss": 1.0379,
+      "step": 3560
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1701250523328781,
+      "learning_rate": 0.00029655520646216214,
+      "loss": 0.9512,
+      "step": 3565
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17910198867321014,
+      "learning_rate": 0.00029653324515615544,
+      "loss": 1.0178,
+      "step": 3570
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.15879999101161957,
+      "learning_rate": 0.00029651121488673737,
+      "loss": 0.9555,
+      "step": 3575
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17738574743270874,
+      "learning_rate": 0.00029648911566427605,
+      "loss": 0.9916,
+      "step": 3580
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1952105015516281,
+      "learning_rate": 0.00029646694749917217,
+      "loss": 1.0,
+      "step": 3585
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1668039858341217,
+      "learning_rate": 0.0002964447104018588,
+      "loss": 0.9486,
+      "step": 3590
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.15947400033473969,
+      "learning_rate": 0.0002964224043828014,
+      "loss": 0.9834,
+      "step": 3595
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18073715269565582,
+      "learning_rate": 0.000296400029452498,
+      "loss": 0.9501,
+      "step": 3600
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18462218344211578,
+      "learning_rate": 0.00029637758562147895,
+      "loss": 1.0168,
+      "step": 3605
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1783631145954132,
+      "learning_rate": 0.000296355072900307,
+      "loss": 0.9695,
+      "step": 3610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1757393628358841,
+      "learning_rate": 0.00029633249129957747,
+      "loss": 0.9556,
+      "step": 3615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17686530947685242,
+      "learning_rate": 0.000296309840829918,
+      "loss": 0.9561,
+      "step": 3620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16874971985816956,
+      "learning_rate": 0.00029628712150198865,
+      "loss": 0.9509,
+      "step": 3625
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17538826167583466,
+      "learning_rate": 0.00029626433332648183,
+      "loss": 1.0664,
+      "step": 3630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16366077959537506,
+      "learning_rate": 0.00029624147631412246,
+      "loss": 0.9833,
+      "step": 3635
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17649723589420319,
+      "learning_rate": 0.0002962185504756678,
+      "loss": 0.9077,
+      "step": 3640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1709517389535904,
+      "learning_rate": 0.0002961955558219076,
+      "loss": 0.9386,
+      "step": 3645
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1735381931066513,
+      "learning_rate": 0.0002961724923636637,
+      "loss": 1.0339,
+      "step": 3650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17340226471424103,
+      "learning_rate": 0.00029614936011179076,
+      "loss": 0.9962,
+      "step": 3655
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1836249977350235,
+      "learning_rate": 0.0002961261590771755,
+      "loss": 0.8727,
+      "step": 3660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16805151104927063,
+      "learning_rate": 0.000296102889270737,
+      "loss": 0.986,
+      "step": 3665
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1725672036409378,
+      "learning_rate": 0.00029607955070342685,
+      "loss": 1.0204,
+      "step": 3670
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17753121256828308,
+      "learning_rate": 0.00029605614338622905,
+      "loss": 0.9744,
+      "step": 3675
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18001006543636322,
+      "learning_rate": 0.00029603266733015983,
+      "loss": 0.9984,
+      "step": 3680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18695253133773804,
+      "learning_rate": 0.0002960091225462677,
+      "loss": 1.0663,
+      "step": 3685
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18563799560070038,
+      "learning_rate": 0.00029598550904563374,
+      "loss": 0.9452,
+      "step": 3690
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18646039068698883,
+      "learning_rate": 0.0002959618268393712,
+      "loss": 1.0318,
+      "step": 3695
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.16049382090568542,
+      "learning_rate": 0.00029593807593862565,
+      "loss": 1.0137,
+      "step": 3700
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17187224328517914,
+      "learning_rate": 0.00029591425635457514,
+      "loss": 0.9957,
+      "step": 3705
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18150264024734497,
+      "learning_rate": 0.00029589036809842987,
+      "loss": 0.9998,
+      "step": 3710
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18180853128433228,
+      "learning_rate": 0.0002958664111814326,
+      "loss": 0.9865,
+      "step": 3715
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17173470556735992,
+      "learning_rate": 0.0002958423856148581,
+      "loss": 0.9939,
+      "step": 3720
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18035240471363068,
+      "learning_rate": 0.0002958182914100137,
+      "loss": 0.9806,
+      "step": 3725
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18080519139766693,
+      "learning_rate": 0.00029579412857823887,
+      "loss": 0.9165,
+      "step": 3730
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18667446076869965,
+      "learning_rate": 0.0002957698971309054,
+      "loss": 0.9465,
+      "step": 3735
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17435042560100555,
+      "learning_rate": 0.0002957455970794175,
+      "loss": 0.9562,
+      "step": 3740
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.19074921309947968,
+      "learning_rate": 0.0002957212284352116,
+      "loss": 0.9912,
+      "step": 3745
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1836334615945816,
+      "learning_rate": 0.0002956967912097563,
+      "loss": 0.9995,
+      "step": 3750
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.19965748488903046,
+      "learning_rate": 0.00029567228541455264,
+      "loss": 0.9773,
+      "step": 3755
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.19002239406108856,
+      "learning_rate": 0.0002956477110611338,
+      "loss": 0.978,
+      "step": 3760
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1769300252199173,
+      "learning_rate": 0.00029562306816106535,
+      "loss": 0.9428,
+      "step": 3765
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1742662787437439,
+      "learning_rate": 0.000295598356725945,
+      "loss": 0.9801,
+      "step": 3770
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17609982192516327,
+      "learning_rate": 0.00029557357676740286,
+      "loss": 0.9655,
+      "step": 3775
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.152922123670578,
+      "learning_rate": 0.00029554872829710114,
+      "loss": 1.0099,
+      "step": 3780
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18085353076457977,
+      "learning_rate": 0.0002955238113267344,
+      "loss": 1.0337,
+      "step": 3785
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17463679611682892,
+      "learning_rate": 0.00029549882586802923,
+      "loss": 0.9719,
+      "step": 3790
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1799176186323166,
+      "learning_rate": 0.0002954737719327448,
+      "loss": 0.9933,
+      "step": 3795
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18026649951934814,
+      "learning_rate": 0.00029544864953267224,
+      "loss": 1.0829,
+      "step": 3800
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18664585053920746,
+      "learning_rate": 0.000295423458679635,
+      "loss": 0.9727,
+      "step": 3805
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1912483125925064,
+      "learning_rate": 0.0002953981993854888,
+      "loss": 0.9602,
+      "step": 3810
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17371563613414764,
+      "learning_rate": 0.00029537287166212146,
+      "loss": 0.9695,
+      "step": 3815
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17529870569705963,
+      "learning_rate": 0.00029534747552145295,
+      "loss": 1.0112,
+      "step": 3820
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.20106241106987,
+      "learning_rate": 0.00029532201097543566,
+      "loss": 1.045,
+      "step": 3825
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.16965335607528687,
+      "learning_rate": 0.00029529647803605406,
+      "loss": 0.9602,
+      "step": 3830
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.16712799668312073,
+      "learning_rate": 0.00029527087671532467,
+      "loss": 0.8681,
+      "step": 3835
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17524246871471405,
+      "learning_rate": 0.00029524520702529645,
+      "loss": 1.0298,
+      "step": 3840
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.16729198396205902,
+      "learning_rate": 0.00029521946897805034,
+      "loss": 1.0393,
+      "step": 3845
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1771865338087082,
+      "learning_rate": 0.00029519366258569954,
+      "loss": 0.9163,
+      "step": 3850
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.19516459107398987,
+      "learning_rate": 0.0002951677878603894,
+      "loss": 1.0929,
+      "step": 3855
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.16571563482284546,
+      "learning_rate": 0.0002951418448142974,
+      "loss": 0.9895,
+      "step": 3860
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17341572046279907,
+      "learning_rate": 0.00029511583345963327,
+      "loss": 0.951,
+      "step": 3865
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.16998711228370667,
+      "learning_rate": 0.00029508975380863867,
+      "loss": 0.9239,
+      "step": 3870
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17954027652740479,
+      "learning_rate": 0.0002950636058735877,
+      "loss": 0.9932,
+      "step": 3875
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17167799174785614,
+      "learning_rate": 0.0002950373896667864,
+      "loss": 1.0185,
+      "step": 3880
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.16167834401130676,
+      "learning_rate": 0.000295011105200573,
+      "loss": 1.0683,
+      "step": 3885
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1931363195180893,
+      "learning_rate": 0.0002949847524873178,
+      "loss": 0.8858,
+      "step": 3890
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1638021469116211,
+      "learning_rate": 0.0002949583315394233,
+      "loss": 0.9769,
+      "step": 3895
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17214910686016083,
+      "learning_rate": 0.00029493184236932405,
+      "loss": 1.0506,
+      "step": 3900
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18750952184200287,
+      "learning_rate": 0.0002949052849894867,
+      "loss": 0.9815,
+      "step": 3905
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.16336201131343842,
+      "learning_rate": 0.00029487865941241014,
+      "loss": 0.9287,
+      "step": 3910
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17103566229343414,
+      "learning_rate": 0.00029485196565062516,
+      "loss": 0.9731,
+      "step": 3915
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.16729870438575745,
+      "learning_rate": 0.0002948252037166948,
+      "loss": 0.9477,
+      "step": 3920
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1882411539554596,
+      "learning_rate": 0.00029479837362321405,
+      "loss": 0.9904,
+      "step": 3925
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18706083297729492,
+      "learning_rate": 0.00029477147538281004,
+      "loss": 0.997,
+      "step": 3930
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18847805261611938,
+      "learning_rate": 0.000294744509008142,
+      "loss": 0.9378,
+      "step": 3935
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1840084046125412,
+      "learning_rate": 0.00029471747451190124,
+      "loss": 0.9559,
+      "step": 3940
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1757187843322754,
+      "learning_rate": 0.000294690371906811,
+      "loss": 1.011,
+      "step": 3945
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.16202908754348755,
+      "learning_rate": 0.00029466320120562683,
+      "loss": 0.9077,
+      "step": 3950
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18490107357501984,
+      "learning_rate": 0.00029463596242113596,
+      "loss": 1.0702,
+      "step": 3955
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1752914935350418,
+      "learning_rate": 0.000294608655566158,
+      "loss": 0.9612,
+      "step": 3960
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17405173182487488,
+      "learning_rate": 0.00029458128065354444,
+      "loss": 1.0131,
+      "step": 3965
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18187612295150757,
+      "learning_rate": 0.0002945538376961788,
+      "loss": 0.9848,
+      "step": 3970
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.19082453846931458,
+      "learning_rate": 0.0002945263267069766,
+      "loss": 0.9559,
+      "step": 3975
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17413316667079926,
+      "learning_rate": 0.0002944987476988855,
+      "loss": 1.0772,
+      "step": 3980
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.15954262018203735,
+      "learning_rate": 0.00029447110068488516,
+      "loss": 0.9037,
+      "step": 3985
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17299918830394745,
+      "learning_rate": 0.000294443385677987,
+      "loss": 0.9624,
+      "step": 3990
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.179169163107872,
+      "learning_rate": 0.00029441560269123483,
+      "loss": 0.983,
+      "step": 3995
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.16181403398513794,
+      "learning_rate": 0.00029438775173770405,
+      "loss": 0.9705,
+      "step": 4000
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1745695024728775,
+      "learning_rate": 0.0002943598328305024,
+      "loss": 0.9664,
+      "step": 4005
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1865403801202774,
+      "learning_rate": 0.0002943318459827693,
+      "loss": 1.0259,
+      "step": 4010
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.21468792855739594,
+      "learning_rate": 0.0002943037912076764,
+      "loss": 0.9311,
+      "step": 4015
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18576794862747192,
+      "learning_rate": 0.0002942756685184272,
+      "loss": 1.07,
+      "step": 4020
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18512584269046783,
+      "learning_rate": 0.0002942474779282571,
+      "loss": 0.9204,
+      "step": 4025
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18075776100158691,
+      "learning_rate": 0.00029421921945043365,
+      "loss": 0.9853,
+      "step": 4030
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17702947556972504,
+      "learning_rate": 0.0002941908930982561,
+      "loss": 1.0537,
+      "step": 4035
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17525263130664825,
+      "learning_rate": 0.0002941624988850558,
+      "loss": 0.9713,
+      "step": 4040
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.16947941482067108,
+      "learning_rate": 0.00029413403682419613,
+      "loss": 0.9988,
+      "step": 4045
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17657147347927094,
+      "learning_rate": 0.0002941055069290721,
+      "loss": 0.9946,
+      "step": 4050
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17668572068214417,
+      "learning_rate": 0.00029407690921311094,
+      "loss": 0.9288,
+      "step": 4055
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18631796538829803,
+      "learning_rate": 0.0002940482436897717,
+      "loss": 0.9218,
+      "step": 4060
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1784496307373047,
+      "learning_rate": 0.00029401951037254524,
+      "loss": 1.059,
+      "step": 4065
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.19037973880767822,
+      "learning_rate": 0.00029399070927495447,
+      "loss": 0.9622,
+      "step": 4070
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17048045992851257,
+      "learning_rate": 0.0002939618404105541,
+      "loss": 0.8903,
+      "step": 4075
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1878838837146759,
+      "learning_rate": 0.00029393290379293085,
+      "loss": 1.0106,
+      "step": 4080
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18040554225444794,
+      "learning_rate": 0.0002939038994357032,
+      "loss": 0.9517,
+      "step": 4085
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.19907140731811523,
+      "learning_rate": 0.0002938748273525216,
+      "loss": 0.9628,
+      "step": 4090
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18067404627799988,
+      "learning_rate": 0.00029384568755706824,
+      "loss": 0.9726,
+      "step": 4095
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.16621145606040955,
+      "learning_rate": 0.0002938164800630574,
+      "loss": 1.0339,
+      "step": 4100
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17213256657123566,
+      "learning_rate": 0.00029378720488423506,
+      "loss": 0.933,
+      "step": 4105
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1835915744304657,
+      "learning_rate": 0.00029375786203437906,
+      "loss": 0.9645,
+      "step": 4110
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.19219228625297546,
+      "learning_rate": 0.00029372845152729916,
+      "loss": 1.0066,
+      "step": 4115
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1753523349761963,
+      "learning_rate": 0.0002936989733768368,
+      "loss": 0.98,
+      "step": 4120
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.16420799493789673,
+      "learning_rate": 0.00029366942759686556,
+      "loss": 0.9817,
+      "step": 4125
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17462868988513947,
+      "learning_rate": 0.0002936398142012906,
+      "loss": 1.024,
+      "step": 4130
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1853259801864624,
+      "learning_rate": 0.0002936101332040489,
+      "loss": 0.9538,
+      "step": 4135
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17520953714847565,
+      "learning_rate": 0.00029358038461910934,
+      "loss": 0.9453,
+      "step": 4140
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18350806832313538,
+      "learning_rate": 0.00029355056846047266,
+      "loss": 0.9445,
+      "step": 4145
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17712032794952393,
+      "learning_rate": 0.0002935206847421713,
+      "loss": 0.9821,
+      "step": 4150
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17345529794692993,
+      "learning_rate": 0.00029349073347826953,
+      "loss": 1.0079,
+      "step": 4155
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18202108144760132,
+      "learning_rate": 0.0002934607146828634,
+      "loss": 1.0103,
+      "step": 4160
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17905019223690033,
+      "learning_rate": 0.00029343062837008076,
+      "loss": 0.9744,
+      "step": 4165
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.16872918605804443,
+      "learning_rate": 0.0002934004745540812,
+      "loss": 1.0276,
+      "step": 4170
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1752139776945114,
+      "learning_rate": 0.00029337025324905616,
+      "loss": 0.9822,
+      "step": 4175
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18080003559589386,
+      "learning_rate": 0.0002933399644692287,
+      "loss": 1.0102,
+      "step": 4180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.22213223576545715,
+      "learning_rate": 0.00029330960822885385,
+      "loss": 0.9913,
+      "step": 4185
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18527017533779144,
+      "learning_rate": 0.0002932791845422182,
+      "loss": 0.9575,
+      "step": 4190
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18203973770141602,
+      "learning_rate": 0.00029324869342364014,
+      "loss": 0.9891,
+      "step": 4195
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17969951033592224,
+      "learning_rate": 0.00029321813488746983,
+      "loss": 0.994,
+      "step": 4200
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18661588430404663,
+      "learning_rate": 0.0002931875089480891,
+      "loss": 1.0039,
+      "step": 4205
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18751640617847443,
+      "learning_rate": 0.0002931568156199115,
+      "loss": 0.9442,
+      "step": 4210
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.16617530584335327,
+      "learning_rate": 0.0002931260549173825,
+      "loss": 0.885,
+      "step": 4215
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1970120072364807,
+      "learning_rate": 0.00029309522685497886,
+      "loss": 1.0264,
+      "step": 4220
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1805301010608673,
+      "learning_rate": 0.00029306433144720947,
+      "loss": 1.0025,
+      "step": 4225
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1933399736881256,
+      "learning_rate": 0.0002930333687086147,
+      "loss": 0.9908,
+      "step": 4230
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18400885164737701,
+      "learning_rate": 0.0002930023386537666,
+      "loss": 0.988,
+      "step": 4235
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18429554998874664,
+      "learning_rate": 0.000292971241297269,
+      "loss": 0.9955,
+      "step": 4240
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17350837588310242,
+      "learning_rate": 0.0002929400766537573,
+      "loss": 0.9956,
+      "step": 4245
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17750461399555206,
+      "learning_rate": 0.0002929088447378986,
+      "loss": 0.9968,
+      "step": 4250
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.16923023760318756,
+      "learning_rate": 0.0002928775455643917,
+      "loss": 0.9371,
+      "step": 4255
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17826791107654572,
+      "learning_rate": 0.0002928461791479671,
+      "loss": 0.999,
+      "step": 4260
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1948651373386383,
+      "learning_rate": 0.0002928147455033868,
+      "loss": 0.9618,
+      "step": 4265
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.16941259801387787,
+      "learning_rate": 0.00029278324464544455,
+      "loss": 0.9693,
+      "step": 4270
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1716768741607666,
+      "learning_rate": 0.00029275167658896563,
+      "loss": 1.0408,
+      "step": 4275
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17600005865097046,
+      "learning_rate": 0.00029272004134880714,
+      "loss": 1.0234,
+      "step": 4280
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18008339405059814,
+      "learning_rate": 0.00029268833893985756,
+      "loss": 0.9356,
+      "step": 4285
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18551653623580933,
+      "learning_rate": 0.00029265656937703713,
+      "loss": 0.9183,
+      "step": 4290
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18553385138511658,
+      "learning_rate": 0.0002926247326752977,
+      "loss": 0.9736,
+      "step": 4295
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18531882762908936,
+      "learning_rate": 0.00029259282884962266,
+      "loss": 1.0085,
+      "step": 4300
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17965206503868103,
+      "learning_rate": 0.000292560857915027,
+      "loss": 0.9765,
+      "step": 4305
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1778474599123001,
+      "learning_rate": 0.0002925288198865573,
+      "loss": 0.9646,
+      "step": 4310
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.19227425754070282,
+      "learning_rate": 0.0002924967147792917,
+      "loss": 0.996,
+      "step": 4315
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.20739661157131195,
+      "learning_rate": 0.00029246454260833997,
+      "loss": 0.9635,
+      "step": 4320
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17460639774799347,
+      "learning_rate": 0.0002924323033888434,
+      "loss": 0.9606,
+      "step": 4325
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.20777195692062378,
+      "learning_rate": 0.0002923999971359748,
+      "loss": 1.0319,
+      "step": 4330
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.19790644943714142,
+      "learning_rate": 0.00029236762386493863,
+      "loss": 1.0417,
+      "step": 4335
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17253664135932922,
+      "learning_rate": 0.0002923351835909707,
+      "loss": 0.9815,
+      "step": 4340
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18922512233257294,
+      "learning_rate": 0.0002923026763293387,
+      "loss": 1.0032,
+      "step": 4345
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1988437920808792,
+      "learning_rate": 0.0002922701020953414,
+      "loss": 0.9955,
+      "step": 4350
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.19901318848133087,
+      "learning_rate": 0.00029223746090430944,
+      "loss": 0.9678,
+      "step": 4355
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1846579760313034,
+      "learning_rate": 0.0002922047527716048,
+      "loss": 0.8918,
+      "step": 4360
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18826933205127716,
+      "learning_rate": 0.000292171977712621,
+      "loss": 0.9736,
+      "step": 4365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1800205409526825,
+      "learning_rate": 0.00029213913574278324,
+      "loss": 0.9471,
+      "step": 4370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.184101402759552,
+      "learning_rate": 0.00029210622687754777,
+      "loss": 1.0363,
+      "step": 4375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18093040585517883,
+      "learning_rate": 0.0002920732511324028,
+      "loss": 0.8811,
+      "step": 4380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.16744713485240936,
+      "learning_rate": 0.0002920402085228677,
+      "loss": 0.9148,
+      "step": 4385
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3018034100532532,
+      "learning_rate": 0.0002920070990644935,
+      "loss": 1.0029,
+      "step": 4390
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2019168734550476,
+      "learning_rate": 0.0002919739227728625,
+      "loss": 0.8603,
+      "step": 4395
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.20707763731479645,
+      "learning_rate": 0.0002919406796635887,
+      "loss": 1.0115,
+      "step": 4400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18676406145095825,
+      "learning_rate": 0.00029190736975231736,
+      "loss": 1.0022,
+      "step": 4405
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1945050209760666,
+      "learning_rate": 0.00029187399305472515,
+      "loss": 1.068,
+      "step": 4410
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.17335373163223267,
+      "learning_rate": 0.0002918405495865203,
+      "loss": 0.9242,
+      "step": 4415
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.17637088894844055,
+      "learning_rate": 0.0002918070393634425,
+      "loss": 0.9241,
+      "step": 4420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18394945561885834,
+      "learning_rate": 0.00029177346240126273,
+      "loss": 1.0194,
+      "step": 4425
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.21276423335075378,
+      "learning_rate": 0.0002917398187157834,
+      "loss": 1.0077,
+      "step": 4430
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.19257494807243347,
+      "learning_rate": 0.0002917061083228383,
+      "loss": 0.982,
+      "step": 4435
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2055777907371521,
+      "learning_rate": 0.0002916723312382927,
+      "loss": 1.0397,
+      "step": 4440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18082697689533234,
+      "learning_rate": 0.00029163848747804327,
+      "loss": 0.963,
+      "step": 4445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18980878591537476,
+      "learning_rate": 0.00029160457705801796,
+      "loss": 0.9314,
+      "step": 4450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1710338592529297,
+      "learning_rate": 0.0002915705999941761,
+      "loss": 1.0261,
+      "step": 4455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17998263239860535,
+      "learning_rate": 0.0002915365563025085,
+      "loss": 0.9295,
+      "step": 4460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19266332685947418,
+      "learning_rate": 0.00029150244599903725,
+      "loss": 0.8856,
+      "step": 4465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1885053515434265,
+      "learning_rate": 0.0002914682690998157,
+      "loss": 0.9803,
+      "step": 4470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19506719708442688,
+      "learning_rate": 0.00029143402562092875,
+      "loss": 0.9873,
+      "step": 4475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.20998205244541168,
+      "learning_rate": 0.0002913997155784924,
+      "loss": 0.9647,
+      "step": 4480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.21117228269577026,
+      "learning_rate": 0.00029136533898865423,
+      "loss": 0.9814,
+      "step": 4485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18329034745693207,
+      "learning_rate": 0.0002913308958675929,
+      "loss": 0.9783,
+      "step": 4490
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18641537427902222,
+      "learning_rate": 0.0002912963862315185,
+      "loss": 0.8872,
+      "step": 4495
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19384218752384186,
+      "learning_rate": 0.0002912618100966725,
+      "loss": 0.9641,
+      "step": 4500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19041642546653748,
+      "learning_rate": 0.00029122716747932747,
+      "loss": 0.9446,
+      "step": 4505
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.20249037444591522,
+      "learning_rate": 0.0002911924583957874,
+      "loss": 0.9998,
+      "step": 4510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1947818547487259,
+      "learning_rate": 0.00029115768286238757,
+      "loss": 0.9414,
+      "step": 4515
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1864546537399292,
+      "learning_rate": 0.00029112284089549445,
+      "loss": 1.0308,
+      "step": 4520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18546593189239502,
+      "learning_rate": 0.0002910879325115059,
+      "loss": 0.9426,
+      "step": 4525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.19408635795116425,
+      "learning_rate": 0.0002910529577268509,
+      "loss": 0.9535,
+      "step": 4530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.19161821901798248,
+      "learning_rate": 0.0002910179165579898,
+      "loss": 0.9453,
+      "step": 4535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1930609792470932,
+      "learning_rate": 0.00029098280902141406,
+      "loss": 1.051,
+      "step": 4540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.17416320741176605,
+      "learning_rate": 0.0002909476351336465,
+      "loss": 0.9667,
+      "step": 4545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1923515945672989,
+      "learning_rate": 0.0002909123949112412,
+      "loss": 0.9336,
+      "step": 4550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.19930413365364075,
+      "learning_rate": 0.0002908770883707832,
+      "loss": 1.032,
+      "step": 4555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.19108958542346954,
+      "learning_rate": 0.00029084171552888914,
+      "loss": 0.9452,
+      "step": 4560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.20645657181739807,
+      "learning_rate": 0.00029080627640220647,
+      "loss": 0.9045,
+      "step": 4565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.21397635340690613,
+      "learning_rate": 0.0002907707710074141,
+      "loss": 0.8976,
+      "step": 4570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1848032921552658,
+      "learning_rate": 0.000290735199361222,
+      "loss": 0.8218,
+      "step": 4575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18126817047595978,
+      "learning_rate": 0.0002906995614803715,
+      "loss": 1.0663,
+      "step": 4580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1907825469970703,
+      "learning_rate": 0.0002906638573816348,
+      "loss": 0.9277,
+      "step": 4585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18289636075496674,
+      "learning_rate": 0.0002906280870818156,
+      "loss": 1.0149,
+      "step": 4590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.17690010368824005,
+      "learning_rate": 0.00029059225059774844,
+      "loss": 1.0116,
+      "step": 4595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.19581712782382965,
+      "learning_rate": 0.00029055634794629924,
+      "loss": 0.9966,
+      "step": 4600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.20098170638084412,
+      "learning_rate": 0.00029052037914436494,
+      "loss": 1.0375,
+      "step": 4605
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.20016787946224213,
+      "learning_rate": 0.00029048434420887373,
+      "loss": 1.0045,
+      "step": 4610
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1856658011674881,
+      "learning_rate": 0.0002904482431567847,
+      "loss": 1.0191,
+      "step": 4615
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.19540125131607056,
+      "learning_rate": 0.0002904120760050884,
+      "loss": 1.0104,
+      "step": 4620
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.20216308534145355,
+      "learning_rate": 0.00029037584277080616,
+      "loss": 1.0659,
+      "step": 4625
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.224759042263031,
+      "learning_rate": 0.00029033954347099057,
+      "loss": 0.9496,
+      "step": 4630
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.16096115112304688,
+      "learning_rate": 0.0002903031781227253,
+      "loss": 0.9887,
+      "step": 4635
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.20530791580677032,
+      "learning_rate": 0.00029026674674312503,
+      "loss": 0.9417,
+      "step": 4640
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17820000648498535,
+      "learning_rate": 0.0002902302493493357,
+      "loss": 0.9157,
+      "step": 4645
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.20749807357788086,
+      "learning_rate": 0.00029019368595853407,
+      "loss": 0.958,
+      "step": 4650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.19266118109226227,
+      "learning_rate": 0.00029015705658792817,
+      "loss": 1.05,
+      "step": 4655
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2041333168745041,
+      "learning_rate": 0.00029012036125475695,
+      "loss": 0.9891,
+      "step": 4660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.19730497896671295,
+      "learning_rate": 0.00029008359997629045,
+      "loss": 1.0394,
+      "step": 4665
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17330877482891083,
+      "learning_rate": 0.00029004677276982986,
+      "loss": 1.0221,
+      "step": 4670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1951434314250946,
+      "learning_rate": 0.0002900098796527071,
+      "loss": 1.0018,
+      "step": 4675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.21202173829078674,
+      "learning_rate": 0.00028997292064228544,
+      "loss": 0.9171,
+      "step": 4680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.19243790209293365,
+      "learning_rate": 0.00028993589575595894,
+      "loss": 0.9833,
+      "step": 4685
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1826474815607071,
+      "learning_rate": 0.00028989880501115276,
+      "loss": 0.875,
+      "step": 4690
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.20149312913417816,
+      "learning_rate": 0.0002898616484253231,
+      "loss": 1.0055,
+      "step": 4695
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.19697706401348114,
+      "learning_rate": 0.000289824426015957,
+      "loss": 0.9464,
+      "step": 4700
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.19809427857398987,
+      "learning_rate": 0.00028978713780057256,
+      "loss": 0.9814,
+      "step": 4705
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.206287682056427,
+      "learning_rate": 0.00028974978379671894,
+      "loss": 0.9652,
+      "step": 4710
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18428125977516174,
+      "learning_rate": 0.0002897123640219761,
+      "loss": 0.942,
+      "step": 4715
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.16364052891731262,
+      "learning_rate": 0.000289674878493955,
+      "loss": 0.9741,
+      "step": 4720
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.20780855417251587,
+      "learning_rate": 0.0002896373272302977,
+      "loss": 0.9654,
+      "step": 4725
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2168758362531662,
+      "learning_rate": 0.0002895997102486769,
+      "loss": 0.9247,
+      "step": 4730
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1811356246471405,
+      "learning_rate": 0.0002895620275667966,
+      "loss": 0.9142,
+      "step": 4735
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18936580419540405,
+      "learning_rate": 0.00028952427920239134,
+      "loss": 0.9091,
+      "step": 4740
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.19724291563034058,
+      "learning_rate": 0.00028948646517322686,
+      "loss": 0.9266,
+      "step": 4745
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.17332862317562103,
+      "learning_rate": 0.0002894485854970997,
+      "loss": 1.0734,
+      "step": 4750
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1730174571275711,
+      "learning_rate": 0.00028941064019183713,
+      "loss": 1.0088,
+      "step": 4755
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18203985691070557,
+      "learning_rate": 0.0002893726292752977,
+      "loss": 0.9595,
+      "step": 4760
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18378064036369324,
+      "learning_rate": 0.0002893345527653705,
+      "loss": 0.9354,
+      "step": 4765
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18141499161720276,
+      "learning_rate": 0.0002892964106799757,
+      "loss": 0.9606,
+      "step": 4770
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.19755259156227112,
+      "learning_rate": 0.000289258203037064,
+      "loss": 1.054,
+      "step": 4775
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.213453009724617,
+      "learning_rate": 0.0002892199298546174,
+      "loss": 0.8363,
+      "step": 4780
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18598608672618866,
+      "learning_rate": 0.00028918159115064846,
+      "loss": 0.9375,
+      "step": 4785
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.19149407744407654,
+      "learning_rate": 0.0002891431869432006,
+      "loss": 0.97,
+      "step": 4790
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.20365601778030396,
+      "learning_rate": 0.0002891047172503482,
+      "loss": 0.9405,
+      "step": 4795
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18206772208213806,
+      "learning_rate": 0.0002890661820901963,
+      "loss": 1.011,
+      "step": 4800
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18795382976531982,
+      "learning_rate": 0.00028902758148088094,
+      "loss": 0.854,
+      "step": 4805
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18891260027885437,
+      "learning_rate": 0.00028898891544056873,
+      "loss": 0.9468,
+      "step": 4810
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.19970501959323883,
+      "learning_rate": 0.0002889501839874572,
+      "loss": 0.9918,
+      "step": 4815
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.20287778973579407,
+      "learning_rate": 0.00028891138713977476,
+      "loss": 0.9039,
+      "step": 4820
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17449329793453217,
+      "learning_rate": 0.0002888725249157804,
+      "loss": 0.928,
+      "step": 4825
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.22723489999771118,
+      "learning_rate": 0.000288833597333764,
+      "loss": 0.9622,
+      "step": 4830
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17067822813987732,
+      "learning_rate": 0.0002887946044120461,
+      "loss": 0.9349,
+      "step": 4835
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.19400516152381897,
+      "learning_rate": 0.00028875554616897823,
+      "loss": 0.9223,
+      "step": 4840
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1980266571044922,
+      "learning_rate": 0.00028871642262294234,
+      "loss": 0.9518,
+      "step": 4845
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1934380978345871,
+      "learning_rate": 0.0002886772337923514,
+      "loss": 1.0257,
+      "step": 4850
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1748974621295929,
+      "learning_rate": 0.0002886379796956489,
+      "loss": 1.0209,
+      "step": 4855
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17561843991279602,
+      "learning_rate": 0.0002885986603513091,
+      "loss": 0.9646,
+      "step": 4860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19227540493011475,
+      "learning_rate": 0.00028855927577783706,
+      "loss": 0.9638,
+      "step": 4865
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.20389790832996368,
+      "learning_rate": 0.0002885198259937684,
+      "loss": 0.9399,
+      "step": 4870
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18054336309432983,
+      "learning_rate": 0.0002884803110176695,
+      "loss": 0.9318,
+      "step": 4875
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1733674854040146,
+      "learning_rate": 0.00028844073086813744,
+      "loss": 0.8768,
+      "step": 4880
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.185361847281456,
+      "learning_rate": 0.00028840108556380006,
+      "loss": 0.9853,
+      "step": 4885
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1938917487859726,
+      "learning_rate": 0.00028836137512331555,
+      "loss": 0.9558,
+      "step": 4890
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18514370918273926,
+      "learning_rate": 0.00028832159956537306,
+      "loss": 0.9225,
+      "step": 4895
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1769241839647293,
+      "learning_rate": 0.0002882817589086924,
+      "loss": 1.0125,
+      "step": 4900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18880169093608856,
+      "learning_rate": 0.0002882418531720237,
+      "loss": 0.9748,
+      "step": 4905
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.20545773208141327,
+      "learning_rate": 0.0002882018823741481,
+      "loss": 0.972,
+      "step": 4910
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17919191718101501,
+      "learning_rate": 0.00028816184653387706,
+      "loss": 0.9201,
+      "step": 4915
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17469099164009094,
+      "learning_rate": 0.00028812174567005285,
+      "loss": 0.9619,
+      "step": 4920
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1765175461769104,
+      "learning_rate": 0.0002880815798015483,
+      "loss": 0.9707,
+      "step": 4925
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18801696598529816,
+      "learning_rate": 0.00028804134894726665,
+      "loss": 0.9369,
+      "step": 4930
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19199238717556,
+      "learning_rate": 0.00028800105312614196,
+      "loss": 0.8942,
+      "step": 4935
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17167691886425018,
+      "learning_rate": 0.00028796069235713893,
+      "loss": 0.8596,
+      "step": 4940
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18981756269931793,
+      "learning_rate": 0.00028792026665925245,
+      "loss": 1.0821,
+      "step": 4945
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18595005571842194,
+      "learning_rate": 0.0002878797760515083,
+      "loss": 1.0417,
+      "step": 4950
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.20380659401416779,
+      "learning_rate": 0.0002878392205529627,
+      "loss": 0.9396,
+      "step": 4955
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.21177223324775696,
+      "learning_rate": 0.0002877986001827024,
+      "loss": 0.9777,
+      "step": 4960
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18329915404319763,
+      "learning_rate": 0.00028775791495984474,
+      "loss": 1.068,
+      "step": 4965
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2002885490655899,
+      "learning_rate": 0.0002877171649035375,
+      "loss": 0.9144,
+      "step": 4970
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18435192108154297,
+      "learning_rate": 0.00028767635003295895,
+      "loss": 0.8893,
+      "step": 4975
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18662065267562866,
+      "learning_rate": 0.000287635470367318,
+      "loss": 0.9352,
+      "step": 4980
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19780193269252777,
+      "learning_rate": 0.000287594525925854,
+      "loss": 0.9593,
+      "step": 4985
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18069545924663544,
+      "learning_rate": 0.0002875535167278367,
+      "loss": 1.0398,
+      "step": 4990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1925516277551651,
+      "learning_rate": 0.00028751244279256647,
+      "loss": 0.9382,
+      "step": 4995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1789090633392334,
+      "learning_rate": 0.0002874713041393739,
+      "loss": 1.0408,
+      "step": 5000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.20351684093475342,
+      "learning_rate": 0.0002874301007876204,
+      "loss": 1.0124,
+      "step": 5005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.20231549441814423,
+      "learning_rate": 0.00028738883275669755,
+      "loss": 1.0305,
+      "step": 5010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17243053019046783,
+      "learning_rate": 0.00028734750006602746,
+      "loss": 0.9991,
+      "step": 5015
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17361502349376678,
+      "learning_rate": 0.0002873061027350627,
+      "loss": 0.971,
+      "step": 5020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1816028356552124,
+      "learning_rate": 0.00028726464078328615,
+      "loss": 0.9863,
+      "step": 5025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18330328166484833,
+      "learning_rate": 0.00028722311423021125,
+      "loss": 1.0478,
+      "step": 5030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.20598195493221283,
+      "learning_rate": 0.00028718152309538175,
+      "loss": 0.9304,
+      "step": 5035
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.19323958456516266,
+      "learning_rate": 0.00028713986739837183,
+      "loss": 0.9557,
+      "step": 5040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.17913362383842468,
+      "learning_rate": 0.0002870981471587861,
+      "loss": 0.8907,
+      "step": 5045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1934191733598709,
+      "learning_rate": 0.0002870563623962593,
+      "loss": 0.9355,
+      "step": 5050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1913587599992752,
+      "learning_rate": 0.00028701451313045695,
+      "loss": 0.9905,
+      "step": 5055
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.19491787254810333,
+      "learning_rate": 0.0002869725993810746,
+      "loss": 0.9745,
+      "step": 5060
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.20128677785396576,
+      "learning_rate": 0.00028693062116783816,
+      "loss": 0.9302,
+      "step": 5065
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18914251029491425,
+      "learning_rate": 0.00028688857851050416,
+      "loss": 0.891,
+      "step": 5070
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.19719275832176208,
+      "learning_rate": 0.0002868464714288592,
+      "loss": 0.9716,
+      "step": 5075
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1894945651292801,
+      "learning_rate": 0.0002868042999427202,
+      "loss": 0.9554,
+      "step": 5080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1891942322254181,
+      "learning_rate": 0.0002867620640719346,
+      "loss": 0.8918,
+      "step": 5085
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1876845359802246,
+      "learning_rate": 0.00028671976383637984,
+      "loss": 0.98,
+      "step": 5090
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.17926909029483795,
+      "learning_rate": 0.00028667739925596395,
+      "loss": 0.8958,
+      "step": 5095
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.203196182847023,
+      "learning_rate": 0.000286634970350625,
+      "loss": 0.9356,
+      "step": 5100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17668934166431427,
+      "learning_rate": 0.00028659247714033154,
+      "loss": 0.9355,
+      "step": 5105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1789340376853943,
+      "learning_rate": 0.00028654991964508224,
+      "loss": 1.0171,
+      "step": 5110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17405730485916138,
+      "learning_rate": 0.00028650729788490606,
+      "loss": 0.9634,
+      "step": 5115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1780971735715866,
+      "learning_rate": 0.0002864646118798622,
+      "loss": 0.9728,
+      "step": 5120
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2179042249917984,
+      "learning_rate": 0.0002864218616500402,
+      "loss": 1.0805,
+      "step": 5125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.20955954492092133,
+      "learning_rate": 0.00028637904721555966,
+      "loss": 0.9261,
+      "step": 5130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18898724019527435,
+      "learning_rate": 0.00028633616859657045,
+      "loss": 1.0134,
+      "step": 5135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.20803610980510712,
+      "learning_rate": 0.0002862932258132527,
+      "loss": 0.9414,
+      "step": 5140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.21270735561847687,
+      "learning_rate": 0.00028625021888581685,
+      "loss": 0.9779,
+      "step": 5145
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.20544709265232086,
+      "learning_rate": 0.0002862071478345031,
+      "loss": 0.9707,
+      "step": 5150
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.19069305062294006,
+      "learning_rate": 0.00028616401267958237,
+      "loss": 0.9954,
+      "step": 5155
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.19101087749004364,
+      "learning_rate": 0.00028612081344135546,
+      "loss": 0.9348,
+      "step": 5160
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17791791260242462,
+      "learning_rate": 0.0002860775501401532,
+      "loss": 0.9205,
+      "step": 5165
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.21044310927391052,
+      "learning_rate": 0.00028603422279633694,
+      "loss": 0.9649,
+      "step": 5170
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1965063214302063,
+      "learning_rate": 0.0002859908314302978,
+      "loss": 0.9648,
+      "step": 5175
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.19343961775302887,
+      "learning_rate": 0.00028594737606245726,
+      "loss": 0.932,
+      "step": 5180
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18944412469863892,
+      "learning_rate": 0.00028590385671326695,
+      "loss": 0.9144,
+      "step": 5185
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.20036911964416504,
+      "learning_rate": 0.0002858602734032084,
+      "loss": 0.965,
+      "step": 5190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.21332001686096191,
+      "learning_rate": 0.00028581662615279345,
+      "loss": 0.9766,
+      "step": 5195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.17393194139003754,
+      "learning_rate": 0.00028577291498256384,
+      "loss": 0.9954,
+      "step": 5200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1827279031276703,
+      "learning_rate": 0.0002857291399130916,
+      "loss": 0.9646,
+      "step": 5205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.190265491604805,
+      "learning_rate": 0.0002856853009649787,
+      "loss": 1.0151,
+      "step": 5210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.18640612065792084,
+      "learning_rate": 0.0002856413981588572,
+      "loss": 1.0085,
+      "step": 5215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.21478848159313202,
+      "learning_rate": 0.00028559743151538913,
+      "loss": 0.9588,
+      "step": 5220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.18051999807357788,
+      "learning_rate": 0.00028555340105526676,
+      "loss": 0.9699,
+      "step": 5225
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1969325840473175,
+      "learning_rate": 0.0002855093067992123,
+      "loss": 0.9939,
+      "step": 5230
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2040921300649643,
+      "learning_rate": 0.00028546514876797796,
+      "loss": 0.8854,
+      "step": 5235
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1841362863779068,
+      "learning_rate": 0.0002854209269823459,
+      "loss": 0.9255,
+      "step": 5240
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.19539733231067657,
+      "learning_rate": 0.0002853766414631285,
+      "loss": 1.0226,
+      "step": 5245
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.19817125797271729,
+      "learning_rate": 0.0002853322922311678,
+      "loss": 0.9626,
+      "step": 5250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.21229805052280426,
+      "learning_rate": 0.0002852878793073362,
+      "loss": 0.957,
+      "step": 5255
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.19763095676898956,
+      "learning_rate": 0.0002852434027125358,
+      "loss": 0.8457,
+      "step": 5260
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.19969575107097626,
+      "learning_rate": 0.00028519886246769884,
+      "loss": 1.0371,
+      "step": 5265
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1981430947780609,
+      "learning_rate": 0.0002851542585937873,
+      "loss": 0.9438,
+      "step": 5270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.20760881900787354,
+      "learning_rate": 0.0002851095911117934,
+      "loss": 1.0244,
+      "step": 5275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.21313069760799408,
+      "learning_rate": 0.00028506486004273903,
+      "loss": 1.0125,
+      "step": 5280
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18353518843650818,
+      "learning_rate": 0.00028502006540767616,
+      "loss": 0.9716,
+      "step": 5285
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1756833791732788,
+      "learning_rate": 0.0002849752072276867,
+      "loss": 1.0393,
+      "step": 5290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.20792143046855927,
+      "learning_rate": 0.00028493028552388223,
+      "loss": 1.0315,
+      "step": 5295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2053552269935608,
+      "learning_rate": 0.00028488530031740454,
+      "loss": 1.0352,
+      "step": 5300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1828060746192932,
+      "learning_rate": 0.00028484025162942516,
+      "loss": 0.8927,
+      "step": 5305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19744126498699188,
+      "learning_rate": 0.0002847951394811454,
+      "loss": 0.9503,
+      "step": 5310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.22091755270957947,
+      "learning_rate": 0.0002847499638937966,
+      "loss": 0.9118,
+      "step": 5315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19735994935035706,
+      "learning_rate": 0.0002847047248886399,
+      "loss": 0.9983,
+      "step": 5320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.17844171822071075,
+      "learning_rate": 0.00028465942248696624,
+      "loss": 0.9512,
+      "step": 5325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.20808811485767365,
+      "learning_rate": 0.00028461405671009645,
+      "loss": 0.9539,
+      "step": 5330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18508221209049225,
+      "learning_rate": 0.00028456862757938117,
+      "loss": 0.9328,
+      "step": 5335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1841544657945633,
+      "learning_rate": 0.0002845231351162009,
+      "loss": 0.9095,
+      "step": 5340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1901453733444214,
+      "learning_rate": 0.0002844775793419659,
+      "loss": 0.8944,
+      "step": 5345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2000284641981125,
+      "learning_rate": 0.00028443196027811617,
+      "loss": 0.9468,
+      "step": 5350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2096180021762848,
+      "learning_rate": 0.0002843862779461216,
+      "loss": 0.9388,
+      "step": 5355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18580293655395508,
+      "learning_rate": 0.00028434053236748175,
+      "loss": 0.9352,
+      "step": 5360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.21264111995697021,
+      "learning_rate": 0.00028429472356372606,
+      "loss": 1.0138,
+      "step": 5365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19658392667770386,
+      "learning_rate": 0.00028424885155641373,
+      "loss": 0.9843,
+      "step": 5370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19093391299247742,
+      "learning_rate": 0.00028420291636713354,
+      "loss": 0.9271,
+      "step": 5375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.21088851988315582,
+      "learning_rate": 0.00028415691801750417,
+      "loss": 1.0146,
+      "step": 5380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17928850650787354,
+      "learning_rate": 0.000284110856529174,
+      "loss": 0.9973,
+      "step": 5385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18797273933887482,
+      "learning_rate": 0.000284064731923821,
+      "loss": 1.0008,
+      "step": 5390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19398821890354156,
+      "learning_rate": 0.00028401854422315306,
+      "loss": 0.8467,
+      "step": 5395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.20564605295658112,
+      "learning_rate": 0.0002839722934489076,
+      "loss": 0.9809,
+      "step": 5400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2124805599451065,
+      "learning_rate": 0.0002839259796228517,
+      "loss": 0.9094,
+      "step": 5405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.16906094551086426,
+      "learning_rate": 0.0002838796027667823,
+      "loss": 0.9083,
+      "step": 5410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.21367835998535156,
+      "learning_rate": 0.0002838331629025258,
+      "loss": 0.9278,
+      "step": 5415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.20226390659809113,
+      "learning_rate": 0.00028378666005193846,
+      "loss": 0.9629,
+      "step": 5420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.21728989481925964,
+      "learning_rate": 0.0002837400942369059,
+      "loss": 0.952,
+      "step": 5425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19252964854240417,
+      "learning_rate": 0.0002836934654793436,
+      "loss": 0.9497,
+      "step": 5430
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.21155524253845215,
+      "learning_rate": 0.00028364677380119665,
+      "loss": 0.9298,
+      "step": 5435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18080827593803406,
+      "learning_rate": 0.0002836000192244397,
+      "loss": 0.9502,
+      "step": 5440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19508054852485657,
+      "learning_rate": 0.000283553201771077,
+      "loss": 0.9299,
+      "step": 5445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19555960595607758,
+      "learning_rate": 0.00028350632146314234,
+      "loss": 0.8338,
+      "step": 5450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19276316463947296,
+      "learning_rate": 0.00028345937832269924,
+      "loss": 1.0148,
+      "step": 5455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19060495495796204,
+      "learning_rate": 0.0002834123723718406,
+      "loss": 0.9975,
+      "step": 5460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18638461828231812,
+      "learning_rate": 0.00028336530363268903,
+      "loss": 1.1051,
+      "step": 5465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.21244901418685913,
+      "learning_rate": 0.00028331817212739666,
+      "loss": 0.9824,
+      "step": 5470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1781817078590393,
+      "learning_rate": 0.0002832709778781451,
+      "loss": 0.9437,
+      "step": 5475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.21162748336791992,
+      "learning_rate": 0.00028322372090714565,
+      "loss": 0.8513,
+      "step": 5480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.20514048635959625,
+      "learning_rate": 0.00028317640123663886,
+      "loss": 0.9249,
+      "step": 5485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1924484670162201,
+      "learning_rate": 0.000283129018888895,
+      "loss": 0.9966,
+      "step": 5490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2287726253271103,
+      "learning_rate": 0.00028308157388621375,
+      "loss": 1.0219,
+      "step": 5495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1756589561700821,
+      "learning_rate": 0.00028303406625092433,
+      "loss": 0.9488,
+      "step": 5500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.23016351461410522,
+      "learning_rate": 0.00028298649600538546,
+      "loss": 0.9064,
+      "step": 5505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17148062586784363,
+      "learning_rate": 0.0002829388631719852,
+      "loss": 1.0173,
+      "step": 5510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2095944881439209,
+      "learning_rate": 0.0002828911677731411,
+      "loss": 0.9834,
+      "step": 5515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1759607195854187,
+      "learning_rate": 0.00028284340983130037,
+      "loss": 0.935,
+      "step": 5520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2132364809513092,
+      "learning_rate": 0.0002827955893689393,
+      "loss": 0.9313,
+      "step": 5525
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18752391636371613,
+      "learning_rate": 0.00028274770640856394,
+      "loss": 0.8606,
+      "step": 5530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1804865151643753,
+      "learning_rate": 0.00028269976097270946,
+      "loss": 1.0291,
+      "step": 5535
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17835399508476257,
+      "learning_rate": 0.0002826517530839407,
+      "loss": 0.9643,
+      "step": 5540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.21218432486057281,
+      "learning_rate": 0.0002826036827648517,
+      "loss": 0.9989,
+      "step": 5545
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.20228761434555054,
+      "learning_rate": 0.0002825555500380659,
+      "loss": 0.9734,
+      "step": 5550
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19665905833244324,
+      "learning_rate": 0.0002825073549262363,
+      "loss": 0.9381,
+      "step": 5555
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.20172518491744995,
+      "learning_rate": 0.000282459097452045,
+      "loss": 1.0319,
+      "step": 5560
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19320924580097198,
+      "learning_rate": 0.0002824107776382036,
+      "loss": 1.0349,
+      "step": 5565
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1959851086139679,
+      "learning_rate": 0.00028236239550745305,
+      "loss": 0.9762,
+      "step": 5570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18430382013320923,
+      "learning_rate": 0.00028231395108256353,
+      "loss": 1.0082,
+      "step": 5575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18709762394428253,
+      "learning_rate": 0.0002822654443863346,
+      "loss": 0.9993,
+      "step": 5580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.20324410498142242,
+      "learning_rate": 0.00028221687544159515,
+      "loss": 0.9235,
+      "step": 5585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1890849769115448,
+      "learning_rate": 0.0002821682442712033,
+      "loss": 1.0188,
+      "step": 5590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19836723804473877,
+      "learning_rate": 0.00028211955089804664,
+      "loss": 0.9934,
+      "step": 5595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17568111419677734,
+      "learning_rate": 0.00028207079534504166,
+      "loss": 0.8919,
+      "step": 5600
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20651723444461823,
+      "learning_rate": 0.0002820219776351345,
+      "loss": 0.9338,
+      "step": 5605
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18715310096740723,
+      "learning_rate": 0.00028197309779130037,
+      "loss": 0.9869,
+      "step": 5610
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20862345397472382,
+      "learning_rate": 0.0002819241558365437,
+      "loss": 0.9206,
+      "step": 5615
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19458311796188354,
+      "learning_rate": 0.00028187515179389826,
+      "loss": 0.9267,
+      "step": 5620
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.17458118498325348,
+      "learning_rate": 0.0002818260856864269,
+      "loss": 0.879,
+      "step": 5625
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19164584577083588,
+      "learning_rate": 0.0002817769575372218,
+      "loss": 0.9952,
+      "step": 5630
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1988450139760971,
+      "learning_rate": 0.00028172776736940436,
+      "loss": 0.8229,
+      "step": 5635
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2000226527452469,
+      "learning_rate": 0.000281678515206125,
+      "loss": 0.9074,
+      "step": 5640
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.21437396109104156,
+      "learning_rate": 0.0002816292010705635,
+      "loss": 0.9317,
+      "step": 5645
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18313203752040863,
+      "learning_rate": 0.0002815798249859287,
+      "loss": 0.9046,
+      "step": 5650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.17462490499019623,
+      "learning_rate": 0.00028153038697545867,
+      "loss": 1.0031,
+      "step": 5655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2010510116815567,
+      "learning_rate": 0.0002814808870624205,
+      "loss": 0.9259,
+      "step": 5660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.21494421362876892,
+      "learning_rate": 0.00028143132527011055,
+      "loss": 0.961,
+      "step": 5665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20886385440826416,
+      "learning_rate": 0.00028138170162185424,
+      "loss": 0.863,
+      "step": 5670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.204929918050766,
+      "learning_rate": 0.00028133201614100604,
+      "loss": 0.948,
+      "step": 5675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.21892932057380676,
+      "learning_rate": 0.0002812822688509497,
+      "loss": 0.9707,
+      "step": 5680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2091946005821228,
+      "learning_rate": 0.00028123245977509784,
+      "loss": 0.9417,
+      "step": 5685
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21841882169246674,
+      "learning_rate": 0.0002811825889368924,
+      "loss": 1.0073,
+      "step": 5690
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.17423883080482483,
+      "learning_rate": 0.0002811326563598041,
+      "loss": 0.9885,
+      "step": 5695
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.22584623098373413,
+      "learning_rate": 0.0002810826620673329,
+      "loss": 0.9202,
+      "step": 5700
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2022750824689865,
+      "learning_rate": 0.00028103260608300785,
+      "loss": 1.1009,
+      "step": 5705
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.20764531195163727,
+      "learning_rate": 0.00028098248843038694,
+      "loss": 0.9689,
+      "step": 5710
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2145623415708542,
+      "learning_rate": 0.00028093230913305716,
+      "loss": 0.9474,
+      "step": 5715
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1971713900566101,
+      "learning_rate": 0.00028088206821463456,
+      "loss": 0.8979,
+      "step": 5720
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18897688388824463,
+      "learning_rate": 0.00028083176569876426,
+      "loss": 0.9777,
+      "step": 5725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.20765312016010284,
+      "learning_rate": 0.00028078140160912017,
+      "loss": 0.8788,
+      "step": 5730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2097761631011963,
+      "learning_rate": 0.00028073097596940537,
+      "loss": 0.9324,
+      "step": 5735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.19260716438293457,
+      "learning_rate": 0.0002806804888033519,
+      "loss": 0.8419,
+      "step": 5740
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.209737628698349,
+      "learning_rate": 0.0002806299401347206,
+      "loss": 0.8578,
+      "step": 5745
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1828380972146988,
+      "learning_rate": 0.00028057932998730136,
+      "loss": 0.9546,
+      "step": 5750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.19403524696826935,
+      "learning_rate": 0.00028052865838491304,
+      "loss": 1.0046,
+      "step": 5755
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.21876457333564758,
+      "learning_rate": 0.00028047792535140336,
+      "loss": 0.9358,
+      "step": 5760
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.182986319065094,
+      "learning_rate": 0.00028042713091064894,
+      "loss": 0.9362,
+      "step": 5765
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1893165558576584,
+      "learning_rate": 0.00028037627508655546,
+      "loss": 0.9647,
+      "step": 5770
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20949877798557281,
+      "learning_rate": 0.0002803253579030572,
+      "loss": 0.9254,
+      "step": 5775
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20711813867092133,
+      "learning_rate": 0.0002802743793841175,
+      "loss": 0.9423,
+      "step": 5780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20353513956069946,
+      "learning_rate": 0.00028022333955372863,
+      "loss": 0.9021,
+      "step": 5785
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.17357411980628967,
+      "learning_rate": 0.0002801722384359116,
+      "loss": 1.0085,
+      "step": 5790
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20761722326278687,
+      "learning_rate": 0.0002801210760547162,
+      "loss": 0.9346,
+      "step": 5795
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2074555903673172,
+      "learning_rate": 0.00028006985243422133,
+      "loss": 0.9919,
+      "step": 5800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.17981496453285217,
+      "learning_rate": 0.00028001856759853435,
+      "loss": 0.9016,
+      "step": 5805
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19092728197574615,
+      "learning_rate": 0.0002799672215717917,
+      "loss": 0.9511,
+      "step": 5810
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20729589462280273,
+      "learning_rate": 0.0002799158143781585,
+      "loss": 0.9397,
+      "step": 5815
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19325435161590576,
+      "learning_rate": 0.0002798643460418286,
+      "loss": 1.0048,
+      "step": 5820
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2112487405538559,
+      "learning_rate": 0.0002798128165870249,
+      "loss": 1.0018,
+      "step": 5825
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20866331458091736,
+      "learning_rate": 0.0002797612260379987,
+      "loss": 1.0327,
+      "step": 5830
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2057882696390152,
+      "learning_rate": 0.0002797095744190303,
+      "loss": 0.9205,
+      "step": 5835
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22236908972263336,
+      "learning_rate": 0.00027965786175442866,
+      "loss": 0.9787,
+      "step": 5840
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22123856842517853,
+      "learning_rate": 0.0002796060880685315,
+      "loss": 0.9991,
+      "step": 5845
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19182252883911133,
+      "learning_rate": 0.0002795542533857052,
+      "loss": 0.9982,
+      "step": 5850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1859193593263626,
+      "learning_rate": 0.00027950235773034485,
+      "loss": 0.8689,
+      "step": 5855
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.20256547629833221,
+      "learning_rate": 0.0002794504011268744,
+      "loss": 0.9262,
+      "step": 5860
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19035175442695618,
+      "learning_rate": 0.0002793983835997462,
+      "loss": 0.8788,
+      "step": 5865
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22139105200767517,
+      "learning_rate": 0.00027934630517344145,
+      "loss": 0.9814,
+      "step": 5870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19183175265789032,
+      "learning_rate": 0.00027929416587247014,
+      "loss": 0.8917,
+      "step": 5875
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2111898511648178,
+      "learning_rate": 0.00027924196572137055,
+      "loss": 0.9915,
+      "step": 5880
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.21190297603607178,
+      "learning_rate": 0.00027918970474470996,
+      "loss": 1.0738,
+      "step": 5885
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18288253247737885,
+      "learning_rate": 0.00027913738296708404,
+      "loss": 0.9588,
+      "step": 5890
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2102237343788147,
+      "learning_rate": 0.0002790850004131171,
+      "loss": 0.8978,
+      "step": 5895
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.22349348664283752,
+      "learning_rate": 0.0002790325571074623,
+      "loss": 0.8554,
+      "step": 5900
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19603189826011658,
+      "learning_rate": 0.000278980053074801,
+      "loss": 0.8539,
+      "step": 5905
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19490835070610046,
+      "learning_rate": 0.0002789274883398435,
+      "loss": 0.9656,
+      "step": 5910
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19723361730575562,
+      "learning_rate": 0.0002788748629273284,
+      "loss": 0.8647,
+      "step": 5915
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.224490687251091,
+      "learning_rate": 0.00027882217686202304,
+      "loss": 1.0071,
+      "step": 5920
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19954900443553925,
+      "learning_rate": 0.0002787694301687231,
+      "loss": 1.0931,
+      "step": 5925
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.20417670905590057,
+      "learning_rate": 0.00027871662287225316,
+      "loss": 0.9013,
+      "step": 5930
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20717184245586395,
+      "learning_rate": 0.00027866375499746587,
+      "loss": 1.0245,
+      "step": 5935
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1946130394935608,
+      "learning_rate": 0.0002786108265692427,
+      "loss": 1.0055,
+      "step": 5940
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.19675332307815552,
+      "learning_rate": 0.0002785578376124935,
+      "loss": 0.9285,
+      "step": 5945
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20337584614753723,
+      "learning_rate": 0.00027850478815215666,
+      "loss": 0.9615,
+      "step": 5950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2144574224948883,
+      "learning_rate": 0.000278451678213199,
+      "loss": 0.9641,
+      "step": 5955
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.19476327300071716,
+      "learning_rate": 0.0002783985078206158,
+      "loss": 0.8523,
+      "step": 5960
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.21178047358989716,
+      "learning_rate": 0.00027834527699943087,
+      "loss": 0.9104,
+      "step": 5965
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20542141795158386,
+      "learning_rate": 0.00027829198577469636,
+      "loss": 0.9744,
+      "step": 5970
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18979620933532715,
+      "learning_rate": 0.0002782386341714929,
+      "loss": 0.9429,
+      "step": 5975
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18054763972759247,
+      "learning_rate": 0.00027818522221492953,
+      "loss": 0.9203,
+      "step": 5980
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1968124806880951,
+      "learning_rate": 0.0002781317499301437,
+      "loss": 0.974,
+      "step": 5985
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2041236311197281,
+      "learning_rate": 0.0002780782173423012,
+      "loss": 0.9027,
+      "step": 5990
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.19817684590816498,
+      "learning_rate": 0.0002780246244765963,
+      "loss": 0.9473,
+      "step": 5995
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.19076892733573914,
+      "learning_rate": 0.0002779709713582515,
+      "loss": 0.9915,
+      "step": 6000
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1970764398574829,
+      "learning_rate": 0.00027791725801251785,
+      "loss": 0.9481,
+      "step": 6005
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20971329510211945,
+      "learning_rate": 0.00027786348446467453,
+      "loss": 0.996,
+      "step": 6010
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18908850848674774,
+      "learning_rate": 0.00027780965074002925,
+      "loss": 0.9819,
+      "step": 6015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1883561760187149,
+      "learning_rate": 0.0002777557568639179,
+      "loss": 0.915,
+      "step": 6020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.20280209183692932,
+      "learning_rate": 0.0002777018028617047,
+      "loss": 0.8848,
+      "step": 6025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.23292230069637299,
+      "learning_rate": 0.00027764778875878225,
+      "loss": 1.0372,
+      "step": 6030
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.20740561187267303,
+      "learning_rate": 0.0002775937145805714,
+      "loss": 0.8555,
+      "step": 6035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22304105758666992,
+      "learning_rate": 0.0002775395803525211,
+      "loss": 1.0331,
+      "step": 6040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.24742744863033295,
+      "learning_rate": 0.0002774853861001089,
+      "loss": 0.9875,
+      "step": 6045
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19389699399471283,
+      "learning_rate": 0.0002774311318488404,
+      "loss": 0.9054,
+      "step": 6050
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.20694305002689362,
+      "learning_rate": 0.00027737681762424936,
+      "loss": 0.9755,
+      "step": 6055
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.21031732857227325,
+      "learning_rate": 0.0002773224434518978,
+      "loss": 0.9667,
+      "step": 6060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22336505353450775,
+      "learning_rate": 0.0002772680093573762,
+      "loss": 0.9697,
+      "step": 6065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2182660698890686,
+      "learning_rate": 0.0002772135153663029,
+      "loss": 1.0385,
+      "step": 6070
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1784467101097107,
+      "learning_rate": 0.00027715896150432473,
+      "loss": 0.868,
+      "step": 6075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2188597470521927,
+      "learning_rate": 0.0002771043477971164,
+      "loss": 0.8924,
+      "step": 6080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2050686776638031,
+      "learning_rate": 0.000277049674270381,
+      "loss": 0.9832,
+      "step": 6085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1878337264060974,
+      "learning_rate": 0.00027699494094984964,
+      "loss": 0.935,
+      "step": 6090
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.20712505280971527,
+      "learning_rate": 0.00027694014786128175,
+      "loss": 0.9392,
+      "step": 6095
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2205507606267929,
+      "learning_rate": 0.00027688529503046473,
+      "loss": 0.9543,
+      "step": 6100
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.21329660713672638,
+      "learning_rate": 0.0002768303824832141,
+      "loss": 0.9957,
+      "step": 6105
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22187063097953796,
+      "learning_rate": 0.00027677541024537363,
+      "loss": 0.8959,
+      "step": 6110
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2496599406003952,
+      "learning_rate": 0.00027672037834281497,
+      "loss": 0.9699,
+      "step": 6115
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.20457853376865387,
+      "learning_rate": 0.000276665286801438,
+      "loss": 0.9096,
+      "step": 6120
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2040635645389557,
+      "learning_rate": 0.00027661013564717063,
+      "loss": 0.9936,
+      "step": 6125
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.203216552734375,
+      "learning_rate": 0.00027655492490596885,
+      "loss": 0.9246,
+      "step": 6130
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18687255680561066,
+      "learning_rate": 0.0002764996546038167,
+      "loss": 0.8891,
+      "step": 6135
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2031632959842682,
+      "learning_rate": 0.00027644432476672614,
+      "loss": 1.0533,
+      "step": 6140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19905006885528564,
+      "learning_rate": 0.00027638893542073726,
+      "loss": 0.9701,
+      "step": 6145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.21555280685424805,
+      "learning_rate": 0.0002763334865919181,
+      "loss": 0.8389,
+      "step": 6150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2201685756444931,
+      "learning_rate": 0.00027627797830636475,
+      "loss": 0.9749,
+      "step": 6155
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.20311476290225983,
+      "learning_rate": 0.00027622241059020123,
+      "loss": 0.914,
+      "step": 6160
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2248125523328781,
+      "learning_rate": 0.0002761667834695796,
+      "loss": 0.9121,
+      "step": 6165
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19576215744018555,
+      "learning_rate": 0.00027611109697067975,
+      "loss": 0.8951,
+      "step": 6170
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22087323665618896,
+      "learning_rate": 0.00027605535111970974,
+      "loss": 0.8982,
+      "step": 6175
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.19735217094421387,
+      "learning_rate": 0.0002759995459429053,
+      "loss": 1.0688,
+      "step": 6180
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20120544731616974,
+      "learning_rate": 0.00027594368146653013,
+      "loss": 0.8408,
+      "step": 6185
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18868596851825714,
+      "learning_rate": 0.0002758877577168761,
+      "loss": 0.9014,
+      "step": 6190
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19567087292671204,
+      "learning_rate": 0.00027583177472026264,
+      "loss": 0.9612,
+      "step": 6195
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19370317459106445,
+      "learning_rate": 0.00027577573250303725,
+      "loss": 0.9728,
+      "step": 6200
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20244130492210388,
+      "learning_rate": 0.00027571963109157533,
+      "loss": 1.0169,
+      "step": 6205
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19868247210979462,
+      "learning_rate": 0.0002756634705122799,
+      "loss": 0.9771,
+      "step": 6210
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20510122179985046,
+      "learning_rate": 0.00027560725079158214,
+      "loss": 0.958,
+      "step": 6215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.21019263565540314,
+      "learning_rate": 0.00027555097195594086,
+      "loss": 0.9748,
+      "step": 6220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.21073760092258453,
+      "learning_rate": 0.0002754946340318428,
+      "loss": 0.9534,
+      "step": 6225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.23775875568389893,
+      "learning_rate": 0.0002754382370458024,
+      "loss": 0.9429,
+      "step": 6230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19734768569469452,
+      "learning_rate": 0.00027538178102436194,
+      "loss": 0.9563,
+      "step": 6235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20335164666175842,
+      "learning_rate": 0.00027532526599409154,
+      "loss": 0.9426,
+      "step": 6240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.20909333229064941,
+      "learning_rate": 0.000275268691981589,
+      "loss": 0.9798,
+      "step": 6245
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.21021656692028046,
+      "learning_rate": 0.00027521205901348,
+      "loss": 0.9096,
+      "step": 6250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.17861585319042206,
+      "learning_rate": 0.0002751553671164179,
+      "loss": 0.9817,
+      "step": 6255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2392309159040451,
+      "learning_rate": 0.00027509861631708373,
+      "loss": 0.9609,
+      "step": 6260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2181014120578766,
+      "learning_rate": 0.0002750418066421863,
+      "loss": 0.9862,
+      "step": 6265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.21585321426391602,
+      "learning_rate": 0.0002749849381184622,
+      "loss": 0.9432,
+      "step": 6270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.203776016831398,
+      "learning_rate": 0.00027492801077267563,
+      "loss": 0.9256,
+      "step": 6275
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19271011650562286,
+      "learning_rate": 0.00027487102463161835,
+      "loss": 0.9535,
+      "step": 6280
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20827655494213104,
+      "learning_rate": 0.00027481397972211013,
+      "loss": 1.0035,
+      "step": 6285
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2051791101694107,
+      "learning_rate": 0.0002747568760709981,
+      "loss": 0.9953,
+      "step": 6290
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2057846039533615,
+      "learning_rate": 0.0002746997137051571,
+      "loss": 0.9733,
+      "step": 6295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1999838501214981,
+      "learning_rate": 0.00027464249265148966,
+      "loss": 0.9024,
+      "step": 6300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.22202569246292114,
+      "learning_rate": 0.000274585212936926,
+      "loss": 1.0046,
+      "step": 6305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20919722318649292,
+      "learning_rate": 0.00027452787458842376,
+      "loss": 0.9343,
+      "step": 6310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19470633566379547,
+      "learning_rate": 0.0002744704776329683,
+      "loss": 0.9979,
+      "step": 6315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.21001891791820526,
+      "learning_rate": 0.0002744130220975725,
+      "loss": 0.9753,
+      "step": 6320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.21184568107128143,
+      "learning_rate": 0.0002743555080092769,
+      "loss": 1.0369,
+      "step": 6325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18605372309684753,
+      "learning_rate": 0.00027429793539514953,
+      "loss": 0.9481,
+      "step": 6330
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.23986585438251495,
+      "learning_rate": 0.0002742403042822859,
+      "loss": 0.9951,
+      "step": 6335
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20543424785137177,
+      "learning_rate": 0.00027418261469780924,
+      "loss": 0.9394,
+      "step": 6340
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2040044069290161,
+      "learning_rate": 0.00027412486666887007,
+      "loss": 0.9354,
+      "step": 6345
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.22892433404922485,
+      "learning_rate": 0.0002740670602226466,
+      "loss": 0.9778,
+      "step": 6350
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19521504640579224,
+      "learning_rate": 0.00027400919538634444,
+      "loss": 0.958,
+      "step": 6355
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.23373307287693024,
+      "learning_rate": 0.00027395127218719675,
+      "loss": 0.9722,
+      "step": 6360
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1768098771572113,
+      "learning_rate": 0.0002738932906524641,
+      "loss": 0.9762,
+      "step": 6365
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.22005042433738708,
+      "learning_rate": 0.00027383525080943447,
+      "loss": 0.8488,
+      "step": 6370
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1887035220861435,
+      "learning_rate": 0.00027377715268542334,
+      "loss": 0.9401,
+      "step": 6375
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.192814439535141,
+      "learning_rate": 0.0002737189963077737,
+      "loss": 0.8912,
+      "step": 6380
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2182081937789917,
+      "learning_rate": 0.00027366078170385573,
+      "loss": 0.9403,
+      "step": 6385
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1897241324186325,
+      "learning_rate": 0.0002736025089010673,
+      "loss": 0.9351,
+      "step": 6390
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.21122263371944427,
+      "learning_rate": 0.0002735441779268335,
+      "loss": 1.0048,
+      "step": 6395
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19820435345172882,
+      "learning_rate": 0.00027348578880860677,
+      "loss": 0.9342,
+      "step": 6400
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1979910135269165,
+      "learning_rate": 0.0002734273415738669,
+      "loss": 0.9044,
+      "step": 6405
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19765061140060425,
+      "learning_rate": 0.0002733688362501213,
+      "loss": 1.0196,
+      "step": 6410
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19544093310832977,
+      "learning_rate": 0.0002733102728649044,
+      "loss": 0.8445,
+      "step": 6415
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.20427776873111725,
+      "learning_rate": 0.00027325165144577804,
+      "loss": 0.924,
+      "step": 6420
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.26921671628952026,
+      "learning_rate": 0.0002731929720203315,
+      "loss": 0.9489,
+      "step": 6425
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.21001920104026794,
+      "learning_rate": 0.00027313423461618116,
+      "loss": 0.9385,
+      "step": 6430
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.21902385354042053,
+      "learning_rate": 0.0002730754392609708,
+      "loss": 0.9673,
+      "step": 6435
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22075380384922028,
+      "learning_rate": 0.0002730165859823716,
+      "loss": 0.8973,
+      "step": 6440
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22786179184913635,
+      "learning_rate": 0.0002729576748080818,
+      "loss": 0.9234,
+      "step": 6445
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.20489779114723206,
+      "learning_rate": 0.00027289870576582695,
+      "loss": 0.92,
+      "step": 6450
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19824951887130737,
+      "learning_rate": 0.0002728396788833598,
+      "loss": 0.9119,
+      "step": 6455
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.23278149962425232,
+      "learning_rate": 0.0002727805941884603,
+      "loss": 1.0518,
+      "step": 6460
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22010761499404907,
+      "learning_rate": 0.00027272145170893585,
+      "loss": 0.9255,
+      "step": 6465
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22987917065620422,
+      "learning_rate": 0.00027266225147262073,
+      "loss": 0.9481,
+      "step": 6470
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19295156002044678,
+      "learning_rate": 0.00027260299350737656,
+      "loss": 0.9497,
+      "step": 6475
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18979798257350922,
+      "learning_rate": 0.0002725436778410922,
+      "loss": 0.8827,
+      "step": 6480
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.21442236006259918,
+      "learning_rate": 0.00027248430450168345,
+      "loss": 0.8832,
+      "step": 6485
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2078474760055542,
+      "learning_rate": 0.0002724248735170934,
+      "loss": 0.9101,
+      "step": 6490
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1883726716041565,
+      "learning_rate": 0.00027236538491529235,
+      "loss": 0.8445,
+      "step": 6495
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1775909662246704,
+      "learning_rate": 0.0002723058387242775,
+      "loss": 0.9911,
+      "step": 6500
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2114265412092209,
+      "learning_rate": 0.0002722462349720733,
+      "loss": 1.0032,
+      "step": 6505
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18201524019241333,
+      "learning_rate": 0.0002721865736867312,
+      "loss": 1.0234,
+      "step": 6510
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2337150126695633,
+      "learning_rate": 0.00027212685489632986,
+      "loss": 0.9353,
+      "step": 6515
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19889463484287262,
+      "learning_rate": 0.0002720670786289749,
+      "loss": 0.8353,
+      "step": 6520
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19570806622505188,
+      "learning_rate": 0.000272007244912799,
+      "loss": 0.9103,
+      "step": 6525
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.216067835688591,
+      "learning_rate": 0.0002719473537759619,
+      "loss": 0.9829,
+      "step": 6530
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.23183125257492065,
+      "learning_rate": 0.0002718874052466504,
+      "loss": 0.9251,
+      "step": 6535
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.22221823036670685,
+      "learning_rate": 0.00027182739935307826,
+      "loss": 0.9561,
+      "step": 6540
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.22014914453029633,
+      "learning_rate": 0.00027176733612348616,
+      "loss": 0.9561,
+      "step": 6545
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20240789651870728,
+      "learning_rate": 0.0002717072155861419,
+      "loss": 0.9529,
+      "step": 6550
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.181975319981575,
+      "learning_rate": 0.0002716470377693403,
+      "loss": 0.8618,
+      "step": 6555
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18764451146125793,
+      "learning_rate": 0.0002715868027014029,
+      "loss": 1.0196,
+      "step": 6560
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20951347053050995,
+      "learning_rate": 0.0002715265104106784,
+      "loss": 0.952,
+      "step": 6565
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2056897133588791,
+      "learning_rate": 0.0002714661609255423,
+      "loss": 0.9513,
+      "step": 6570
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.22014079988002777,
+      "learning_rate": 0.0002714057542743971,
+      "loss": 0.9351,
+      "step": 6575
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20651625096797943,
+      "learning_rate": 0.00027134529048567223,
+      "loss": 0.9362,
+      "step": 6580
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2232450395822525,
+      "learning_rate": 0.00027128476958782386,
+      "loss": 1.0102,
+      "step": 6585
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2110331654548645,
+      "learning_rate": 0.00027122419160933515,
+      "loss": 0.9414,
+      "step": 6590
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.20717830955982208,
+      "learning_rate": 0.0002711635565787162,
+      "loss": 0.9147,
+      "step": 6595
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.20119911432266235,
+      "learning_rate": 0.00027110286452450375,
+      "loss": 0.892,
+      "step": 6600
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22192524373531342,
+      "learning_rate": 0.0002710421154752616,
+      "loss": 0.9916,
+      "step": 6605
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19518253207206726,
+      "learning_rate": 0.0002709813094595802,
+      "loss": 0.8616,
+      "step": 6610
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.21804095804691315,
+      "learning_rate": 0.0002709204465060769,
+      "loss": 0.939,
+      "step": 6615
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.23485125601291656,
+      "learning_rate": 0.00027085952664339586,
+      "loss": 0.9235,
+      "step": 6620
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.21770937740802765,
+      "learning_rate": 0.00027079854990020793,
+      "loss": 0.9558,
+      "step": 6625
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19482004642486572,
+      "learning_rate": 0.00027073751630521083,
+      "loss": 0.9604,
+      "step": 6630
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1984790414571762,
+      "learning_rate": 0.000270676425887129,
+      "loss": 0.9706,
+      "step": 6635
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1953192949295044,
+      "learning_rate": 0.0002706152786747136,
+      "loss": 1.0016,
+      "step": 6640
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19475625455379486,
+      "learning_rate": 0.00027055407469674255,
+      "loss": 0.9146,
+      "step": 6645
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2099728286266327,
+      "learning_rate": 0.00027049281398202046,
+      "loss": 0.9263,
+      "step": 6650
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2410629391670227,
+      "learning_rate": 0.00027043149655937864,
+      "loss": 0.9945,
+      "step": 6655
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.21162976324558258,
+      "learning_rate": 0.0002703701224576752,
+      "loss": 0.8995,
+      "step": 6660
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.20428724586963654,
+      "learning_rate": 0.0002703086917057947,
+      "loss": 0.899,
+      "step": 6665
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22366644442081451,
+      "learning_rate": 0.0002702472043326486,
+      "loss": 1.0311,
+      "step": 6670
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.20753750205039978,
+      "learning_rate": 0.00027018566036717483,
+      "loss": 0.9162,
+      "step": 6675
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.20095977187156677,
+      "learning_rate": 0.00027012405983833806,
+      "loss": 0.9609,
+      "step": 6680
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1888018697500229,
+      "learning_rate": 0.00027006240277512955,
+      "loss": 0.9741,
+      "step": 6685
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.21836860477924347,
+      "learning_rate": 0.0002700006892065671,
+      "loss": 0.9016,
+      "step": 6690
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.23613397777080536,
+      "learning_rate": 0.0002699389191616952,
+      "loss": 0.8804,
+      "step": 6695
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2188320904970169,
+      "learning_rate": 0.00026987709266958497,
+      "loss": 0.8731,
+      "step": 6700
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.21378405392169952,
+      "learning_rate": 0.0002698152097593339,
+      "loss": 1.0018,
+      "step": 6705
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2068474143743515,
+      "learning_rate": 0.0002697532704600662,
+      "loss": 0.8768,
+      "step": 6710
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.20965330302715302,
+      "learning_rate": 0.0002696912748009325,
+      "loss": 1.071,
+      "step": 6715
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2051534354686737,
+      "learning_rate": 0.00026962922281111,
+      "loss": 0.8599,
+      "step": 6720
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.20214852690696716,
+      "learning_rate": 0.0002695671145198026,
+      "loss": 0.974,
+      "step": 6725
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.23140518367290497,
+      "learning_rate": 0.00026950494995624035,
+      "loss": 0.8391,
+      "step": 6730
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.22693967819213867,
+      "learning_rate": 0.00026944272914968,
+      "loss": 0.9452,
+      "step": 6735
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19972123205661774,
+      "learning_rate": 0.00026938045212940477,
+      "loss": 0.8612,
+      "step": 6740
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1956896334886551,
+      "learning_rate": 0.00026931811892472423,
+      "loss": 0.9701,
+      "step": 6745
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.21115869283676147,
+      "learning_rate": 0.00026925572956497455,
+      "loss": 0.8717,
+      "step": 6750
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.22253437340259552,
+      "learning_rate": 0.00026919328407951814,
+      "loss": 0.9054,
+      "step": 6755
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2643227279186249,
+      "learning_rate": 0.0002691307824977439,
+      "loss": 0.961,
+      "step": 6760
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20802360773086548,
+      "learning_rate": 0.0002690682248490673,
+      "loss": 0.8992,
+      "step": 6765
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.25810569524765015,
+      "learning_rate": 0.00026900561116292995,
+      "loss": 0.9011,
+      "step": 6770
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20976845920085907,
+      "learning_rate": 0.0002689429414687999,
+      "loss": 0.9485,
+      "step": 6775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.17168466746807098,
+      "learning_rate": 0.0002688802157961716,
+      "loss": 1.0081,
+      "step": 6780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21908418834209442,
+      "learning_rate": 0.00026881743417456585,
+      "loss": 0.9562,
+      "step": 6785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2043706625699997,
+      "learning_rate": 0.0002687545966335298,
+      "loss": 1.0096,
+      "step": 6790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20396995544433594,
+      "learning_rate": 0.00026869170320263685,
+      "loss": 1.0249,
+      "step": 6795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2096068263053894,
+      "learning_rate": 0.00026862875391148676,
+      "loss": 1.0383,
+      "step": 6800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.23566322028636932,
+      "learning_rate": 0.0002685657487897055,
+      "loss": 0.9589,
+      "step": 6805
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20193132758140564,
+      "learning_rate": 0.0002685026878669455,
+      "loss": 0.9232,
+      "step": 6810
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21133491396903992,
+      "learning_rate": 0.0002684395711728851,
+      "loss": 0.8744,
+      "step": 6815
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2161945253610611,
+      "learning_rate": 0.00026837639873722934,
+      "loss": 0.9788,
+      "step": 6820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21006715297698975,
+      "learning_rate": 0.0002683131705897092,
+      "loss": 0.9007,
+      "step": 6825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19389329850673676,
+      "learning_rate": 0.0002682498867600819,
+      "loss": 0.9501,
+      "step": 6830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2149161845445633,
+      "learning_rate": 0.00026818654727813086,
+      "loss": 0.9366,
+      "step": 6835
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19821734726428986,
+      "learning_rate": 0.0002681231521736659,
+      "loss": 0.9437,
+      "step": 6840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21618692576885223,
+      "learning_rate": 0.00026805970147652277,
+      "loss": 0.8672,
+      "step": 6845
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21498163044452667,
+      "learning_rate": 0.00026799619521656346,
+      "loss": 1.0109,
+      "step": 6850
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2292798012495041,
+      "learning_rate": 0.00026793263342367616,
+      "loss": 0.9201,
+      "step": 6855
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19892308115959167,
+      "learning_rate": 0.0002678690161277752,
+      "loss": 0.9206,
+      "step": 6860
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18660472333431244,
+      "learning_rate": 0.00026780534335880084,
+      "loss": 0.9114,
+      "step": 6865
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2231675237417221,
+      "learning_rate": 0.00026774161514671975,
+      "loss": 0.9933,
+      "step": 6870
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2067699432373047,
+      "learning_rate": 0.0002676778315215245,
+      "loss": 1.0153,
+      "step": 6875
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19282564520835876,
+      "learning_rate": 0.00026761399251323375,
+      "loss": 0.8656,
+      "step": 6880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21129785478115082,
+      "learning_rate": 0.0002675500981518923,
+      "loss": 0.9499,
+      "step": 6885
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19059066474437714,
+      "learning_rate": 0.0002674861484675709,
+      "loss": 1.0097,
+      "step": 6890
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.23170576989650726,
+      "learning_rate": 0.0002674221434903665,
+      "loss": 0.8877,
+      "step": 6895
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21638576686382294,
+      "learning_rate": 0.0002673580832504019,
+      "loss": 0.987,
+      "step": 6900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.22118517756462097,
+      "learning_rate": 0.00026729396777782597,
+      "loss": 0.9849,
+      "step": 6905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.20184585452079773,
+      "learning_rate": 0.0002672297971028136,
+      "loss": 0.9106,
+      "step": 6910
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21392621099948883,
+      "learning_rate": 0.0002671655712555656,
+      "loss": 1.0239,
+      "step": 6915
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2100532501935959,
+      "learning_rate": 0.0002671012902663089,
+      "loss": 0.994,
+      "step": 6920
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.21326220035552979,
+      "learning_rate": 0.0002670369541652961,
+      "loss": 0.9117,
+      "step": 6925
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2225876897573471,
+      "learning_rate": 0.000266972562982806,
+      "loss": 0.9071,
+      "step": 6930
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18916994333267212,
+      "learning_rate": 0.00026690811674914323,
+      "loss": 0.9641,
+      "step": 6935
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2092379480600357,
+      "learning_rate": 0.0002668436154946383,
+      "loss": 1.061,
+      "step": 6940
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18153078854084015,
+      "learning_rate": 0.0002667790592496477,
+      "loss": 1.0148,
+      "step": 6945
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.20708273351192474,
+      "learning_rate": 0.00026671444804455375,
+      "loss": 0.9417,
+      "step": 6950
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2154036909341812,
+      "learning_rate": 0.0002666497819097645,
+      "loss": 0.8497,
+      "step": 6955
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.22335532307624817,
+      "learning_rate": 0.0002665850608757141,
+      "loss": 0.884,
+      "step": 6960
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2135249823331833,
+      "learning_rate": 0.0002665202849728624,
+      "loss": 0.9287,
+      "step": 6965
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2231944501399994,
+      "learning_rate": 0.000266455454231695,
+      "loss": 0.9385,
+      "step": 6970
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19477687776088715,
+      "learning_rate": 0.0002663905686827235,
+      "loss": 0.9468,
+      "step": 6975
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.21901415288448334,
+      "learning_rate": 0.0002663256283564852,
+      "loss": 0.8921,
+      "step": 6980
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.21255704760551453,
+      "learning_rate": 0.00026626063328354316,
+      "loss": 0.9014,
+      "step": 6985
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.20495054125785828,
+      "learning_rate": 0.0002661955834944861,
+      "loss": 1.0034,
+      "step": 6990
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1916796863079071,
+      "learning_rate": 0.0002661304790199288,
+      "loss": 1.0317,
+      "step": 6995
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.21294309198856354,
+      "learning_rate": 0.0002660653198905115,
+      "loss": 0.9128,
+      "step": 7000
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.22985853254795074,
+      "learning_rate": 0.00026600010613690023,
+      "loss": 0.9521,
+      "step": 7005
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24661943316459656,
+      "learning_rate": 0.00026593483778978677,
+      "loss": 0.9456,
+      "step": 7010
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.21790087223052979,
+      "learning_rate": 0.0002658695148798886,
+      "loss": 0.9558,
+      "step": 7015
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18873614072799683,
+      "learning_rate": 0.00026580413743794874,
+      "loss": 0.914,
+      "step": 7020
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.21589027345180511,
+      "learning_rate": 0.0002657387054947361,
+      "loss": 0.9285,
+      "step": 7025
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1877526193857193,
+      "learning_rate": 0.0002656732190810451,
+      "loss": 0.9537,
+      "step": 7030
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22506104409694672,
+      "learning_rate": 0.0002656076782276958,
+      "loss": 0.8938,
+      "step": 7035
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23874376714229584,
+      "learning_rate": 0.000265542082965534,
+      "loss": 0.8486,
+      "step": 7040
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.20064063370227814,
+      "learning_rate": 0.00026547643332543077,
+      "loss": 0.98,
+      "step": 7045
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.22124479711055756,
+      "learning_rate": 0.0002654107293382833,
+      "loss": 0.9002,
+      "step": 7050
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1762978434562683,
+      "learning_rate": 0.00026534497103501383,
+      "loss": 0.9397,
+      "step": 7055
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.21579968929290771,
+      "learning_rate": 0.0002652791584465706,
+      "loss": 0.9606,
+      "step": 7060
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2093055099248886,
+      "learning_rate": 0.000265213291603927,
+      "loss": 0.9967,
+      "step": 7065
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2238216996192932,
+      "learning_rate": 0.00026514737053808234,
+      "loss": 0.9985,
+      "step": 7070
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.20095250010490417,
+      "learning_rate": 0.0002650813952800611,
+      "loss": 0.8892,
+      "step": 7075
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2064996361732483,
+      "learning_rate": 0.00026501536586091357,
+      "loss": 0.9211,
+      "step": 7080
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.20481903851032257,
+      "learning_rate": 0.0002649492823117153,
+      "loss": 0.8601,
+      "step": 7085
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19760817289352417,
+      "learning_rate": 0.0002648831446635674,
+      "loss": 0.9275,
+      "step": 7090
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2292260080575943,
+      "learning_rate": 0.0002648169529475965,
+      "loss": 1.0424,
+      "step": 7095
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2114565074443817,
+      "learning_rate": 0.0002647507071949546,
+      "loss": 0.874,
+      "step": 7100
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19804421067237854,
+      "learning_rate": 0.00026468440743681915,
+      "loss": 0.9876,
+      "step": 7105
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.21388956904411316,
+      "learning_rate": 0.000264618053704393,
+      "loss": 0.8863,
+      "step": 7110
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19273947179317474,
+      "learning_rate": 0.0002645516460289044,
+      "loss": 1.0307,
+      "step": 7115
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22840675711631775,
+      "learning_rate": 0.000264485184441607,
+      "loss": 0.8876,
+      "step": 7120
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.22079254686832428,
+      "learning_rate": 0.00026441866897378,
+      "loss": 0.9526,
+      "step": 7125
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1948045939207077,
+      "learning_rate": 0.00026435209965672756,
+      "loss": 0.9624,
+      "step": 7130
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19133198261260986,
+      "learning_rate": 0.0002642854765217795,
+      "loss": 0.9241,
+      "step": 7135
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2186049222946167,
+      "learning_rate": 0.00026421879960029096,
+      "loss": 0.9452,
+      "step": 7140
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19075119495391846,
+      "learning_rate": 0.00026415206892364216,
+      "loss": 0.9007,
+      "step": 7145
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20891107618808746,
+      "learning_rate": 0.00026408528452323885,
+      "loss": 0.9266,
+      "step": 7150
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20901614427566528,
+      "learning_rate": 0.000264018446430512,
+      "loss": 0.9023,
+      "step": 7155
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20765335857868195,
+      "learning_rate": 0.0002639515546769179,
+      "loss": 0.9053,
+      "step": 7160
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.21316955983638763,
+      "learning_rate": 0.0002638846092939379,
+      "loss": 0.8916,
+      "step": 7165
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20992538332939148,
+      "learning_rate": 0.00026381761031307873,
+      "loss": 0.968,
+      "step": 7170
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2142067402601242,
+      "learning_rate": 0.0002637505577658725,
+      "loss": 0.9669,
+      "step": 7175
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2190820276737213,
+      "learning_rate": 0.0002636834516838761,
+      "loss": 0.9586,
+      "step": 7180
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18383167684078217,
+      "learning_rate": 0.0002636162920986721,
+      "loss": 0.9392,
+      "step": 7185
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.24156947433948517,
+      "learning_rate": 0.00026354907904186796,
+      "loss": 0.942,
+      "step": 7190
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2070099264383316,
+      "learning_rate": 0.00026348181254509635,
+      "loss": 0.9375,
+      "step": 7195
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22052322328090668,
+      "learning_rate": 0.00026341449264001516,
+      "loss": 0.8665,
+      "step": 7200
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.20919539034366608,
+      "learning_rate": 0.00026334711935830735,
+      "loss": 0.8609,
+      "step": 7205
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.24203932285308838,
+      "learning_rate": 0.00026327969273168104,
+      "loss": 0.9312,
+      "step": 7210
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19879262149333954,
+      "learning_rate": 0.00026321221279186944,
+      "loss": 0.7982,
+      "step": 7215
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.190317302942276,
+      "learning_rate": 0.0002631446795706308,
+      "loss": 0.9479,
+      "step": 7220
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.20319460332393646,
+      "learning_rate": 0.0002630770930997486,
+      "loss": 0.9041,
+      "step": 7225
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.20878246426582336,
+      "learning_rate": 0.00026300945341103113,
+      "loss": 0.9042,
+      "step": 7230
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.208158478140831,
+      "learning_rate": 0.000262941760536312,
+      "loss": 0.91,
+      "step": 7235
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2144281417131424,
+      "learning_rate": 0.0002628740145074497,
+      "loss": 0.8517,
+      "step": 7240
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1934666633605957,
+      "learning_rate": 0.0002628062153563277,
+      "loss": 0.9417,
+      "step": 7245
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2240288108587265,
+      "learning_rate": 0.0002627383631148546,
+      "loss": 0.9912,
+      "step": 7250
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2270900011062622,
+      "learning_rate": 0.00026267045781496384,
+      "loss": 0.8963,
+      "step": 7255
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2111656367778778,
+      "learning_rate": 0.00026260249948861406,
+      "loss": 0.9598,
+      "step": 7260
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.21797975897789001,
+      "learning_rate": 0.0002625344881677885,
+      "loss": 0.9216,
+      "step": 7265
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.20986701548099518,
+      "learning_rate": 0.00026246642388449575,
+      "loss": 0.8921,
+      "step": 7270
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.21102838218212128,
+      "learning_rate": 0.00026239830667076897,
+      "loss": 1.0191,
+      "step": 7275
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2326158732175827,
+      "learning_rate": 0.00026233013655866646,
+      "loss": 0.9563,
+      "step": 7280
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1865602284669876,
+      "learning_rate": 0.0002622619135802713,
+      "loss": 0.974,
+      "step": 7285
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1943947672843933,
+      "learning_rate": 0.00026219363776769155,
+      "loss": 0.9563,
+      "step": 7290
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.21199694275856018,
+      "learning_rate": 0.00026212530915306,
+      "loss": 0.9788,
+      "step": 7295
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.20368671417236328,
+      "learning_rate": 0.0002620569277685344,
+      "loss": 0.9533,
+      "step": 7300
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2417079508304596,
+      "learning_rate": 0.00026198849364629723,
+      "loss": 0.9253,
+      "step": 7305
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.20021148025989532,
+      "learning_rate": 0.00026192000681855604,
+      "loss": 0.9328,
+      "step": 7310
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.21492332220077515,
+      "learning_rate": 0.00026185146731754285,
+      "loss": 0.9173,
+      "step": 7315
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.22613677382469177,
+      "learning_rate": 0.00026178287517551464,
+      "loss": 0.9277,
+      "step": 7320
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.21383176743984222,
+      "learning_rate": 0.0002617142304247532,
+      "loss": 0.9311,
+      "step": 7325
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.20949169993400574,
+      "learning_rate": 0.00026164553309756497,
+      "loss": 0.9854,
+      "step": 7330
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.23615580797195435,
+      "learning_rate": 0.00026157678322628127,
+      "loss": 0.9295,
+      "step": 7335
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18629109859466553,
+      "learning_rate": 0.00026150798084325803,
+      "loss": 0.9045,
+      "step": 7340
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2189890593290329,
+      "learning_rate": 0.00026143912598087593,
+      "loss": 0.9756,
+      "step": 7345
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1957770138978958,
+      "learning_rate": 0.00026137021867154043,
+      "loss": 0.8535,
+      "step": 7350
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20852722227573395,
+      "learning_rate": 0.00026130125894768146,
+      "loss": 0.8852,
+      "step": 7355
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2296961098909378,
+      "learning_rate": 0.0002612322468417538,
+      "loss": 0.929,
+      "step": 7360
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.21166066825389862,
+      "learning_rate": 0.00026116318238623694,
+      "loss": 1.0121,
+      "step": 7365
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20803911983966827,
+      "learning_rate": 0.0002610940656136348,
+      "loss": 0.9132,
+      "step": 7370
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.26100292801856995,
+      "learning_rate": 0.0002610248965564761,
+      "loss": 0.9956,
+      "step": 7375
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19871069490909576,
+      "learning_rate": 0.000260955675247314,
+      "loss": 0.904,
+      "step": 7380
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.21508464217185974,
+      "learning_rate": 0.0002608864017187264,
+      "loss": 0.8962,
+      "step": 7385
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.22008737921714783,
+      "learning_rate": 0.0002608170760033158,
+      "loss": 0.8767,
+      "step": 7390
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2107255458831787,
+      "learning_rate": 0.0002607476981337091,
+      "loss": 0.9574,
+      "step": 7395
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2372090220451355,
+      "learning_rate": 0.00026067826814255777,
+      "loss": 0.9463,
+      "step": 7400
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20857949554920197,
+      "learning_rate": 0.000260608786062538,
+      "loss": 0.9372,
+      "step": 7405
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.21467188000679016,
+      "learning_rate": 0.0002605392519263503,
+      "loss": 0.9589,
+      "step": 7410
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20547367632389069,
+      "learning_rate": 0.0002604696657667197,
+      "loss": 0.9673,
+      "step": 7415
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.23634222149848938,
+      "learning_rate": 0.00026040002761639586,
+      "loss": 0.8479,
+      "step": 7420
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19742855429649353,
+      "learning_rate": 0.0002603303375081527,
+      "loss": 0.9117,
+      "step": 7425
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2211044728755951,
+      "learning_rate": 0.0002602605954747888,
+      "loss": 0.9479,
+      "step": 7430
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20085452497005463,
+      "learning_rate": 0.000260190801549127,
+      "loss": 0.96,
+      "step": 7435
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20444773137569427,
+      "learning_rate": 0.0002601209557640147,
+      "loss": 0.8778,
+      "step": 7440
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.21985496580600739,
+      "learning_rate": 0.00026005105815232364,
+      "loss": 0.9513,
+      "step": 7445
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22137680649757385,
+      "learning_rate": 0.0002599811087469498,
+      "loss": 1.0321,
+      "step": 7450
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20409676432609558,
+      "learning_rate": 0.0002599111075808139,
+      "loss": 0.9039,
+      "step": 7455
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20805245637893677,
+      "learning_rate": 0.0002598410546868608,
+      "loss": 0.9834,
+      "step": 7460
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23089753091335297,
+      "learning_rate": 0.00025977095009805957,
+      "loss": 0.9394,
+      "step": 7465
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2073964923620224,
+      "learning_rate": 0.0002597007938474039,
+      "loss": 0.8757,
+      "step": 7470
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22039000689983368,
+      "learning_rate": 0.00025963058596791157,
+      "loss": 0.8823,
+      "step": 7475
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20339925587177277,
+      "learning_rate": 0.00025956032649262475,
+      "loss": 0.9434,
+      "step": 7480
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2380754053592682,
+      "learning_rate": 0.0002594900154546099,
+      "loss": 0.9431,
+      "step": 7485
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22166424989700317,
+      "learning_rate": 0.00025941965288695776,
+      "loss": 0.8625,
+      "step": 7490
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19690659642219543,
+      "learning_rate": 0.00025934923882278325,
+      "loss": 0.8757,
+      "step": 7495
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.225230872631073,
+      "learning_rate": 0.00025927877329522554,
+      "loss": 0.9952,
+      "step": 7500
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23189792037010193,
+      "learning_rate": 0.00025920825633744815,
+      "loss": 1.0034,
+      "step": 7505
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22297634184360504,
+      "learning_rate": 0.0002591376879826386,
+      "loss": 0.9525,
+      "step": 7510
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2259618490934372,
+      "learning_rate": 0.00025906706826400863,
+      "loss": 0.9325,
+      "step": 7515
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21782998740673065,
+      "learning_rate": 0.0002589963972147945,
+      "loss": 0.9086,
+      "step": 7520
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2169419527053833,
+      "learning_rate": 0.000258925674868256,
+      "loss": 0.9848,
+      "step": 7525
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.20681601762771606,
+      "learning_rate": 0.00025885490125767774,
+      "loss": 0.9753,
+      "step": 7530
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21519158780574799,
+      "learning_rate": 0.00025878407641636794,
+      "loss": 0.9679,
+      "step": 7535
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.22115248441696167,
+      "learning_rate": 0.00025871320037765917,
+      "loss": 0.8951,
+      "step": 7540
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21856175363063812,
+      "learning_rate": 0.0002586422731749081,
+      "loss": 0.9302,
+      "step": 7545
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21080365777015686,
+      "learning_rate": 0.0002585712948414953,
+      "loss": 0.9291,
+      "step": 7550
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.210664302110672,
+      "learning_rate": 0.0002585002654108257,
+      "loss": 0.9287,
+      "step": 7555
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19781038165092468,
+      "learning_rate": 0.0002584291849163279,
+      "loss": 0.9384,
+      "step": 7560
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.22275352478027344,
+      "learning_rate": 0.0002583580533914549,
+      "loss": 0.824,
+      "step": 7565
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2083931863307953,
+      "learning_rate": 0.00025828687086968354,
+      "loss": 0.9764,
+      "step": 7570
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.20877456665039062,
+      "learning_rate": 0.00025821563738451464,
+      "loss": 0.9347,
+      "step": 7575
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1873340606689453,
+      "learning_rate": 0.00025814435296947307,
+      "loss": 0.9142,
+      "step": 7580
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2477707713842392,
+      "learning_rate": 0.0002580730176581076,
+      "loss": 0.9027,
+      "step": 7585
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23362557590007782,
+      "learning_rate": 0.000258001631483991,
+      "loss": 0.8038,
+      "step": 7590
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.22810354828834534,
+      "learning_rate": 0.00025793019448072007,
+      "loss": 0.9889,
+      "step": 7595
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2291039526462555,
+      "learning_rate": 0.0002578587066819153,
+      "loss": 0.9264,
+      "step": 7600
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.22235728800296783,
+      "learning_rate": 0.00025778716812122136,
+      "loss": 0.9701,
+      "step": 7605
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.21946272253990173,
+      "learning_rate": 0.00025771557883230657,
+      "loss": 0.8935,
+      "step": 7610
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.21064096689224243,
+      "learning_rate": 0.00025764393884886324,
+      "loss": 0.9104,
+      "step": 7615
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2479812055826187,
+      "learning_rate": 0.00025757224820460755,
+      "loss": 0.8872,
+      "step": 7620
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2186814844608307,
+      "learning_rate": 0.0002575005069332795,
+      "loss": 0.9364,
+      "step": 7625
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2108563780784607,
+      "learning_rate": 0.00025742871506864295,
+      "loss": 0.8766,
+      "step": 7630
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.213986337184906,
+      "learning_rate": 0.0002573568726444854,
+      "loss": 0.9053,
+      "step": 7635
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.23144182562828064,
+      "learning_rate": 0.0002572849796946184,
+      "loss": 0.9933,
+      "step": 7640
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2001713067293167,
+      "learning_rate": 0.00025721303625287717,
+      "loss": 0.9321,
+      "step": 7645
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19496455788612366,
+      "learning_rate": 0.00025714104235312064,
+      "loss": 0.861,
+      "step": 7650
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19990824162960052,
+      "learning_rate": 0.0002570689980292315,
+      "loss": 0.9007,
+      "step": 7655
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2162327915430069,
+      "learning_rate": 0.0002569969033151163,
+      "loss": 0.9936,
+      "step": 7660
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.20773164927959442,
+      "learning_rate": 0.00025692475824470504,
+      "loss": 0.9124,
+      "step": 7665
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19005447626113892,
+      "learning_rate": 0.0002568525628519518,
+      "loss": 1.0014,
+      "step": 7670
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18565884232521057,
+      "learning_rate": 0.00025678031717083394,
+      "loss": 0.8888,
+      "step": 7675
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.21810932457447052,
+      "learning_rate": 0.0002567080212353528,
+      "loss": 0.8517,
+      "step": 7680
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.21368250250816345,
+      "learning_rate": 0.00025663567507953314,
+      "loss": 0.9296,
+      "step": 7685
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19680459797382355,
+      "learning_rate": 0.0002565632787374236,
+      "loss": 1.0182,
+      "step": 7690
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.23982003331184387,
+      "learning_rate": 0.00025649083224309617,
+      "loss": 0.9503,
+      "step": 7695
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.24432729184627533,
+      "learning_rate": 0.00025641833563064666,
+      "loss": 0.903,
+      "step": 7700
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.20875149965286255,
+      "learning_rate": 0.00025634578893419434,
+      "loss": 0.8951,
+      "step": 7705
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.21281814575195312,
+      "learning_rate": 0.00025627319218788215,
+      "loss": 0.9059,
+      "step": 7710
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.226210355758667,
+      "learning_rate": 0.0002562005454258765,
+      "loss": 0.9195,
+      "step": 7715
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22133582830429077,
+      "learning_rate": 0.0002561278486823673,
+      "loss": 0.8901,
+      "step": 7720
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.24061179161071777,
+      "learning_rate": 0.00025605510199156817,
+      "loss": 0.9548,
+      "step": 7725
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.20201779901981354,
+      "learning_rate": 0.000255982305387716,
+      "loss": 0.8928,
+      "step": 7730
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.23118892312049866,
+      "learning_rate": 0.00025590945890507146,
+      "loss": 0.9098,
+      "step": 7735
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3037243187427521,
+      "learning_rate": 0.00025583656257791834,
+      "loss": 0.8707,
+      "step": 7740
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2243233621120453,
+      "learning_rate": 0.0002557636164405641,
+      "loss": 0.8642,
+      "step": 7745
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1916697919368744,
+      "learning_rate": 0.0002556906205273398,
+      "loss": 0.9297,
+      "step": 7750
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18961936235427856,
+      "learning_rate": 0.00025561757487259953,
+      "loss": 0.8665,
+      "step": 7755
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.21484404802322388,
+      "learning_rate": 0.000255544479510721,
+      "loss": 0.8957,
+      "step": 7760
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2340548038482666,
+      "learning_rate": 0.0002554713344761055,
+      "loss": 0.8395,
+      "step": 7765
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.20856556296348572,
+      "learning_rate": 0.00025539813980317733,
+      "loss": 0.9225,
+      "step": 7770
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.21993188560009003,
+      "learning_rate": 0.00025532489552638446,
+      "loss": 0.9975,
+      "step": 7775
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.21606536209583282,
+      "learning_rate": 0.000255251601680198,
+      "loss": 0.9617,
+      "step": 7780
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23577193915843964,
+      "learning_rate": 0.00025517825829911246,
+      "loss": 0.9713,
+      "step": 7785
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2203446924686432,
+      "learning_rate": 0.0002551048654176457,
+      "loss": 0.984,
+      "step": 7790
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23059968650341034,
+      "learning_rate": 0.0002550314230703389,
+      "loss": 0.9334,
+      "step": 7795
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2229011207818985,
+      "learning_rate": 0.0002549579312917564,
+      "loss": 1.0343,
+      "step": 7800
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18573442101478577,
+      "learning_rate": 0.0002548843901164859,
+      "loss": 0.8456,
+      "step": 7805
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2154776155948639,
+      "learning_rate": 0.00025481079957913826,
+      "loss": 0.983,
+      "step": 7810
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.20241975784301758,
+      "learning_rate": 0.0002547371597143477,
+      "loss": 1.0003,
+      "step": 7815
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2103583961725235,
+      "learning_rate": 0.0002546634705567716,
+      "loss": 0.9616,
+      "step": 7820
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1959349662065506,
+      "learning_rate": 0.0002545897321410905,
+      "loss": 1.0201,
+      "step": 7825
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.22472964227199554,
+      "learning_rate": 0.00025451594450200804,
+      "loss": 0.997,
+      "step": 7830
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2131294161081314,
+      "learning_rate": 0.0002544421076742513,
+      "loss": 0.9673,
+      "step": 7835
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.21508903801441193,
+      "learning_rate": 0.00025436822169257027,
+      "loss": 0.9775,
+      "step": 7840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2155522108078003,
+      "learning_rate": 0.00025429428659173815,
+      "loss": 0.7657,
+      "step": 7845
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.22796915471553802,
+      "learning_rate": 0.00025422030240655123,
+      "loss": 0.9161,
+      "step": 7850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19367897510528564,
+      "learning_rate": 0.000254146269171829,
+      "loss": 0.9141,
+      "step": 7855
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2098265439271927,
+      "learning_rate": 0.00025407218692241384,
+      "loss": 1.0136,
+      "step": 7860
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.20742662250995636,
+      "learning_rate": 0.00025399805569317145,
+      "loss": 1.0097,
+      "step": 7865
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19661657512187958,
+      "learning_rate": 0.00025392387551899034,
+      "loss": 0.8654,
+      "step": 7870
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2003818154335022,
+      "learning_rate": 0.0002538496464347822,
+      "loss": 0.9325,
+      "step": 7875
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.21950916945934296,
+      "learning_rate": 0.0002537753684754817,
+      "loss": 0.9218,
+      "step": 7880
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2205251157283783,
+      "learning_rate": 0.00025370104167604657,
+      "loss": 0.9599,
+      "step": 7885
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.21949486434459686,
+      "learning_rate": 0.00025362666607145744,
+      "loss": 0.8522,
+      "step": 7890
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.20743881165981293,
+      "learning_rate": 0.00025355224169671786,
+      "loss": 0.9354,
+      "step": 7895
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19251248240470886,
+      "learning_rate": 0.0002534777685868545,
+      "loss": 0.9454,
+      "step": 7900
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.22613319754600525,
+      "learning_rate": 0.00025340324677691685,
+      "loss": 0.929,
+      "step": 7905
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.206687793135643,
+      "learning_rate": 0.00025332867630197735,
+      "loss": 0.8953,
+      "step": 7910
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19443422555923462,
+      "learning_rate": 0.0002532540571971313,
+      "loss": 0.9731,
+      "step": 7915
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.20105896890163422,
+      "learning_rate": 0.00025317938949749705,
+      "loss": 0.924,
+      "step": 7920
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2265363335609436,
+      "learning_rate": 0.0002531046732382156,
+      "loss": 0.9835,
+      "step": 7925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19356288015842438,
+      "learning_rate": 0.00025302990845445087,
+      "loss": 1.0531,
+      "step": 7930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1920010894536972,
+      "learning_rate": 0.00025295509518138975,
+      "loss": 0.9466,
+      "step": 7935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.22366103529930115,
+      "learning_rate": 0.00025288023345424176,
+      "loss": 0.9589,
+      "step": 7940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.22013331949710846,
+      "learning_rate": 0.00025280532330823944,
+      "loss": 1.0145,
+      "step": 7945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24709923565387726,
+      "learning_rate": 0.00025273036477863785,
+      "loss": 1.0086,
+      "step": 7950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2514353096485138,
+      "learning_rate": 0.00025265535790071505,
+      "loss": 0.9726,
+      "step": 7955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2117491513490677,
+      "learning_rate": 0.0002525803027097717,
+      "loss": 0.8978,
+      "step": 7960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.21775074303150177,
+      "learning_rate": 0.0002525051992411314,
+      "loss": 0.8716,
+      "step": 7965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.20569385588169098,
+      "learning_rate": 0.0002524300475301402,
+      "loss": 0.8722,
+      "step": 7970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19833216071128845,
+      "learning_rate": 0.00025235484761216697,
+      "loss": 0.859,
+      "step": 7975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19562938809394836,
+      "learning_rate": 0.00025227959952260344,
+      "loss": 0.9029,
+      "step": 7980
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19179534912109375,
+      "learning_rate": 0.00025220430329686377,
+      "loss": 0.9367,
+      "step": 7985
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.20224545896053314,
+      "learning_rate": 0.0002521289589703848,
+      "loss": 0.8961,
+      "step": 7990
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2402665615081787,
+      "learning_rate": 0.0002520535665786262,
+      "loss": 0.8604,
+      "step": 7995
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24251817166805267,
+      "learning_rate": 0.00025197812615707007,
+      "loss": 0.9031,
+      "step": 8000
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.23049946129322052,
+      "learning_rate": 0.00025190263774122113,
+      "loss": 0.8619,
+      "step": 8005
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2142931967973709,
+      "learning_rate": 0.0002518271013666068,
+      "loss": 0.8672,
+      "step": 8010
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21534845232963562,
+      "learning_rate": 0.0002517515170687771,
+      "loss": 0.9955,
+      "step": 8015
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19614648818969727,
+      "learning_rate": 0.0002516758848833043,
+      "loss": 0.8943,
+      "step": 8020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21292729675769806,
+      "learning_rate": 0.0002516002048457835,
+      "loss": 0.9365,
+      "step": 8025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20765379071235657,
+      "learning_rate": 0.0002515244769918323,
+      "loss": 0.8914,
+      "step": 8030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.24531885981559753,
+      "learning_rate": 0.00025144870135709077,
+      "loss": 0.9454,
+      "step": 8035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21342940628528595,
+      "learning_rate": 0.0002513728779772213,
+      "loss": 0.9629,
+      "step": 8040
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20636995136737823,
+      "learning_rate": 0.00025129700688790896,
+      "loss": 0.9343,
+      "step": 8045
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20889271795749664,
+      "learning_rate": 0.00025122108812486124,
+      "loss": 0.9349,
+      "step": 8050
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2028619647026062,
+      "learning_rate": 0.000251145121723808,
+      "loss": 0.9775,
+      "step": 8055
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20538374781608582,
+      "learning_rate": 0.0002510691077205015,
+      "loss": 0.8841,
+      "step": 8060
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.23155823349952698,
+      "learning_rate": 0.0002509930461507166,
+      "loss": 0.9182,
+      "step": 8065
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.21618753671646118,
+      "learning_rate": 0.00025091693705025023,
+      "loss": 0.8398,
+      "step": 8070
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2337218075990677,
+      "learning_rate": 0.00025084078045492194,
+      "loss": 0.9557,
+      "step": 8075
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2220161408185959,
+      "learning_rate": 0.0002507645764005736,
+      "loss": 1.008,
+      "step": 8080
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20406211912631989,
+      "learning_rate": 0.00025068832492306924,
+      "loss": 0.8859,
+      "step": 8085
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.22060614824295044,
+      "learning_rate": 0.0002506120260582955,
+      "loss": 0.936,
+      "step": 8090
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2314573973417282,
+      "learning_rate": 0.000250535679842161,
+      "loss": 0.9284,
+      "step": 8095
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.22508397698402405,
+      "learning_rate": 0.00025045928631059694,
+      "loss": 0.9302,
+      "step": 8100
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2177351713180542,
+      "learning_rate": 0.00025038284549955655,
+      "loss": 0.9366,
+      "step": 8105
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2541705369949341,
+      "learning_rate": 0.0002503063574450155,
+      "loss": 0.9312,
+      "step": 8110
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.22795630991458893,
+      "learning_rate": 0.0002502298221829715,
+      "loss": 0.8796,
+      "step": 8115
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.23147457838058472,
+      "learning_rate": 0.0002501532397494447,
+      "loss": 0.9392,
+      "step": 8120
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.21217665076255798,
+      "learning_rate": 0.0002500766101804773,
+      "loss": 0.9313,
+      "step": 8125
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2149830311536789,
+      "learning_rate": 0.0002499999335121337,
+      "loss": 0.8611,
+      "step": 8130
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.22693344950675964,
+      "learning_rate": 0.0002499232097805004,
+      "loss": 0.9367,
+      "step": 8135
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.23371291160583496,
+      "learning_rate": 0.0002498464390216864,
+      "loss": 0.9207,
+      "step": 8140
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2417604625225067,
+      "learning_rate": 0.00024976962127182224,
+      "loss": 0.9299,
+      "step": 8145
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2232208102941513,
+      "learning_rate": 0.00024969275656706115,
+      "loss": 0.9032,
+      "step": 8150
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.23492024838924408,
+      "learning_rate": 0.0002496158449435781,
+      "loss": 0.8478,
+      "step": 8155
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.22294215857982635,
+      "learning_rate": 0.00024953888643757026,
+      "loss": 0.9258,
+      "step": 8160
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.20344236493110657,
+      "learning_rate": 0.0002494618810852569,
+      "loss": 0.8963,
+      "step": 8165
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1954520046710968,
+      "learning_rate": 0.0002493848289228793,
+      "loss": 0.8931,
+      "step": 8170
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2296024113893509,
+      "learning_rate": 0.00024930772998670074,
+      "loss": 0.931,
+      "step": 8175
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19687330722808838,
+      "learning_rate": 0.00024923058431300653,
+      "loss": 0.8976,
+      "step": 8180
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.20918093621730804,
+      "learning_rate": 0.000249153391938104,
+      "loss": 0.9273,
+      "step": 8185
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2444303333759308,
+      "learning_rate": 0.0002490761528983224,
+      "loss": 0.9052,
+      "step": 8190
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2560185194015503,
+      "learning_rate": 0.00024899886723001307,
+      "loss": 0.8845,
+      "step": 8195
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.20742972195148468,
+      "learning_rate": 0.00024892153496954917,
+      "loss": 0.8958,
+      "step": 8200
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.25156137347221375,
+      "learning_rate": 0.0002488441561533258,
+      "loss": 0.862,
+      "step": 8205
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22066602110862732,
+      "learning_rate": 0.00024876673081776,
+      "loss": 0.9099,
+      "step": 8210
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2297365367412567,
+      "learning_rate": 0.00024868925899929084,
+      "loss": 0.8339,
+      "step": 8215
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.23268838226795197,
+      "learning_rate": 0.0002486117407343789,
+      "loss": 0.8629,
+      "step": 8220
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.20442204177379608,
+      "learning_rate": 0.0002485341760595071,
+      "loss": 0.929,
+      "step": 8225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2177409529685974,
+      "learning_rate": 0.0002484565650111798,
+      "loss": 0.9028,
+      "step": 8230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2090945988893509,
+      "learning_rate": 0.0002483789076259233,
+      "loss": 0.9519,
+      "step": 8235
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.20811286568641663,
+      "learning_rate": 0.00024830120394028586,
+      "loss": 0.9343,
+      "step": 8240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.21751175820827484,
+      "learning_rate": 0.0002482234539908374,
+      "loss": 0.9037,
+      "step": 8245
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.20378360152244568,
+      "learning_rate": 0.0002481456578141695,
+      "loss": 0.9319,
+      "step": 8250
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.21400323510169983,
+      "learning_rate": 0.00024806781544689575,
+      "loss": 0.8737,
+      "step": 8255
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19827169179916382,
+      "learning_rate": 0.00024798992692565136,
+      "loss": 0.9123,
+      "step": 8260
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.23193618655204773,
+      "learning_rate": 0.00024791199228709317,
+      "loss": 0.9127,
+      "step": 8265
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.24364005029201508,
+      "learning_rate": 0.00024783401156789985,
+      "loss": 0.9055,
+      "step": 8270
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.23433059453964233,
+      "learning_rate": 0.00024775598480477175,
+      "loss": 1.01,
+      "step": 8275
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.20962095260620117,
+      "learning_rate": 0.0002476779120344308,
+      "loss": 0.8084,
+      "step": 8280
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2257934808731079,
+      "learning_rate": 0.00024759979329362067,
+      "loss": 0.9176,
+      "step": 8285
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.21964751183986664,
+      "learning_rate": 0.0002475216286191067,
+      "loss": 0.8642,
+      "step": 8290
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.21779011189937592,
+      "learning_rate": 0.00024744341804767555,
+      "loss": 0.9282,
+      "step": 8295
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.21334509551525116,
+      "learning_rate": 0.000247365161616136,
+      "loss": 0.8526,
+      "step": 8300
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.21431876718997955,
+      "learning_rate": 0.00024728685936131794,
+      "loss": 0.8295,
+      "step": 8305
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.22023451328277588,
+      "learning_rate": 0.0002472085113200731,
+      "loss": 0.9086,
+      "step": 8310
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.20858727395534515,
+      "learning_rate": 0.0002471301175292746,
+      "loss": 0.9956,
+      "step": 8315
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19900618493556976,
+      "learning_rate": 0.00024705167802581727,
+      "loss": 0.9715,
+      "step": 8320
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.21103012561798096,
+      "learning_rate": 0.0002469731928466172,
+      "loss": 0.8762,
+      "step": 8325
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2619311809539795,
+      "learning_rate": 0.0002468946620286122,
+      "loss": 0.9099,
+      "step": 8330
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.22611823678016663,
+      "learning_rate": 0.0002468160856087615,
+      "loss": 0.9334,
+      "step": 8335
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.20762339234352112,
+      "learning_rate": 0.0002467374636240458,
+      "loss": 0.9715,
+      "step": 8340
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.20186099410057068,
+      "learning_rate": 0.0002466587961114671,
+      "loss": 0.8772,
+      "step": 8345
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2584278881549835,
+      "learning_rate": 0.000246580083108049,
+      "loss": 0.9342,
+      "step": 8350
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.20932774245738983,
+      "learning_rate": 0.0002465013246508365,
+      "loss": 0.9424,
+      "step": 8355
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.22265595197677612,
+      "learning_rate": 0.0002464225207768959,
+      "loss": 0.8854,
+      "step": 8360
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.24274423718452454,
+      "learning_rate": 0.0002463436715233149,
+      "loss": 0.9185,
+      "step": 8365
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.22588522732257843,
+      "learning_rate": 0.0002462647769272027,
+      "loss": 0.8415,
+      "step": 8370
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.21932215988636017,
+      "learning_rate": 0.00024618583702568954,
+      "loss": 0.9426,
+      "step": 8375
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.2024673968553543,
+      "learning_rate": 0.0002461068518559273,
+      "loss": 0.833,
+      "step": 8380
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.22020402550697327,
+      "learning_rate": 0.00024602782145508885,
+      "loss": 0.8884,
+      "step": 8385
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.24328754842281342,
+      "learning_rate": 0.00024594874586036876,
+      "loss": 0.8266,
+      "step": 8390
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.23017850518226624,
+      "learning_rate": 0.00024586962510898244,
+      "loss": 0.96,
+      "step": 8395
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.22810372710227966,
+      "learning_rate": 0.0002457904592381668,
+      "loss": 0.9248,
+      "step": 8400
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.23655059933662415,
+      "learning_rate": 0.00024571124828518003,
+      "loss": 0.9204,
+      "step": 8405
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.24605585634708405,
+      "learning_rate": 0.0002456319922873013,
+      "loss": 0.8909,
+      "step": 8410
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.22107426822185516,
+      "learning_rate": 0.00024555269128183116,
+      "loss": 0.8827,
+      "step": 8415
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.24414613842964172,
+      "learning_rate": 0.00024547334530609124,
+      "loss": 0.8416,
+      "step": 8420
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.23650889098644257,
+      "learning_rate": 0.00024539395439742453,
+      "loss": 0.8868,
+      "step": 8425
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.23410624265670776,
+      "learning_rate": 0.000245314518593195,
+      "loss": 1.0127,
+      "step": 8430
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2250613421201706,
+      "learning_rate": 0.0002452350379307876,
+      "loss": 0.9244,
+      "step": 8435
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.22231152653694153,
+      "learning_rate": 0.00024515551244760865,
+      "loss": 0.9534,
+      "step": 8440
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2265365868806839,
+      "learning_rate": 0.0002450759421810856,
+      "loss": 0.8905,
+      "step": 8445
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.21347838640213013,
+      "learning_rate": 0.0002449963271686668,
+      "loss": 0.9037,
+      "step": 8450
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.23790526390075684,
+      "learning_rate": 0.0002449166674478217,
+      "loss": 0.8819,
+      "step": 8455
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2057124823331833,
+      "learning_rate": 0.0002448369630560408,
+      "loss": 0.9506,
+      "step": 8460
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.22661954164505005,
+      "learning_rate": 0.00024475721403083566,
+      "loss": 0.8741,
+      "step": 8465
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2240198701620102,
+      "learning_rate": 0.0002446774204097388,
+      "loss": 0.8722,
+      "step": 8470
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.19682292640209198,
+      "learning_rate": 0.0002445975822303038,
+      "loss": 0.8477,
+      "step": 8475
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.20487092435359955,
+      "learning_rate": 0.00024451769953010504,
+      "loss": 0.8088,
+      "step": 8480
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.23260267078876495,
+      "learning_rate": 0.00024443777234673807,
+      "loss": 0.8667,
+      "step": 8485
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2150842696428299,
+      "learning_rate": 0.00024435780071781926,
+      "loss": 0.8324,
+      "step": 8490
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.24172092974185944,
+      "learning_rate": 0.00024427778468098587,
+      "loss": 0.8618,
+      "step": 8495
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.23562046885490417,
+      "learning_rate": 0.0002441977242738962,
+      "loss": 0.9769,
+      "step": 8500
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.2472938448190689,
+      "learning_rate": 0.00024411761953422922,
+      "loss": 0.8776,
+      "step": 8505
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.23211072385311127,
+      "learning_rate": 0.000244037470499685,
+      "loss": 0.8499,
+      "step": 8510
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2152828425168991,
+      "learning_rate": 0.00024395727720798424,
+      "loss": 0.8764,
+      "step": 8515
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.21999038755893707,
+      "learning_rate": 0.0002438770396968686,
+      "loss": 0.8729,
+      "step": 8520
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.21073397994041443,
+      "learning_rate": 0.0002437967580041005,
+      "loss": 0.8665,
+      "step": 8525
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.21868467330932617,
+      "learning_rate": 0.00024371643216746324,
+      "loss": 0.8653,
+      "step": 8530
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.22046944499015808,
+      "learning_rate": 0.00024363606222476075,
+      "loss": 0.8832,
+      "step": 8535
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.23462894558906555,
+      "learning_rate": 0.00024355564821381784,
+      "loss": 0.8693,
+      "step": 8540
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.20430710911750793,
+      "learning_rate": 0.00024347519017247995,
+      "loss": 0.9157,
+      "step": 8545
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2538154721260071,
+      "learning_rate": 0.00024339468813861336,
+      "loss": 0.8323,
+      "step": 8550
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2316758930683136,
+      "learning_rate": 0.000243314142150105,
+      "loss": 0.8574,
+      "step": 8555
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.239527627825737,
+      "learning_rate": 0.0002432335522448625,
+      "loss": 0.9551,
+      "step": 8560
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.21243155002593994,
+      "learning_rate": 0.00024315291846081406,
+      "loss": 0.9939,
+      "step": 8565
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.23001927137374878,
+      "learning_rate": 0.00024307224083590874,
+      "loss": 0.8392,
+      "step": 8570
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.21947364509105682,
+      "learning_rate": 0.00024299151940811606,
+      "loss": 0.8661,
+      "step": 8575
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.23186787962913513,
+      "learning_rate": 0.0002429107542154261,
+      "loss": 0.8679,
+      "step": 8580
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.2559497058391571,
+      "learning_rate": 0.00024282994529584983,
+      "loss": 0.8598,
+      "step": 8585
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.23815637826919556,
+      "learning_rate": 0.00024274909268741848,
+      "loss": 0.8349,
+      "step": 8590
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2084992676973343,
+      "learning_rate": 0.00024266819642818405,
+      "loss": 0.8345,
+      "step": 8595
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2263403832912445,
+      "learning_rate": 0.0002425872565562189,
+      "loss": 0.8879,
+      "step": 8600
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.23247919976711273,
+      "learning_rate": 0.00024250627310961614,
+      "loss": 0.7877,
+      "step": 8605
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.21611854434013367,
+      "learning_rate": 0.00024242524612648917,
+      "loss": 0.8724,
+      "step": 8610
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2197057604789734,
+      "learning_rate": 0.00024234417564497206,
+      "loss": 0.8138,
+      "step": 8615
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2542968988418579,
+      "learning_rate": 0.00024226306170321924,
+      "loss": 0.9583,
+      "step": 8620
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.21506857872009277,
+      "learning_rate": 0.00024218190433940558,
+      "loss": 0.9456,
+      "step": 8625
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.21320503950119019,
+      "learning_rate": 0.0002421007035917265,
+      "loss": 0.9033,
+      "step": 8630
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.23590131103992462,
+      "learning_rate": 0.00024201945949839775,
+      "loss": 0.9297,
+      "step": 8635
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.23142153024673462,
+      "learning_rate": 0.0002419381720976555,
+      "loss": 0.9025,
+      "step": 8640
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2024351954460144,
+      "learning_rate": 0.00024185684142775623,
+      "loss": 0.9095,
+      "step": 8645
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2539699077606201,
+      "learning_rate": 0.00024177546752697697,
+      "loss": 0.8968,
+      "step": 8650
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.22273805737495422,
+      "learning_rate": 0.00024169405043361492,
+      "loss": 0.8366,
+      "step": 8655
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.2413063645362854,
+      "learning_rate": 0.00024161259018598764,
+      "loss": 0.8781,
+      "step": 8660
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.21893925964832306,
+      "learning_rate": 0.00024153108682243307,
+      "loss": 0.8827,
+      "step": 8665
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.23175576329231262,
+      "learning_rate": 0.00024144954038130936,
+      "loss": 0.8959,
+      "step": 8670
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2520297169685364,
+      "learning_rate": 0.00024136795090099502,
+      "loss": 0.9922,
+      "step": 8675
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.21685057878494263,
+      "learning_rate": 0.00024128631841988877,
+      "loss": 0.9727,
+      "step": 8680
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.22948361933231354,
+      "learning_rate": 0.00024120464297640955,
+      "loss": 0.8525,
+      "step": 8685
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.24964188039302826,
+      "learning_rate": 0.00024112292460899652,
+      "loss": 0.8168,
+      "step": 8690
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.27692463994026184,
+      "learning_rate": 0.00024104116335610905,
+      "loss": 0.957,
+      "step": 8695
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.24565111100673676,
+      "learning_rate": 0.00024095935925622675,
+      "loss": 0.8839,
+      "step": 8700
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.20882341265678406,
+      "learning_rate": 0.00024087751234784933,
+      "loss": 0.9486,
+      "step": 8705
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.23391057550907135,
+      "learning_rate": 0.0002407956226694966,
+      "loss": 0.9281,
+      "step": 8710
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2399512678384781,
+      "learning_rate": 0.00024071369025970867,
+      "loss": 0.7803,
+      "step": 8715
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.21823789179325104,
+      "learning_rate": 0.00024063171515704556,
+      "loss": 0.8684,
+      "step": 8720
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.21568119525909424,
+      "learning_rate": 0.0002405496974000875,
+      "loss": 0.7765,
+      "step": 8725
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2163381576538086,
+      "learning_rate": 0.00024046763702743478,
+      "loss": 0.8652,
+      "step": 8730
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.22184047102928162,
+      "learning_rate": 0.00024038553407770778,
+      "loss": 0.8446,
+      "step": 8735
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.22127588093280792,
+      "learning_rate": 0.00024030338858954678,
+      "loss": 0.8153,
+      "step": 8740
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.25217580795288086,
+      "learning_rate": 0.0002402212006016123,
+      "loss": 0.8407,
+      "step": 8745
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.22362253069877625,
+      "learning_rate": 0.0002401389701525846,
+      "loss": 0.7756,
+      "step": 8750
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.2407907396554947,
+      "learning_rate": 0.00024005669728116417,
+      "loss": 0.8441,
+      "step": 8755
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.23379315435886383,
+      "learning_rate": 0.00023997438202607124,
+      "loss": 0.8697,
+      "step": 8760
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.23524151742458344,
+      "learning_rate": 0.00023989202442604621,
+      "loss": 0.9377,
+      "step": 8765
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.22088152170181274,
+      "learning_rate": 0.00023980962451984928,
+      "loss": 0.8814,
+      "step": 8770
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.24777436256408691,
+      "learning_rate": 0.00023972718234626056,
+      "loss": 0.7183,
+      "step": 8775
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.22386722266674042,
+      "learning_rate": 0.00023964469794408006,
+      "loss": 0.9437,
+      "step": 8780
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2572818100452423,
+      "learning_rate": 0.0002395621713521277,
+      "loss": 0.9118,
+      "step": 8785
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.23130157589912415,
+      "learning_rate": 0.00023947960260924326,
+      "loss": 0.8401,
+      "step": 8790
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.26878032088279724,
+      "learning_rate": 0.0002393969917542863,
+      "loss": 0.9485,
+      "step": 8795
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.21913166344165802,
+      "learning_rate": 0.00023931433882613617,
+      "loss": 0.8719,
+      "step": 8800
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2042398601770401,
+      "learning_rate": 0.00023923164386369225,
+      "loss": 0.8471,
+      "step": 8805
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": NaN,
+      "learning_rate": 0.00023916545765519838,
+      "loss": 1.0099,
+      "step": 8810
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2109871208667755,
+      "learning_rate": 0.0002390826871291146,
+      "loss": 0.9514,
+      "step": 8815
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.24620996415615082,
+      "learning_rate": 0.00023899987467775985,
+      "loss": 0.794,
+      "step": 8820
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.21983186900615692,
+      "learning_rate": 0.00023891702034010856,
+      "loss": 0.8971,
+      "step": 8825
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.23902229964733124,
+      "learning_rate": 0.00023883412415515458,
+      "loss": 0.874,
+      "step": 8830
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2428964078426361,
+      "learning_rate": 0.0002387511861619117,
+      "loss": 0.9071,
+      "step": 8835
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.22053857147693634,
+      "learning_rate": 0.00023866820639941328,
+      "loss": 0.8574,
+      "step": 8840
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.20824211835861206,
+      "learning_rate": 0.0002385851849067124,
+      "loss": 0.9146,
+      "step": 8845
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.21827155351638794,
+      "learning_rate": 0.0002385021217228816,
+      "loss": 0.8853,
+      "step": 8850
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.24931156635284424,
+      "learning_rate": 0.0002384190168870133,
+      "loss": 0.8394,
+      "step": 8855
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.23989886045455933,
+      "learning_rate": 0.00023833587043821933,
+      "loss": 0.8461,
+      "step": 8860
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.25101611018180847,
+      "learning_rate": 0.00023825268241563121,
+      "loss": 0.8247,
+      "step": 8865
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.22489690780639648,
+      "learning_rate": 0.00023816945285839994,
+      "loss": 0.918,
+      "step": 8870
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.22989100217819214,
+      "learning_rate": 0.00023808618180569613,
+      "loss": 0.8747,
+      "step": 8875
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.22614827752113342,
+      "learning_rate": 0.00023800286929670996,
+      "loss": 0.8885,
+      "step": 8880
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.213409423828125,
+      "learning_rate": 0.00023791951537065098,
+      "loss": 0.9846,
+      "step": 8885
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.22785250842571259,
+      "learning_rate": 0.00023783612006674835,
+      "loss": 0.9043,
+      "step": 8890
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2558668553829193,
+      "learning_rate": 0.00023775268342425071,
+      "loss": 0.8575,
+      "step": 8895
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2358073592185974,
+      "learning_rate": 0.00023766920548242597,
+      "loss": 0.8788,
+      "step": 8900
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.2330961525440216,
+      "learning_rate": 0.00023758568628056185,
+      "loss": 0.8552,
+      "step": 8905
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.26288071274757385,
+      "learning_rate": 0.00023750212585796506,
+      "loss": 0.9431,
+      "step": 8910
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.22953157126903534,
+      "learning_rate": 0.000237418524253962,
+      "loss": 0.8415,
+      "step": 8915
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.23770654201507568,
+      "learning_rate": 0.00023733488150789832,
+      "loss": 0.9387,
+      "step": 8920
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.24671205878257751,
+      "learning_rate": 0.00023725119765913915,
+      "loss": 0.8804,
+      "step": 8925
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2374102622270584,
+      "learning_rate": 0.00023716747274706886,
+      "loss": 1.0063,
+      "step": 8930
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.24478651583194733,
+      "learning_rate": 0.0002370837068110911,
+      "loss": 0.9604,
+      "step": 8935
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.22993561625480652,
+      "learning_rate": 0.00023699989989062892,
+      "loss": 0.861,
+      "step": 8940
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2684665620326996,
+      "learning_rate": 0.0002369160520251248,
+      "loss": 0.8511,
+      "step": 8945
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.25119537115097046,
+      "learning_rate": 0.00023683216325404009,
+      "loss": 0.8144,
+      "step": 8950
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.22731202840805054,
+      "learning_rate": 0.0002367482336168558,
+      "loss": 0.8949,
+      "step": 8955
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.23801319301128387,
+      "learning_rate": 0.00023666426315307188,
+      "loss": 0.9336,
+      "step": 8960
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.24626471102237701,
+      "learning_rate": 0.00023658025190220774,
+      "loss": 0.8794,
+      "step": 8965
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.235481858253479,
+      "learning_rate": 0.00023649619990380184,
+      "loss": 0.8264,
+      "step": 8970
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2578223645687103,
+      "learning_rate": 0.00023641210719741175,
+      "loss": 0.9651,
+      "step": 8975
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.24087263643741608,
+      "learning_rate": 0.00023632797382261441,
+      "loss": 0.8688,
+      "step": 8980
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2264793962240219,
+      "learning_rate": 0.00023624379981900572,
+      "loss": 0.8988,
+      "step": 8985
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.21337303519248962,
+      "learning_rate": 0.00023615958522620078,
+      "loss": 0.8212,
+      "step": 8990
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.24498280882835388,
+      "learning_rate": 0.00023607533008383373,
+      "loss": 0.8061,
+      "step": 8995
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.23145338892936707,
+      "learning_rate": 0.00023599103443155788,
+      "loss": 0.8282,
+      "step": 9000
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2749665379524231,
+      "learning_rate": 0.00023590669830904554,
+      "loss": 0.9067,
+      "step": 9005
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.23750852048397064,
+      "learning_rate": 0.00023582232175598812,
+      "loss": 0.9132,
+      "step": 9010
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.22623078525066376,
+      "learning_rate": 0.000235737904812096,
+      "loss": 0.9233,
+      "step": 9015
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.24327704310417175,
+      "learning_rate": 0.00023567034220227885,
+      "loss": 0.9254,
+      "step": 9020
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.23452426493167877,
+      "learning_rate": 0.00023558585265501518,
+      "loss": 0.789,
+      "step": 9025
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.24461832642555237,
+      "learning_rate": 0.00023550132282820706,
+      "loss": 0.9243,
+      "step": 9030
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.22294314205646515,
+      "learning_rate": 0.00023541675276163697,
+      "loss": 0.8918,
+      "step": 9035
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2230527549982071,
+      "learning_rate": 0.00023533214249510647,
+      "loss": 0.8376,
+      "step": 9040
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.26348450779914856,
+      "learning_rate": 0.00023524749206843586,
+      "loss": 0.9469,
+      "step": 9045
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2523627281188965,
+      "learning_rate": 0.00023516280152146454,
+      "loss": 0.881,
+      "step": 9050
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2316828817129135,
+      "learning_rate": 0.00023507807089405064,
+      "loss": 0.8915,
+      "step": 9055
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.21704962849617004,
+      "learning_rate": 0.00023499330022607124,
+      "loss": 0.8677,
+      "step": 9060
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.26538321375846863,
+      "learning_rate": 0.0002349084895574222,
+      "loss": 0.8538,
+      "step": 9065
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.21919500827789307,
+      "learning_rate": 0.00023482363892801827,
+      "loss": 0.8222,
+      "step": 9070
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.27171579003334045,
+      "learning_rate": 0.00023473874837779294,
+      "loss": 0.8996,
+      "step": 9075
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.20920461416244507,
+      "learning_rate": 0.0002346538179466985,
+      "loss": 0.8708,
+      "step": 9080
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.23751945793628693,
+      "learning_rate": 0.00023456884767470614,
+      "loss": 0.9058,
+      "step": 9085
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.23487117886543274,
+      "learning_rate": 0.0002344838376018056,
+      "loss": 0.9029,
+      "step": 9090
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.20502355694770813,
+      "learning_rate": 0.00023439878776800542,
+      "loss": 0.9057,
+      "step": 9095
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.26424431800842285,
+      "learning_rate": 0.00023431369821333293,
+      "loss": 0.9132,
+      "step": 9100
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2559512257575989,
+      "learning_rate": 0.00023422856897783412,
+      "loss": 0.8964,
+      "step": 9105
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.21677064895629883,
+      "learning_rate": 0.0002341434001015736,
+      "loss": 0.851,
+      "step": 9110
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2509647607803345,
+      "learning_rate": 0.00023405819162463466,
+      "loss": 0.8652,
+      "step": 9115
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.20928482711315155,
+      "learning_rate": 0.00023397294358711924,
+      "loss": 0.8989,
+      "step": 9120
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2391272932291031,
+      "learning_rate": 0.00023388765602914792,
+      "loss": 1.0029,
+      "step": 9125
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.21468204259872437,
+      "learning_rate": 0.0002338023289908599,
+      "loss": 0.817,
+      "step": 9130
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2177235335111618,
+      "learning_rate": 0.00023371696251241279,
+      "loss": 0.8765,
+      "step": 9135
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2142833024263382,
+      "learning_rate": 0.000233631556633983,
+      "loss": 0.8559,
+      "step": 9140
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.22310085594654083,
+      "learning_rate": 0.00023354611139576536,
+      "loss": 0.8448,
+      "step": 9145
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2540875971317291,
+      "learning_rate": 0.00023346062683797324,
+      "loss": 0.7871,
+      "step": 9150
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2475668340921402,
+      "learning_rate": 0.0002333751030008384,
+      "loss": 0.923,
+      "step": 9155
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.24676388502120972,
+      "learning_rate": 0.0002332895399246114,
+      "loss": 0.792,
+      "step": 9160
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.22518321871757507,
+      "learning_rate": 0.00023320393764956086,
+      "loss": 0.8875,
+      "step": 9165
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.24502132833003998,
+      "learning_rate": 0.00023311829621597418,
+      "loss": 0.9119,
+      "step": 9170
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.21606354415416718,
+      "learning_rate": 0.00023303261566415704,
+      "loss": 0.908,
+      "step": 9175
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2341851145029068,
+      "learning_rate": 0.00023294689603443352,
+      "loss": 0.8705,
+      "step": 9180
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.23818925023078918,
+      "learning_rate": 0.00023286113736714612,
+      "loss": 0.9589,
+      "step": 9185
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2910539507865906,
+      "learning_rate": 0.0002327753397026558,
+      "loss": 0.8729,
+      "step": 9190
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.24293243885040283,
+      "learning_rate": 0.0002326895030813417,
+      "loss": 0.8785,
+      "step": 9195
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.22408191859722137,
+      "learning_rate": 0.0002326036275436014,
+      "loss": 0.8641,
+      "step": 9200
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2291467785835266,
+      "learning_rate": 0.00023251771312985084,
+      "loss": 0.8185,
+      "step": 9205
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.24368716776371002,
+      "learning_rate": 0.0002324317598805241,
+      "loss": 0.8201,
+      "step": 9210
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.24222686886787415,
+      "learning_rate": 0.00023234576783607373,
+      "loss": 0.9468,
+      "step": 9215
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.23429812490940094,
+      "learning_rate": 0.00023225973703697037,
+      "loss": 0.9275,
+      "step": 9220
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.23759447038173676,
+      "learning_rate": 0.000232173667523703,
+      "loss": 0.8471,
+      "step": 9225
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2510417401790619,
+      "learning_rate": 0.00023208755933677881,
+      "loss": 0.8939,
+      "step": 9230
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.23719042539596558,
+      "learning_rate": 0.00023200141251672314,
+      "loss": 0.8509,
+      "step": 9235
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.21512798964977264,
+      "learning_rate": 0.0002319152271040796,
+      "loss": 0.8789,
+      "step": 9240
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2508089542388916,
+      "learning_rate": 0.00023182900313940979,
+      "loss": 0.8954,
+      "step": 9245
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.22558368742465973,
+      "learning_rate": 0.00023174274066329367,
+      "loss": 0.9115,
+      "step": 9250
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2333495169878006,
+      "learning_rate": 0.00023165643971632924,
+      "loss": 0.9437,
+      "step": 9255
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.27782970666885376,
+      "learning_rate": 0.00023157010033913252,
+      "loss": 0.9699,
+      "step": 9260
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2564479410648346,
+      "learning_rate": 0.0002314837225723377,
+      "loss": 0.8747,
+      "step": 9265
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2404770702123642,
+      "learning_rate": 0.0002313973064565971,
+      "loss": 0.891,
+      "step": 9270
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.22722746431827545,
+      "learning_rate": 0.00023131085203258092,
+      "loss": 0.8386,
+      "step": 9275
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.21286796033382416,
+      "learning_rate": 0.00023122435934097755,
+      "loss": 0.9262,
+      "step": 9280
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.22385799884796143,
+      "learning_rate": 0.00023113782842249328,
+      "loss": 0.857,
+      "step": 9285
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.222075417637825,
+      "learning_rate": 0.00023105125931785245,
+      "loss": 0.894,
+      "step": 9290
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.23088407516479492,
+      "learning_rate": 0.00023096465206779736,
+      "loss": 0.8705,
+      "step": 9295
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2188183218240738,
+      "learning_rate": 0.00023087800671308826,
+      "loss": 0.8935,
+      "step": 9300
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2047409862279892,
+      "learning_rate": 0.0002307913232945033,
+      "loss": 0.8515,
+      "step": 9305
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.23035292327404022,
+      "learning_rate": 0.00023070460185283862,
+      "loss": 0.9546,
+      "step": 9310
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2521415054798126,
+      "learning_rate": 0.00023061784242890817,
+      "loss": 0.8708,
+      "step": 9315
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.2409847378730774,
+      "learning_rate": 0.00023053104506354387,
+      "loss": 0.857,
+      "step": 9320
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.31675606966018677,
+      "learning_rate": 0.0002304442097975954,
+      "loss": 0.9459,
+      "step": 9325
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.30147692561149597,
+      "learning_rate": 0.00023035733667193034,
+      "loss": 0.8456,
+      "step": 9330
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.24489833414554596,
+      "learning_rate": 0.00023027042572743405,
+      "loss": 0.8259,
+      "step": 9335
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.2613428831100464,
+      "learning_rate": 0.00023018347700500973,
+      "loss": 1.0109,
+      "step": 9340
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.24886786937713623,
+      "learning_rate": 0.00023009649054557828,
+      "loss": 0.8927,
+      "step": 9345
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.2450346052646637,
+      "learning_rate": 0.00023000946639007848,
+      "loss": 0.9878,
+      "step": 9350
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.2285212278366089,
+      "learning_rate": 0.00022992240457946674,
+      "loss": 0.8819,
+      "step": 9355
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.22266454994678497,
+      "learning_rate": 0.00022983530515471732,
+      "loss": 0.8673,
+      "step": 9360
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.228920578956604,
+      "learning_rate": 0.000229748168156822,
+      "loss": 0.9104,
+      "step": 9365
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.22747859358787537,
+      "learning_rate": 0.00022966099362679038,
+      "loss": 0.8806,
+      "step": 9370
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.24505122005939484,
+      "learning_rate": 0.00022957378160564973,
+      "loss": 0.9054,
+      "step": 9375
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.22118686139583588,
+      "learning_rate": 0.00022948653213444487,
+      "loss": 0.8939,
+      "step": 9380
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.21411919593811035,
+      "learning_rate": 0.00022939924525423834,
+      "loss": 0.9658,
+      "step": 9385
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.25308582186698914,
+      "learning_rate": 0.00022931192100611022,
+      "loss": 0.9179,
+      "step": 9390
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.24361754953861237,
+      "learning_rate": 0.0002292245594311582,
+      "loss": 0.95,
+      "step": 9395
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.23733997344970703,
+      "learning_rate": 0.00022913716057049757,
+      "loss": 0.8376,
+      "step": 9400
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.23433391749858856,
+      "learning_rate": 0.0002290497244652611,
+      "loss": 0.9291,
+      "step": 9405
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.22158537805080414,
+      "learning_rate": 0.00022896225115659913,
+      "loss": 0.8927,
+      "step": 9410
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.21899032592773438,
+      "learning_rate": 0.00022887474068567952,
+      "loss": 0.8168,
+      "step": 9415
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.23201830685138702,
+      "learning_rate": 0.0002287871930936876,
+      "loss": 0.9301,
+      "step": 9420
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.22301717102527618,
+      "learning_rate": 0.00022869960842182614,
+      "loss": 0.8611,
+      "step": 9425
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2502361238002777,
+      "learning_rate": 0.00022861198671131542,
+      "loss": 0.9485,
+      "step": 9430
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.24774372577667236,
+      "learning_rate": 0.00022852432800339313,
+      "loss": 0.8949,
+      "step": 9435
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2235736846923828,
+      "learning_rate": 0.00022843663233931442,
+      "loss": 0.8333,
+      "step": 9440
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.23626188933849335,
+      "learning_rate": 0.0002283488997603517,
+      "loss": 0.8102,
+      "step": 9445
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.25719600915908813,
+      "learning_rate": 0.00022826113030779486,
+      "loss": 0.8408,
+      "step": 9450
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2473081797361374,
+      "learning_rate": 0.00022817332402295113,
+      "loss": 0.81,
+      "step": 9455
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.20688697695732117,
+      "learning_rate": 0.00022808548094714506,
+      "loss": 0.896,
+      "step": 9460
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.26973938941955566,
+      "learning_rate": 0.0002279976011217185,
+      "loss": 0.878,
+      "step": 9465
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.24812696874141693,
+      "learning_rate": 0.00022790968458803065,
+      "loss": 0.7833,
+      "step": 9470
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.22464975714683533,
+      "learning_rate": 0.00022782173138745793,
+      "loss": 0.8938,
+      "step": 9475
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.23503178358078003,
+      "learning_rate": 0.00022773374156139406,
+      "loss": 0.8603,
+      "step": 9480
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.236893430352211,
+      "learning_rate": 0.00022764571515124994,
+      "loss": 0.8637,
+      "step": 9485
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.24112752079963684,
+      "learning_rate": 0.00022755765219845372,
+      "loss": 0.893,
+      "step": 9490
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2311786562204361,
+      "learning_rate": 0.00022746955274445077,
+      "loss": 0.8524,
+      "step": 9495
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.22384856641292572,
+      "learning_rate": 0.0002273814168307036,
+      "loss": 0.8552,
+      "step": 9500
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2306017130613327,
+      "learning_rate": 0.00022729324449869198,
+      "loss": 0.9581,
+      "step": 9505
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.23824800550937653,
+      "learning_rate": 0.0002272050357899126,
+      "loss": 0.8444,
+      "step": 9510
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.22070693969726562,
+      "learning_rate": 0.00022711679074587953,
+      "loss": 0.9218,
+      "step": 9515
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.25660625100135803,
+      "learning_rate": 0.0002270285094081237,
+      "loss": 0.9183,
+      "step": 9520
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.22683514654636383,
+      "learning_rate": 0.00022694019181819337,
+      "loss": 0.8944,
+      "step": 9525
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2356630563735962,
+      "learning_rate": 0.0002268518380176536,
+      "loss": 0.86,
+      "step": 9530
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2331274300813675,
+      "learning_rate": 0.00022676344804808675,
+      "loss": 0.9143,
+      "step": 9535
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.22305701673030853,
+      "learning_rate": 0.00022667502195109198,
+      "loss": 0.8527,
+      "step": 9540
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.23889246582984924,
+      "learning_rate": 0.00022658655976828557,
+      "loss": 0.9313,
+      "step": 9545
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.25855275988578796,
+      "learning_rate": 0.00022649806154130078,
+      "loss": 0.9187,
+      "step": 9550
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.21481330692768097,
+      "learning_rate": 0.00022640952731178786,
+      "loss": 0.8408,
+      "step": 9555
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2593211531639099,
+      "learning_rate": 0.00022632095712141382,
+      "loss": 0.8638,
+      "step": 9560
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2373592108488083,
+      "learning_rate": 0.0002262323510118629,
+      "loss": 0.8963,
+      "step": 9565
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2230243980884552,
+      "learning_rate": 0.0002261437090248359,
+      "loss": 0.8791,
+      "step": 9570
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.21514588594436646,
+      "learning_rate": 0.00022605503120205086,
+      "loss": 0.8647,
+      "step": 9575
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.22944727540016174,
+      "learning_rate": 0.0002259663175852424,
+      "loss": 0.9185,
+      "step": 9580
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.2357485294342041,
+      "learning_rate": 0.00022587756821616214,
+      "loss": 0.8481,
+      "step": 9585
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.21694447100162506,
+      "learning_rate": 0.00022578878313657844,
+      "loss": 0.9085,
+      "step": 9590
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.23980222642421722,
+      "learning_rate": 0.00022569996238827654,
+      "loss": 0.8727,
+      "step": 9595
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.21908119320869446,
+      "learning_rate": 0.00022561110601305838,
+      "loss": 0.8425,
+      "step": 9600
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.2655925452709198,
+      "learning_rate": 0.0002255222140527428,
+      "loss": 0.9436,
+      "step": 9605
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.25513389706611633,
+      "learning_rate": 0.0002254332865491653,
+      "loss": 0.9691,
+      "step": 9610
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.2476007342338562,
+      "learning_rate": 0.000225344323544178,
+      "loss": 0.9306,
+      "step": 9615
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.2203802466392517,
+      "learning_rate": 0.00022525532507965004,
+      "loss": 0.9796,
+      "step": 9620
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.21193300187587738,
+      "learning_rate": 0.0002251662911974669,
+      "loss": 0.8913,
+      "step": 9625
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.21106500923633575,
+      "learning_rate": 0.0002250772219395309,
+      "loss": 0.8984,
+      "step": 9630
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.24506857991218567,
+      "learning_rate": 0.00022498811734776103,
+      "loss": 0.8861,
+      "step": 9635
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.22095774114131927,
+      "learning_rate": 0.0002248989774640929,
+      "loss": 0.8482,
+      "step": 9640
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.24854597449302673,
+      "learning_rate": 0.0002248098023304786,
+      "loss": 0.8741,
+      "step": 9645
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.22818954288959503,
+      "learning_rate": 0.00022472059198888698,
+      "loss": 0.8486,
+      "step": 9650
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.24147124588489532,
+      "learning_rate": 0.0002246313464813034,
+      "loss": 0.9675,
+      "step": 9655
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.22634311020374298,
+      "learning_rate": 0.00022454206584972971,
+      "loss": 0.8038,
+      "step": 9660
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.23374037444591522,
+      "learning_rate": 0.00022445275013618444,
+      "loss": 0.9025,
+      "step": 9665
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.22729557752609253,
+      "learning_rate": 0.00022436339938270236,
+      "loss": 0.8884,
+      "step": 9670
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.24079249799251556,
+      "learning_rate": 0.00022427401363133502,
+      "loss": 0.8695,
+      "step": 9675
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.22827574610710144,
+      "learning_rate": 0.0002241845929241503,
+      "loss": 0.8461,
+      "step": 9680
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.22621886432170868,
+      "learning_rate": 0.00022409513730323256,
+      "loss": 1.0162,
+      "step": 9685
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.24320857226848602,
+      "learning_rate": 0.00022400564681068264,
+      "loss": 0.899,
+      "step": 9690
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.24638378620147705,
+      "learning_rate": 0.00022391612148861764,
+      "loss": 0.8847,
+      "step": 9695
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.25771698355674744,
+      "learning_rate": 0.00022382656137917117,
+      "loss": 0.9079,
+      "step": 9700
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2110559642314911,
+      "learning_rate": 0.0002237369665244932,
+      "loss": 0.9551,
+      "step": 9705
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.220799520611763,
+      "learning_rate": 0.00022364733696675007,
+      "loss": 0.9265,
+      "step": 9710
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.23551765084266663,
+      "learning_rate": 0.00022355767274812442,
+      "loss": 0.9568,
+      "step": 9715
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.22385844588279724,
+      "learning_rate": 0.0002234679739108152,
+      "loss": 0.8029,
+      "step": 9720
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.22753435373306274,
+      "learning_rate": 0.00022337824049703764,
+      "loss": 0.9675,
+      "step": 9725
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2253807932138443,
+      "learning_rate": 0.00022328847254902333,
+      "loss": 0.8846,
+      "step": 9730
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.20921418070793152,
+      "learning_rate": 0.00022319867010901998,
+      "loss": 0.8575,
+      "step": 9735
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.22992300987243652,
+      "learning_rate": 0.0002231088332192916,
+      "loss": 0.8715,
+      "step": 9740
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.23045524954795837,
+      "learning_rate": 0.00022301896192211847,
+      "loss": 0.8612,
+      "step": 9745
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.2747400999069214,
+      "learning_rate": 0.00022292905625979694,
+      "loss": 0.8996,
+      "step": 9750
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2662562131881714,
+      "learning_rate": 0.0002228391162746397,
+      "loss": 0.8955,
+      "step": 9755
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2570030689239502,
+      "learning_rate": 0.00022274914200897533,
+      "loss": 0.9389,
+      "step": 9760
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.21170185506343842,
+      "learning_rate": 0.0002226591335051489,
+      "loss": 0.8996,
+      "step": 9765
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2330474704504013,
+      "learning_rate": 0.00022256909080552127,
+      "loss": 0.855,
+      "step": 9770
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.23178677260875702,
+      "learning_rate": 0.00022247901395246956,
+      "loss": 0.908,
+      "step": 9775
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2605157494544983,
+      "learning_rate": 0.00022238890298838696,
+      "loss": 0.9045,
+      "step": 9780
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2496998906135559,
+      "learning_rate": 0.00022229875795568262,
+      "loss": 0.8768,
+      "step": 9785
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.24557536840438843,
+      "learning_rate": 0.00022220857889678177,
+      "loss": 0.8546,
+      "step": 9790
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.22727739810943604,
+      "learning_rate": 0.00022211836585412582,
+      "loss": 0.8804,
+      "step": 9795
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.21401318907737732,
+      "learning_rate": 0.00022202811887017188,
+      "loss": 0.8564,
+      "step": 9800
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.25362133979797363,
+      "learning_rate": 0.00022193783798739325,
+      "loss": 0.925,
+      "step": 9805
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.24248819053173065,
+      "learning_rate": 0.00022184752324827902,
+      "loss": 0.8071,
+      "step": 9810
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2690879702568054,
+      "learning_rate": 0.0002217571746953344,
+      "loss": 0.926,
+      "step": 9815
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.24345597624778748,
+      "learning_rate": 0.00022166679237108037,
+      "loss": 0.9282,
+      "step": 9820
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.26057952642440796,
+      "learning_rate": 0.0002215763763180539,
+      "loss": 0.9823,
+      "step": 9825
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.23440931737422943,
+      "learning_rate": 0.00022148592657880768,
+      "loss": 0.8077,
+      "step": 9830
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.258858323097229,
+      "learning_rate": 0.00022139544319591052,
+      "loss": 0.8348,
+      "step": 9835
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.24155326187610626,
+      "learning_rate": 0.00022130492621194681,
+      "loss": 0.854,
+      "step": 9840
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.23018091917037964,
+      "learning_rate": 0.00022121437566951686,
+      "loss": 1.018,
+      "step": 9845
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.21002855896949768,
+      "learning_rate": 0.00022112379161123673,
+      "loss": 0.8137,
+      "step": 9850
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.28612980246543884,
+      "learning_rate": 0.00022103317407973837,
+      "loss": 0.8465,
+      "step": 9855
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.20760983228683472,
+      "learning_rate": 0.00022094252311766929,
+      "loss": 0.9207,
+      "step": 9860
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.23730003833770752,
+      "learning_rate": 0.00022085183876769293,
+      "loss": 0.9247,
+      "step": 9865
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.2215205430984497,
+      "learning_rate": 0.00022076112107248833,
+      "loss": 0.901,
+      "step": 9870
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.2527283728122711,
+      "learning_rate": 0.00022067037007475026,
+      "loss": 0.9053,
+      "step": 9875
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.20407399535179138,
+      "learning_rate": 0.00022057958581718915,
+      "loss": 0.9464,
+      "step": 9880
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.23334871232509613,
+      "learning_rate": 0.00022048876834253103,
+      "loss": 0.8626,
+      "step": 9885
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.22306807339191437,
+      "learning_rate": 0.00022039791769351772,
+      "loss": 0.8706,
+      "step": 9890
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.21715085208415985,
+      "learning_rate": 0.00022030703391290646,
+      "loss": 0.847,
+      "step": 9895
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.24487926065921783,
+      "learning_rate": 0.00022021611704347026,
+      "loss": 0.9132,
+      "step": 9900
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.24505330622196198,
+      "learning_rate": 0.00022012516712799756,
+      "loss": 0.9286,
+      "step": 9905
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.2754554748535156,
+      "learning_rate": 0.00022003418420929243,
+      "loss": 0.885,
+      "step": 9910
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.2651885449886322,
+      "learning_rate": 0.00021994316833017443,
+      "loss": 0.9331,
+      "step": 9915
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.25558674335479736,
+      "learning_rate": 0.00021987033192416743,
+      "loss": 1.0044,
+      "step": 9920
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2470509111881256,
+      "learning_rate": 0.00021977925682426118,
+      "loss": 0.8667,
+      "step": 9925
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.22508101165294647,
+      "learning_rate": 0.00021968814888391935,
+      "loss": 0.9218,
+      "step": 9930
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.22299574315547943,
+      "learning_rate": 0.00021959700814602035,
+      "loss": 0.9041,
+      "step": 9935
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.25101420283317566,
+      "learning_rate": 0.0002195058346534581,
+      "loss": 0.8797,
+      "step": 9940
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.23478543758392334,
+      "learning_rate": 0.00021941462844914182,
+      "loss": 0.9141,
+      "step": 9945
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.25883424282073975,
+      "learning_rate": 0.00021932338957599625,
+      "loss": 0.9015,
+      "step": 9950
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.22861811518669128,
+      "learning_rate": 0.00021923211807696133,
+      "loss": 0.8923,
+      "step": 9955
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.24520093202590942,
+      "learning_rate": 0.00021914081399499258,
+      "loss": 0.8958,
+      "step": 9960
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.23438018560409546,
+      "learning_rate": 0.00021904947737306065,
+      "loss": 1.023,
+      "step": 9965
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.24439287185668945,
+      "learning_rate": 0.0002189581082541516,
+      "loss": 0.9076,
+      "step": 9970
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.23474149405956268,
+      "learning_rate": 0.00021886670668126674,
+      "loss": 0.8026,
+      "step": 9975
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.25041329860687256,
+      "learning_rate": 0.00021877527269742277,
+      "loss": 0.9282,
+      "step": 9980
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2247781604528427,
+      "learning_rate": 0.00021868380634565147,
+      "loss": 0.907,
+      "step": 9985
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.22819451987743378,
+      "learning_rate": 0.00021859230766900004,
+      "loss": 0.8245,
+      "step": 9990
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.2385624200105667,
+      "learning_rate": 0.00021850077671053072,
+      "loss": 0.9177,
+      "step": 9995
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.21760748326778412,
+      "learning_rate": 0.00021840921351332107,
+      "loss": 0.8566,
+      "step": 10000
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.22546535730361938,
+      "learning_rate": 0.00021831761812046363,
+      "loss": 0.8671,
+      "step": 10005
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2704804539680481,
+      "learning_rate": 0.00021822599057506648,
+      "loss": 0.8348,
+      "step": 10010
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2751447558403015,
+      "learning_rate": 0.00021813433092025236,
+      "loss": 0.9228,
+      "step": 10015
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.241837278008461,
+      "learning_rate": 0.00021804263919915947,
+      "loss": 0.9471,
+      "step": 10020
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.24710319936275482,
+      "learning_rate": 0.00021795091545494092,
+      "loss": 0.959,
+      "step": 10025
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2651970684528351,
+      "learning_rate": 0.000217859159730765,
+      "loss": 0.8134,
+      "step": 10030
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2539055347442627,
+      "learning_rate": 0.00021776737206981498,
+      "loss": 0.9223,
+      "step": 10035
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.31177031993865967,
+      "learning_rate": 0.0002176755525152892,
+      "loss": 0.9064,
+      "step": 10040
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.22030460834503174,
+      "learning_rate": 0.00021758370111040094,
+      "loss": 0.8532,
+      "step": 10045
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2701607346534729,
+      "learning_rate": 0.00021749181789837858,
+      "loss": 0.8357,
+      "step": 10050
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.24372340738773346,
+      "learning_rate": 0.00021739990292246535,
+      "loss": 0.929,
+      "step": 10055
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.26125940680503845,
+      "learning_rate": 0.00021730795622591952,
+      "loss": 0.9623,
+      "step": 10060
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.23417995870113373,
+      "learning_rate": 0.00021721597785201427,
+      "loss": 0.7909,
+      "step": 10065
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.22926466166973114,
+      "learning_rate": 0.00021712396784403772,
+      "loss": 0.8785,
+      "step": 10070
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.24475359916687012,
+      "learning_rate": 0.00021703192624529272,
+      "loss": 0.902,
+      "step": 10075
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.2308359593153,
+      "learning_rate": 0.0002169398530990972,
+      "loss": 0.9083,
+      "step": 10080
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.27559104561805725,
+      "learning_rate": 0.00021684774844878376,
+      "loss": 0.8948,
+      "step": 10085
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.22297874093055725,
+      "learning_rate": 0.0002167556123377,
+      "loss": 0.8743,
+      "step": 10090
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.25241953134536743,
+      "learning_rate": 0.00021666344480920818,
+      "loss": 0.8831,
+      "step": 10095
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.2630932033061981,
+      "learning_rate": 0.0002165712459066854,
+      "loss": 0.8839,
+      "step": 10100
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.24986648559570312,
+      "learning_rate": 0.00021647901567352357,
+      "loss": 0.9109,
+      "step": 10105
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.24380378425121307,
+      "learning_rate": 0.00021638675415312924,
+      "loss": 0.9033,
+      "step": 10110
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.22641584277153015,
+      "learning_rate": 0.00021629446138892377,
+      "loss": 0.9176,
+      "step": 10115
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.2435508668422699,
+      "learning_rate": 0.0002162021374243432,
+      "loss": 0.8718,
+      "step": 10120
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.25140294432640076,
+      "learning_rate": 0.00021610978230283823,
+      "loss": 0.9289,
+      "step": 10125
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.23567943274974823,
+      "learning_rate": 0.0002160173960678743,
+      "loss": 0.8465,
+      "step": 10130
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.2589253783226013,
+      "learning_rate": 0.00021592497876293137,
+      "loss": 0.9386,
+      "step": 10135
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.2508923411369324,
+      "learning_rate": 0.0002158325304315042,
+      "loss": 0.8488,
+      "step": 10140
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.2547130584716797,
+      "learning_rate": 0.00021574005111710192,
+      "loss": 0.8462,
+      "step": 10145
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.22627414762973785,
+      "learning_rate": 0.00021564754086324844,
+      "loss": 0.8633,
+      "step": 10150
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.21264758706092834,
+      "learning_rate": 0.00021555499971348215,
+      "loss": 0.8839,
+      "step": 10155
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.21670518815517426,
+      "learning_rate": 0.00021546242771135597,
+      "loss": 0.8753,
+      "step": 10160
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.21713611483573914,
+      "learning_rate": 0.0002153698249004374,
+      "loss": 0.838,
+      "step": 10165
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.20740459859371185,
+      "learning_rate": 0.00021527719132430833,
+      "loss": 0.9157,
+      "step": 10170
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.23185713589191437,
+      "learning_rate": 0.00021518452702656528,
+      "loss": 0.8678,
+      "step": 10175
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.24864669144153595,
+      "learning_rate": 0.00021509183205081905,
+      "loss": 0.8134,
+      "step": 10180
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.21126706898212433,
+      "learning_rate": 0.00021499910644069502,
+      "loss": 0.8509,
+      "step": 10185
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.21727898716926575,
+      "learning_rate": 0.000214906350239833,
+      "loss": 0.8654,
+      "step": 10190
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.2389363795518875,
+      "learning_rate": 0.00021481356349188705,
+      "loss": 0.8315,
+      "step": 10195
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.22045078873634338,
+      "learning_rate": 0.00021472074624052573,
+      "loss": 0.851,
+      "step": 10200
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.22910571098327637,
+      "learning_rate": 0.0002146278985294319,
+      "loss": 0.9826,
+      "step": 10205
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.2421579509973526,
+      "learning_rate": 0.0002145350204023028,
+      "loss": 0.8577,
+      "step": 10210
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.2479030340909958,
+      "learning_rate": 0.00021444211190285001,
+      "loss": 0.9059,
+      "step": 10215
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.22180233895778656,
+      "learning_rate": 0.00021434917307479927,
+      "loss": 0.9248,
+      "step": 10220
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.23422615230083466,
+      "learning_rate": 0.00021425620396189071,
+      "loss": 0.8011,
+      "step": 10225
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.23456484079360962,
+      "learning_rate": 0.0002141632046078787,
+      "loss": 0.8772,
+      "step": 10230
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.24434050917625427,
+      "learning_rate": 0.00021407017505653176,
+      "loss": 0.8983,
+      "step": 10235
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.26233410835266113,
+      "learning_rate": 0.00021397711535163275,
+      "loss": 0.7669,
+      "step": 10240
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.23086436092853546,
+      "learning_rate": 0.00021388402553697863,
+      "loss": 0.85,
+      "step": 10245
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2340136170387268,
+      "learning_rate": 0.00021379090565638064,
+      "loss": 0.8472,
+      "step": 10250
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2060890644788742,
+      "learning_rate": 0.00021369775575366397,
+      "loss": 0.791,
+      "step": 10255
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.22005440294742584,
+      "learning_rate": 0.00021360457587266812,
+      "loss": 0.7792,
+      "step": 10260
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.220989391207695,
+      "learning_rate": 0.00021351136605724658,
+      "loss": 0.9087,
+      "step": 10265
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.20939388871192932,
+      "learning_rate": 0.00021341812635126706,
+      "loss": 0.8171,
+      "step": 10270
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.22090451419353485,
+      "learning_rate": 0.00021332485679861123,
+      "loss": 0.831,
+      "step": 10275
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.25128301978111267,
+      "learning_rate": 0.0002132315574431748,
+      "loss": 0.8665,
+      "step": 10280
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.23706404864788055,
+      "learning_rate": 0.00021313822832886762,
+      "loss": 0.8653,
+      "step": 10285
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.21622657775878906,
+      "learning_rate": 0.00021304486949961344,
+      "loss": 0.9074,
+      "step": 10290
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.20798739790916443,
+      "learning_rate": 0.00021295148099935,
+      "loss": 0.9282,
+      "step": 10295
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.226746067404747,
+      "learning_rate": 0.00021285806287202902,
+      "loss": 0.8985,
+      "step": 10300
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2492750883102417,
+      "learning_rate": 0.00021276461516161622,
+      "loss": 0.8985,
+      "step": 10305
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.252853125333786,
+      "learning_rate": 0.0002126711379120912,
+      "loss": 0.9173,
+      "step": 10310
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.23097646236419678,
+      "learning_rate": 0.00021257763116744744,
+      "loss": 0.9335,
+      "step": 10315
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.25283390283584595,
+      "learning_rate": 0.0002124840949716923,
+      "loss": 0.9115,
+      "step": 10320
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2866900861263275,
+      "learning_rate": 0.00021239052936884703,
+      "loss": 0.9016,
+      "step": 10325
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.23940923810005188,
+      "learning_rate": 0.0002122969344029467,
+      "loss": 0.8282,
+      "step": 10330
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.26236215233802795,
+      "learning_rate": 0.00021220331011804022,
+      "loss": 0.8244,
+      "step": 10335
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.23621851205825806,
+      "learning_rate": 0.00021210965655819031,
+      "loss": 0.8655,
+      "step": 10340
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.21117086708545685,
+      "learning_rate": 0.00021201597376747344,
+      "loss": 0.9017,
+      "step": 10345
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2675582468509674,
+      "learning_rate": 0.00021192226178997977,
+      "loss": 0.8432,
+      "step": 10350
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2619032561779022,
+      "learning_rate": 0.0002118285206698134,
+      "loss": 0.8931,
+      "step": 10355
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2394731044769287,
+      "learning_rate": 0.0002117347504510919,
+      "loss": 0.8431,
+      "step": 10360
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.22042964398860931,
+      "learning_rate": 0.00021164095117794674,
+      "loss": 0.8538,
+      "step": 10365
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.22833536565303802,
+      "learning_rate": 0.00021154712289452285,
+      "loss": 0.9355,
+      "step": 10370
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.24898435175418854,
+      "learning_rate": 0.00021145326564497903,
+      "loss": 0.9127,
+      "step": 10375
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2540437579154968,
+      "learning_rate": 0.00021135937947348757,
+      "loss": 0.9235,
+      "step": 10380
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.26430103182792664,
+      "learning_rate": 0.00021126546442423453,
+      "loss": 0.8293,
+      "step": 10385
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.24272888898849487,
+      "learning_rate": 0.0002111715205414193,
+      "loss": 0.9249,
+      "step": 10390
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2834715247154236,
+      "learning_rate": 0.00021107754786925512,
+      "loss": 0.8638,
+      "step": 10395
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.23184886574745178,
+      "learning_rate": 0.0002109835464519685,
+      "loss": 0.8379,
+      "step": 10400
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2805626392364502,
+      "learning_rate": 0.00021088951633379982,
+      "loss": 0.9148,
+      "step": 10405
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.2405596673488617,
+      "learning_rate": 0.0002107954575590026,
+      "loss": 0.9326,
+      "step": 10410
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.21941906213760376,
+      "learning_rate": 0.00021070137017184415,
+      "loss": 0.8786,
+      "step": 10415
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.24544595181941986,
+      "learning_rate": 0.000210607254216605,
+      "loss": 0.8619,
+      "step": 10420
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.22791095077991486,
+      "learning_rate": 0.00021051310973757936,
+      "loss": 0.8595,
+      "step": 10425
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2604474127292633,
+      "learning_rate": 0.00021041893677907473,
+      "loss": 0.8809,
+      "step": 10430
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.22478339076042175,
+      "learning_rate": 0.00021032473538541195,
+      "loss": 0.8569,
+      "step": 10435
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.25225207209587097,
+      "learning_rate": 0.0002102305056009254,
+      "loss": 0.8429,
+      "step": 10440
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.24379763007164001,
+      "learning_rate": 0.00021013624746996272,
+      "loss": 0.8249,
+      "step": 10445
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.22271931171417236,
+      "learning_rate": 0.00021004196103688487,
+      "loss": 0.9641,
+      "step": 10450
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2445717453956604,
+      "learning_rate": 0.00020994764634606628,
+      "loss": 0.8815,
+      "step": 10455
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.22217918932437897,
+      "learning_rate": 0.0002098533034418945,
+      "loss": 0.8908,
+      "step": 10460
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.24231913685798645,
+      "learning_rate": 0.00020975893236877048,
+      "loss": 0.864,
+      "step": 10465
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.23367081582546234,
+      "learning_rate": 0.0002096645331711083,
+      "loss": 0.8971,
+      "step": 10470
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2660883963108063,
+      "learning_rate": 0.00020957010589333546,
+      "loss": 0.8435,
+      "step": 10475
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2160717397928238,
+      "learning_rate": 0.00020947565057989249,
+      "loss": 0.8609,
+      "step": 10480
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.26367098093032837,
+      "learning_rate": 0.00020938116727523324,
+      "loss": 0.9108,
+      "step": 10485
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.23195737600326538,
+      "learning_rate": 0.0002092866560238247,
+      "loss": 0.8359,
+      "step": 10490
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2410704344511032,
+      "learning_rate": 0.00020919211687014697,
+      "loss": 0.8616,
+      "step": 10495
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.23946736752986908,
+      "learning_rate": 0.00020909754985869335,
+      "loss": 0.7911,
+      "step": 10500
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.23445159196853638,
+      "learning_rate": 0.0002090029550339702,
+      "loss": 0.8216,
+      "step": 10505
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2318773865699768,
+      "learning_rate": 0.00020890833244049695,
+      "loss": 0.7892,
+      "step": 10510
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.22648349404335022,
+      "learning_rate": 0.0002088136821228062,
+      "loss": 0.9097,
+      "step": 10515
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.23851384222507477,
+      "learning_rate": 0.00020871900412544345,
+      "loss": 0.8725,
+      "step": 10520
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.22477130591869354,
+      "learning_rate": 0.00020862429849296743,
+      "loss": 0.8717,
+      "step": 10525
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.22048017382621765,
+      "learning_rate": 0.00020852956526994963,
+      "loss": 0.8755,
+      "step": 10530
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2534314692020416,
+      "learning_rate": 0.00020843480450097473,
+      "loss": 0.8789,
+      "step": 10535
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2241785079240799,
+      "learning_rate": 0.0002083400162306403,
+      "loss": 0.818,
+      "step": 10540
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.25126171112060547,
+      "learning_rate": 0.00020824520050355681,
+      "loss": 0.8867,
+      "step": 10545
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.24423153698444366,
+      "learning_rate": 0.00020815035736434766,
+      "loss": 0.9067,
+      "step": 10550
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2558138072490692,
+      "learning_rate": 0.00020805548685764923,
+      "loss": 0.8161,
+      "step": 10555
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.23357480764389038,
+      "learning_rate": 0.00020796058902811075,
+      "loss": 0.8864,
+      "step": 10560
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2539251446723938,
+      "learning_rate": 0.00020786566392039428,
+      "loss": 0.881,
+      "step": 10565
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2158096432685852,
+      "learning_rate": 0.00020777071157917468,
+      "loss": 0.8784,
+      "step": 10570
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.22970278561115265,
+      "learning_rate": 0.0002076757320491397,
+      "loss": 0.9085,
+      "step": 10575
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.2667364776134491,
+      "learning_rate": 0.00020758072537498985,
+      "loss": 0.896,
+      "step": 10580
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.22878745198249817,
+      "learning_rate": 0.00020748569160143845,
+      "loss": 0.9321,
+      "step": 10585
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2596031725406647,
+      "learning_rate": 0.00020739063077321155,
+      "loss": 0.9034,
+      "step": 10590
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.23091496527194977,
+      "learning_rate": 0.00020729554293504794,
+      "loss": 0.9254,
+      "step": 10595
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.258047491312027,
+      "learning_rate": 0.00020720042813169906,
+      "loss": 0.7867,
+      "step": 10600
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.23771882057189941,
+      "learning_rate": 0.00020710528640792916,
+      "loss": 0.8768,
+      "step": 10605
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.22498424351215363,
+      "learning_rate": 0.0002070101178085151,
+      "loss": 0.9163,
+      "step": 10610
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2267744243144989,
+      "learning_rate": 0.0002069149223782463,
+      "loss": 0.872,
+      "step": 10615
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.23287959396839142,
+      "learning_rate": 0.00020681970016192495,
+      "loss": 0.9396,
+      "step": 10620
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.23463228344917297,
+      "learning_rate": 0.00020672445120436582,
+      "loss": 0.9393,
+      "step": 10625
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.24994787573814392,
+      "learning_rate": 0.00020662917555039616,
+      "loss": 0.827,
+      "step": 10630
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.24779953062534332,
+      "learning_rate": 0.00020653387324485588,
+      "loss": 0.9111,
+      "step": 10635
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.27712857723236084,
+      "learning_rate": 0.00020643854433259742,
+      "loss": 0.9207,
+      "step": 10640
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.25297531485557556,
+      "learning_rate": 0.0002063431888584858,
+      "loss": 0.7711,
+      "step": 10645
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.22088488936424255,
+      "learning_rate": 0.0002062478068673983,
+      "loss": 0.8474,
+      "step": 10650
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2587786018848419,
+      "learning_rate": 0.00020615239840422506,
+      "loss": 0.9111,
+      "step": 10655
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2736605107784271,
+      "learning_rate": 0.00020605696351386828,
+      "loss": 0.9088,
+      "step": 10660
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.275921493768692,
+      "learning_rate": 0.00020596150224124293,
+      "loss": 0.8737,
+      "step": 10665
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.24181650578975677,
+      "learning_rate": 0.00020586601463127611,
+      "loss": 0.9555,
+      "step": 10670
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.25569936633110046,
+      "learning_rate": 0.00020577050072890764,
+      "loss": 0.9392,
+      "step": 10675
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.22888949513435364,
+      "learning_rate": 0.00020567496057908948,
+      "loss": 0.8406,
+      "step": 10680
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.22984679043293,
+      "learning_rate": 0.0002055793942267859,
+      "loss": 0.7473,
+      "step": 10685
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.21859197318553925,
+      "learning_rate": 0.00020548380171697366,
+      "loss": 0.9393,
+      "step": 10690
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.23983286321163177,
+      "learning_rate": 0.00020538818309464178,
+      "loss": 0.7875,
+      "step": 10695
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.23376217484474182,
+      "learning_rate": 0.00020529253840479155,
+      "loss": 0.9696,
+      "step": 10700
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2256166785955429,
+      "learning_rate": 0.00020519686769243653,
+      "loss": 0.9457,
+      "step": 10705
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.24345454573631287,
+      "learning_rate": 0.00020510117100260255,
+      "loss": 0.77,
+      "step": 10710
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2544359862804413,
+      "learning_rate": 0.00020500544838032765,
+      "loss": 0.895,
+      "step": 10715
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2471429854631424,
+      "learning_rate": 0.00020490969987066207,
+      "loss": 0.9008,
+      "step": 10720
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.22353385388851166,
+      "learning_rate": 0.00020481392551866827,
+      "loss": 0.9069,
+      "step": 10725
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.24593974649906158,
+      "learning_rate": 0.00020471812536942074,
+      "loss": 1.057,
+      "step": 10730
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.2564097046852112,
+      "learning_rate": 0.00020462229946800634,
+      "loss": 0.8346,
+      "step": 10735
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.22804971039295197,
+      "learning_rate": 0.0002045264478595238,
+      "loss": 0.8141,
+      "step": 10740
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.25203844904899597,
+      "learning_rate": 0.00020443057058908417,
+      "loss": 0.8955,
+      "step": 10745
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.24008525907993317,
+      "learning_rate": 0.0002043346677018104,
+      "loss": 0.8148,
+      "step": 10750
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.27215099334716797,
+      "learning_rate": 0.00020423873924283763,
+      "loss": 0.8765,
+      "step": 10755
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2237570732831955,
+      "learning_rate": 0.000204142785257313,
+      "loss": 0.9221,
+      "step": 10760
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.24160721898078918,
+      "learning_rate": 0.00020404680579039558,
+      "loss": 0.8637,
+      "step": 10765
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.23897403478622437,
+      "learning_rate": 0.0002039508008872565,
+      "loss": 0.7704,
+      "step": 10770
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.23746059834957123,
+      "learning_rate": 0.00020385477059307885,
+      "loss": 0.8945,
+      "step": 10775
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.23488637804985046,
+      "learning_rate": 0.0002037587149530577,
+      "loss": 0.8504,
+      "step": 10780
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.26902300119400024,
+      "learning_rate": 0.00020366263401240005,
+      "loss": 0.8331,
+      "step": 10785
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.25453630089759827,
+      "learning_rate": 0.00020356652781632477,
+      "loss": 0.913,
+      "step": 10790
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.23822472989559174,
+      "learning_rate": 0.00020347039641006257,
+      "loss": 0.8158,
+      "step": 10795
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.25680580735206604,
+      "learning_rate": 0.00020337423983885617,
+      "loss": 0.7857,
+      "step": 10800
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2617832124233246,
+      "learning_rate": 0.00020327805814795993,
+      "loss": 0.9449,
+      "step": 10805
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.23507827520370483,
+      "learning_rate": 0.00020318185138264026,
+      "loss": 0.9813,
+      "step": 10810
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2346469610929489,
+      "learning_rate": 0.00020308561958817518,
+      "loss": 0.8555,
+      "step": 10815
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2674165666103363,
+      "learning_rate": 0.00020298936280985466,
+      "loss": 0.9193,
+      "step": 10820
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.21533791720867157,
+      "learning_rate": 0.00020289308109298028,
+      "loss": 0.8939,
+      "step": 10825
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.23552724719047546,
+      "learning_rate": 0.0002027967744828654,
+      "loss": 0.8654,
+      "step": 10830
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.26331043243408203,
+      "learning_rate": 0.00020270044302483505,
+      "loss": 0.801,
+      "step": 10835
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.240268275141716,
+      "learning_rate": 0.00020260408676422615,
+      "loss": 0.7792,
+      "step": 10840
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.23092767596244812,
+      "learning_rate": 0.00020250770574638708,
+      "loss": 0.9452,
+      "step": 10845
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2468583583831787,
+      "learning_rate": 0.00020241130001667797,
+      "loss": 0.8753,
+      "step": 10850
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.27682459354400635,
+      "learning_rate": 0.0002023148696204705,
+      "loss": 0.8244,
+      "step": 10855
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.27396732568740845,
+      "learning_rate": 0.00020221841460314814,
+      "loss": 0.8651,
+      "step": 10860
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.24095529317855835,
+      "learning_rate": 0.00020212193501010573,
+      "loss": 0.8664,
+      "step": 10865
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.24401700496673584,
+      "learning_rate": 0.00020202543088674977,
+      "loss": 0.8406,
+      "step": 10870
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.22764812409877777,
+      "learning_rate": 0.0002019289022784983,
+      "loss": 0.9253,
+      "step": 10875
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2579391598701477,
+      "learning_rate": 0.00020183234923078092,
+      "loss": 0.9783,
+      "step": 10880
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.23062071204185486,
+      "learning_rate": 0.0002017357717890387,
+      "loss": 0.8766,
+      "step": 10885
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.23495079576969147,
+      "learning_rate": 0.00020163916999872418,
+      "loss": 0.8313,
+      "step": 10890
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2631068527698517,
+      "learning_rate": 0.00020154254390530142,
+      "loss": 0.9304,
+      "step": 10895
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2535402178764343,
+      "learning_rate": 0.00020144589355424578,
+      "loss": 0.8889,
+      "step": 10900
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2067861258983612,
+      "learning_rate": 0.00020134921899104416,
+      "loss": 0.8527,
+      "step": 10905
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.2404996007680893,
+      "learning_rate": 0.00020125252026119487,
+      "loss": 0.8413,
+      "step": 10910
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.27104899287223816,
+      "learning_rate": 0.00020115579741020745,
+      "loss": 0.8226,
+      "step": 10915
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.21938621997833252,
+      "learning_rate": 0.000201059050483603,
+      "loss": 0.85,
+      "step": 10920
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.2280419021844864,
+      "learning_rate": 0.00020096227952691366,
+      "loss": 0.8882,
+      "step": 10925
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.23701699078083038,
+      "learning_rate": 0.00020086548458568326,
+      "loss": 0.8571,
+      "step": 10930
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.22839820384979248,
+      "learning_rate": 0.00020076866570546662,
+      "loss": 0.797,
+      "step": 10935
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.23596982657909393,
+      "learning_rate": 0.00020067182293182994,
+      "loss": 0.9124,
+      "step": 10940
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.22004413604736328,
+      "learning_rate": 0.00020057495631035056,
+      "loss": 0.9069,
+      "step": 10945
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.23677165806293488,
+      "learning_rate": 0.00020047806588661726,
+      "loss": 0.8417,
+      "step": 10950
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.2435624748468399,
+      "learning_rate": 0.00020038115170622982,
+      "loss": 0.7833,
+      "step": 10955
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.2496495395898819,
+      "learning_rate": 0.00020028421381479926,
+      "loss": 0.8015,
+      "step": 10960
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.23711077868938446,
+      "learning_rate": 0.0002001872522579478,
+      "loss": 0.8105,
+      "step": 10965
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.23245902359485626,
+      "learning_rate": 0.0002000902670813088,
+      "loss": 0.973,
+      "step": 10970
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.23252607882022858,
+      "learning_rate": 0.0001999932583305266,
+      "loss": 0.8141,
+      "step": 10975
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.22429288923740387,
+      "learning_rate": 0.00019989622605125684,
+      "loss": 0.9489,
+      "step": 10980
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.2248774915933609,
+      "learning_rate": 0.00019979917028916606,
+      "loss": 0.9385,
+      "step": 10985
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.2719941735267639,
+      "learning_rate": 0.00019970209108993202,
+      "loss": 0.8649,
+      "step": 10990
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.2678004503250122,
+      "learning_rate": 0.0001996049884992433,
+      "loss": 0.8,
+      "step": 10995
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2112882286310196,
+      "learning_rate": 0.0001995078625627997,
+      "loss": 0.8478,
+      "step": 11000
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2869125306606293,
+      "learning_rate": 0.00019941071332631188,
+      "loss": 0.8296,
+      "step": 11005
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.21323542296886444,
+      "learning_rate": 0.00019931354083550147,
+      "loss": 0.8881,
+      "step": 11010
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.23022828996181488,
+      "learning_rate": 0.0001992163451361011,
+      "loss": 0.8604,
+      "step": 11015
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.23515096306800842,
+      "learning_rate": 0.00019911912627385426,
+      "loss": 0.9111,
+      "step": 11020
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2547999918460846,
+      "learning_rate": 0.00019902188429451542,
+      "loss": 0.881,
+      "step": 11025
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.28279218077659607,
+      "learning_rate": 0.0001989246192438499,
+      "loss": 0.908,
+      "step": 11030
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.24193325638771057,
+      "learning_rate": 0.00019882733116763376,
+      "loss": 0.8769,
+      "step": 11035
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2332753986120224,
+      "learning_rate": 0.00019873002011165409,
+      "loss": 0.803,
+      "step": 11040
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.24493838846683502,
+      "learning_rate": 0.00019863268612170873,
+      "loss": 0.825,
+      "step": 11045
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.22939081490039825,
+      "learning_rate": 0.00019853532924360618,
+      "loss": 0.9395,
+      "step": 11050
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.24580281972885132,
+      "learning_rate": 0.0001984379495231659,
+      "loss": 0.8964,
+      "step": 11055
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.22950120270252228,
+      "learning_rate": 0.00019834054700621802,
+      "loss": 0.8883,
+      "step": 11060
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2450743019580841,
+      "learning_rate": 0.00019824312173860332,
+      "loss": 0.8721,
+      "step": 11065
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.2725335955619812,
+      "learning_rate": 0.00019814567376617346,
+      "loss": 0.8063,
+      "step": 11070
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.27576860785484314,
+      "learning_rate": 0.00019804820313479066,
+      "loss": 0.9648,
+      "step": 11075
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.23889948427677155,
+      "learning_rate": 0.00019795070989032788,
+      "loss": 0.8958,
+      "step": 11080
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.25442391633987427,
+      "learning_rate": 0.00019785319407866853,
+      "loss": 0.838,
+      "step": 11085
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.26583176851272583,
+      "learning_rate": 0.00019775565574570698,
+      "loss": 0.8578,
+      "step": 11090
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.25874045491218567,
+      "learning_rate": 0.00019765809493734786,
+      "loss": 0.8606,
+      "step": 11095
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.24192367494106293,
+      "learning_rate": 0.00019756051169950663,
+      "loss": 0.7974,
+      "step": 11100
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.23584434390068054,
+      "learning_rate": 0.0001974629060781091,
+      "loss": 0.8465,
+      "step": 11105
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.28704845905303955,
+      "learning_rate": 0.00019736527811909185,
+      "loss": 0.9114,
+      "step": 11110
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.21719807386398315,
+      "learning_rate": 0.00019726762786840177,
+      "loss": 0.9526,
+      "step": 11115
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.23290511965751648,
+      "learning_rate": 0.00019716995537199624,
+      "loss": 0.8695,
+      "step": 11120
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.26695218682289124,
+      "learning_rate": 0.00019707226067584326,
+      "loss": 0.8904,
+      "step": 11125
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.225587397813797,
+      "learning_rate": 0.0001969745438259212,
+      "loss": 0.7754,
+      "step": 11130
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2279675006866455,
+      "learning_rate": 0.00019687680486821883,
+      "loss": 0.9336,
+      "step": 11135
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2307017594575882,
+      "learning_rate": 0.00019677904384873536,
+      "loss": 0.8192,
+      "step": 11140
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2410012036561966,
+      "learning_rate": 0.00019668126081348036,
+      "loss": 0.8694,
+      "step": 11145
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.26555535197257996,
+      "learning_rate": 0.00019658345580847382,
+      "loss": 0.8598,
+      "step": 11150
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.225142240524292,
+      "learning_rate": 0.00019648562887974598,
+      "loss": 0.7988,
+      "step": 11155
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2711564600467682,
+      "learning_rate": 0.00019638778007333743,
+      "loss": 0.8018,
+      "step": 11160
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2621377408504486,
+      "learning_rate": 0.00019628990943529909,
+      "loss": 0.8682,
+      "step": 11165
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.23893457651138306,
+      "learning_rate": 0.00019619201701169217,
+      "loss": 0.8165,
+      "step": 11170
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2528732419013977,
+      "learning_rate": 0.00019609410284858797,
+      "loss": 0.8636,
+      "step": 11175
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.24171330034732819,
+      "learning_rate": 0.0001959961669920683,
+      "loss": 0.9089,
+      "step": 11180
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.24632695317268372,
+      "learning_rate": 0.00019589820948822493,
+      "loss": 0.92,
+      "step": 11185
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2687559723854065,
+      "learning_rate": 0.00019580023038316,
+      "loss": 0.8794,
+      "step": 11190
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2685549259185791,
+      "learning_rate": 0.0001957022297229856,
+      "loss": 0.8629,
+      "step": 11195
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.23067381978034973,
+      "learning_rate": 0.00019560420755382416,
+      "loss": 0.836,
+      "step": 11200
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2574182450771332,
+      "learning_rate": 0.0001955061639218082,
+      "loss": 0.8133,
+      "step": 11205
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2490427941083908,
+      "learning_rate": 0.00019540809887308032,
+      "loss": 0.7957,
+      "step": 11210
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.24531954526901245,
+      "learning_rate": 0.00019531001245379312,
+      "loss": 0.9497,
+      "step": 11215
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2284311205148697,
+      "learning_rate": 0.00019521190471010936,
+      "loss": 0.8048,
+      "step": 11220
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.23735016584396362,
+      "learning_rate": 0.00019511377568820184,
+      "loss": 0.7849,
+      "step": 11225
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.24056896567344666,
+      "learning_rate": 0.00019501562543425329,
+      "loss": 0.8545,
+      "step": 11230
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2434719055891037,
+      "learning_rate": 0.00019491745399445644,
+      "loss": 0.8467,
+      "step": 11235
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.24002546072006226,
+      "learning_rate": 0.0001948192614150141,
+      "loss": 0.8546,
+      "step": 11240
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2705593705177307,
+      "learning_rate": 0.00019472104774213893,
+      "loss": 0.9254,
+      "step": 11245
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2447061389684677,
+      "learning_rate": 0.00019462281302205355,
+      "loss": 0.946,
+      "step": 11250
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.24118636548519135,
+      "learning_rate": 0.0001945245573009905,
+      "loss": 0.8431,
+      "step": 11255
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.22377237677574158,
+      "learning_rate": 0.0001944262806251922,
+      "loss": 0.9679,
+      "step": 11260
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.24068385362625122,
+      "learning_rate": 0.00019432798304091085,
+      "loss": 0.928,
+      "step": 11265
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2337529957294464,
+      "learning_rate": 0.0001942296645944086,
+      "loss": 0.9368,
+      "step": 11270
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.23204030096530914,
+      "learning_rate": 0.00019413132533195737,
+      "loss": 0.9575,
+      "step": 11275
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.23348142206668854,
+      "learning_rate": 0.00019403296529983888,
+      "loss": 0.8327,
+      "step": 11280
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.22873573005199432,
+      "learning_rate": 0.00019393458454434464,
+      "loss": 0.7914,
+      "step": 11285
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2327519953250885,
+      "learning_rate": 0.00019383618311177587,
+      "loss": 0.8401,
+      "step": 11290
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2520465850830078,
+      "learning_rate": 0.00019373776104844362,
+      "loss": 0.8922,
+      "step": 11295
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2645134925842285,
+      "learning_rate": 0.00019363931840066847,
+      "loss": 0.8806,
+      "step": 11300
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2500244677066803,
+      "learning_rate": 0.00019354085521478088,
+      "loss": 0.9306,
+      "step": 11305
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2344416230916977,
+      "learning_rate": 0.0001934423715371209,
+      "loss": 0.9232,
+      "step": 11310
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.2218599170446396,
+      "learning_rate": 0.0001933438674140382,
+      "loss": 0.8308,
+      "step": 11315
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.25029993057250977,
+      "learning_rate": 0.00019324534289189203,
+      "loss": 0.861,
+      "step": 11320
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.25338882207870483,
+      "learning_rate": 0.00019314679801705144,
+      "loss": 0.8986,
+      "step": 11325
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.22876478731632233,
+      "learning_rate": 0.00019304823283589482,
+      "loss": 1.0461,
+      "step": 11330
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2699691355228424,
+      "learning_rate": 0.00019294964739481024,
+      "loss": 0.8773,
+      "step": 11335
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.23028436303138733,
+      "learning_rate": 0.00019285104174019527,
+      "loss": 0.9431,
+      "step": 11340
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.21594227850437164,
+      "learning_rate": 0.00019275241591845704,
+      "loss": 0.904,
+      "step": 11345
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.25375184416770935,
+      "learning_rate": 0.00019265376997601205,
+      "loss": 0.9063,
+      "step": 11350
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.24144120514392853,
+      "learning_rate": 0.0001925551039592865,
+      "loss": 0.8952,
+      "step": 11355
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.23259896039962769,
+      "learning_rate": 0.00019245641791471577,
+      "loss": 0.9056,
+      "step": 11360
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.26159223914146423,
+      "learning_rate": 0.00019235771188874485,
+      "loss": 0.8856,
+      "step": 11365
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.22818419337272644,
+      "learning_rate": 0.0001922589859278281,
+      "loss": 0.8179,
+      "step": 11370
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.24243588745594025,
+      "learning_rate": 0.00019216024007842915,
+      "loss": 0.8159,
+      "step": 11375
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.22796067595481873,
+      "learning_rate": 0.00019206147438702108,
+      "loss": 0.9193,
+      "step": 11380
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.25152865052223206,
+      "learning_rate": 0.00019196268890008642,
+      "loss": 0.938,
+      "step": 11385
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.22121083736419678,
+      "learning_rate": 0.0001918638836641168,
+      "loss": 0.82,
+      "step": 11390
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.22968332469463348,
+      "learning_rate": 0.00019176505872561326,
+      "loss": 0.9606,
+      "step": 11395
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.2671717405319214,
+      "learning_rate": 0.00019166621413108613,
+      "loss": 0.9147,
+      "step": 11400
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.24766571819782257,
+      "learning_rate": 0.00019156734992705496,
+      "loss": 0.8639,
+      "step": 11405
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.244247704744339,
+      "learning_rate": 0.00019146846616004842,
+      "loss": 0.7505,
+      "step": 11410
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.23957106471061707,
+      "learning_rate": 0.00019136956287660464,
+      "loss": 0.8708,
+      "step": 11415
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2419036477804184,
+      "learning_rate": 0.00019127064012327072,
+      "loss": 0.9644,
+      "step": 11420
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2512461245059967,
+      "learning_rate": 0.000191171697946603,
+      "loss": 0.8277,
+      "step": 11425
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.24636131525039673,
+      "learning_rate": 0.00019107273639316696,
+      "loss": 0.877,
+      "step": 11430
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.21014370024204254,
+      "learning_rate": 0.0001909737555095372,
+      "loss": 0.8611,
+      "step": 11435
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2295524775981903,
+      "learning_rate": 0.0001908747553422974,
+      "loss": 0.9243,
+      "step": 11440
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.24215233325958252,
+      "learning_rate": 0.0001907757359380403,
+      "loss": 0.9948,
+      "step": 11445
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2857617735862732,
+      "learning_rate": 0.00019067669734336776,
+      "loss": 0.8859,
+      "step": 11450
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.25199002027511597,
+      "learning_rate": 0.00019057763960489063,
+      "loss": 0.9734,
+      "step": 11455
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2548152506351471,
+      "learning_rate": 0.00019047856276922873,
+      "loss": 0.8183,
+      "step": 11460
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.23145854473114014,
+      "learning_rate": 0.00019037946688301097,
+      "loss": 0.9237,
+      "step": 11465
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.24691297113895416,
+      "learning_rate": 0.00019028035199287512,
+      "loss": 0.8724,
+      "step": 11470
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2465231567621231,
+      "learning_rate": 0.00019018121814546799,
+      "loss": 0.8638,
+      "step": 11475
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2617138922214508,
+      "learning_rate": 0.00019008206538744516,
+      "loss": 0.9152,
+      "step": 11480
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.23858489096164703,
+      "learning_rate": 0.00018998289376547135,
+      "loss": 0.7177,
+      "step": 11485
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2825676202774048,
+      "learning_rate": 0.00018988370332621987,
+      "loss": 0.9155,
+      "step": 11490
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.21846085786819458,
+      "learning_rate": 0.00018978449411637317,
+      "loss": 0.8946,
+      "step": 11495
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.23341843485832214,
+      "learning_rate": 0.00018968526618262226,
+      "loss": 0.833,
+      "step": 11500
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.24220559000968933,
+      "learning_rate": 0.00018958601957166725,
+      "loss": 0.8047,
+      "step": 11505
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2510910630226135,
+      "learning_rate": 0.0001894867543302168,
+      "loss": 0.8861,
+      "step": 11510
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.24948173761367798,
+      "learning_rate": 0.00018938747050498847,
+      "loss": 0.8644,
+      "step": 11515
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.25031885504722595,
+      "learning_rate": 0.00018928816814270844,
+      "loss": 0.8713,
+      "step": 11520
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2391744703054428,
+      "learning_rate": 0.0001891888472901118,
+      "loss": 0.8059,
+      "step": 11525
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.26208725571632385,
+      "learning_rate": 0.0001890895079939422,
+      "loss": 0.8532,
+      "step": 11530
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2613973021507263,
+      "learning_rate": 0.00018899015030095201,
+      "loss": 0.8387,
+      "step": 11535
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2578382194042206,
+      "learning_rate": 0.00018889077425790225,
+      "loss": 0.9407,
+      "step": 11540
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.22165289521217346,
+      "learning_rate": 0.00018879137991156263,
+      "loss": 0.817,
+      "step": 11545
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2351749688386917,
+      "learning_rate": 0.0001886919673087114,
+      "loss": 0.8507,
+      "step": 11550
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2208326905965805,
+      "learning_rate": 0.00018859253649613545,
+      "loss": 0.7727,
+      "step": 11555
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.21380893886089325,
+      "learning_rate": 0.00018849308752063017,
+      "loss": 0.829,
+      "step": 11560
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.25803694128990173,
+      "learning_rate": 0.00018839362042899962,
+      "loss": 0.8363,
+      "step": 11565
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.28940629959106445,
+      "learning_rate": 0.0001882941352680562,
+      "loss": 0.9299,
+      "step": 11570
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2648991346359253,
+      "learning_rate": 0.00018819463208462112,
+      "loss": 0.8416,
+      "step": 11575
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.22129850089550018,
+      "learning_rate": 0.00018809511092552372,
+      "loss": 0.8448,
+      "step": 11580
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.26116523146629333,
+      "learning_rate": 0.00018799557183760206,
+      "loss": 0.9321,
+      "step": 11585
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2547585964202881,
+      "learning_rate": 0.00018789601486770245,
+      "loss": 0.8947,
+      "step": 11590
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.28379014134407043,
+      "learning_rate": 0.0001877964400626798,
+      "loss": 0.8113,
+      "step": 11595
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.21314746141433716,
+      "learning_rate": 0.00018769684746939723,
+      "loss": 0.89,
+      "step": 11600
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2194174975156784,
+      "learning_rate": 0.00018759723713472642,
+      "loss": 0.8691,
+      "step": 11605
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2298017144203186,
+      "learning_rate": 0.00018749760910554715,
+      "loss": 0.8259,
+      "step": 11610
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2847272455692291,
+      "learning_rate": 0.0001873979634287479,
+      "loss": 0.8773,
+      "step": 11615
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.23975825309753418,
+      "learning_rate": 0.00018729830015122508,
+      "loss": 0.9002,
+      "step": 11620
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.22761796414852142,
+      "learning_rate": 0.00018719861931988354,
+      "loss": 0.8456,
+      "step": 11625
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.24914436042308807,
+      "learning_rate": 0.0001870989209816364,
+      "loss": 0.8834,
+      "step": 11630
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2555813491344452,
+      "learning_rate": 0.0001869992051834051,
+      "loss": 0.7416,
+      "step": 11635
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2689943015575409,
+      "learning_rate": 0.00018689947197211908,
+      "loss": 0.85,
+      "step": 11640
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.21290338039398193,
+      "learning_rate": 0.00018679972139471615,
+      "loss": 0.9248,
+      "step": 11645
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.21777038276195526,
+      "learning_rate": 0.00018669995349814228,
+      "loss": 0.9654,
+      "step": 11650
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2189740687608719,
+      "learning_rate": 0.0001866001683293515,
+      "loss": 0.8877,
+      "step": 11655
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2227373570203781,
+      "learning_rate": 0.00018650036593530607,
+      "loss": 0.9332,
+      "step": 11660
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2457219511270523,
+      "learning_rate": 0.00018640054636297632,
+      "loss": 0.7944,
+      "step": 11665
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2539142072200775,
+      "learning_rate": 0.00018630070965934057,
+      "loss": 0.7875,
+      "step": 11670
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.23336252570152283,
+      "learning_rate": 0.0001862008558713854,
+      "loss": 0.7474,
+      "step": 11675
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.33268702030181885,
+      "learning_rate": 0.00018610098504610523,
+      "loss": 0.944,
+      "step": 11680
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.23532938957214355,
+      "learning_rate": 0.0001860010972305026,
+      "loss": 0.998,
+      "step": 11685
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.24364040791988373,
+      "learning_rate": 0.0001859011924715881,
+      "loss": 0.8007,
+      "step": 11690
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.26851868629455566,
+      "learning_rate": 0.00018580127081638018,
+      "loss": 0.8973,
+      "step": 11695
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.27191677689552307,
+      "learning_rate": 0.00018570133231190524,
+      "loss": 0.9319,
+      "step": 11700
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.24935273826122284,
+      "learning_rate": 0.00018560137700519776,
+      "loss": 0.8381,
+      "step": 11705
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.19971820712089539,
+      "learning_rate": 0.00018550140494329995,
+      "loss": 0.9152,
+      "step": 11710
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2567872703075409,
+      "learning_rate": 0.000185401416173262,
+      "loss": 0.8468,
+      "step": 11715
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2599542438983917,
+      "learning_rate": 0.00018530141074214195,
+      "loss": 0.8679,
+      "step": 11720
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.2495017796754837,
+      "learning_rate": 0.00018520138869700573,
+      "loss": 0.8317,
+      "step": 11725
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.25889188051223755,
+      "learning_rate": 0.00018510135008492696,
+      "loss": 0.9075,
+      "step": 11730
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.24831044673919678,
+      "learning_rate": 0.00018500129495298718,
+      "loss": 0.8721,
+      "step": 11735
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.23168979585170746,
+      "learning_rate": 0.0001849012233482756,
+      "loss": 0.9592,
+      "step": 11740
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.27312368154525757,
+      "learning_rate": 0.0001848011353178893,
+      "loss": 0.8826,
+      "step": 11745
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.23228977620601654,
+      "learning_rate": 0.00018470103090893297,
+      "loss": 0.8053,
+      "step": 11750
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.23857393860816956,
+      "learning_rate": 0.00018460091016851915,
+      "loss": 0.8638,
+      "step": 11755
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.24902932345867157,
+      "learning_rate": 0.00018450077314376793,
+      "loss": 0.8719,
+      "step": 11760
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.29293930530548096,
+      "learning_rate": 0.0001844006198818072,
+      "loss": 0.9337,
+      "step": 11765
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2380332350730896,
+      "learning_rate": 0.00018430045042977224,
+      "loss": 0.8136,
+      "step": 11770
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2724967896938324,
+      "learning_rate": 0.0001842002648348063,
+      "loss": 0.9473,
+      "step": 11775
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.24302633106708527,
+      "learning_rate": 0.00018410006314405992,
+      "loss": 0.7744,
+      "step": 11780
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.242751345038414,
+      "learning_rate": 0.0001839998454046914,
+      "loss": 0.8238,
+      "step": 11785
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.23499782383441925,
+      "learning_rate": 0.00018389961166386657,
+      "loss": 0.8673,
+      "step": 11790
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.22136856615543365,
+      "learning_rate": 0.00018379936196875868,
+      "loss": 0.9232,
+      "step": 11795
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.22664988040924072,
+      "learning_rate": 0.00018369909636654867,
+      "loss": 0.8296,
+      "step": 11800
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2627767026424408,
+      "learning_rate": 0.00018359881490442471,
+      "loss": 0.8778,
+      "step": 11805
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.22755451500415802,
+      "learning_rate": 0.00018349851762958274,
+      "loss": 0.949,
+      "step": 11810
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2317892462015152,
+      "learning_rate": 0.00018339820458922589,
+      "loss": 0.8828,
+      "step": 11815
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.23136401176452637,
+      "learning_rate": 0.00018329787583056486,
+      "loss": 0.7622,
+      "step": 11820
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.24994751811027527,
+      "learning_rate": 0.00018319753140081765,
+      "loss": 0.8811,
+      "step": 11825
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2566693425178528,
+      "learning_rate": 0.00018309717134720974,
+      "loss": 0.9607,
+      "step": 11830
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2577987313270569,
+      "learning_rate": 0.00018299679571697383,
+      "loss": 0.8902,
+      "step": 11835
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2891867160797119,
+      "learning_rate": 0.00018289640455735012,
+      "loss": 0.8107,
+      "step": 11840
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.24823874235153198,
+      "learning_rate": 0.0001827959979155859,
+      "loss": 0.8493,
+      "step": 11845
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.23691007494926453,
+      "learning_rate": 0.00018269557583893602,
+      "loss": 0.8704,
+      "step": 11850
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2653874158859253,
+      "learning_rate": 0.00018259513837466228,
+      "loss": 0.8665,
+      "step": 11855
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3187112510204315,
+      "learning_rate": 0.00018249468557003404,
+      "loss": 0.7894,
+      "step": 11860
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2333473116159439,
+      "learning_rate": 0.00018239421747232758,
+      "loss": 0.7926,
+      "step": 11865
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2679731547832489,
+      "learning_rate": 0.0001822937341288267,
+      "loss": 0.848,
+      "step": 11870
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2512493431568146,
+      "learning_rate": 0.00018219323558682203,
+      "loss": 0.9605,
+      "step": 11875
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.241338312625885,
+      "learning_rate": 0.0001820927218936116,
+      "loss": 0.8209,
+      "step": 11880
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.2403215616941452,
+      "learning_rate": 0.0001819921930965005,
+      "loss": 0.808,
+      "step": 11885
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.24776634573936462,
+      "learning_rate": 0.00018189164924280088,
+      "loss": 0.8508,
+      "step": 11890
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.23758898675441742,
+      "learning_rate": 0.00018179109037983203,
+      "loss": 0.9911,
+      "step": 11895
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.24971432983875275,
+      "learning_rate": 0.00018169051655492034,
+      "loss": 0.8894,
+      "step": 11900
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.22912085056304932,
+      "learning_rate": 0.0001815899278153991,
+      "loss": 0.8912,
+      "step": 11905
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.271433562040329,
+      "learning_rate": 0.0001814893242086088,
+      "loss": 0.8364,
+      "step": 11910
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2268996387720108,
+      "learning_rate": 0.00018138870578189676,
+      "loss": 0.8419,
+      "step": 11915
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2744608521461487,
+      "learning_rate": 0.0001812880725826174,
+      "loss": 0.8018,
+      "step": 11920
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2408338189125061,
+      "learning_rate": 0.00018118742465813206,
+      "loss": 0.84,
+      "step": 11925
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.27505186200141907,
+      "learning_rate": 0.00018108676205580895,
+      "loss": 0.8555,
+      "step": 11930
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.23788927495479584,
+      "learning_rate": 0.00018098608482302328,
+      "loss": 0.8613,
+      "step": 11935
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2505749762058258,
+      "learning_rate": 0.00018088539300715705,
+      "loss": 0.7991,
+      "step": 11940
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.22999998927116394,
+      "learning_rate": 0.00018078468665559924,
+      "loss": 0.7935,
+      "step": 11945
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2383863478899002,
+      "learning_rate": 0.00018068396581574553,
+      "loss": 0.878,
+      "step": 11950
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.22946636378765106,
+      "learning_rate": 0.00018058323053499854,
+      "loss": 0.8047,
+      "step": 11955
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.23430262506008148,
+      "learning_rate": 0.0001804824808607676,
+      "loss": 0.8075,
+      "step": 11960
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2360125333070755,
+      "learning_rate": 0.0001803817168404689,
+      "loss": 0.8581,
+      "step": 11965
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.24007384479045868,
+      "learning_rate": 0.00018028093852152528,
+      "loss": 0.8953,
+      "step": 11970
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2502374053001404,
+      "learning_rate": 0.00018018014595136644,
+      "loss": 0.8884,
+      "step": 11975
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.24452511966228485,
+      "learning_rate": 0.00018007933917742864,
+      "loss": 1.0623,
+      "step": 11980
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.23154529929161072,
+      "learning_rate": 0.0001799785182471549,
+      "loss": 0.8695,
+      "step": 11985
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.2401302307844162,
+      "learning_rate": 0.00017987768320799495,
+      "loss": 0.9001,
+      "step": 11990
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.37644556164741516,
+      "learning_rate": 0.00017977683410740503,
+      "loss": 0.8526,
+      "step": 11995
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.2637287974357605,
+      "learning_rate": 0.00017967597099284813,
+      "loss": 0.9424,
+      "step": 12000
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.23815138638019562,
+      "learning_rate": 0.0001795750939117938,
+      "loss": 0.8184,
+      "step": 12005
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.2669789493083954,
+      "learning_rate": 0.00017947420291171813,
+      "loss": 0.9174,
+      "step": 12010
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3020137548446655,
+      "learning_rate": 0.00017937329804010372,
+      "loss": 0.9216,
+      "step": 12015
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.22852839529514313,
+      "learning_rate": 0.00017927237934443983,
+      "loss": 0.8869,
+      "step": 12020
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.27569183707237244,
+      "learning_rate": 0.00017917144687222206,
+      "loss": 0.8327,
+      "step": 12025
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.2452722042798996,
+      "learning_rate": 0.0001790705006709527,
+      "loss": 0.9104,
+      "step": 12030
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.24271848797798157,
+      "learning_rate": 0.00017896954078814028,
+      "loss": 0.9341,
+      "step": 12035
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.21451996266841888,
+      "learning_rate": 0.00017886856727129994,
+      "loss": 0.8143,
+      "step": 12040
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.23725618422031403,
+      "learning_rate": 0.00017876758016795313,
+      "loss": 0.8801,
+      "step": 12045
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.24408486485481262,
+      "learning_rate": 0.00017866657952562778,
+      "loss": 0.8149,
+      "step": 12050
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.22583889961242676,
+      "learning_rate": 0.0001785655653918581,
+      "loss": 0.8196,
+      "step": 12055
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.23950889706611633,
+      "learning_rate": 0.00017846453781418474,
+      "loss": 0.7369,
+      "step": 12060
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.24430805444717407,
+      "learning_rate": 0.00017836349684015456,
+      "loss": 0.9414,
+      "step": 12065
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.22092491388320923,
+      "learning_rate": 0.00017826244251732088,
+      "loss": 0.9288,
+      "step": 12070
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.23350967466831207,
+      "learning_rate": 0.00017816137489324314,
+      "loss": 0.8764,
+      "step": 12075
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.25949931144714355,
+      "learning_rate": 0.0001780602940154872,
+      "loss": 0.7207,
+      "step": 12080
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.23288941383361816,
+      "learning_rate": 0.00017795919993162504,
+      "loss": 0.8304,
+      "step": 12085
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.261445015668869,
+      "learning_rate": 0.00017785809268923493,
+      "loss": 0.8202,
+      "step": 12090
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.26035887002944946,
+      "learning_rate": 0.0001777569723359012,
+      "loss": 0.8119,
+      "step": 12095
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.24832282960414886,
+      "learning_rate": 0.00017765583891921454,
+      "loss": 0.8939,
+      "step": 12100
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.24832552671432495,
+      "learning_rate": 0.00017755469248677163,
+      "loss": 0.8043,
+      "step": 12105
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.24635636806488037,
+      "learning_rate": 0.0001774535330861754,
+      "loss": 0.9287,
+      "step": 12110
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.26285749673843384,
+      "learning_rate": 0.0001773523607650348,
+      "loss": 0.903,
+      "step": 12115
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.23266449570655823,
+      "learning_rate": 0.0001772511755709649,
+      "loss": 0.8647,
+      "step": 12120
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.23919335007667542,
+      "learning_rate": 0.00017714997755158675,
+      "loss": 0.8978,
+      "step": 12125
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.247425839304924,
+      "learning_rate": 0.0001770487667545276,
+      "loss": 0.8336,
+      "step": 12130
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.28573018312454224,
+      "learning_rate": 0.00017694754322742048,
+      "loss": 0.9016,
+      "step": 12135
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.23751410841941833,
+      "learning_rate": 0.00017684630701790468,
+      "loss": 0.8116,
+      "step": 12140
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2392253428697586,
+      "learning_rate": 0.0001767450581736252,
+      "loss": 0.9136,
+      "step": 12145
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.23572461307048798,
+      "learning_rate": 0.00017664379674223318,
+      "loss": 0.8182,
+      "step": 12150
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2536904513835907,
+      "learning_rate": 0.0001765425227713856,
+      "loss": 0.8735,
+      "step": 12155
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2494330257177353,
+      "learning_rate": 0.00017644123630874539,
+      "loss": 0.888,
+      "step": 12160
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2419360876083374,
+      "learning_rate": 0.00017633993740198128,
+      "loss": 0.8367,
+      "step": 12165
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.22948452830314636,
+      "learning_rate": 0.0001762386260987679,
+      "loss": 0.903,
+      "step": 12170
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2524193227291107,
+      "learning_rate": 0.0001761373024467857,
+      "loss": 0.8555,
+      "step": 12175
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2861385941505432,
+      "learning_rate": 0.000176035966493721,
+      "loss": 0.9068,
+      "step": 12180
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2484591007232666,
+      "learning_rate": 0.0001759346182872658,
+      "loss": 0.9578,
+      "step": 12185
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.24529612064361572,
+      "learning_rate": 0.000175833257875118,
+      "loss": 0.8648,
+      "step": 12190
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.23602719604969025,
+      "learning_rate": 0.00017573188530498117,
+      "loss": 0.9002,
+      "step": 12195
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.24331901967525482,
+      "learning_rate": 0.0001756305006245646,
+      "loss": 0.8624,
+      "step": 12200
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2478303611278534,
+      "learning_rate": 0.0001755291038815832,
+      "loss": 0.8435,
+      "step": 12205
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.26991724967956543,
+      "learning_rate": 0.0001754276951237578,
+      "loss": 0.8139,
+      "step": 12210
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2541549503803253,
+      "learning_rate": 0.00017532627439881458,
+      "loss": 0.85,
+      "step": 12215
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.24864061176776886,
+      "learning_rate": 0.0001752248417544856,
+      "loss": 0.8312,
+      "step": 12220
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.26080724596977234,
+      "learning_rate": 0.00017512339723850835,
+      "loss": 0.8774,
+      "step": 12225
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.27303215861320496,
+      "learning_rate": 0.00017502194089862608,
+      "loss": 0.8132,
+      "step": 12230
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.24449607729911804,
+      "learning_rate": 0.00017492047278258748,
+      "loss": 0.9219,
+      "step": 12235
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.24350832402706146,
+      "learning_rate": 0.0001748189929381468,
+      "loss": 0.814,
+      "step": 12240
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.22966676950454712,
+      "learning_rate": 0.0001747175014130638,
+      "loss": 0.8465,
+      "step": 12245
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2714262008666992,
+      "learning_rate": 0.00017461599825510386,
+      "loss": 0.885,
+      "step": 12250
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2457001507282257,
+      "learning_rate": 0.00017451448351203758,
+      "loss": 0.8376,
+      "step": 12255
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.22185452282428741,
+      "learning_rate": 0.00017441295723164132,
+      "loss": 0.8311,
+      "step": 12260
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2851621210575104,
+      "learning_rate": 0.00017431141946169662,
+      "loss": 0.7891,
+      "step": 12265
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2442905604839325,
+      "learning_rate": 0.00017420987024999065,
+      "loss": 0.9465,
+      "step": 12270
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.26960885524749756,
+      "learning_rate": 0.00017410830964431566,
+      "loss": 0.8436,
+      "step": 12275
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.23894372582435608,
+      "learning_rate": 0.0001740067376924696,
+      "loss": 0.8469,
+      "step": 12280
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.295200914144516,
+      "learning_rate": 0.00017390515444225548,
+      "loss": 0.907,
+      "step": 12285
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.24998031556606293,
+      "learning_rate": 0.00017380355994148187,
+      "loss": 0.8642,
+      "step": 12290
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.25072336196899414,
+      "learning_rate": 0.0001737019542379624,
+      "loss": 0.8299,
+      "step": 12295
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.23040717840194702,
+      "learning_rate": 0.00017360033737951622,
+      "loss": 0.8464,
+      "step": 12300
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.21226970851421356,
+      "learning_rate": 0.0001734987094139675,
+      "loss": 0.8759,
+      "step": 12305
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.26856541633605957,
+      "learning_rate": 0.0001733970703891457,
+      "loss": 0.8438,
+      "step": 12310
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2336094230413437,
+      "learning_rate": 0.00017329542035288565,
+      "loss": 0.8019,
+      "step": 12315
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2525746524333954,
+      "learning_rate": 0.00017319375935302713,
+      "loss": 0.9302,
+      "step": 12320
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.21425200998783112,
+      "learning_rate": 0.00017309208743741526,
+      "loss": 0.835,
+      "step": 12325
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2605019509792328,
+      "learning_rate": 0.00017299040465390013,
+      "loss": 0.9248,
+      "step": 12330
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.22529718279838562,
+      "learning_rate": 0.00017288871105033713,
+      "loss": 0.8703,
+      "step": 12335
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.24195663630962372,
+      "learning_rate": 0.00017278700667458657,
+      "loss": 0.836,
+      "step": 12340
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.25822916626930237,
+      "learning_rate": 0.00017268529157451394,
+      "loss": 0.8901,
+      "step": 12345
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2364194244146347,
+      "learning_rate": 0.00017258356579798973,
+      "loss": 0.8438,
+      "step": 12350
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.26082637906074524,
+      "learning_rate": 0.0001724818293928895,
+      "loss": 0.9269,
+      "step": 12355
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2297302931547165,
+      "learning_rate": 0.00017238008240709374,
+      "loss": 0.9076,
+      "step": 12360
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2835592031478882,
+      "learning_rate": 0.00017227832488848799,
+      "loss": 0.8485,
+      "step": 12365
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2650296986103058,
+      "learning_rate": 0.0001721765568849627,
+      "loss": 0.9593,
+      "step": 12370
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.22768686711788177,
+      "learning_rate": 0.00017207477844441335,
+      "loss": 0.8921,
+      "step": 12375
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.22558462619781494,
+      "learning_rate": 0.00017197298961474006,
+      "loss": 0.8421,
+      "step": 12380
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.224419966340065,
+      "learning_rate": 0.00017187119044384823,
+      "loss": 0.8392,
+      "step": 12385
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.2719701826572418,
+      "learning_rate": 0.00017176938097964784,
+      "loss": 0.9365,
+      "step": 12390
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.24385976791381836,
+      "learning_rate": 0.00017166756127005384,
+      "loss": 0.8264,
+      "step": 12395
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.23817993700504303,
+      "learning_rate": 0.00017156573136298592,
+      "loss": 0.9154,
+      "step": 12400
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.25745269656181335,
+      "learning_rate": 0.00017146389130636864,
+      "loss": 0.8911,
+      "step": 12405
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.29501572251319885,
+      "learning_rate": 0.0001713620411481314,
+      "loss": 0.8841,
+      "step": 12410
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.263081431388855,
+      "learning_rate": 0.00017126018093620808,
+      "loss": 0.8334,
+      "step": 12415
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.22736036777496338,
+      "learning_rate": 0.0001711583107185376,
+      "loss": 0.88,
+      "step": 12420
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.23410391807556152,
+      "learning_rate": 0.00017105643054306352,
+      "loss": 0.8257,
+      "step": 12425
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.26745110750198364,
+      "learning_rate": 0.00017095454045773387,
+      "loss": 0.9252,
+      "step": 12430
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.24385030567646027,
+      "learning_rate": 0.00017085264051050166,
+      "loss": 0.8395,
+      "step": 12435
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2342156320810318,
+      "learning_rate": 0.0001707507307493243,
+      "loss": 0.8266,
+      "step": 12440
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2477523386478424,
+      "learning_rate": 0.00017064881122216398,
+      "loss": 0.8427,
+      "step": 12445
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.252871572971344,
+      "learning_rate": 0.00017054688197698736,
+      "loss": 0.8368,
+      "step": 12450
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.22697529196739197,
+      "learning_rate": 0.00017044494306176576,
+      "loss": 0.7964,
+      "step": 12455
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.27553001046180725,
+      "learning_rate": 0.00017034299452447493,
+      "loss": 0.8546,
+      "step": 12460
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.26646605134010315,
+      "learning_rate": 0.00017024103641309537,
+      "loss": 0.8383,
+      "step": 12465
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.255669504404068,
+      "learning_rate": 0.00017013906877561187,
+      "loss": 0.7589,
+      "step": 12470
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.25287455320358276,
+      "learning_rate": 0.0001700370916600138,
+      "loss": 0.8203,
+      "step": 12475
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2946148216724396,
+      "learning_rate": 0.000169935105114295,
+      "loss": 0.9016,
+      "step": 12480
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2425968050956726,
+      "learning_rate": 0.0001698331091864537,
+      "loss": 0.786,
+      "step": 12485
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.24818405508995056,
+      "learning_rate": 0.00016973110392449255,
+      "loss": 0.8466,
+      "step": 12490
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2572382688522339,
+      "learning_rate": 0.00016962908937641873,
+      "loss": 0.8854,
+      "step": 12495
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.22933310270309448,
+      "learning_rate": 0.0001695270655902435,
+      "loss": 0.8843,
+      "step": 12500
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2477482110261917,
+      "learning_rate": 0.00016942503261398276,
+      "loss": 0.8412,
+      "step": 12505
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2406783103942871,
+      "learning_rate": 0.00016932299049565657,
+      "loss": 0.7979,
+      "step": 12510
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2689076364040375,
+      "learning_rate": 0.00016922093928328937,
+      "loss": 0.8078,
+      "step": 12515
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2702324092388153,
+      "learning_rate": 0.00016911887902490986,
+      "loss": 0.8463,
+      "step": 12520
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.27444717288017273,
+      "learning_rate": 0.00016901680976855096,
+      "loss": 0.8428,
+      "step": 12525
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.24861744046211243,
+      "learning_rate": 0.00016891473156224976,
+      "loss": 0.9324,
+      "step": 12530
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2355334758758545,
+      "learning_rate": 0.00016881264445404786,
+      "loss": 0.8497,
+      "step": 12535
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.24266460537910461,
+      "learning_rate": 0.00016871054849199068,
+      "loss": 0.8034,
+      "step": 12540
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.24508504569530487,
+      "learning_rate": 0.00016860844372412802,
+      "loss": 0.8216,
+      "step": 12545
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.23630234599113464,
+      "learning_rate": 0.00016850633019851378,
+      "loss": 0.8422,
+      "step": 12550
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2652592658996582,
+      "learning_rate": 0.00016840420796320602,
+      "loss": 0.9161,
+      "step": 12555
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.21832112967967987,
+      "learning_rate": 0.00016830207706626675,
+      "loss": 0.7847,
+      "step": 12560
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2525995671749115,
+      "learning_rate": 0.00016819993755576225,
+      "loss": 0.8049,
+      "step": 12565
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.2583009898662567,
+      "learning_rate": 0.00016809778947976273,
+      "loss": 0.8428,
+      "step": 12570
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2520254850387573,
+      "learning_rate": 0.00016799563288634247,
+      "loss": 0.8095,
+      "step": 12575
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2578343152999878,
+      "learning_rate": 0.00016789346782357975,
+      "loss": 0.7919,
+      "step": 12580
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.26357656717300415,
+      "learning_rate": 0.00016779129433955686,
+      "loss": 0.8591,
+      "step": 12585
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.256157249212265,
+      "learning_rate": 0.00016768911248236001,
+      "loss": 0.8812,
+      "step": 12590
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.25076690316200256,
+      "learning_rate": 0.00016758692230007939,
+      "loss": 0.8109,
+      "step": 12595
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.26250481605529785,
+      "learning_rate": 0.00016748472384080912,
+      "loss": 0.944,
+      "step": 12600
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.25288984179496765,
+      "learning_rate": 0.0001673825171526471,
+      "loss": 0.9143,
+      "step": 12605
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.26439496874809265,
+      "learning_rate": 0.0001672803022836953,
+      "loss": 0.9372,
+      "step": 12610
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.23811744153499603,
+      "learning_rate": 0.00016717807928205936,
+      "loss": 0.816,
+      "step": 12615
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2801905870437622,
+      "learning_rate": 0.00016707584819584885,
+      "loss": 0.7269,
+      "step": 12620
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2476351410150528,
+      "learning_rate": 0.00016697360907317712,
+      "loss": 0.7855,
+      "step": 12625
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2882719933986664,
+      "learning_rate": 0.0001668713619621613,
+      "loss": 0.9298,
+      "step": 12630
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.27787694334983826,
+      "learning_rate": 0.00016676910691092224,
+      "loss": 0.8385,
+      "step": 12635
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.28809016942977905,
+      "learning_rate": 0.00016666684396758459,
+      "loss": 0.7968,
+      "step": 12640
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.24893060326576233,
+      "learning_rate": 0.00016656457318027667,
+      "loss": 0.8045,
+      "step": 12645
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.22097425162792206,
+      "learning_rate": 0.0001664622945971305,
+      "loss": 0.8332,
+      "step": 12650
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.241326704621315,
+      "learning_rate": 0.00016636000826628184,
+      "loss": 0.8432,
+      "step": 12655
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2555682063102722,
+      "learning_rate": 0.00016625771423586991,
+      "loss": 0.998,
+      "step": 12660
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2659325897693634,
+      "learning_rate": 0.00016615541255403786,
+      "loss": 0.796,
+      "step": 12665
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.24846947193145752,
+      "learning_rate": 0.0001660531032689321,
+      "loss": 0.8544,
+      "step": 12670
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2357734739780426,
+      "learning_rate": 0.0001659507864287029,
+      "loss": 0.8947,
+      "step": 12675
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.22256316244602203,
+      "learning_rate": 0.00016584846208150383,
+      "loss": 0.7654,
+      "step": 12680
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2255926877260208,
+      "learning_rate": 0.00016574613027549217,
+      "loss": 0.9368,
+      "step": 12685
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2199205905199051,
+      "learning_rate": 0.00016564379105882873,
+      "loss": 0.8522,
+      "step": 12690
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.28501272201538086,
+      "learning_rate": 0.0001655414444796777,
+      "loss": 0.9038,
+      "step": 12695
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.24840682744979858,
+      "learning_rate": 0.0001654390905862068,
+      "loss": 0.8622,
+      "step": 12700
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.25936436653137207,
+      "learning_rate": 0.00016533672942658717,
+      "loss": 0.7946,
+      "step": 12705
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2659285068511963,
+      "learning_rate": 0.0001652343610489933,
+      "loss": 0.8187,
+      "step": 12710
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.24160555005073547,
+      "learning_rate": 0.00016513198550160326,
+      "loss": 0.7751,
+      "step": 12715
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.26907095313072205,
+      "learning_rate": 0.00016502960283259823,
+      "loss": 0.807,
+      "step": 12720
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.2532917559146881,
+      "learning_rate": 0.00016492721309016307,
+      "loss": 0.8671,
+      "step": 12725
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.22019241750240326,
+      "learning_rate": 0.00016482481632248568,
+      "loss": 0.8272,
+      "step": 12730
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.22375644743442535,
+      "learning_rate": 0.00016472241257775743,
+      "loss": 0.7704,
+      "step": 12735
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2382332682609558,
+      "learning_rate": 0.00016462000190417292,
+      "loss": 0.9716,
+      "step": 12740
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.279604971408844,
+      "learning_rate": 0.00016451758434992997,
+      "loss": 0.8371,
+      "step": 12745
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2584107220172882,
+      "learning_rate": 0.00016441515996322973,
+      "loss": 0.8307,
+      "step": 12750
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.24332301318645477,
+      "learning_rate": 0.00016431272879227648,
+      "loss": 0.8213,
+      "step": 12755
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2555186450481415,
+      "learning_rate": 0.00016421029088527775,
+      "loss": 0.8662,
+      "step": 12760
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.21609874069690704,
+      "learning_rate": 0.00016410784629044422,
+      "loss": 0.9062,
+      "step": 12765
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.23835386335849762,
+      "learning_rate": 0.00016400539505598974,
+      "loss": 0.9274,
+      "step": 12770
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.24308866262435913,
+      "learning_rate": 0.00016390293723013124,
+      "loss": 0.9382,
+      "step": 12775
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2690812647342682,
+      "learning_rate": 0.00016380047286108874,
+      "loss": 0.8704,
+      "step": 12780
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.24690578877925873,
+      "learning_rate": 0.00016369800199708546,
+      "loss": 0.9505,
+      "step": 12785
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.27407118678092957,
+      "learning_rate": 0.00016359552468634748,
+      "loss": 0.806,
+      "step": 12790
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.23908410966396332,
+      "learning_rate": 0.00016349304097710416,
+      "loss": 0.8412,
+      "step": 12795
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.21888162195682526,
+      "learning_rate": 0.00016339055091758764,
+      "loss": 0.91,
+      "step": 12800
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2692570984363556,
+      "learning_rate": 0.00016328805455603315,
+      "loss": 0.8517,
+      "step": 12805
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.21654078364372253,
+      "learning_rate": 0.00016318555194067892,
+      "loss": 0.9092,
+      "step": 12810
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.22260205447673798,
+      "learning_rate": 0.00016308304311976604,
+      "loss": 0.7538,
+      "step": 12815
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2802688479423523,
+      "learning_rate": 0.00016298052814153866,
+      "loss": 0.9275,
+      "step": 12820
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.24824702739715576,
+      "learning_rate": 0.00016287800705424362,
+      "loss": 0.8563,
+      "step": 12825
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2404843270778656,
+      "learning_rate": 0.00016277547990613083,
+      "loss": 0.9009,
+      "step": 12830
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.24702101945877075,
+      "learning_rate": 0.00016267294674545286,
+      "loss": 0.9061,
+      "step": 12835
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.24450421333312988,
+      "learning_rate": 0.0001625704076204654,
+      "loss": 0.9004,
+      "step": 12840
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2563953399658203,
+      "learning_rate": 0.00016246786257942658,
+      "loss": 0.9063,
+      "step": 12845
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2544556260108948,
+      "learning_rate": 0.00016236531167059762,
+      "loss": 0.8174,
+      "step": 12850
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.23205184936523438,
+      "learning_rate": 0.00016226275494224233,
+      "loss": 0.8916,
+      "step": 12855
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2657812237739563,
+      "learning_rate": 0.00016216019244262735,
+      "loss": 0.8149,
+      "step": 12860
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.21930214762687683,
+      "learning_rate": 0.00016205762422002198,
+      "loss": 0.8223,
+      "step": 12865
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2352408766746521,
+      "learning_rate": 0.00016195505032269821,
+      "loss": 0.8929,
+      "step": 12870
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.24339045584201813,
+      "learning_rate": 0.00016185247079893075,
+      "loss": 0.919,
+      "step": 12875
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.27827414870262146,
+      "learning_rate": 0.00016174988569699696,
+      "loss": 0.8859,
+      "step": 12880
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2651593089103699,
+      "learning_rate": 0.0001616472950651767,
+      "loss": 0.7699,
+      "step": 12885
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.24380792677402496,
+      "learning_rate": 0.00016154469895175266,
+      "loss": 0.9605,
+      "step": 12890
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2639029920101166,
+      "learning_rate": 0.00016144209740500982,
+      "loss": 0.8723,
+      "step": 12895
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2522783577442169,
+      "learning_rate": 0.000161339490473236,
+      "loss": 0.8958,
+      "step": 12900
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.2770918309688568,
+      "learning_rate": 0.00016123687820472139,
+      "loss": 0.797,
+      "step": 12905
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.22492796182632446,
+      "learning_rate": 0.00016113426064775875,
+      "loss": 0.8496,
+      "step": 12910
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2644830644130707,
+      "learning_rate": 0.00016103163785064327,
+      "loss": 0.8458,
+      "step": 12915
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2638452649116516,
+      "learning_rate": 0.0001609290098616727,
+      "loss": 0.8408,
+      "step": 12920
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.23867619037628174,
+      "learning_rate": 0.0001608263767291471,
+      "loss": 0.8522,
+      "step": 12925
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2165214717388153,
+      "learning_rate": 0.00016072373850136912,
+      "loss": 0.9328,
+      "step": 12930
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2513158619403839,
+      "learning_rate": 0.00016062109522664366,
+      "loss": 0.8523,
+      "step": 12935
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.22370438277721405,
+      "learning_rate": 0.00016051844695327806,
+      "loss": 0.8542,
+      "step": 12940
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2550356090068817,
+      "learning_rate": 0.000160415793729582,
+      "loss": 0.9122,
+      "step": 12945
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.23933614790439606,
+      "learning_rate": 0.00016031313560386758,
+      "loss": 0.8236,
+      "step": 12950
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2359176129102707,
+      "learning_rate": 0.000160210472624449,
+      "loss": 0.8815,
+      "step": 12955
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2673303186893463,
+      "learning_rate": 0.00016010780483964295,
+      "loss": 0.8111,
+      "step": 12960
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2548558712005615,
+      "learning_rate": 0.00016000513229776826,
+      "loss": 0.919,
+      "step": 12965
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.27557799220085144,
+      "learning_rate": 0.00015990245504714608,
+      "loss": 0.7909,
+      "step": 12970
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.285546213388443,
+      "learning_rate": 0.00015979977313609965,
+      "loss": 0.8933,
+      "step": 12975
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.2725813090801239,
+      "learning_rate": 0.00015969708661295456,
+      "loss": 0.8644,
+      "step": 12980
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.23822170495986938,
+      "learning_rate": 0.0001595943955260385,
+      "loss": 0.8707,
+      "step": 12985
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2242768406867981,
+      "learning_rate": 0.0001594916999236813,
+      "loss": 0.7392,
+      "step": 12990
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.21174441277980804,
+      "learning_rate": 0.00015938899985421486,
+      "loss": 0.794,
+      "step": 12995
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.24703176319599152,
+      "learning_rate": 0.00015928629536597332,
+      "loss": 0.8355,
+      "step": 13000
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.22385214269161224,
+      "learning_rate": 0.00015918358650729276,
+      "loss": 0.763,
+      "step": 13005
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.24205328524112701,
+      "learning_rate": 0.00015908087332651142,
+      "loss": 0.86,
+      "step": 13010
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.256166011095047,
+      "learning_rate": 0.00015897815587196954,
+      "loss": 0.8247,
+      "step": 13015
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2543538212776184,
+      "learning_rate": 0.00015887543419200936,
+      "loss": 0.9216,
+      "step": 13020
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2570268511772156,
+      "learning_rate": 0.0001587727083349751,
+      "loss": 0.8575,
+      "step": 13025
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.29525551199913025,
+      "learning_rate": 0.000158669978349213,
+      "loss": 0.9212,
+      "step": 13030
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.23510275781154633,
+      "learning_rate": 0.0001585672442830711,
+      "loss": 0.8684,
+      "step": 13035
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.23046576976776123,
+      "learning_rate": 0.00015846450618489958,
+      "loss": 0.7974,
+      "step": 13040
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2453136444091797,
+      "learning_rate": 0.0001583617641030503,
+      "loss": 0.832,
+      "step": 13045
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.27528902888298035,
+      "learning_rate": 0.0001582590180858772,
+      "loss": 0.7578,
+      "step": 13050
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.2460784763097763,
+      "learning_rate": 0.0001581562681817359,
+      "loss": 0.8231,
+      "step": 13055
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.29714435338974,
+      "learning_rate": 0.00015805351443898388,
+      "loss": 0.7481,
+      "step": 13060
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.22743813693523407,
+      "learning_rate": 0.0001579507569059806,
+      "loss": 0.7635,
+      "step": 13065
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.24308447539806366,
+      "learning_rate": 0.00015784799563108706,
+      "loss": 0.8822,
+      "step": 13070
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2288735806941986,
+      "learning_rate": 0.00015774523066266612,
+      "loss": 0.8427,
+      "step": 13075
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.23703986406326294,
+      "learning_rate": 0.00015764246204908245,
+      "loss": 0.9294,
+      "step": 13080
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2489616721868515,
+      "learning_rate": 0.0001575396898387023,
+      "loss": 0.9377,
+      "step": 13085
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.25076568126678467,
+      "learning_rate": 0.00015743691407989378,
+      "loss": 0.9284,
+      "step": 13090
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.247811421751976,
+      "learning_rate": 0.00015733413482102652,
+      "loss": 0.867,
+      "step": 13095
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2610926330089569,
+      "learning_rate": 0.00015723135211047186,
+      "loss": 0.8203,
+      "step": 13100
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2661215662956238,
+      "learning_rate": 0.00015712856599660267,
+      "loss": 0.9256,
+      "step": 13105
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.22630742192268372,
+      "learning_rate": 0.00015702577652779368,
+      "loss": 0.8556,
+      "step": 13110
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.3097597360610962,
+      "learning_rate": 0.00015692298375242087,
+      "loss": 0.8452,
+      "step": 13115
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2603763937950134,
+      "learning_rate": 0.00015682018771886203,
+      "loss": 0.8142,
+      "step": 13120
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2540077865123749,
+      "learning_rate": 0.00015671738847549633,
+      "loss": 0.9639,
+      "step": 13125
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.24496349692344666,
+      "learning_rate": 0.0001566145860707046,
+      "loss": 1.0137,
+      "step": 13130
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2169584482908249,
+      "learning_rate": 0.00015651178055286897,
+      "loss": 0.9417,
+      "step": 13135
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.23268476128578186,
+      "learning_rate": 0.0001564089719703732,
+      "loss": 0.91,
+      "step": 13140
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.21907760202884674,
+      "learning_rate": 0.0001563061603716023,
+      "loss": 0.8338,
+      "step": 13145
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.24328762292861938,
+      "learning_rate": 0.00015620334580494297,
+      "loss": 0.872,
+      "step": 13150
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.26593923568725586,
+      "learning_rate": 0.00015610052831878304,
+      "loss": 0.9121,
+      "step": 13155
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2493075728416443,
+      "learning_rate": 0.00015599770796151196,
+      "loss": 0.8829,
+      "step": 13160
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2412213832139969,
+      "learning_rate": 0.00015589488478152027,
+      "loss": 0.8247,
+      "step": 13165
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2626934349536896,
+      "learning_rate": 0.00015579205882720014,
+      "loss": 0.8603,
+      "step": 13170
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2839190363883972,
+      "learning_rate": 0.0001556892301469447,
+      "loss": 0.8257,
+      "step": 13175
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.28261008858680725,
+      "learning_rate": 0.0001555863987891486,
+      "loss": 0.7699,
+      "step": 13180
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2586536109447479,
+      "learning_rate": 0.00015548356480220773,
+      "loss": 0.9528,
+      "step": 13185
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.25561654567718506,
+      "learning_rate": 0.0001553807282345192,
+      "loss": 0.809,
+      "step": 13190
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.24702797830104828,
+      "learning_rate": 0.00015527788913448124,
+      "loss": 0.9084,
+      "step": 13195
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.22921974956989288,
+      "learning_rate": 0.0001551750475504934,
+      "loss": 0.8321,
+      "step": 13200
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.2360827922821045,
+      "learning_rate": 0.0001550722035309563,
+      "loss": 0.8252,
+      "step": 13205
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.24088476598262787,
+      "learning_rate": 0.00015496935712427183,
+      "loss": 0.7994,
+      "step": 13210
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.22839505970478058,
+      "learning_rate": 0.00015486650837884277,
+      "loss": 0.891,
+      "step": 13215
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.257959246635437,
+      "learning_rate": 0.00015476365734307335,
+      "loss": 0.853,
+      "step": 13220
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.23323047161102295,
+      "learning_rate": 0.00015466080406536853,
+      "loss": 0.8453,
+      "step": 13225
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.24002352356910706,
+      "learning_rate": 0.00015455794859413458,
+      "loss": 0.7921,
+      "step": 13230
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.270093709230423,
+      "learning_rate": 0.0001544550909777786,
+      "loss": 0.8082,
+      "step": 13235
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.25525951385498047,
+      "learning_rate": 0.0001543522312647089,
+      "loss": 0.9131,
+      "step": 13240
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.26229140162467957,
+      "learning_rate": 0.00015424936950333463,
+      "loss": 0.9213,
+      "step": 13245
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.25442442297935486,
+      "learning_rate": 0.00015414650574206595,
+      "loss": 0.8611,
+      "step": 13250
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2250799685716629,
+      "learning_rate": 0.00015404364002931397,
+      "loss": 0.8195,
+      "step": 13255
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2337421178817749,
+      "learning_rate": 0.00015394077241349073,
+      "loss": 0.8752,
+      "step": 13260
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2822083830833435,
+      "learning_rate": 0.00015383790294300908,
+      "loss": 0.8054,
+      "step": 13265
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.25542697310447693,
+      "learning_rate": 0.00015373503166628288,
+      "loss": 0.8523,
+      "step": 13270
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.23330925405025482,
+      "learning_rate": 0.00015363215863172671,
+      "loss": 0.9308,
+      "step": 13275
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3322324752807617,
+      "learning_rate": 0.00015352928388775612,
+      "loss": 0.8854,
+      "step": 13280
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2551574110984802,
+      "learning_rate": 0.00015342640748278725,
+      "loss": 0.8225,
+      "step": 13285
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2288314402103424,
+      "learning_rate": 0.00015332352946523733,
+      "loss": 0.8141,
+      "step": 13290
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.28480198979377747,
+      "learning_rate": 0.000153220649883524,
+      "loss": 0.7063,
+      "step": 13295
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.26448380947113037,
+      "learning_rate": 0.0001531177687860659,
+      "loss": 0.8755,
+      "step": 13300
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.29811567068099976,
+      "learning_rate": 0.00015301488622128224,
+      "loss": 0.867,
+      "step": 13305
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2261502593755722,
+      "learning_rate": 0.00015291200223759306,
+      "loss": 0.881,
+      "step": 13310
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.26991018652915955,
+      "learning_rate": 0.0001528091168834189,
+      "loss": 0.835,
+      "step": 13315
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.2542611062526703,
+      "learning_rate": 0.00015270623020718102,
+      "loss": 0.8976,
+      "step": 13320
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2485206127166748,
+      "learning_rate": 0.00015260334225730137,
+      "loss": 0.8939,
+      "step": 13325
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.26317083835601807,
+      "learning_rate": 0.00015250045308220236,
+      "loss": 0.8537,
+      "step": 13330
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.23447948694229126,
+      "learning_rate": 0.00015239756273030715,
+      "loss": 0.8983,
+      "step": 13335
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.25760382413864136,
+      "learning_rate": 0.00015229467125003925,
+      "loss": 0.7641,
+      "step": 13340
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.24973775446414948,
+      "learning_rate": 0.00015219177868982286,
+      "loss": 0.8753,
+      "step": 13345
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.26316988468170166,
+      "learning_rate": 0.00015208888509808267,
+      "loss": 0.9377,
+      "step": 13350
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.23706527054309845,
+      "learning_rate": 0.00015198599052324377,
+      "loss": 0.8637,
+      "step": 13355
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2416650950908661,
+      "learning_rate": 0.00015188309501373175,
+      "loss": 0.8773,
+      "step": 13360
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.24447450041770935,
+      "learning_rate": 0.0001517801986179727,
+      "loss": 0.8913,
+      "step": 13365
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.24984611570835114,
+      "learning_rate": 0.00015167730138439305,
+      "loss": 0.8512,
+      "step": 13370
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2462444007396698,
+      "learning_rate": 0.00015157440336141967,
+      "loss": 0.934,
+      "step": 13375
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.23950566351413727,
+      "learning_rate": 0.0001514715045974798,
+      "loss": 0.8259,
+      "step": 13380
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.23663043975830078,
+      "learning_rate": 0.000151368605141001,
+      "loss": 0.7744,
+      "step": 13385
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2686599791049957,
+      "learning_rate": 0.00015126570504041115,
+      "loss": 0.8617,
+      "step": 13390
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.22998297214508057,
+      "learning_rate": 0.0001511628043441385,
+      "loss": 0.8539,
+      "step": 13395
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2632577121257782,
+      "learning_rate": 0.00015105990310061146,
+      "loss": 0.8452,
+      "step": 13400
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2295655459165573,
+      "learning_rate": 0.00015095700135825887,
+      "loss": 0.9384,
+      "step": 13405
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.24616779386997223,
+      "learning_rate": 0.00015085409916550961,
+      "loss": 0.8459,
+      "step": 13410
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.25910162925720215,
+      "learning_rate": 0.00015075119657079298,
+      "loss": 0.8525,
+      "step": 13415
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.22823968529701233,
+      "learning_rate": 0.00015064829362253828,
+      "loss": 0.8893,
+      "step": 13420
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2842884063720703,
+      "learning_rate": 0.0001505453903691751,
+      "loss": 0.806,
+      "step": 13425
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.24456340074539185,
+      "learning_rate": 0.00015044248685913304,
+      "loss": 0.871,
+      "step": 13430
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.23443590104579926,
+      "learning_rate": 0.00015033958314084202,
+      "loss": 0.8712,
+      "step": 13435
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.28807491064071655,
+      "learning_rate": 0.00015023667926273183,
+      "loss": 0.8278,
+      "step": 13440
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.28618142008781433,
+      "learning_rate": 0.00015013377527323257,
+      "loss": 0.9076,
+      "step": 13445
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2491266131401062,
+      "learning_rate": 0.0001500308712207742,
+      "loss": 0.8047,
+      "step": 13450
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.25134310126304626,
+      "learning_rate": 0.00014992796715378686,
+      "loss": 0.8135,
+      "step": 13455
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2570821940898895,
+      "learning_rate": 0.00014982506312070053,
+      "loss": 0.8727,
+      "step": 13460
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.23932205140590668,
+      "learning_rate": 0.0001497221591699453,
+      "loss": 0.8599,
+      "step": 13465
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.274517685174942,
+      "learning_rate": 0.00014961925534995118,
+      "loss": 0.8436,
+      "step": 13470
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.24874992668628693,
+      "learning_rate": 0.0001495163517091482,
+      "loss": 0.8423,
+      "step": 13475
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.23968589305877686,
+      "learning_rate": 0.00014941344829596612,
+      "loss": 0.9005,
+      "step": 13480
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.25740721821784973,
+      "learning_rate": 0.00014931054515883473,
+      "loss": 0.8737,
+      "step": 13485
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.22568397223949432,
+      "learning_rate": 0.00014920764234618373,
+      "loss": 0.9345,
+      "step": 13490
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2677795886993408,
+      "learning_rate": 0.00014910473990644254,
+      "loss": 0.8573,
+      "step": 13495
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2688808739185333,
+      "learning_rate": 0.00014900183788804048,
+      "loss": 0.8759,
+      "step": 13500
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.25429514050483704,
+      "learning_rate": 0.0001488989363394066,
+      "loss": 0.7835,
+      "step": 13505
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.3000656068325043,
+      "learning_rate": 0.00014879603530896992,
+      "loss": 0.8517,
+      "step": 13510
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2686077356338501,
+      "learning_rate": 0.00014869313484515897,
+      "loss": 0.9386,
+      "step": 13515
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2108505368232727,
+      "learning_rate": 0.0001485902349964022,
+      "loss": 0.8535,
+      "step": 13520
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2740882635116577,
+      "learning_rate": 0.0001484873358111276,
+      "loss": 0.9123,
+      "step": 13525
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.234098881483078,
+      "learning_rate": 0.00014838443733776306,
+      "loss": 0.8044,
+      "step": 13530
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2779216468334198,
+      "learning_rate": 0.00014828153962473593,
+      "loss": 0.8323,
+      "step": 13535
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.24527031183242798,
+      "learning_rate": 0.00014817864272047334,
+      "loss": 0.8261,
+      "step": 13540
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2469957321882248,
+      "learning_rate": 0.00014807574667340188,
+      "loss": 0.836,
+      "step": 13545
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.24512287974357605,
+      "learning_rate": 0.00014797285153194805,
+      "loss": 0.799,
+      "step": 13550
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.26496610045433044,
+      "learning_rate": 0.00014786995734453756,
+      "loss": 0.8458,
+      "step": 13555
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.2510302662849426,
+      "learning_rate": 0.0001477670641595959,
+      "loss": 0.8904,
+      "step": 13560
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.22821520268917084,
+      "learning_rate": 0.00014766417202554798,
+      "loss": 0.8772,
+      "step": 13565
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.27513188123703003,
+      "learning_rate": 0.0001475612809908183,
+      "loss": 0.81,
+      "step": 13570
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.24558843672275543,
+      "learning_rate": 0.00014745839110383077,
+      "loss": 0.7985,
+      "step": 13575
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.26508501172065735,
+      "learning_rate": 0.0001473555024130088,
+      "loss": 0.8178,
+      "step": 13580
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.23892903327941895,
+      "learning_rate": 0.00014725261496677513,
+      "loss": 0.8101,
+      "step": 13585
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.24258942902088165,
+      "learning_rate": 0.00014714972881355216,
+      "loss": 0.875,
+      "step": 13590
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.26198121905326843,
+      "learning_rate": 0.0001470468440017615,
+      "loss": 0.8671,
+      "step": 13595
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.22335876524448395,
+      "learning_rate": 0.0001469439605798241,
+      "loss": 0.7895,
+      "step": 13600
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.21281731128692627,
+      "learning_rate": 0.0001468410785961603,
+      "loss": 0.8186,
+      "step": 13605
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.26256877183914185,
+      "learning_rate": 0.00014673819809918985,
+      "loss": 0.8658,
+      "step": 13610
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2902090549468994,
+      "learning_rate": 0.0001466353191373317,
+      "loss": 0.8785,
+      "step": 13615
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2701566815376282,
+      "learning_rate": 0.0001465324417590041,
+      "loss": 0.9397,
+      "step": 13620
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2777699828147888,
+      "learning_rate": 0.00014642956601262452,
+      "loss": 0.8809,
+      "step": 13625
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.24040934443473816,
+      "learning_rate": 0.0001463266919466098,
+      "loss": 0.9082,
+      "step": 13630
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.2828356623649597,
+      "learning_rate": 0.0001462238196093758,
+      "loss": 0.8458,
+      "step": 13635
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.24517838656902313,
+      "learning_rate": 0.00014612094904933772,
+      "loss": 0.9284,
+      "step": 13640
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.23123225569725037,
+      "learning_rate": 0.00014601808031490982,
+      "loss": 0.7244,
+      "step": 13645
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.23790529370307922,
+      "learning_rate": 0.00014591521345450558,
+      "loss": 0.7898,
+      "step": 13650
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2609568238258362,
+      "learning_rate": 0.00014581234851653753,
+      "loss": 0.8663,
+      "step": 13655
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.36788272857666016,
+      "learning_rate": 0.00014570948554941735,
+      "loss": 0.8454,
+      "step": 13660
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.23802486062049866,
+      "learning_rate": 0.0001456066246015557,
+      "loss": 0.9379,
+      "step": 13665
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.26445284485816956,
+      "learning_rate": 0.00014550376572136246,
+      "loss": 0.9054,
+      "step": 13670
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.23677241802215576,
+      "learning_rate": 0.0001454009089572464,
+      "loss": 0.7262,
+      "step": 13675
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2473301738500595,
+      "learning_rate": 0.0001452980543576153,
+      "loss": 0.7921,
+      "step": 13680
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.25561195611953735,
+      "learning_rate": 0.0001451952019708759,
+      "loss": 0.8822,
+      "step": 13685
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.27183815836906433,
+      "learning_rate": 0.0001450923518454341,
+      "loss": 0.9218,
+      "step": 13690
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.22998422384262085,
+      "learning_rate": 0.0001449895040296945,
+      "loss": 0.909,
+      "step": 13695
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2734430730342865,
+      "learning_rate": 0.00014488665857206065,
+      "loss": 0.8962,
+      "step": 13700
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.23932689428329468,
+      "learning_rate": 0.0001447838155209351,
+      "loss": 0.8714,
+      "step": 13705
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.25987210869789124,
+      "learning_rate": 0.0001446809749247192,
+      "loss": 0.7435,
+      "step": 13710
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2543168365955353,
+      "learning_rate": 0.00014457813683181316,
+      "loss": 0.8738,
+      "step": 13715
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3010767698287964,
+      "learning_rate": 0.00014447530129061597,
+      "loss": 0.7546,
+      "step": 13720
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.24551789462566376,
+      "learning_rate": 0.00014437246834952537,
+      "loss": 0.8564,
+      "step": 13725
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.22315555810928345,
+      "learning_rate": 0.00014426963805693816,
+      "loss": 0.8453,
+      "step": 13730
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2511034905910492,
+      "learning_rate": 0.00014416681046124953,
+      "loss": 0.7888,
+      "step": 13735
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.28927719593048096,
+      "learning_rate": 0.00014406398561085364,
+      "loss": 0.8495,
+      "step": 13740
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.23084275424480438,
+      "learning_rate": 0.00014396116355414322,
+      "loss": 0.8212,
+      "step": 13745
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.29675766825675964,
+      "learning_rate": 0.0001438583443395098,
+      "loss": 0.97,
+      "step": 13750
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.27987009286880493,
+      "learning_rate": 0.00014375552801534352,
+      "loss": 0.8982,
+      "step": 13755
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.21575410664081573,
+      "learning_rate": 0.00014365271463003307,
+      "loss": 0.9205,
+      "step": 13760
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.27890732884407043,
+      "learning_rate": 0.0001435499042319659,
+      "loss": 0.8773,
+      "step": 13765
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.23023445904254913,
+      "learning_rate": 0.00014344709686952802,
+      "loss": 0.9438,
+      "step": 13770
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.2396959662437439,
+      "learning_rate": 0.000143344292591104,
+      "loss": 0.8498,
+      "step": 13775
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.2347070574760437,
+      "learning_rate": 0.0001432414914450769,
+      "loss": 0.8571,
+      "step": 13780
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.2968361973762512,
+      "learning_rate": 0.00014313869347982831,
+      "loss": 0.7974,
+      "step": 13785
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.26392439007759094,
+      "learning_rate": 0.0001430358987437385,
+      "loss": 0.8104,
+      "step": 13790
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.25702106952667236,
+      "learning_rate": 0.000142933107285186,
+      "loss": 0.8631,
+      "step": 13795
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.25434088706970215,
+      "learning_rate": 0.0001428303191525479,
+      "loss": 0.8464,
+      "step": 13800
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.23187103867530823,
+      "learning_rate": 0.00014272753439419962,
+      "loss": 0.8209,
+      "step": 13805
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.2695891261100769,
+      "learning_rate": 0.00014262475305851523,
+      "loss": 0.8192,
+      "step": 13810
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.24711039662361145,
+      "learning_rate": 0.000142521975193867,
+      "loss": 0.9271,
+      "step": 13815
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.25541651248931885,
+      "learning_rate": 0.00014241920084862554,
+      "loss": 0.8063,
+      "step": 13820
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.22565728425979614,
+      "learning_rate": 0.00014231643007115994,
+      "loss": 0.8552,
+      "step": 13825
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.24936147034168243,
+      "learning_rate": 0.0001422136629098375,
+      "loss": 0.872,
+      "step": 13830
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.29638171195983887,
+      "learning_rate": 0.0001421108994130239,
+      "loss": 0.9048,
+      "step": 13835
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.23596297204494476,
+      "learning_rate": 0.00014200813962908293,
+      "loss": 0.7751,
+      "step": 13840
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.23561915755271912,
+      "learning_rate": 0.00014190538360637695,
+      "loss": 0.8732,
+      "step": 13845
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.24926547706127167,
+      "learning_rate": 0.00014180263139326624,
+      "loss": 0.8895,
+      "step": 13850
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.30811989307403564,
+      "learning_rate": 0.00014169988303810942,
+      "loss": 0.8066,
+      "step": 13855
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.243475079536438,
+      "learning_rate": 0.00014159713858926323,
+      "loss": 0.8425,
+      "step": 13860
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.23428189754486084,
+      "learning_rate": 0.00014149439809508273,
+      "loss": 0.8981,
+      "step": 13865
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.26283276081085205,
+      "learning_rate": 0.00014139166160392094,
+      "loss": 0.8915,
+      "step": 13870
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2469039112329483,
+      "learning_rate": 0.00014128892916412907,
+      "loss": 0.7645,
+      "step": 13875
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2410871535539627,
+      "learning_rate": 0.00014118620082405637,
+      "loss": 0.811,
+      "step": 13880
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.23918049037456512,
+      "learning_rate": 0.00014108347663205033,
+      "loss": 0.7641,
+      "step": 13885
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2121327817440033,
+      "learning_rate": 0.00014098075663645628,
+      "loss": 0.9609,
+      "step": 13890
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2228885442018509,
+      "learning_rate": 0.0001408780408856177,
+      "loss": 0.9062,
+      "step": 13895
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.23560738563537598,
+      "learning_rate": 0.0001407753294278759,
+      "loss": 0.7888,
+      "step": 13900
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.25033679604530334,
+      "learning_rate": 0.0001406726223115705,
+      "loss": 0.8781,
+      "step": 13905
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.26739415526390076,
+      "learning_rate": 0.00014056991958503882,
+      "loss": 0.8365,
+      "step": 13910
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.24519668519496918,
+      "learning_rate": 0.0001404672212966161,
+      "loss": 0.9437,
+      "step": 13915
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.232150599360466,
+      "learning_rate": 0.0001403645274946356,
+      "loss": 0.7994,
+      "step": 13920
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.25735652446746826,
+      "learning_rate": 0.00014026183822742847,
+      "loss": 0.8427,
+      "step": 13925
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2597762644290924,
+      "learning_rate": 0.00014015915354332367,
+      "loss": 0.8662,
+      "step": 13930
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2447414994239807,
+      "learning_rate": 0.000140056473490648,
+      "loss": 0.8525,
+      "step": 13935
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.24776887893676758,
+      "learning_rate": 0.0001399537981177261,
+      "loss": 0.7445,
+      "step": 13940
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2430184930562973,
+      "learning_rate": 0.00013985112747288048,
+      "loss": 0.8798,
+      "step": 13945
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.26686951518058777,
+      "learning_rate": 0.00013974846160443128,
+      "loss": 0.8167,
+      "step": 13950
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2600416839122772,
+      "learning_rate": 0.0001396458005606965,
+      "loss": 0.7542,
+      "step": 13955
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.262658953666687,
+      "learning_rate": 0.0001395431443899918,
+      "loss": 0.8568,
+      "step": 13960
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.24480225145816803,
+      "learning_rate": 0.00013944049314063063,
+      "loss": 0.8143,
+      "step": 13965
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.25501397252082825,
+      "learning_rate": 0.0001393378468609241,
+      "loss": 0.7804,
+      "step": 13970
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.24708352982997894,
+      "learning_rate": 0.00013923520559918086,
+      "loss": 0.8356,
+      "step": 13975
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2388540506362915,
+      "learning_rate": 0.00013913256940370733,
+      "loss": 0.876,
+      "step": 13980
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2332463413476944,
+      "learning_rate": 0.00013902993832280757,
+      "loss": 0.7916,
+      "step": 13985
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2795470058917999,
+      "learning_rate": 0.00013892731240478317,
+      "loss": 0.8881,
+      "step": 13990
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2751898169517517,
+      "learning_rate": 0.00013882469169793324,
+      "loss": 0.7779,
+      "step": 13995
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.27980801463127136,
+      "learning_rate": 0.00013872207625055449,
+      "loss": 0.8326,
+      "step": 14000
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2469402402639389,
+      "learning_rate": 0.00013861946611094125,
+      "loss": 0.85,
+      "step": 14005
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.23618659377098083,
+      "learning_rate": 0.00013851686132738516,
+      "loss": 0.7895,
+      "step": 14010
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.25577905774116516,
+      "learning_rate": 0.00013841426194817548,
+      "loss": 0.8099,
+      "step": 14015
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.23130281269550323,
+      "learning_rate": 0.0001383116680215988,
+      "loss": 0.9164,
+      "step": 14020
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2609565258026123,
+      "learning_rate": 0.00013820907959593938,
+      "loss": 0.8081,
+      "step": 14025
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.24102197587490082,
+      "learning_rate": 0.00013810649671947868,
+      "loss": 0.7962,
+      "step": 14030
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.23725546896457672,
+      "learning_rate": 0.0001380039194404956,
+      "loss": 0.813,
+      "step": 14035
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.24073052406311035,
+      "learning_rate": 0.00013790134780726634,
+      "loss": 0.7779,
+      "step": 14040
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2828931510448456,
+      "learning_rate": 0.00013779878186806463,
+      "loss": 0.8943,
+      "step": 14045
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.28520870208740234,
+      "learning_rate": 0.00013769622167116138,
+      "loss": 0.8855,
+      "step": 14050
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2628406882286072,
+      "learning_rate": 0.0001375936672648248,
+      "loss": 0.796,
+      "step": 14055
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.23413865268230438,
+      "learning_rate": 0.00013749111869732034,
+      "loss": 0.9203,
+      "step": 14060
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.24607300758361816,
+      "learning_rate": 0.0001373885760169109,
+      "loss": 0.8389,
+      "step": 14065
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.23336441814899445,
+      "learning_rate": 0.00013728603927185644,
+      "loss": 0.8227,
+      "step": 14070
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.284675270318985,
+      "learning_rate": 0.00013718350851041407,
+      "loss": 0.8254,
+      "step": 14075
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2564062178134918,
+      "learning_rate": 0.00013708098378083813,
+      "loss": 0.7736,
+      "step": 14080
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2292274683713913,
+      "learning_rate": 0.00013697846513138035,
+      "loss": 0.8294,
+      "step": 14085
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2765873074531555,
+      "learning_rate": 0.0001368759526102893,
+      "loss": 0.8284,
+      "step": 14090
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2501067519187927,
+      "learning_rate": 0.0001367734462658108,
+      "loss": 0.9865,
+      "step": 14095
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.26736894249916077,
+      "learning_rate": 0.00013667094614618766,
+      "loss": 0.7624,
+      "step": 14100
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.24033088982105255,
+      "learning_rate": 0.00013656845229965996,
+      "loss": 0.8987,
+      "step": 14105
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2716580927371979,
+      "learning_rate": 0.00013646596477446467,
+      "loss": 0.9425,
+      "step": 14110
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2522694766521454,
+      "learning_rate": 0.00013636348361883578,
+      "loss": 0.9069,
+      "step": 14115
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.25670236349105835,
+      "learning_rate": 0.00013626100888100432,
+      "loss": 0.9238,
+      "step": 14120
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.26054003834724426,
+      "learning_rate": 0.00013615854060919838,
+      "loss": 0.8625,
+      "step": 14125
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.23610499501228333,
+      "learning_rate": 0.0001360560788516429,
+      "loss": 0.7802,
+      "step": 14130
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.2758851647377014,
+      "learning_rate": 0.0001359536236565598,
+      "loss": 0.8364,
+      "step": 14135
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.26497259736061096,
+      "learning_rate": 0.0001358511750721678,
+      "loss": 0.8287,
+      "step": 14140
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.24857574701309204,
+      "learning_rate": 0.0001357487331466827,
+      "loss": 0.862,
+      "step": 14145
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.28877443075180054,
+      "learning_rate": 0.0001356462979283171,
+      "loss": 0.8711,
+      "step": 14150
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2738126218318939,
+      "learning_rate": 0.00013554386946528033,
+      "loss": 0.9011,
+      "step": 14155
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.26807695627212524,
+      "learning_rate": 0.0001354414478057786,
+      "loss": 0.7915,
+      "step": 14160
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.269781231880188,
+      "learning_rate": 0.0001353390329980151,
+      "loss": 0.8746,
+      "step": 14165
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2851022183895111,
+      "learning_rate": 0.00013523662509018952,
+      "loss": 0.8851,
+      "step": 14170
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.26294824481010437,
+      "learning_rate": 0.00013513422413049847,
+      "loss": 0.7998,
+      "step": 14175
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.23697425425052643,
+      "learning_rate": 0.00013503183016713518,
+      "loss": 0.8183,
+      "step": 14180
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.22184935212135315,
+      "learning_rate": 0.0001349294432482897,
+      "loss": 0.7641,
+      "step": 14185
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.23600788414478302,
+      "learning_rate": 0.00013482706342214873,
+      "loss": 0.8828,
+      "step": 14190
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.26765936613082886,
+      "learning_rate": 0.0001347246907368956,
+      "loss": 0.8785,
+      "step": 14195
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2510226368904114,
+      "learning_rate": 0.00013462232524071022,
+      "loss": 0.9082,
+      "step": 14200
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2888829708099365,
+      "learning_rate": 0.0001345199669817693,
+      "loss": 0.8412,
+      "step": 14205
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.22700832784175873,
+      "learning_rate": 0.00013441761600824602,
+      "loss": 0.8552,
+      "step": 14210
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2521451711654663,
+      "learning_rate": 0.0001343152723683101,
+      "loss": 0.7209,
+      "step": 14215
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.24360665678977966,
+      "learning_rate": 0.00013421293611012784,
+      "loss": 0.8414,
+      "step": 14220
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2193845957517624,
+      "learning_rate": 0.00013411060728186217,
+      "loss": 0.7894,
+      "step": 14225
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.2685592472553253,
+      "learning_rate": 0.00013400828593167238,
+      "loss": 0.8794,
+      "step": 14230
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.23406338691711426,
+      "learning_rate": 0.0001339059721077143,
+      "loss": 0.8955,
+      "step": 14235
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2340015023946762,
+      "learning_rate": 0.00013380366585814016,
+      "loss": 0.7859,
+      "step": 14240
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2922716438770294,
+      "learning_rate": 0.00013370136723109876,
+      "loss": 0.7737,
+      "step": 14245
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.26283466815948486,
+      "learning_rate": 0.0001335990762747352,
+      "loss": 0.8702,
+      "step": 14250
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2611430287361145,
+      "learning_rate": 0.00013349679303719105,
+      "loss": 0.7799,
+      "step": 14255
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2265879213809967,
+      "learning_rate": 0.00013339451756660408,
+      "loss": 0.8453,
+      "step": 14260
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2740434408187866,
+      "learning_rate": 0.00013329224991110865,
+      "loss": 0.854,
+      "step": 14265
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.23583091795444489,
+      "learning_rate": 0.00013318999011883526,
+      "loss": 0.8633,
+      "step": 14270
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.21470926702022552,
+      "learning_rate": 0.00013308773823791074,
+      "loss": 0.8227,
+      "step": 14275
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2593861520290375,
+      "learning_rate": 0.0001329854943164582,
+      "loss": 0.7781,
+      "step": 14280
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.27231550216674805,
+      "learning_rate": 0.00013288325840259715,
+      "loss": 0.8061,
+      "step": 14285
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2605496644973755,
+      "learning_rate": 0.0001327810305444431,
+      "loss": 0.9153,
+      "step": 14290
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.280644953250885,
+      "learning_rate": 0.0001326788107901079,
+      "loss": 0.8612,
+      "step": 14295
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2497917264699936,
+      "learning_rate": 0.0001325765991876995,
+      "loss": 0.7707,
+      "step": 14300
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.26758888363838196,
+      "learning_rate": 0.0001324743957853222,
+      "loss": 0.8533,
+      "step": 14305
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2774585485458374,
+      "learning_rate": 0.00013237220063107625,
+      "loss": 0.9308,
+      "step": 14310
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.21268567442893982,
+      "learning_rate": 0.00013227001377305806,
+      "loss": 0.8137,
+      "step": 14315
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2505970001220703,
+      "learning_rate": 0.0001321678352593602,
+      "loss": 0.7499,
+      "step": 14320
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2653083801269531,
+      "learning_rate": 0.00013206566513807125,
+      "loss": 0.8927,
+      "step": 14325
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2813301086425781,
+      "learning_rate": 0.0001319635034572759,
+      "loss": 0.8058,
+      "step": 14330
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.23401866853237152,
+      "learning_rate": 0.0001318613502650547,
+      "loss": 0.8581,
+      "step": 14335
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.24355721473693848,
+      "learning_rate": 0.0001317592056094845,
+      "loss": 0.9558,
+      "step": 14340
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.23587583005428314,
+      "learning_rate": 0.0001316570695386379,
+      "loss": 0.8461,
+      "step": 14345
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2816319763660431,
+      "learning_rate": 0.00013155494210058353,
+      "loss": 0.8245,
+      "step": 14350
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2452281266450882,
+      "learning_rate": 0.0001314528233433859,
+      "loss": 0.8102,
+      "step": 14355
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2361225038766861,
+      "learning_rate": 0.0001313507133151056,
+      "loss": 0.7548,
+      "step": 14360
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2706858217716217,
+      "learning_rate": 0.0001312486120637989,
+      "loss": 0.8989,
+      "step": 14365
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.25683704018592834,
+      "learning_rate": 0.0001311465196375181,
+      "loss": 0.87,
+      "step": 14370
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2547888457775116,
+      "learning_rate": 0.0001310444360843112,
+      "loss": 0.8649,
+      "step": 14375
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.28074026107788086,
+      "learning_rate": 0.00013094236145222223,
+      "loss": 0.8258,
+      "step": 14380
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.2704260051250458,
+      "learning_rate": 0.00013084029578929086,
+      "loss": 0.8283,
+      "step": 14385
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.22389455139636993,
+      "learning_rate": 0.00013073823914355257,
+      "loss": 0.7883,
+      "step": 14390
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.29739248752593994,
+      "learning_rate": 0.00013063619156303854,
+      "loss": 0.9081,
+      "step": 14395
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.25847524404525757,
+      "learning_rate": 0.00013053415309577588,
+      "loss": 0.8948,
+      "step": 14400
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2561867833137512,
+      "learning_rate": 0.0001304321237897872,
+      "loss": 0.9301,
+      "step": 14405
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.23929399251937866,
+      "learning_rate": 0.00013033010369309088,
+      "loss": 0.853,
+      "step": 14410
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2782881557941437,
+      "learning_rate": 0.0001302280928537009,
+      "loss": 0.8784,
+      "step": 14415
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2864670753479004,
+      "learning_rate": 0.00013012609131962712,
+      "loss": 0.888,
+      "step": 14420
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.25334519147872925,
+      "learning_rate": 0.00013002409913887475,
+      "loss": 0.8426,
+      "step": 14425
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2505394518375397,
+      "learning_rate": 0.00012992211635944474,
+      "loss": 0.7904,
+      "step": 14430
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2567066252231598,
+      "learning_rate": 0.00012982014302933347,
+      "loss": 0.838,
+      "step": 14435
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2711067199707031,
+      "learning_rate": 0.00012971817919653307,
+      "loss": 0.8792,
+      "step": 14440
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.25377967953681946,
+      "learning_rate": 0.00012961622490903108,
+      "loss": 0.7812,
+      "step": 14445
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2506578862667084,
+      "learning_rate": 0.00012951428021481056,
+      "loss": 0.9277,
+      "step": 14450
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.22411175072193146,
+      "learning_rate": 0.00012941234516185003,
+      "loss": 0.9082,
+      "step": 14455
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.252632200717926,
+      "learning_rate": 0.00012931041979812364,
+      "loss": 0.7627,
+      "step": 14460
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2694997489452362,
+      "learning_rate": 0.00012920850417160078,
+      "loss": 0.9178,
+      "step": 14465
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2513391077518463,
+      "learning_rate": 0.0001291065983302463,
+      "loss": 0.8679,
+      "step": 14470
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2643163800239563,
+      "learning_rate": 0.00012900470232202045,
+      "loss": 0.8087,
+      "step": 14475
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2633654773235321,
+      "learning_rate": 0.00012890281619487898,
+      "loss": 0.8016,
+      "step": 14480
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.24543441832065582,
+      "learning_rate": 0.00012880093999677282,
+      "loss": 0.8115,
+      "step": 14485
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2534711956977844,
+      "learning_rate": 0.00012869907377564827,
+      "loss": 0.7686,
+      "step": 14490
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.24713997542858124,
+      "learning_rate": 0.00012859721757944696,
+      "loss": 0.8995,
+      "step": 14495
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2603427469730377,
+      "learning_rate": 0.00012849537145610587,
+      "loss": 0.9065,
+      "step": 14500
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.26300033926963806,
+      "learning_rate": 0.00012839353545355712,
+      "loss": 0.8299,
+      "step": 14505
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2567319869995117,
+      "learning_rate": 0.0001282917096197281,
+      "loss": 0.8677,
+      "step": 14510
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2563308775424957,
+      "learning_rate": 0.0001281898940025414,
+      "loss": 0.7826,
+      "step": 14515
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2692480683326721,
+      "learning_rate": 0.0001280880886499149,
+      "loss": 0.8732,
+      "step": 14520
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.26520323753356934,
+      "learning_rate": 0.0001279862936097616,
+      "loss": 0.8239,
+      "step": 14525
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2513013780117035,
+      "learning_rate": 0.00012788450892998952,
+      "loss": 0.7887,
+      "step": 14530
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2527408003807068,
+      "learning_rate": 0.000127782734658502,
+      "loss": 0.9332,
+      "step": 14535
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2598322629928589,
+      "learning_rate": 0.00012768097084319736,
+      "loss": 0.8357,
+      "step": 14540
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.257907509803772,
+      "learning_rate": 0.00012757921753196906,
+      "loss": 0.7879,
+      "step": 14545
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2565610110759735,
+      "learning_rate": 0.00012747747477270552,
+      "loss": 0.9305,
+      "step": 14550
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.25154614448547363,
+      "learning_rate": 0.00012737574261329027,
+      "loss": 0.734,
+      "step": 14555
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.24832646548748016,
+      "learning_rate": 0.00012727402110160194,
+      "loss": 0.853,
+      "step": 14560
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.2888548672199249,
+      "learning_rate": 0.00012717231028551397,
+      "loss": 0.8503,
+      "step": 14565
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2529507279396057,
+      "learning_rate": 0.00012707061021289485,
+      "loss": 0.8343,
+      "step": 14570
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.27451494336128235,
+      "learning_rate": 0.00012696892093160803,
+      "loss": 0.7783,
+      "step": 14575
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2568528652191162,
+      "learning_rate": 0.00012686724248951189,
+      "loss": 0.9118,
+      "step": 14580
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2533772587776184,
+      "learning_rate": 0.00012676557493445962,
+      "loss": 0.7751,
+      "step": 14585
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.272686630487442,
+      "learning_rate": 0.0001266639183142994,
+      "loss": 0.8448,
+      "step": 14590
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.25127992033958435,
+      "learning_rate": 0.0001265622726768741,
+      "loss": 0.8768,
+      "step": 14595
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2625642418861389,
+      "learning_rate": 0.00012646063807002168,
+      "loss": 0.9275,
+      "step": 14600
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.26558226346969604,
+      "learning_rate": 0.00012635901454157472,
+      "loss": 0.7573,
+      "step": 14605
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2596949338912964,
+      "learning_rate": 0.00012625740213936064,
+      "loss": 0.8072,
+      "step": 14610
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.25480490922927856,
+      "learning_rate": 0.0001261558009112015,
+      "loss": 0.9022,
+      "step": 14615
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2669465243816376,
+      "learning_rate": 0.00012605421090491434,
+      "loss": 0.978,
+      "step": 14620
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.22800852358341217,
+      "learning_rate": 0.00012595263216831076,
+      "loss": 0.6918,
+      "step": 14625
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2479008287191391,
+      "learning_rate": 0.00012585106474919704,
+      "loss": 0.8508,
+      "step": 14630
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.24926097691059113,
+      "learning_rate": 0.00012574950869537419,
+      "loss": 0.7538,
+      "step": 14635
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.23529042303562164,
+      "learning_rate": 0.0001256479640546379,
+      "loss": 0.8676,
+      "step": 14640
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.24101610481739044,
+      "learning_rate": 0.00012554643087477844,
+      "loss": 0.9266,
+      "step": 14645
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.28825363516807556,
+      "learning_rate": 0.00012544490920358072,
+      "loss": 0.9265,
+      "step": 14650
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2545327842235565,
+      "learning_rate": 0.00012534339908882412,
+      "loss": 0.8619,
+      "step": 14655
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.3079511523246765,
+      "learning_rate": 0.00012524190057828277,
+      "loss": 0.853,
+      "step": 14660
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2799226939678192,
+      "learning_rate": 0.00012514041371972518,
+      "loss": 0.9301,
+      "step": 14665
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.25301575660705566,
+      "learning_rate": 0.00012503893856091448,
+      "loss": 0.8581,
+      "step": 14670
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.26482102274894714,
+      "learning_rate": 0.00012493747514960815,
+      "loss": 0.9563,
+      "step": 14675
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.23120246827602386,
+      "learning_rate": 0.0001248360235335584,
+      "loss": 0.8965,
+      "step": 14680
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2709689736366272,
+      "learning_rate": 0.00012473458376051163,
+      "loss": 0.7622,
+      "step": 14685
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.254367470741272,
+      "learning_rate": 0.00012463315587820878,
+      "loss": 0.9004,
+      "step": 14690
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.25970759987831116,
+      "learning_rate": 0.0001245317399343851,
+      "loss": 0.7974,
+      "step": 14695
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.23521548509597778,
+      "learning_rate": 0.00012443033597677047,
+      "loss": 0.7991,
+      "step": 14700
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2456759810447693,
+      "learning_rate": 0.00012432894405308887,
+      "loss": 0.9068,
+      "step": 14705
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.3220599889755249,
+      "learning_rate": 0.00012422756421105868,
+      "loss": 0.8967,
+      "step": 14710
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.25704309344291687,
+      "learning_rate": 0.00012412619649839263,
+      "loss": 0.9185,
+      "step": 14715
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2799650728702545,
+      "learning_rate": 0.0001240248409627978,
+      "loss": 0.8382,
+      "step": 14720
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.22531536221504211,
+      "learning_rate": 0.00012392349765197541,
+      "loss": 0.7933,
+      "step": 14725
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2780819535255432,
+      "learning_rate": 0.00012382216661362098,
+      "loss": 0.8583,
+      "step": 14730
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2472156137228012,
+      "learning_rate": 0.00012372084789542424,
+      "loss": 0.7335,
+      "step": 14735
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.22671552002429962,
+      "learning_rate": 0.00012361954154506926,
+      "loss": 0.8889,
+      "step": 14740
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.23616662621498108,
+      "learning_rate": 0.00012351824761023405,
+      "loss": 0.9454,
+      "step": 14745
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.24710243940353394,
+      "learning_rate": 0.00012341696613859098,
+      "loss": 0.7842,
+      "step": 14750
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.27012568712234497,
+      "learning_rate": 0.0001233156971778064,
+      "loss": 0.7432,
+      "step": 14755
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.30002421140670776,
+      "learning_rate": 0.00012321444077554095,
+      "loss": 0.9517,
+      "step": 14760
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2851850688457489,
+      "learning_rate": 0.00012311319697944913,
+      "loss": 0.8708,
+      "step": 14765
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2681834101676941,
+      "learning_rate": 0.00012301196583717973,
+      "loss": 0.9541,
+      "step": 14770
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2770148515701294,
+      "learning_rate": 0.0001229107473963754,
+      "loss": 0.7985,
+      "step": 14775
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.23922573029994965,
+      "learning_rate": 0.000122809541704673,
+      "loss": 0.8304,
+      "step": 14780
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.38129690289497375,
+      "learning_rate": 0.00012270834880970323,
+      "loss": 0.8315,
+      "step": 14785
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2657111585140228,
+      "learning_rate": 0.00012260716875909085,
+      "loss": 0.792,
+      "step": 14790
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2710582911968231,
+      "learning_rate": 0.0001225060016004545,
+      "loss": 0.7308,
+      "step": 14795
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2484636902809143,
+      "learning_rate": 0.0001224048473814069,
+      "loss": 0.852,
+      "step": 14800
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.24950478971004486,
+      "learning_rate": 0.0001223037061495545,
+      "loss": 0.8909,
+      "step": 14805
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2482597976922989,
+      "learning_rate": 0.00012220257795249778,
+      "loss": 0.8531,
+      "step": 14810
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2781403362751007,
+      "learning_rate": 0.00012210146283783092,
+      "loss": 0.855,
+      "step": 14815
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.26472207903862,
+      "learning_rate": 0.00012200036085314218,
+      "loss": 0.8358,
+      "step": 14820
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.25163668394088745,
+      "learning_rate": 0.00012189927204601348,
+      "loss": 0.8767,
+      "step": 14825
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.23350991308689117,
+      "learning_rate": 0.00012179819646402052,
+      "loss": 0.8239,
+      "step": 14830
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.275762677192688,
+      "learning_rate": 0.00012169713415473288,
+      "loss": 0.7539,
+      "step": 14835
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2697180509567261,
+      "learning_rate": 0.00012159608516571383,
+      "loss": 0.8765,
+      "step": 14840
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.25511398911476135,
+      "learning_rate": 0.00012149504954452036,
+      "loss": 0.9266,
+      "step": 14845
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.26254042983055115,
+      "learning_rate": 0.0001213940273387031,
+      "loss": 0.8991,
+      "step": 14850
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.27790650725364685,
+      "learning_rate": 0.00012129301859580665,
+      "loss": 0.8492,
+      "step": 14855
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2523384690284729,
+      "learning_rate": 0.00012119202336336897,
+      "loss": 0.8993,
+      "step": 14860
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.23431278765201569,
+      "learning_rate": 0.00012109104168892177,
+      "loss": 0.9201,
+      "step": 14865
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2700154185295105,
+      "learning_rate": 0.00012099007361999037,
+      "loss": 0.8558,
+      "step": 14870
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2271653562784195,
+      "learning_rate": 0.00012088911920409374,
+      "loss": 0.8236,
+      "step": 14875
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.24218647181987762,
+      "learning_rate": 0.00012078817848874434,
+      "loss": 0.7495,
+      "step": 14880
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.25467637181282043,
+      "learning_rate": 0.00012068725152144827,
+      "loss": 0.8275,
+      "step": 14885
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.277286559343338,
+      "learning_rate": 0.00012058633834970502,
+      "loss": 0.8756,
+      "step": 14890
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.26092687249183655,
+      "learning_rate": 0.00012048543902100779,
+      "loss": 0.8228,
+      "step": 14895
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.24140042066574097,
+      "learning_rate": 0.00012038455358284309,
+      "loss": 0.8216,
+      "step": 14900
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2441730499267578,
+      "learning_rate": 0.00012028368208269097,
+      "loss": 0.873,
+      "step": 14905
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.242695614695549,
+      "learning_rate": 0.00012018282456802487,
+      "loss": 0.734,
+      "step": 14910
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.27291175723075867,
+      "learning_rate": 0.00012008198108631176,
+      "loss": 1.0009,
+      "step": 14915
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.25119420886039734,
+      "learning_rate": 0.00011998115168501192,
+      "loss": 1.0065,
+      "step": 14920
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2412206530570984,
+      "learning_rate": 0.00011988033641157898,
+      "loss": 0.8785,
+      "step": 14925
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.24067923426628113,
+      "learning_rate": 0.00011977953531345996,
+      "loss": 0.9148,
+      "step": 14930
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2587399184703827,
+      "learning_rate": 0.00011967874843809522,
+      "loss": 0.8226,
+      "step": 14935
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.26349955797195435,
+      "learning_rate": 0.00011957797583291841,
+      "loss": 0.7778,
+      "step": 14940
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2319829910993576,
+      "learning_rate": 0.00011947721754535645,
+      "loss": 0.7828,
+      "step": 14945
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.27520623803138733,
+      "learning_rate": 0.00011937647362282948,
+      "loss": 0.8186,
+      "step": 14950
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.23423904180526733,
+      "learning_rate": 0.00011927574411275107,
+      "loss": 0.773,
+      "step": 14955
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2540885806083679,
+      "learning_rate": 0.0001191750290625278,
+      "loss": 0.7941,
+      "step": 14960
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.26657721400260925,
+      "learning_rate": 0.00011907432851955952,
+      "loss": 0.848,
+      "step": 14965
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.22718775272369385,
+      "learning_rate": 0.00011897364253123921,
+      "loss": 0.7955,
+      "step": 14970
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.252336323261261,
+      "learning_rate": 0.00011887297114495312,
+      "loss": 0.894,
+      "step": 14975
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2825542390346527,
+      "learning_rate": 0.0001187723144080805,
+      "loss": 0.9111,
+      "step": 14980
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2752092182636261,
+      "learning_rate": 0.00011867167236799376,
+      "loss": 0.839,
+      "step": 14985
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2676568925380707,
+      "learning_rate": 0.00011857104507205831,
+      "loss": 0.9122,
+      "step": 14990
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.24729391932487488,
+      "learning_rate": 0.00011847043256763285,
+      "loss": 0.7432,
+      "step": 14995
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.24606819450855255,
+      "learning_rate": 0.00011836983490206889,
+      "loss": 0.7616,
+      "step": 15000
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.27244672179222107,
+      "learning_rate": 0.00011826925212271102,
+      "loss": 0.8588,
+      "step": 15005
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.26637476682662964,
+      "learning_rate": 0.00011816868427689683,
+      "loss": 0.9134,
+      "step": 15010
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.27990347146987915,
+      "learning_rate": 0.00011806813141195691,
+      "loss": 0.8513,
+      "step": 15015
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.33465734124183655,
+      "learning_rate": 0.0001179675935752148,
+      "loss": 0.8542,
+      "step": 15020
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.22314803302288055,
+      "learning_rate": 0.0001178670708139869,
+      "loss": 0.8547,
+      "step": 15025
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2743472754955292,
+      "learning_rate": 0.00011776656317558251,
+      "loss": 0.8579,
+      "step": 15030
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.23580728471279144,
+      "learning_rate": 0.000117666070707304,
+      "loss": 0.9977,
+      "step": 15035
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2850353419780731,
+      "learning_rate": 0.0001175655934564464,
+      "loss": 0.7897,
+      "step": 15040
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2732098698616028,
+      "learning_rate": 0.00011746513147029762,
+      "loss": 0.6902,
+      "step": 15045
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.26247236132621765,
+      "learning_rate": 0.00011736468479613841,
+      "loss": 0.74,
+      "step": 15050
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.23917360603809357,
+      "learning_rate": 0.00011726425348124232,
+      "loss": 0.8132,
+      "step": 15055
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.2416449338197708,
+      "learning_rate": 0.00011716383757287568,
+      "loss": 0.94,
+      "step": 15060
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.26987558603286743,
+      "learning_rate": 0.00011706343711829753,
+      "loss": 0.8469,
+      "step": 15065
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2562698721885681,
+      "learning_rate": 0.0001169630521647596,
+      "loss": 0.7888,
+      "step": 15070
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2376549392938614,
+      "learning_rate": 0.0001168626827595065,
+      "loss": 0.8179,
+      "step": 15075
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.23913566768169403,
+      "learning_rate": 0.0001167623289497754,
+      "loss": 0.9149,
+      "step": 15080
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2533036172389984,
+      "learning_rate": 0.00011666199078279604,
+      "loss": 0.8483,
+      "step": 15085
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.23935401439666748,
+      "learning_rate": 0.00011656166830579087,
+      "loss": 0.8704,
+      "step": 15090
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.243647962808609,
+      "learning_rate": 0.00011646136156597513,
+      "loss": 0.8779,
+      "step": 15095
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.27903953194618225,
+      "learning_rate": 0.0001163610706105564,
+      "loss": 0.8851,
+      "step": 15100
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.25441214442253113,
+      "learning_rate": 0.00011626079548673496,
+      "loss": 0.8484,
+      "step": 15105
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.23863151669502258,
+      "learning_rate": 0.00011616053624170359,
+      "loss": 0.8271,
+      "step": 15110
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.271961510181427,
+      "learning_rate": 0.00011606029292264766,
+      "loss": 0.855,
+      "step": 15115
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.25115951895713806,
+      "learning_rate": 0.00011596006557674497,
+      "loss": 0.8599,
+      "step": 15120
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.21527209877967834,
+      "learning_rate": 0.00011585985425116589,
+      "loss": 0.8878,
+      "step": 15125
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2371513992547989,
+      "learning_rate": 0.00011575965899307306,
+      "loss": 0.8347,
+      "step": 15130
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.26192623376846313,
+      "learning_rate": 0.00011565947984962187,
+      "loss": 0.814,
+      "step": 15135
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2839601933956146,
+      "learning_rate": 0.00011555931686795987,
+      "loss": 0.8186,
+      "step": 15140
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2582460045814514,
+      "learning_rate": 0.0001154591700952271,
+      "loss": 0.7673,
+      "step": 15145
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2260724902153015,
+      "learning_rate": 0.0001153590395785559,
+      "loss": 0.8031,
+      "step": 15150
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2650599777698517,
+      "learning_rate": 0.00011525892536507111,
+      "loss": 0.8469,
+      "step": 15155
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.280758798122406,
+      "learning_rate": 0.00011515882750188976,
+      "loss": 0.8937,
+      "step": 15160
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.23013721406459808,
+      "learning_rate": 0.00011505874603612122,
+      "loss": 0.8758,
+      "step": 15165
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.23622775077819824,
+      "learning_rate": 0.0001149586810148671,
+      "loss": 0.9219,
+      "step": 15170
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.23313015699386597,
+      "learning_rate": 0.00011485863248522144,
+      "loss": 0.8072,
+      "step": 15175
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.27392974495887756,
+      "learning_rate": 0.00011475860049427036,
+      "loss": 0.7727,
+      "step": 15180
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.24279291927814484,
+      "learning_rate": 0.00011465858508909219,
+      "loss": 0.8076,
+      "step": 15185
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.22247040271759033,
+      "learning_rate": 0.00011455858631675752,
+      "loss": 0.8597,
+      "step": 15190
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.23884879052639008,
+      "learning_rate": 0.00011445860422432913,
+      "loss": 0.9097,
+      "step": 15195
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.22139222919940948,
+      "learning_rate": 0.00011435863885886188,
+      "loss": 0.8262,
+      "step": 15200
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.20830343663692474,
+      "learning_rate": 0.00011425869026740278,
+      "loss": 0.8878,
+      "step": 15205
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.23654043674468994,
+      "learning_rate": 0.00011415875849699094,
+      "loss": 0.8093,
+      "step": 15210
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2731797993183136,
+      "learning_rate": 0.00011405884359465766,
+      "loss": 0.8693,
+      "step": 15215
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2721179723739624,
+      "learning_rate": 0.00011395894560742612,
+      "loss": 0.7557,
+      "step": 15220
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2314218133687973,
+      "learning_rate": 0.0001138590645823117,
+      "loss": 0.911,
+      "step": 15225
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2549746036529541,
+      "learning_rate": 0.00011375920056632164,
+      "loss": 0.8295,
+      "step": 15230
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2613559365272522,
+      "learning_rate": 0.00011365935360645536,
+      "loss": 0.8176,
+      "step": 15235
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.24376095831394196,
+      "learning_rate": 0.0001135595237497041,
+      "loss": 0.8712,
+      "step": 15240
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2682994604110718,
+      "learning_rate": 0.00011345971104305111,
+      "loss": 0.8861,
+      "step": 15245
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2671447992324829,
+      "learning_rate": 0.0001133599155334715,
+      "loss": 0.9091,
+      "step": 15250
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2641143500804901,
+      "learning_rate": 0.00011326013726793249,
+      "loss": 0.8653,
+      "step": 15255
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2844027280807495,
+      "learning_rate": 0.00011316037629339299,
+      "loss": 0.8379,
+      "step": 15260
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.25467634201049805,
+      "learning_rate": 0.00011306063265680384,
+      "loss": 0.7981,
+      "step": 15265
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.27292025089263916,
+      "learning_rate": 0.00011296090640510758,
+      "loss": 0.9155,
+      "step": 15270
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.27698391675949097,
+      "learning_rate": 0.00011286119758523885,
+      "loss": 0.8058,
+      "step": 15275
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2688562870025635,
+      "learning_rate": 0.00011276150624412388,
+      "loss": 0.8207,
+      "step": 15280
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2935139238834381,
+      "learning_rate": 0.00011266183242868073,
+      "loss": 0.8824,
+      "step": 15285
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.26238155364990234,
+      "learning_rate": 0.00011256217618581916,
+      "loss": 0.8231,
+      "step": 15290
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.23228612542152405,
+      "learning_rate": 0.00011246253756244079,
+      "loss": 0.8126,
+      "step": 15295
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2749589681625366,
+      "learning_rate": 0.00011236291660543881,
+      "loss": 0.8331,
+      "step": 15300
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.26084092259407043,
+      "learning_rate": 0.0001122633133616982,
+      "loss": 0.9848,
+      "step": 15305
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.2661038041114807,
+      "learning_rate": 0.0001121637278780954,
+      "loss": 0.873,
+      "step": 15310
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.27569863200187683,
+      "learning_rate": 0.00011206416020149887,
+      "loss": 0.8332,
+      "step": 15315
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.23671133816242218,
+      "learning_rate": 0.00011196461037876834,
+      "loss": 0.8793,
+      "step": 15320
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.26371756196022034,
+      "learning_rate": 0.00011186507845675527,
+      "loss": 0.8082,
+      "step": 15325
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2549302875995636,
+      "learning_rate": 0.00011176556448230271,
+      "loss": 0.7892,
+      "step": 15330
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2468237727880478,
+      "learning_rate": 0.00011166606850224524,
+      "loss": 0.7797,
+      "step": 15335
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2780306041240692,
+      "learning_rate": 0.00011156659056340898,
+      "loss": 0.7513,
+      "step": 15340
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.264596551656723,
+      "learning_rate": 0.00011146713071261145,
+      "loss": 0.8805,
+      "step": 15345
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2658274173736572,
+      "learning_rate": 0.00011136768899666191,
+      "loss": 0.8718,
+      "step": 15350
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.24554196000099182,
+      "learning_rate": 0.00011126826546236087,
+      "loss": 0.943,
+      "step": 15355
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.24591578543186188,
+      "learning_rate": 0.00011116886015650035,
+      "loss": 0.7976,
+      "step": 15360
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2368362545967102,
+      "learning_rate": 0.00011106947312586373,
+      "loss": 0.867,
+      "step": 15365
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.2802934944629669,
+      "learning_rate": 0.00011097010441722595,
+      "loss": 0.8927,
+      "step": 15370
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.25117719173431396,
+      "learning_rate": 0.00011087075407735316,
+      "loss": 0.8307,
+      "step": 15375
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.23962007462978363,
+      "learning_rate": 0.00011077142215300297,
+      "loss": 0.8692,
+      "step": 15380
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.22337029874324799,
+      "learning_rate": 0.00011067210869092417,
+      "loss": 0.7974,
+      "step": 15385
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.25178852677345276,
+      "learning_rate": 0.00011057281373785712,
+      "loss": 0.9266,
+      "step": 15390
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.25446346402168274,
+      "learning_rate": 0.00011047353734053327,
+      "loss": 0.8743,
+      "step": 15395
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.26338401436805725,
+      "learning_rate": 0.0001103742795456754,
+      "loss": 0.8066,
+      "step": 15400
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2573286294937134,
+      "learning_rate": 0.00011027504039999744,
+      "loss": 0.8641,
+      "step": 15405
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2608354091644287,
+      "learning_rate": 0.00011017581995020475,
+      "loss": 0.8588,
+      "step": 15410
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.23720592260360718,
+      "learning_rate": 0.0001100766182429937,
+      "loss": 0.8744,
+      "step": 15415
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.26355910301208496,
+      "learning_rate": 0.00010997743532505192,
+      "loss": 0.8457,
+      "step": 15420
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.3250594139099121,
+      "learning_rate": 0.00010987827124305812,
+      "loss": 0.7379,
+      "step": 15425
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2539902329444885,
+      "learning_rate": 0.00010977912604368232,
+      "loss": 0.7225,
+      "step": 15430
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.23028779029846191,
+      "learning_rate": 0.00010967999977358551,
+      "loss": 0.8119,
+      "step": 15435
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.23592324554920197,
+      "learning_rate": 0.00010958089247941981,
+      "loss": 0.8199,
+      "step": 15440
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2703402638435364,
+      "learning_rate": 0.00010948180420782835,
+      "loss": 0.8367,
+      "step": 15445
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2614709734916687,
+      "learning_rate": 0.00010938273500544543,
+      "loss": 0.7616,
+      "step": 15450
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2527773976325989,
+      "learning_rate": 0.00010928368491889626,
+      "loss": 0.9269,
+      "step": 15455
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2695715129375458,
+      "learning_rate": 0.00010918465399479712,
+      "loss": 0.8234,
+      "step": 15460
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2826976478099823,
+      "learning_rate": 0.00010908564227975518,
+      "loss": 0.8455,
+      "step": 15465
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2363000065088272,
+      "learning_rate": 0.00010898664982036877,
+      "loss": 0.8312,
+      "step": 15470
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.29471728205680847,
+      "learning_rate": 0.000108887676663227,
+      "loss": 0.8394,
+      "step": 15475
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.24590782821178436,
+      "learning_rate": 0.00010878872285490984,
+      "loss": 0.8675,
+      "step": 15480
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.26483914256095886,
+      "learning_rate": 0.00010868978844198827,
+      "loss": 0.7959,
+      "step": 15485
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.25184696912765503,
+      "learning_rate": 0.00010859087347102416,
+      "loss": 0.8902,
+      "step": 15490
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.23830364644527435,
+      "learning_rate": 0.00010849197798857015,
+      "loss": 0.9064,
+      "step": 15495
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.24322542548179626,
+      "learning_rate": 0.00010839310204116975,
+      "loss": 0.854,
+      "step": 15500
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.24347706139087677,
+      "learning_rate": 0.0001082942456753572,
+      "loss": 0.8495,
+      "step": 15505
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.24465703964233398,
+      "learning_rate": 0.0001081954089376577,
+      "loss": 0.8317,
+      "step": 15510
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.24925029277801514,
+      "learning_rate": 0.00010809659187458702,
+      "loss": 0.8258,
+      "step": 15515
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.27092206478118896,
+      "learning_rate": 0.00010799779453265178,
+      "loss": 0.8398,
+      "step": 15520
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2629016041755676,
+      "learning_rate": 0.00010789901695834921,
+      "loss": 0.7819,
+      "step": 15525
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2438107430934906,
+      "learning_rate": 0.00010780025919816748,
+      "loss": 0.8834,
+      "step": 15530
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2341863512992859,
+      "learning_rate": 0.00010770152129858515,
+      "loss": 0.8463,
+      "step": 15535
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2474951595067978,
+      "learning_rate": 0.00010760280330607161,
+      "loss": 0.8869,
+      "step": 15540
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2520921528339386,
+      "learning_rate": 0.00010750410526708675,
+      "loss": 0.8906,
+      "step": 15545
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2502596974372864,
+      "learning_rate": 0.00010740542722808123,
+      "loss": 0.774,
+      "step": 15550
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.2887754440307617,
+      "learning_rate": 0.0001073067692354962,
+      "loss": 0.7609,
+      "step": 15555
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.26220768690109253,
+      "learning_rate": 0.00010720813133576336,
+      "loss": 0.855,
+      "step": 15560
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.264213889837265,
+      "learning_rate": 0.00010710951357530489,
+      "loss": 0.7649,
+      "step": 15565
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.26668137311935425,
+      "learning_rate": 0.00010701091600053379,
+      "loss": 0.8864,
+      "step": 15570
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2364049255847931,
+      "learning_rate": 0.00010691233865785321,
+      "loss": 0.8681,
+      "step": 15575
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2663012444972992,
+      "learning_rate": 0.00010681378159365696,
+      "loss": 0.9484,
+      "step": 15580
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.25167354941368103,
+      "learning_rate": 0.00010671524485432926,
+      "loss": 0.8411,
+      "step": 15585
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2526310384273529,
+      "learning_rate": 0.00010661672848624477,
+      "loss": 0.8295,
+      "step": 15590
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.25531667470932007,
+      "learning_rate": 0.0001065182325357686,
+      "loss": 0.8014,
+      "step": 15595
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.24380475282669067,
+      "learning_rate": 0.00010641975704925615,
+      "loss": 0.9029,
+      "step": 15600
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.25942888855934143,
+      "learning_rate": 0.00010632130207305324,
+      "loss": 0.8727,
+      "step": 15605
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.26547759771347046,
+      "learning_rate": 0.00010622286765349618,
+      "loss": 0.7635,
+      "step": 15610
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.27288028597831726,
+      "learning_rate": 0.00010612445383691137,
+      "loss": 0.7731,
+      "step": 15615
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2501557171344757,
+      "learning_rate": 0.00010602606066961564,
+      "loss": 0.8859,
+      "step": 15620
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.25001439452171326,
+      "learning_rate": 0.00010592768819791608,
+      "loss": 0.8516,
+      "step": 15625
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2799443006515503,
+      "learning_rate": 0.00010582933646811008,
+      "loss": 0.8835,
+      "step": 15630
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2613951861858368,
+      "learning_rate": 0.00010573100552648517,
+      "loss": 0.9239,
+      "step": 15635
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.27016937732696533,
+      "learning_rate": 0.00010563269541931922,
+      "loss": 0.8181,
+      "step": 15640
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.22996383905410767,
+      "learning_rate": 0.00010553440619288014,
+      "loss": 0.8646,
+      "step": 15645
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.29044246673583984,
+      "learning_rate": 0.00010543613789342621,
+      "loss": 0.8249,
+      "step": 15650
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2618498206138611,
+      "learning_rate": 0.00010533789056720571,
+      "loss": 0.7699,
+      "step": 15655
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.25093433260917664,
+      "learning_rate": 0.00010523966426045709,
+      "loss": 0.7474,
+      "step": 15660
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2828831076622009,
+      "learning_rate": 0.00010514145901940887,
+      "loss": 0.7825,
+      "step": 15665
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2531161308288574,
+      "learning_rate": 0.0001050432748902798,
+      "loss": 0.8004,
+      "step": 15670
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.23352687060832977,
+      "learning_rate": 0.0001049451119192785,
+      "loss": 0.7407,
+      "step": 15675
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.22876708209514618,
+      "learning_rate": 0.00010484697015260379,
+      "loss": 0.7868,
+      "step": 15680
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.27175286412239075,
+      "learning_rate": 0.00010474884963644434,
+      "loss": 0.7846,
+      "step": 15685
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2837650775909424,
+      "learning_rate": 0.00010465075041697908,
+      "loss": 0.7969,
+      "step": 15690
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2851541340351105,
+      "learning_rate": 0.00010455267254037663,
+      "loss": 0.7862,
+      "step": 15695
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.2439732849597931,
+      "learning_rate": 0.00010445461605279579,
+      "loss": 0.8859,
+      "step": 15700
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.24269287288188934,
+      "learning_rate": 0.00010435658100038505,
+      "loss": 0.7997,
+      "step": 15705
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.24573330581188202,
+      "learning_rate": 0.00010425856742928313,
+      "loss": 1.0228,
+      "step": 15710
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.25582510232925415,
+      "learning_rate": 0.00010416057538561842,
+      "loss": 0.8363,
+      "step": 15715
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.21897505223751068,
+      "learning_rate": 0.00010406260491550918,
+      "loss": 0.8839,
+      "step": 15720
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.23649723827838898,
+      "learning_rate": 0.0001039646560650636,
+      "loss": 0.8441,
+      "step": 15725
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2502078413963318,
+      "learning_rate": 0.00010386672888037969,
+      "loss": 0.8262,
+      "step": 15730
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2571605145931244,
+      "learning_rate": 0.00010376882340754519,
+      "loss": 0.8712,
+      "step": 15735
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.26057517528533936,
+      "learning_rate": 0.0001036709396926377,
+      "loss": 0.8395,
+      "step": 15740
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2550731301307678,
+      "learning_rate": 0.00010357307778172445,
+      "loss": 0.8821,
+      "step": 15745
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2894580066204071,
+      "learning_rate": 0.00010347523772086268,
+      "loss": 0.8224,
+      "step": 15750
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.24666161835193634,
+      "learning_rate": 0.00010337741955609907,
+      "loss": 0.7983,
+      "step": 15755
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.26421642303466797,
+      "learning_rate": 0.00010327962333347008,
+      "loss": 0.7837,
+      "step": 15760
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.27278557419776917,
+      "learning_rate": 0.00010318184909900188,
+      "loss": 0.757,
+      "step": 15765
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.25606653094291687,
+      "learning_rate": 0.00010308409689871029,
+      "loss": 0.9554,
+      "step": 15770
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2995527982711792,
+      "learning_rate": 0.00010298636677860074,
+      "loss": 0.8754,
+      "step": 15775
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2641645669937134,
+      "learning_rate": 0.00010288865878466825,
+      "loss": 0.841,
+      "step": 15780
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2527436912059784,
+      "learning_rate": 0.00010279097296289741,
+      "loss": 0.7688,
+      "step": 15785
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.25887179374694824,
+      "learning_rate": 0.0001026933093592625,
+      "loss": 0.7526,
+      "step": 15790
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2504650354385376,
+      "learning_rate": 0.00010259566801972721,
+      "loss": 0.8904,
+      "step": 15795
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.235322505235672,
+      "learning_rate": 0.00010249804899024482,
+      "loss": 0.7883,
+      "step": 15800
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.250555157661438,
+      "learning_rate": 0.00010240045231675802,
+      "loss": 0.8137,
+      "step": 15805
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2679235637187958,
+      "learning_rate": 0.00010230287804519914,
+      "loss": 0.8408,
+      "step": 15810
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2516213357448578,
+      "learning_rate": 0.00010220532622148982,
+      "loss": 0.8198,
+      "step": 15815
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.26022079586982727,
+      "learning_rate": 0.00010210779689154118,
+      "loss": 0.8253,
+      "step": 15820
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.3123502731323242,
+      "learning_rate": 0.0001020102901012537,
+      "loss": 0.8452,
+      "step": 15825
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.24394842982292175,
+      "learning_rate": 0.00010191280589651746,
+      "loss": 0.8889,
+      "step": 15830
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.24881669878959656,
+      "learning_rate": 0.00010181534432321171,
+      "loss": 0.8135,
+      "step": 15835
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2647436261177063,
+      "learning_rate": 0.00010171790542720504,
+      "loss": 0.8659,
+      "step": 15840
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2636089324951172,
+      "learning_rate": 0.00010162048925435549,
+      "loss": 0.891,
+      "step": 15845
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2648860216140747,
+      "learning_rate": 0.00010152309585051035,
+      "loss": 0.7962,
+      "step": 15850
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2279883474111557,
+      "learning_rate": 0.00010142572526150616,
+      "loss": 0.9005,
+      "step": 15855
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.297305703163147,
+      "learning_rate": 0.0001013283775331687,
+      "loss": 0.7451,
+      "step": 15860
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2450573891401291,
+      "learning_rate": 0.00010123105271131319,
+      "loss": 0.8414,
+      "step": 15865
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2383514642715454,
+      "learning_rate": 0.00010113375084174382,
+      "loss": 0.8569,
+      "step": 15870
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.23251613974571228,
+      "learning_rate": 0.00010103647197025414,
+      "loss": 0.8139,
+      "step": 15875
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2642490863800049,
+      "learning_rate": 0.0001009392161426267,
+      "loss": 0.8877,
+      "step": 15880
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.25998982787132263,
+      "learning_rate": 0.00010084198340463345,
+      "loss": 0.8697,
+      "step": 15885
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.2662920355796814,
+      "learning_rate": 0.00010074477380203529,
+      "loss": 0.7452,
+      "step": 15890
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.24700447916984558,
+      "learning_rate": 0.00010064758738058231,
+      "loss": 0.856,
+      "step": 15895
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.26831844449043274,
+      "learning_rate": 0.0001005504241860136,
+      "loss": 0.7415,
+      "step": 15900
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.26766031980514526,
+      "learning_rate": 0.00010045328426405749,
+      "loss": 0.7633,
+      "step": 15905
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.24661500751972198,
+      "learning_rate": 0.00010035616766043119,
+      "loss": 0.8742,
+      "step": 15910
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2439432591199875,
+      "learning_rate": 0.00010025907442084102,
+      "loss": 0.7972,
+      "step": 15915
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.236423522233963,
+      "learning_rate": 0.0001001620045909822,
+      "loss": 0.7564,
+      "step": 15920
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.26059049367904663,
+      "learning_rate": 0.00010006495821653914,
+      "loss": 0.824,
+      "step": 15925
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2574262022972107,
+      "learning_rate": 9.996793534318505e-05,
+      "loss": 0.8156,
+      "step": 15930
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2577226758003235,
+      "learning_rate": 9.987093601658209e-05,
+      "loss": 0.8548,
+      "step": 15935
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2414378523826599,
+      "learning_rate": 9.977396028238136e-05,
+      "loss": 0.8627,
+      "step": 15940
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2435644119977951,
+      "learning_rate": 9.969639671222088e-05,
+      "loss": 0.8817,
+      "step": 15945
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.25886785984039307,
+      "learning_rate": 9.959946355934948e-05,
+      "loss": 0.7624,
+      "step": 15950
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.22503872215747833,
+      "learning_rate": 9.950255412664435e-05,
+      "loss": 0.9167,
+      "step": 15955
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.24574321508407593,
+      "learning_rate": 9.940566845971425e-05,
+      "loss": 0.7814,
+      "step": 15960
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.23402521014213562,
+      "learning_rate": 9.93088066041567e-05,
+      "loss": 0.8389,
+      "step": 15965
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.27619630098342896,
+      "learning_rate": 9.921196860555813e-05,
+      "loss": 0.8364,
+      "step": 15970
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2658560872077942,
+      "learning_rate": 9.91151545094938e-05,
+      "loss": 0.8762,
+      "step": 15975
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2881810963153839,
+      "learning_rate": 9.90183643615276e-05,
+      "loss": 0.8234,
+      "step": 15980
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2377784252166748,
+      "learning_rate": 9.892159820721216e-05,
+      "loss": 0.8424,
+      "step": 15985
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.22696682810783386,
+      "learning_rate": 9.882485609208885e-05,
+      "loss": 0.7891,
+      "step": 15990
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2700740098953247,
+      "learning_rate": 9.872813806168778e-05,
+      "loss": 0.8931,
+      "step": 15995
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.23319804668426514,
+      "learning_rate": 9.86314441615276e-05,
+      "loss": 0.7404,
+      "step": 16000
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.23562254011631012,
+      "learning_rate": 9.853477443711572e-05,
+      "loss": 0.9749,
+      "step": 16005
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.23507817089557648,
+      "learning_rate": 9.843812893394801e-05,
+      "loss": 0.8976,
+      "step": 16010
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.25539952516555786,
+      "learning_rate": 9.834150769750921e-05,
+      "loss": 0.8241,
+      "step": 16015
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.24266016483306885,
+      "learning_rate": 9.824491077327242e-05,
+      "loss": 0.8013,
+      "step": 16020
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.29498210549354553,
+      "learning_rate": 9.814833820669934e-05,
+      "loss": 0.8838,
+      "step": 16025
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.26613032817840576,
+      "learning_rate": 9.805179004324022e-05,
+      "loss": 0.8375,
+      "step": 16030
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2595424950122833,
+      "learning_rate": 9.795526632833388e-05,
+      "loss": 0.8382,
+      "step": 16035
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.24384133517742157,
+      "learning_rate": 9.785876710740755e-05,
+      "loss": 0.9724,
+      "step": 16040
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2365197241306305,
+      "learning_rate": 9.776229242587701e-05,
+      "loss": 0.8506,
+      "step": 16045
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2736034095287323,
+      "learning_rate": 9.766584232914633e-05,
+      "loss": 0.818,
+      "step": 16050
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2698809504508972,
+      "learning_rate": 9.756941686260826e-05,
+      "loss": 0.7629,
+      "step": 16055
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.24631649255752563,
+      "learning_rate": 9.747301607164378e-05,
+      "loss": 0.8116,
+      "step": 16060
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.23516754806041718,
+      "learning_rate": 9.737664000162233e-05,
+      "loss": 0.8521,
+      "step": 16065
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.22055235505104065,
+      "learning_rate": 9.728028869790162e-05,
+      "loss": 0.8474,
+      "step": 16070
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.27016615867614746,
+      "learning_rate": 9.718396220582785e-05,
+      "loss": 0.8257,
+      "step": 16075
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.26734715700149536,
+      "learning_rate": 9.708766057073543e-05,
+      "loss": 0.7563,
+      "step": 16080
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2916874587535858,
+      "learning_rate": 9.69913838379471e-05,
+      "loss": 0.7673,
+      "step": 16085
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.23642754554748535,
+      "learning_rate": 9.689513205277387e-05,
+      "loss": 0.8493,
+      "step": 16090
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2527269721031189,
+      "learning_rate": 9.679890526051507e-05,
+      "loss": 0.8293,
+      "step": 16095
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2781146168708801,
+      "learning_rate": 9.670270350645823e-05,
+      "loss": 0.728,
+      "step": 16100
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.24582897126674652,
+      "learning_rate": 9.660652683587907e-05,
+      "loss": 0.8191,
+      "step": 16105
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.28193217515945435,
+      "learning_rate": 9.65103752940415e-05,
+      "loss": 0.8848,
+      "step": 16110
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2869209349155426,
+      "learning_rate": 9.641424892619766e-05,
+      "loss": 0.7661,
+      "step": 16115
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.25964149832725525,
+      "learning_rate": 9.631814777758782e-05,
+      "loss": 0.8795,
+      "step": 16120
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2651737332344055,
+      "learning_rate": 9.622207189344035e-05,
+      "loss": 0.8498,
+      "step": 16125
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.24850745499134064,
+      "learning_rate": 9.612602131897169e-05,
+      "loss": 0.8648,
+      "step": 16130
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2507181167602539,
+      "learning_rate": 9.602999609938658e-05,
+      "loss": 0.7739,
+      "step": 16135
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.24325953423976898,
+      "learning_rate": 9.593399627987757e-05,
+      "loss": 0.8611,
+      "step": 16140
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2649693489074707,
+      "learning_rate": 9.58380219056254e-05,
+      "loss": 0.7544,
+      "step": 16145
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.25593236088752747,
+      "learning_rate": 9.574207302179874e-05,
+      "loss": 0.8856,
+      "step": 16150
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.22325646877288818,
+      "learning_rate": 9.56461496735544e-05,
+      "loss": 0.7696,
+      "step": 16155
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2551780343055725,
+      "learning_rate": 9.555025190603709e-05,
+      "loss": 0.8319,
+      "step": 16160
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2650226354598999,
+      "learning_rate": 9.54543797643794e-05,
+      "loss": 0.873,
+      "step": 16165
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.25608155131340027,
+      "learning_rate": 9.5358533293702e-05,
+      "loss": 0.8617,
+      "step": 16170
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2500768303871155,
+      "learning_rate": 9.526271253911346e-05,
+      "loss": 0.8167,
+      "step": 16175
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.24089059233665466,
+      "learning_rate": 9.516691754571015e-05,
+      "loss": 0.8327,
+      "step": 16180
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.24258366227149963,
+      "learning_rate": 9.50711483585764e-05,
+      "loss": 0.8146,
+      "step": 16185
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2549149990081787,
+      "learning_rate": 9.49754050227843e-05,
+      "loss": 0.8334,
+      "step": 16190
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.23933325707912445,
+      "learning_rate": 9.487968758339395e-05,
+      "loss": 0.8018,
+      "step": 16195
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.24095794558525085,
+      "learning_rate": 9.478399608545314e-05,
+      "loss": 0.8354,
+      "step": 16200
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2553461492061615,
+      "learning_rate": 9.468833057399741e-05,
+      "loss": 0.8267,
+      "step": 16205
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.22408875823020935,
+      "learning_rate": 9.459269109405017e-05,
+      "loss": 0.9216,
+      "step": 16210
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2684914469718933,
+      "learning_rate": 9.44970776906225e-05,
+      "loss": 0.7983,
+      "step": 16215
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.26553064584732056,
+      "learning_rate": 9.440149040871329e-05,
+      "loss": 0.8494,
+      "step": 16220
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.2991175055503845,
+      "learning_rate": 9.430592929330907e-05,
+      "loss": 0.9366,
+      "step": 16225
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2854938209056854,
+      "learning_rate": 9.421039438938399e-05,
+      "loss": 0.7927,
+      "step": 16230
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.23396198451519012,
+      "learning_rate": 9.41148857419001e-05,
+      "loss": 0.8203,
+      "step": 16235
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2335088849067688,
+      "learning_rate": 9.401940339580687e-05,
+      "loss": 0.8674,
+      "step": 16240
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.27896398305892944,
+      "learning_rate": 9.392394739604141e-05,
+      "loss": 0.8955,
+      "step": 16245
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.27797919511795044,
+      "learning_rate": 9.382851778752858e-05,
+      "loss": 0.8088,
+      "step": 16250
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.22393245995044708,
+      "learning_rate": 9.373311461518066e-05,
+      "loss": 0.7955,
+      "step": 16255
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2545322775840759,
+      "learning_rate": 9.363773792389759e-05,
+      "loss": 0.8558,
+      "step": 16260
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.23659881949424744,
+      "learning_rate": 9.354238775856672e-05,
+      "loss": 0.9073,
+      "step": 16265
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2683257460594177,
+      "learning_rate": 9.344706416406312e-05,
+      "loss": 0.7528,
+      "step": 16270
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2514216899871826,
+      "learning_rate": 9.335176718524919e-05,
+      "loss": 0.8392,
+      "step": 16275
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.22830678522586823,
+      "learning_rate": 9.325649686697485e-05,
+      "loss": 0.7843,
+      "step": 16280
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2406618446111679,
+      "learning_rate": 9.316125325407746e-05,
+      "loss": 0.8878,
+      "step": 16285
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.24721916019916534,
+      "learning_rate": 9.306603639138187e-05,
+      "loss": 0.8405,
+      "step": 16290
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.2820783853530884,
+      "learning_rate": 9.297084632370026e-05,
+      "loss": 0.8073,
+      "step": 16295
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.24439501762390137,
+      "learning_rate": 9.287568309583227e-05,
+      "loss": 0.9226,
+      "step": 16300
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.25293993949890137,
+      "learning_rate": 9.278054675256479e-05,
+      "loss": 0.7773,
+      "step": 16305
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.26933521032333374,
+      "learning_rate": 9.268543733867225e-05,
+      "loss": 0.8361,
+      "step": 16310
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.27007099986076355,
+      "learning_rate": 9.259035489891628e-05,
+      "loss": 0.7993,
+      "step": 16315
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.25989168882369995,
+      "learning_rate": 9.24952994780458e-05,
+      "loss": 0.8354,
+      "step": 16320
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2629483640193939,
+      "learning_rate": 9.240027112079702e-05,
+      "loss": 0.7799,
+      "step": 16325
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.24853236973285675,
+      "learning_rate": 9.230526987189351e-05,
+      "loss": 0.7944,
+      "step": 16330
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2612740099430084,
+      "learning_rate": 9.2210295776046e-05,
+      "loss": 0.7517,
+      "step": 16335
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2724323570728302,
+      "learning_rate": 9.211534887795241e-05,
+      "loss": 0.8058,
+      "step": 16340
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2580575942993164,
+      "learning_rate": 9.202042922229788e-05,
+      "loss": 0.8327,
+      "step": 16345
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.29225781559944153,
+      "learning_rate": 9.192553685375488e-05,
+      "loss": 0.847,
+      "step": 16350
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.26023924350738525,
+      "learning_rate": 9.18306718169828e-05,
+      "loss": 0.8871,
+      "step": 16355
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.26399603486061096,
+      "learning_rate": 9.173583415662835e-05,
+      "loss": 0.8788,
+      "step": 16360
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2926609516143799,
+      "learning_rate": 9.164102391732514e-05,
+      "loss": 0.7472,
+      "step": 16365
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2964513897895813,
+      "learning_rate": 9.156519549902523e-05,
+      "loss": 0.88,
+      "step": 16370
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.28954005241394043,
+      "learning_rate": 9.147043473005033e-05,
+      "loss": 0.7696,
+      "step": 16375
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.2357548326253891,
+      "learning_rate": 9.137570150703257e-05,
+      "loss": 0.7794,
+      "step": 16380
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.22681371867656708,
+      "learning_rate": 9.128099587455652e-05,
+      "loss": 0.8403,
+      "step": 16385
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.26432034373283386,
+      "learning_rate": 9.118631787719381e-05,
+      "loss": 0.8527,
+      "step": 16390
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2536108195781708,
+      "learning_rate": 9.109166755950302e-05,
+      "loss": 0.8427,
+      "step": 16395
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.22776862978935242,
+      "learning_rate": 9.101596726464934e-05,
+      "loss": 0.8992,
+      "step": 16400
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.24553075432777405,
+      "learning_rate": 9.092136688261414e-05,
+      "loss": 0.8519,
+      "step": 16405
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.25476738810539246,
+      "learning_rate": 9.082679430494567e-05,
+      "loss": 0.7777,
+      "step": 16410
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2561832368373871,
+      "learning_rate": 9.07322495761529e-05,
+      "loss": 0.8118,
+      "step": 16415
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.26670464873313904,
+      "learning_rate": 9.06377327407317e-05,
+      "loss": 0.873,
+      "step": 16420
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.261623352766037,
+      "learning_rate": 9.05432438431648e-05,
+      "loss": 0.7815,
+      "step": 16425
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.283966600894928,
+      "learning_rate": 9.044878292792187e-05,
+      "loss": 0.8202,
+      "step": 16430
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.23486657440662384,
+      "learning_rate": 9.035435003945933e-05,
+      "loss": 0.802,
+      "step": 16435
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.233073890209198,
+      "learning_rate": 9.025994522222043e-05,
+      "loss": 0.8367,
+      "step": 16440
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.24187245965003967,
+      "learning_rate": 9.016556852063515e-05,
+      "loss": 0.894,
+      "step": 16445
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2719327211380005,
+      "learning_rate": 9.007121997912044e-05,
+      "loss": 0.9086,
+      "step": 16450
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2707170844078064,
+      "learning_rate": 8.997689964207978e-05,
+      "loss": 0.7866,
+      "step": 16455
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.2778245508670807,
+      "learning_rate": 8.988260755390346e-05,
+      "loss": 0.8574,
+      "step": 16460
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.24011406302452087,
+      "learning_rate": 8.978834375896841e-05,
+      "loss": 0.8875,
+      "step": 16465
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.26258188486099243,
+      "learning_rate": 8.96941083016384e-05,
+      "loss": 0.8013,
+      "step": 16470
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.25577783584594727,
+      "learning_rate": 8.95999012262637e-05,
+      "loss": 0.7737,
+      "step": 16475
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.2674656808376312,
+      "learning_rate": 8.950572257718132e-05,
+      "loss": 0.8443,
+      "step": 16480
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.2525770962238312,
+      "learning_rate": 8.941157239871479e-05,
+      "loss": 0.8448,
+      "step": 16485
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.23468473553657532,
+      "learning_rate": 8.931745073517443e-05,
+      "loss": 0.883,
+      "step": 16490
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.23663926124572754,
+      "learning_rate": 8.922335763085696e-05,
+      "loss": 0.8718,
+      "step": 16495
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.22883319854736328,
+      "learning_rate": 8.912929313004572e-05,
+      "loss": 0.9618,
+      "step": 16500
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.24675635993480682,
+      "learning_rate": 8.903525727701054e-05,
+      "loss": 0.8928,
+      "step": 16505
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.2842852771282196,
+      "learning_rate": 8.89412501160079e-05,
+      "loss": 0.8255,
+      "step": 16510
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.26586049795150757,
+      "learning_rate": 8.884727169128066e-05,
+      "loss": 0.9482,
+      "step": 16515
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.3201524615287781,
+      "learning_rate": 8.875332204705818e-05,
+      "loss": 0.8118,
+      "step": 16520
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.2589406967163086,
+      "learning_rate": 8.865940122755623e-05,
+      "loss": 0.8802,
+      "step": 16525
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.28572580218315125,
+      "learning_rate": 8.85655092769772e-05,
+      "loss": 0.9211,
+      "step": 16530
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.24606913328170776,
+      "learning_rate": 8.847164623950965e-05,
+      "loss": 0.8589,
+      "step": 16535
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.2393767237663269,
+      "learning_rate": 8.837781215932862e-05,
+      "loss": 0.8135,
+      "step": 16540
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.26532062888145447,
+      "learning_rate": 8.828400708059567e-05,
+      "loss": 0.7389,
+      "step": 16545
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.25288575887680054,
+      "learning_rate": 8.81902310474585e-05,
+      "loss": 0.8357,
+      "step": 16550
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.23215311765670776,
+      "learning_rate": 8.809648410405123e-05,
+      "loss": 0.7262,
+      "step": 16555
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2512628734111786,
+      "learning_rate": 8.800276629449426e-05,
+      "loss": 0.7932,
+      "step": 16560
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.23022831976413727,
+      "learning_rate": 8.790907766289437e-05,
+      "loss": 0.7226,
+      "step": 16565
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.24034039676189423,
+      "learning_rate": 8.781541825334453e-05,
+      "loss": 0.8625,
+      "step": 16570
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.23891344666481018,
+      "learning_rate": 8.772178810992392e-05,
+      "loss": 0.8287,
+      "step": 16575
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.23461419343948364,
+      "learning_rate": 8.762818727669797e-05,
+      "loss": 0.7968,
+      "step": 16580
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.24658669531345367,
+      "learning_rate": 8.753461579771846e-05,
+      "loss": 0.787,
+      "step": 16585
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.23091553151607513,
+      "learning_rate": 8.744107371702315e-05,
+      "loss": 0.8718,
+      "step": 16590
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.26409897208213806,
+      "learning_rate": 8.734756107863608e-05,
+      "loss": 0.8378,
+      "step": 16595
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.27900955080986023,
+      "learning_rate": 8.725407792656731e-05,
+      "loss": 0.7441,
+      "step": 16600
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2287808060646057,
+      "learning_rate": 8.716062430481328e-05,
+      "loss": 0.8506,
+      "step": 16605
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.26879292726516724,
+      "learning_rate": 8.706720025735627e-05,
+      "loss": 0.8325,
+      "step": 16610
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.28211572766304016,
+      "learning_rate": 8.697380582816476e-05,
+      "loss": 0.8599,
+      "step": 16615
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2926715016365051,
+      "learning_rate": 8.688044106119325e-05,
+      "loss": 0.7717,
+      "step": 16620
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.27610814571380615,
+      "learning_rate": 8.678710600038233e-05,
+      "loss": 0.8317,
+      "step": 16625
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2478606104850769,
+      "learning_rate": 8.669380068965856e-05,
+      "loss": 0.8358,
+      "step": 16630
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2299978882074356,
+      "learning_rate": 8.660052517293448e-05,
+      "loss": 0.8603,
+      "step": 16635
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.27341413497924805,
+      "learning_rate": 8.650727949410867e-05,
+      "loss": 0.9561,
+      "step": 16640
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.25213295221328735,
+      "learning_rate": 8.641406369706572e-05,
+      "loss": 0.8045,
+      "step": 16645
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.2605089843273163,
+      "learning_rate": 8.6320877825676e-05,
+      "loss": 0.8112,
+      "step": 16650
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.27594542503356934,
+      "learning_rate": 8.622772192379588e-05,
+      "loss": 0.8145,
+      "step": 16655
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.2801823318004608,
+      "learning_rate": 8.61345960352676e-05,
+      "loss": 0.6998,
+      "step": 16660
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.2606346011161804,
+      "learning_rate": 8.604150020391937e-05,
+      "loss": 0.7655,
+      "step": 16665
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.23958322405815125,
+      "learning_rate": 8.594843447356517e-05,
+      "loss": 0.8022,
+      "step": 16670
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.26263347268104553,
+      "learning_rate": 8.585539888800475e-05,
+      "loss": 0.7376,
+      "step": 16675
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.23959647119045258,
+      "learning_rate": 8.576239349102375e-05,
+      "loss": 0.763,
+      "step": 16680
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.25928083062171936,
+      "learning_rate": 8.56694183263937e-05,
+      "loss": 0.7902,
+      "step": 16685
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.24009615182876587,
+      "learning_rate": 8.557647343787175e-05,
+      "loss": 0.8473,
+      "step": 16690
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.2364853471517563,
+      "learning_rate": 8.548355886920084e-05,
+      "loss": 0.7173,
+      "step": 16695
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.2518242299556732,
+      "learning_rate": 8.539067466410962e-05,
+      "loss": 0.8219,
+      "step": 16700
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.28275394439697266,
+      "learning_rate": 8.529782086631254e-05,
+      "loss": 0.9054,
+      "step": 16705
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.2858952283859253,
+      "learning_rate": 8.520499751950965e-05,
+      "loss": 0.8016,
+      "step": 16710
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.2402840256690979,
+      "learning_rate": 8.51122046673867e-05,
+      "loss": 0.8295,
+      "step": 16715
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.2762582004070282,
+      "learning_rate": 8.501944235361502e-05,
+      "loss": 0.733,
+      "step": 16720
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.2743135094642639,
+      "learning_rate": 8.492671062185177e-05,
+      "loss": 0.7954,
+      "step": 16725
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.24607881903648376,
+      "learning_rate": 8.483400951573954e-05,
+      "loss": 0.7754,
+      "step": 16730
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.27087563276290894,
+      "learning_rate": 8.474133907890651e-05,
+      "loss": 0.7463,
+      "step": 16735
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.24239374697208405,
+      "learning_rate": 8.464869935496641e-05,
+      "loss": 0.7516,
+      "step": 16740
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.27605438232421875,
+      "learning_rate": 8.455609038751871e-05,
+      "loss": 0.7053,
+      "step": 16745
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.29548242688179016,
+      "learning_rate": 8.446351222014822e-05,
+      "loss": 0.7657,
+      "step": 16750
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.26143786311149597,
+      "learning_rate": 8.437096489642526e-05,
+      "loss": 0.7796,
+      "step": 16755
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.26324915885925293,
+      "learning_rate": 8.42784484599057e-05,
+      "loss": 0.7717,
+      "step": 16760
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.2451925128698349,
+      "learning_rate": 8.418596295413083e-05,
+      "loss": 0.8181,
+      "step": 16765
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.26199206709861755,
+      "learning_rate": 8.409350842262741e-05,
+      "loss": 0.8205,
+      "step": 16770
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.26254212856292725,
+      "learning_rate": 8.400108490890763e-05,
+      "loss": 0.7734,
+      "step": 16775
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.24498294293880463,
+      "learning_rate": 8.390869245646897e-05,
+      "loss": 0.7043,
+      "step": 16780
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.30816173553466797,
+      "learning_rate": 8.381633110879454e-05,
+      "loss": 0.8023,
+      "step": 16785
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.24860282242298126,
+      "learning_rate": 8.372400090935256e-05,
+      "loss": 0.7682,
+      "step": 16790
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.28883734345436096,
+      "learning_rate": 8.363170190159673e-05,
+      "loss": 0.8831,
+      "step": 16795
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.28336241841316223,
+      "learning_rate": 8.353943412896596e-05,
+      "loss": 0.7443,
+      "step": 16800
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.27060964703559875,
+      "learning_rate": 8.34471976348846e-05,
+      "loss": 0.8307,
+      "step": 16805
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.27045443654060364,
+      "learning_rate": 8.33549924627622e-05,
+      "loss": 0.7417,
+      "step": 16810
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.24936868250370026,
+      "learning_rate": 8.326281865599356e-05,
+      "loss": 0.801,
+      "step": 16815
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.28478366136550903,
+      "learning_rate": 8.317067625795867e-05,
+      "loss": 0.7997,
+      "step": 16820
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.28683024644851685,
+      "learning_rate": 8.307856531202295e-05,
+      "loss": 0.8765,
+      "step": 16825
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.2502270042896271,
+      "learning_rate": 8.298648586153676e-05,
+      "loss": 0.8878,
+      "step": 16830
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.25373736023902893,
+      "learning_rate": 8.289443794983578e-05,
+      "loss": 0.7722,
+      "step": 16835
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.2576674818992615,
+      "learning_rate": 8.280242162024079e-05,
+      "loss": 0.8224,
+      "step": 16840
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.26497557759284973,
+      "learning_rate": 8.271043691605778e-05,
+      "loss": 0.8216,
+      "step": 16845
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.2641935646533966,
+      "learning_rate": 8.261848388057775e-05,
+      "loss": 0.755,
+      "step": 16850
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.28792330622673035,
+      "learning_rate": 8.252656255707689e-05,
+      "loss": 0.7801,
+      "step": 16855
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.2565497159957886,
+      "learning_rate": 8.243467298881636e-05,
+      "loss": 0.7938,
+      "step": 16860
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.2724232077598572,
+      "learning_rate": 8.234281521904253e-05,
+      "loss": 0.8549,
+      "step": 16865
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.24661044776439667,
+      "learning_rate": 8.225098929098673e-05,
+      "loss": 0.812,
+      "step": 16870
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.23888921737670898,
+      "learning_rate": 8.215919524786521e-05,
+      "loss": 0.7724,
+      "step": 16875
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.24567218124866486,
+      "learning_rate": 8.206743313287925e-05,
+      "loss": 0.7095,
+      "step": 16880
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.2549126148223877,
+      "learning_rate": 8.197570298921533e-05,
+      "loss": 0.7106,
+      "step": 16885
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.2611244022846222,
+      "learning_rate": 8.18840048600446e-05,
+      "loss": 0.8283,
+      "step": 16890
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.24455487728118896,
+      "learning_rate": 8.179233878852323e-05,
+      "loss": 0.7387,
+      "step": 16895
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.3053877651691437,
+      "learning_rate": 8.170070481779224e-05,
+      "loss": 0.8595,
+      "step": 16900
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.2966110408306122,
+      "learning_rate": 8.160910299097782e-05,
+      "loss": 0.7955,
+      "step": 16905
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.29260462522506714,
+      "learning_rate": 8.15175333511907e-05,
+      "loss": 0.8606,
+      "step": 16910
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.27227169275283813,
+      "learning_rate": 8.14259959415267e-05,
+      "loss": 0.6951,
+      "step": 16915
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.2581891417503357,
+      "learning_rate": 8.133449080506615e-05,
+      "loss": 0.8414,
+      "step": 16920
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.2824326157569885,
+      "learning_rate": 8.124301798487458e-05,
+      "loss": 0.7542,
+      "step": 16925
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.28472477197647095,
+      "learning_rate": 8.115157752400211e-05,
+      "loss": 0.7875,
+      "step": 16930
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.3563947081565857,
+      "learning_rate": 8.106016946548365e-05,
+      "loss": 0.8582,
+      "step": 16935
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.27918270230293274,
+      "learning_rate": 8.096879385233879e-05,
+      "loss": 0.8431,
+      "step": 16940
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.31938403844833374,
+      "learning_rate": 8.087745072757208e-05,
+      "loss": 0.8991,
+      "step": 16945
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.33040347695350647,
+      "learning_rate": 8.078614013417253e-05,
+      "loss": 0.7591,
+      "step": 16950
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.28828245401382446,
+      "learning_rate": 8.069486211511394e-05,
+      "loss": 0.7754,
+      "step": 16955
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.24763058125972748,
+      "learning_rate": 8.060361671335474e-05,
+      "loss": 0.7698,
+      "step": 16960
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.23437534272670746,
+      "learning_rate": 8.051240397183818e-05,
+      "loss": 0.7253,
+      "step": 16965
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.2842622995376587,
+      "learning_rate": 8.042122393349189e-05,
+      "loss": 0.773,
+      "step": 16970
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.26870718598365784,
+      "learning_rate": 8.033007664122827e-05,
+      "loss": 0.7793,
+      "step": 16975
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.26021209359169006,
+      "learning_rate": 8.023896213794425e-05,
+      "loss": 0.8217,
+      "step": 16980
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.26792365312576294,
+      "learning_rate": 8.014788046652135e-05,
+      "loss": 0.8153,
+      "step": 16985
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.2724539637565613,
+      "learning_rate": 8.00568316698256e-05,
+      "loss": 0.7681,
+      "step": 16990
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.27169302105903625,
+      "learning_rate": 7.996581579070762e-05,
+      "loss": 0.7707,
+      "step": 16995
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.25969168543815613,
+      "learning_rate": 7.987483287200243e-05,
+      "loss": 0.8198,
+      "step": 17000
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.279259592294693,
+      "learning_rate": 7.978388295652974e-05,
+      "loss": 0.7626,
+      "step": 17005
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.27438825368881226,
+      "learning_rate": 7.969296608709351e-05,
+      "loss": 0.7247,
+      "step": 17010
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.30393609404563904,
+      "learning_rate": 7.96020823064823e-05,
+      "loss": 0.7484,
+      "step": 17015
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.2659038007259369,
+      "learning_rate": 7.951123165746892e-05,
+      "loss": 0.9497,
+      "step": 17020
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.2514190673828125,
+      "learning_rate": 7.942041418281086e-05,
+      "loss": 0.729,
+      "step": 17025
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.2725908160209656,
+      "learning_rate": 7.932962992524974e-05,
+      "loss": 0.8552,
+      "step": 17030
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.28015610575675964,
+      "learning_rate": 7.923887892751165e-05,
+      "loss": 0.7491,
+      "step": 17035
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.2758082151412964,
+      "learning_rate": 7.914816123230703e-05,
+      "loss": 0.7509,
+      "step": 17040
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.26611337065696716,
+      "learning_rate": 7.905747688233069e-05,
+      "loss": 0.7658,
+      "step": 17045
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.24941769242286682,
+      "learning_rate": 7.896682592026164e-05,
+      "loss": 0.7546,
+      "step": 17050
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.28314536809921265,
+      "learning_rate": 7.887620838876326e-05,
+      "loss": 0.7747,
+      "step": 17055
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.29477396607398987,
+      "learning_rate": 7.878562433048316e-05,
+      "loss": 0.8356,
+      "step": 17060
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.278089314699173,
+      "learning_rate": 7.869507378805321e-05,
+      "loss": 0.8421,
+      "step": 17065
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.27571311593055725,
+      "learning_rate": 7.86045568040895e-05,
+      "loss": 0.8294,
+      "step": 17070
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.27711910009384155,
+      "learning_rate": 7.851407342119226e-05,
+      "loss": 0.8208,
+      "step": 17075
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.27659597992897034,
+      "learning_rate": 7.842362368194611e-05,
+      "loss": 0.7766,
+      "step": 17080
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.28369686007499695,
+      "learning_rate": 7.833320762891964e-05,
+      "loss": 0.8258,
+      "step": 17085
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.2623741626739502,
+      "learning_rate": 7.824282530466562e-05,
+      "loss": 0.836,
+      "step": 17090
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.2581416368484497,
+      "learning_rate": 7.815247675172097e-05,
+      "loss": 0.7441,
+      "step": 17095
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.2673351764678955,
+      "learning_rate": 7.806216201260677e-05,
+      "loss": 0.8585,
+      "step": 17100
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.26765942573547363,
+      "learning_rate": 7.797188112982811e-05,
+      "loss": 0.7769,
+      "step": 17105
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.28835397958755493,
+      "learning_rate": 7.788163414587417e-05,
+      "loss": 0.8654,
+      "step": 17110
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.263345330953598,
+      "learning_rate": 7.779142110321812e-05,
+      "loss": 0.9097,
+      "step": 17115
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.2905290126800537,
+      "learning_rate": 7.770124204431734e-05,
+      "loss": 0.8099,
+      "step": 17120
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.2563280463218689,
+      "learning_rate": 7.761109701161308e-05,
+      "loss": 0.8323,
+      "step": 17125
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.24418336153030396,
+      "learning_rate": 7.752098604753045e-05,
+      "loss": 0.7408,
+      "step": 17130
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.267090767621994,
+      "learning_rate": 7.743090919447869e-05,
+      "loss": 0.6895,
+      "step": 17135
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.2744917869567871,
+      "learning_rate": 7.734086649485109e-05,
+      "loss": 0.7986,
+      "step": 17140
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.2881864011287689,
+      "learning_rate": 7.725085799102464e-05,
+      "loss": 0.7426,
+      "step": 17145
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.32715755701065063,
+      "learning_rate": 7.716088372536035e-05,
+      "loss": 0.779,
+      "step": 17150
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.2890661060810089,
+      "learning_rate": 7.707094374020302e-05,
+      "loss": 0.8839,
+      "step": 17155
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.2906983196735382,
+      "learning_rate": 7.698103807788152e-05,
+      "loss": 0.9254,
+      "step": 17160
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.27394556999206543,
+      "learning_rate": 7.689116678070839e-05,
+      "loss": 0.7887,
+      "step": 17165
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.2804103195667267,
+      "learning_rate": 7.680132989098004e-05,
+      "loss": 0.7343,
+      "step": 17170
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.2799692451953888,
+      "learning_rate": 7.671152745097664e-05,
+      "loss": 0.7547,
+      "step": 17175
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.2532331943511963,
+      "learning_rate": 7.662175950296231e-05,
+      "loss": 0.8438,
+      "step": 17180
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.2761310338973999,
+      "learning_rate": 7.653202608918479e-05,
+      "loss": 0.7216,
+      "step": 17185
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.27380895614624023,
+      "learning_rate": 7.644232725187557e-05,
+      "loss": 0.7755,
+      "step": 17190
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.3025756776332855,
+      "learning_rate": 7.635266303324993e-05,
+      "loss": 0.7884,
+      "step": 17195
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.26098042726516724,
+      "learning_rate": 7.62630334755068e-05,
+      "loss": 0.7618,
+      "step": 17200
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.2765859365463257,
+      "learning_rate": 7.617343862082887e-05,
+      "loss": 0.8837,
+      "step": 17205
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.29430699348449707,
+      "learning_rate": 7.608387851138241e-05,
+      "loss": 0.8484,
+      "step": 17210
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.2927020788192749,
+      "learning_rate": 7.599435318931737e-05,
+      "loss": 0.7411,
+      "step": 17215
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.29646044969558716,
+      "learning_rate": 7.590486269676741e-05,
+      "loss": 0.7927,
+      "step": 17220
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.2691041827201843,
+      "learning_rate": 7.58154070758497e-05,
+      "loss": 0.7709,
+      "step": 17225
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.2560572922229767,
+      "learning_rate": 7.572598636866499e-05,
+      "loss": 0.8836,
+      "step": 17230
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.27019980549812317,
+      "learning_rate": 7.563660061729763e-05,
+      "loss": 0.83,
+      "step": 17235
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.2663953900337219,
+      "learning_rate": 7.554724986381558e-05,
+      "loss": 0.7005,
+      "step": 17240
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.2575012743473053,
+      "learning_rate": 7.545793415027026e-05,
+      "loss": 0.8471,
+      "step": 17245
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.3181961476802826,
+      "learning_rate": 7.53686535186966e-05,
+      "loss": 0.8156,
+      "step": 17250
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.2710973024368286,
+      "learning_rate": 7.527940801111296e-05,
+      "loss": 0.7603,
+      "step": 17255
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.23654502630233765,
+      "learning_rate": 7.519019766952135e-05,
+      "loss": 0.7897,
+      "step": 17260
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.2502569258213043,
+      "learning_rate": 7.51010225359071e-05,
+      "loss": 0.7789,
+      "step": 17265
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.2620236277580261,
+      "learning_rate": 7.501188265223893e-05,
+      "loss": 0.7543,
+      "step": 17270
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.27987682819366455,
+      "learning_rate": 7.492277806046908e-05,
+      "loss": 0.8192,
+      "step": 17275
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.23856423795223236,
+      "learning_rate": 7.483370880253311e-05,
+      "loss": 0.7938,
+      "step": 17280
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.26246753334999084,
+      "learning_rate": 7.474467492034998e-05,
+      "loss": 0.7727,
+      "step": 17285
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.2560681998729706,
+      "learning_rate": 7.465567645582199e-05,
+      "loss": 0.7481,
+      "step": 17290
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.27870872616767883,
+      "learning_rate": 7.45667134508347e-05,
+      "loss": 0.7659,
+      "step": 17295
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.2658903896808624,
+      "learning_rate": 7.447778594725717e-05,
+      "loss": 0.8235,
+      "step": 17300
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.3086778223514557,
+      "learning_rate": 7.438889398694161e-05,
+      "loss": 0.8337,
+      "step": 17305
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.2679433226585388,
+      "learning_rate": 7.430003761172349e-05,
+      "loss": 0.7323,
+      "step": 17310
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.2665143311023712,
+      "learning_rate": 7.421121686342152e-05,
+      "loss": 0.8554,
+      "step": 17315
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.24720510840415955,
+      "learning_rate": 7.412243178383784e-05,
+      "loss": 0.788,
+      "step": 17320
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.28797394037246704,
+      "learning_rate": 7.403368241475757e-05,
+      "loss": 0.7654,
+      "step": 17325
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.24396300315856934,
+      "learning_rate": 7.394496879794911e-05,
+      "loss": 0.8301,
+      "step": 17330
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.26526421308517456,
+      "learning_rate": 7.385629097516407e-05,
+      "loss": 0.9021,
+      "step": 17335
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.2767682671546936,
+      "learning_rate": 7.376764898813714e-05,
+      "loss": 0.7755,
+      "step": 17340
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.3317005932331085,
+      "learning_rate": 7.367904287858618e-05,
+      "loss": 0.8482,
+      "step": 17345
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.2770203948020935,
+      "learning_rate": 7.359047268821219e-05,
+      "loss": 0.8076,
+      "step": 17350
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.2635518014431,
+      "learning_rate": 7.350193845869918e-05,
+      "loss": 0.8716,
+      "step": 17355
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.2668205797672272,
+      "learning_rate": 7.341344023171441e-05,
+      "loss": 0.7338,
+      "step": 17360
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.29583418369293213,
+      "learning_rate": 7.332497804890803e-05,
+      "loss": 0.7572,
+      "step": 17365
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.26507455110549927,
+      "learning_rate": 7.323655195191328e-05,
+      "loss": 0.7844,
+      "step": 17370
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.2781868278980255,
+      "learning_rate": 7.314816198234636e-05,
+      "loss": 0.7465,
+      "step": 17375
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.2629784345626831,
+      "learning_rate": 7.305980818180663e-05,
+      "loss": 0.8417,
+      "step": 17380
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.2834410071372986,
+      "learning_rate": 7.297149059187628e-05,
+      "loss": 0.7415,
+      "step": 17385
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.3003019392490387,
+      "learning_rate": 7.28832092541205e-05,
+      "loss": 0.8308,
+      "step": 17390
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2973310351371765,
+      "learning_rate": 7.279496421008735e-05,
+      "loss": 0.7261,
+      "step": 17395
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2581186890602112,
+      "learning_rate": 7.2706755501308e-05,
+      "loss": 0.6732,
+      "step": 17400
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.276574969291687,
+      "learning_rate": 7.261858316929634e-05,
+      "loss": 0.7984,
+      "step": 17405
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.24236898124217987,
+      "learning_rate": 7.253044725554922e-05,
+      "loss": 0.7923,
+      "step": 17410
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.26904451847076416,
+      "learning_rate": 7.244234780154627e-05,
+      "loss": 0.7706,
+      "step": 17415
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2632739841938019,
+      "learning_rate": 7.235428484875006e-05,
+      "loss": 0.748,
+      "step": 17420
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.3082495629787445,
+      "learning_rate": 7.226625843860595e-05,
+      "loss": 0.8136,
+      "step": 17425
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.25225359201431274,
+      "learning_rate": 7.217826861254208e-05,
+      "loss": 0.7853,
+      "step": 17430
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2693781554698944,
+      "learning_rate": 7.209031541196931e-05,
+      "loss": 0.8022,
+      "step": 17435
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2704066038131714,
+      "learning_rate": 7.200239887828147e-05,
+      "loss": 0.8619,
+      "step": 17440
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2754969000816345,
+      "learning_rate": 7.191451905285494e-05,
+      "loss": 0.7699,
+      "step": 17445
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2515884041786194,
+      "learning_rate": 7.182667597704889e-05,
+      "loss": 0.7937,
+      "step": 17450
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.27935126423835754,
+      "learning_rate": 7.17388696922051e-05,
+      "loss": 0.7801,
+      "step": 17455
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.24923402070999146,
+      "learning_rate": 7.165110023964828e-05,
+      "loss": 0.8244,
+      "step": 17460
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.30999448895454407,
+      "learning_rate": 7.156336766068557e-05,
+      "loss": 0.7159,
+      "step": 17465
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.2763430178165436,
+      "learning_rate": 7.147567199660684e-05,
+      "loss": 0.8173,
+      "step": 17470
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.25914233922958374,
+      "learning_rate": 7.138801328868453e-05,
+      "loss": 0.7264,
+      "step": 17475
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.28961434960365295,
+      "learning_rate": 7.130039157817384e-05,
+      "loss": 0.7886,
+      "step": 17480
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.29315581917762756,
+      "learning_rate": 7.121280690631239e-05,
+      "loss": 0.7655,
+      "step": 17485
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.24615170061588287,
+      "learning_rate": 7.112525931432047e-05,
+      "loss": 0.7012,
+      "step": 17490
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.28100574016571045,
+      "learning_rate": 7.103774884340087e-05,
+      "loss": 0.8085,
+      "step": 17495
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.25875842571258545,
+      "learning_rate": 7.095027553473891e-05,
+      "loss": 0.85,
+      "step": 17500
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.28253424167633057,
+      "learning_rate": 7.086283942950246e-05,
+      "loss": 0.7207,
+      "step": 17505
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.2582673728466034,
+      "learning_rate": 7.077544056884182e-05,
+      "loss": 0.8388,
+      "step": 17510
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.2423962503671646,
+      "learning_rate": 7.068807899388976e-05,
+      "loss": 0.7351,
+      "step": 17515
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.2518567144870758,
+      "learning_rate": 7.060075474576165e-05,
+      "loss": 0.755,
+      "step": 17520
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.2877778708934784,
+      "learning_rate": 7.051346786555513e-05,
+      "loss": 0.7769,
+      "step": 17525
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.2926938235759735,
+      "learning_rate": 7.042621839435029e-05,
+      "loss": 0.8798,
+      "step": 17530
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.295379638671875,
+      "learning_rate": 7.033900637320958e-05,
+      "loss": 0.7104,
+      "step": 17535
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.27940165996551514,
+      "learning_rate": 7.0251831843178e-05,
+      "loss": 0.732,
+      "step": 17540
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.2717758119106293,
+      "learning_rate": 7.01646948452827e-05,
+      "loss": 0.8071,
+      "step": 17545
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.26429474353790283,
+      "learning_rate": 7.007759542053324e-05,
+      "loss": 0.767,
+      "step": 17550
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.3140299916267395,
+      "learning_rate": 6.999053360992155e-05,
+      "loss": 0.7958,
+      "step": 17555
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.2404472976922989,
+      "learning_rate": 6.990350945442173e-05,
+      "loss": 0.8426,
+      "step": 17560
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.2709415555000305,
+      "learning_rate": 6.981652299499032e-05,
+      "loss": 0.7802,
+      "step": 17565
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.2728278338909149,
+      "learning_rate": 6.972957427256594e-05,
+      "loss": 0.7786,
+      "step": 17570
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.27763307094573975,
+      "learning_rate": 6.964266332806966e-05,
+      "loss": 0.735,
+      "step": 17575
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.27662762999534607,
+      "learning_rate": 6.955579020240459e-05,
+      "loss": 0.8434,
+      "step": 17580
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.26812487840652466,
+      "learning_rate": 6.946895493645613e-05,
+      "loss": 0.7325,
+      "step": 17585
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.27408894896507263,
+      "learning_rate": 6.938215757109176e-05,
+      "loss": 0.9187,
+      "step": 17590
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.27132144570350647,
+      "learning_rate": 6.929539814716136e-05,
+      "loss": 0.7242,
+      "step": 17595
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.3133727014064789,
+      "learning_rate": 6.920867670549668e-05,
+      "loss": 0.8118,
+      "step": 17600
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.2706137001514435,
+      "learning_rate": 6.912199328691175e-05,
+      "loss": 0.7838,
+      "step": 17605
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.26105305552482605,
+      "learning_rate": 6.90353479322026e-05,
+      "loss": 0.821,
+      "step": 17610
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.26275086402893066,
+      "learning_rate": 6.894874068214751e-05,
+      "loss": 0.8778,
+      "step": 17615
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.28266021609306335,
+      "learning_rate": 6.88621715775067e-05,
+      "loss": 0.91,
+      "step": 17620
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.2602560520172119,
+      "learning_rate": 6.877564065902245e-05,
+      "loss": 0.8879,
+      "step": 17625
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.2614297568798065,
+      "learning_rate": 6.868914796741907e-05,
+      "loss": 0.7482,
+      "step": 17630
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.2839134633541107,
+      "learning_rate": 6.860269354340292e-05,
+      "loss": 0.7748,
+      "step": 17635
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.2632676064968109,
+      "learning_rate": 6.85162774276623e-05,
+      "loss": 0.8696,
+      "step": 17640
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.24448035657405853,
+      "learning_rate": 6.842989966086751e-05,
+      "loss": 0.7791,
+      "step": 17645
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.289100706577301,
+      "learning_rate": 6.834356028367076e-05,
+      "loss": 0.7354,
+      "step": 17650
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.27051296830177307,
+      "learning_rate": 6.82572593367063e-05,
+      "loss": 0.9232,
+      "step": 17655
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.25651922821998596,
+      "learning_rate": 6.81709968605902e-05,
+      "loss": 0.8086,
+      "step": 17660
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.24785958230495453,
+      "learning_rate": 6.808477289592045e-05,
+      "loss": 0.7559,
+      "step": 17665
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.25814223289489746,
+      "learning_rate": 6.799858748327681e-05,
+      "loss": 0.8003,
+      "step": 17670
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.2958534061908722,
+      "learning_rate": 6.791244066322115e-05,
+      "loss": 0.8522,
+      "step": 17675
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.27761539816856384,
+      "learning_rate": 6.782633247629697e-05,
+      "loss": 0.8069,
+      "step": 17680
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.33598434925079346,
+      "learning_rate": 6.774026296302963e-05,
+      "loss": 0.7955,
+      "step": 17685
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.26065829396247864,
+      "learning_rate": 6.765423216392623e-05,
+      "loss": 0.751,
+      "step": 17690
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.2681087553501129,
+      "learning_rate": 6.756824011947586e-05,
+      "loss": 0.7626,
+      "step": 17695
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.27821341156959534,
+      "learning_rate": 6.748228687014915e-05,
+      "loss": 0.6555,
+      "step": 17700
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.2935525178909302,
+      "learning_rate": 6.739637245639858e-05,
+      "loss": 0.8103,
+      "step": 17705
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.29609110951423645,
+      "learning_rate": 6.73104969186583e-05,
+      "loss": 0.8574,
+      "step": 17710
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.30124762654304504,
+      "learning_rate": 6.722466029734422e-05,
+      "loss": 0.865,
+      "step": 17715
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.26538020372390747,
+      "learning_rate": 6.713886263285388e-05,
+      "loss": 0.7681,
+      "step": 17720
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.26841890811920166,
+      "learning_rate": 6.705310396556651e-05,
+      "loss": 0.7141,
+      "step": 17725
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.2693963050842285,
+      "learning_rate": 6.696738433584295e-05,
+      "loss": 0.7318,
+      "step": 17730
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.3165275752544403,
+      "learning_rate": 6.688170378402581e-05,
+      "loss": 0.7613,
+      "step": 17735
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.2781963050365448,
+      "learning_rate": 6.679606235043913e-05,
+      "loss": 0.7928,
+      "step": 17740
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.2980717122554779,
+      "learning_rate": 6.671046007538862e-05,
+      "loss": 0.7451,
+      "step": 17745
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.26822733879089355,
+      "learning_rate": 6.662489699916153e-05,
+      "loss": 0.8043,
+      "step": 17750
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.2781360447406769,
+      "learning_rate": 6.653937316202675e-05,
+      "loss": 0.8289,
+      "step": 17755
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.28398361802101135,
+      "learning_rate": 6.64538886042346e-05,
+      "loss": 0.7195,
+      "step": 17760
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.28372859954833984,
+      "learning_rate": 6.636844336601695e-05,
+      "loss": 0.8084,
+      "step": 17765
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.2855859696865082,
+      "learning_rate": 6.628303748758719e-05,
+      "loss": 0.7861,
+      "step": 17770
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.2785280644893646,
+      "learning_rate": 6.619767100914013e-05,
+      "loss": 0.8347,
+      "step": 17775
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.2754531800746918,
+      "learning_rate": 6.611234397085207e-05,
+      "loss": 0.8999,
+      "step": 17780
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.2457299828529358,
+      "learning_rate": 6.602705641288078e-05,
+      "loss": 0.6812,
+      "step": 17785
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.2855609357357025,
+      "learning_rate": 6.594180837536533e-05,
+      "loss": 0.7006,
+      "step": 17790
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.23807214200496674,
+      "learning_rate": 6.585659989842641e-05,
+      "loss": 0.7315,
+      "step": 17795
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.2731025815010071,
+      "learning_rate": 6.57714310221659e-05,
+      "loss": 0.8854,
+      "step": 17800
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.2559722363948822,
+      "learning_rate": 6.568630178666706e-05,
+      "loss": 0.7751,
+      "step": 17805
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.28620442748069763,
+      "learning_rate": 6.560121223199455e-05,
+      "loss": 0.85,
+      "step": 17810
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.2663728892803192,
+      "learning_rate": 6.551616239819441e-05,
+      "loss": 0.9119,
+      "step": 17815
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.2895793318748474,
+      "learning_rate": 6.543115232529386e-05,
+      "loss": 0.8386,
+      "step": 17820
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.25901633501052856,
+      "learning_rate": 6.534618205330146e-05,
+      "loss": 0.8167,
+      "step": 17825
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.2944163382053375,
+      "learning_rate": 6.526125162220702e-05,
+      "loss": 0.8255,
+      "step": 17830
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.27901577949523926,
+      "learning_rate": 6.51763610719817e-05,
+      "loss": 0.7202,
+      "step": 17835
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.25898268818855286,
+      "learning_rate": 6.509151044257776e-05,
+      "loss": 0.7927,
+      "step": 17840
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.27605336904525757,
+      "learning_rate": 6.500669977392874e-05,
+      "loss": 0.9295,
+      "step": 17845
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.3069280683994293,
+      "learning_rate": 6.492192910594933e-05,
+      "loss": 0.8324,
+      "step": 17850
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.2427961677312851,
+      "learning_rate": 6.483719847853545e-05,
+      "loss": 0.8261,
+      "step": 17855
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.31651127338409424,
+      "learning_rate": 6.475250793156412e-05,
+      "loss": 0.7048,
+      "step": 17860
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.3037257492542267,
+      "learning_rate": 6.466785750489357e-05,
+      "loss": 0.7998,
+      "step": 17865
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.29183289408683777,
+      "learning_rate": 6.458324723836299e-05,
+      "loss": 0.7498,
+      "step": 17870
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.264967679977417,
+      "learning_rate": 6.449867717179293e-05,
+      "loss": 0.844,
+      "step": 17875
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.319900780916214,
+      "learning_rate": 6.44141473449848e-05,
+      "loss": 0.7544,
+      "step": 17880
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.2805964946746826,
+      "learning_rate": 6.432965779772115e-05,
+      "loss": 0.8429,
+      "step": 17885
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.24578483402729034,
+      "learning_rate": 6.424520856976551e-05,
+      "loss": 0.7295,
+      "step": 17890
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.27586978673934937,
+      "learning_rate": 6.416079970086259e-05,
+      "loss": 0.9162,
+      "step": 17895
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.2891899347305298,
+      "learning_rate": 6.407643123073797e-05,
+      "loss": 0.7682,
+      "step": 17900
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.26478612422943115,
+      "learning_rate": 6.399210319909824e-05,
+      "loss": 0.8028,
+      "step": 17905
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.26194480061531067,
+      "learning_rate": 6.390781564563093e-05,
+      "loss": 0.8431,
+      "step": 17910
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.2496415376663208,
+      "learning_rate": 6.382356861000466e-05,
+      "loss": 0.8244,
+      "step": 17915
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.2828780710697174,
+      "learning_rate": 6.373936213186884e-05,
+      "loss": 0.8356,
+      "step": 17920
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.3063983917236328,
+      "learning_rate": 6.365519625085388e-05,
+      "loss": 0.9442,
+      "step": 17925
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.2884746491909027,
+      "learning_rate": 6.357107100657088e-05,
+      "loss": 0.7947,
+      "step": 17930
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.2872295379638672,
+      "learning_rate": 6.348698643861213e-05,
+      "loss": 0.8708,
+      "step": 17935
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.2698356509208679,
+      "learning_rate": 6.340294258655056e-05,
+      "loss": 0.7472,
+      "step": 17940
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.2601388394832611,
+      "learning_rate": 6.331893948994003e-05,
+      "loss": 0.7562,
+      "step": 17945
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.27124738693237305,
+      "learning_rate": 6.32349771883151e-05,
+      "loss": 0.8861,
+      "step": 17950
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.2834950089454651,
+      "learning_rate": 6.315105572119134e-05,
+      "loss": 0.8054,
+      "step": 17955
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.26736146211624146,
+      "learning_rate": 6.306717512806492e-05,
+      "loss": 0.8403,
+      "step": 17960
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.2949734330177307,
+      "learning_rate": 6.298333544841284e-05,
+      "loss": 0.788,
+      "step": 17965
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.2661183178424835,
+      "learning_rate": 6.289953672169276e-05,
+      "loss": 0.7755,
+      "step": 17970
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.23854653537273407,
+      "learning_rate": 6.281577898734329e-05,
+      "loss": 0.7737,
+      "step": 17975
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.27708199620246887,
+      "learning_rate": 6.273206228478351e-05,
+      "loss": 0.7334,
+      "step": 17980
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.25264453887939453,
+      "learning_rate": 6.264838665341331e-05,
+      "loss": 0.796,
+      "step": 17985
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.2900809645652771,
+      "learning_rate": 6.25647521326132e-05,
+      "loss": 0.78,
+      "step": 17990
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.24564304947853088,
+      "learning_rate": 6.248115876174438e-05,
+      "loss": 0.7526,
+      "step": 17995
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.29007548093795776,
+      "learning_rate": 6.239760658014865e-05,
+      "loss": 0.8331,
+      "step": 18000
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.2651544213294983,
+      "learning_rate": 6.231409562714845e-05,
+      "loss": 0.7203,
+      "step": 18005
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.28148752450942993,
+      "learning_rate": 6.223062594204676e-05,
+      "loss": 0.7299,
+      "step": 18010
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.2735466957092285,
+      "learning_rate": 6.214719756412729e-05,
+      "loss": 0.6825,
+      "step": 18015
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.2654740810394287,
+      "learning_rate": 6.206381053265412e-05,
+      "loss": 0.8268,
+      "step": 18020
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.264364093542099,
+      "learning_rate": 6.198046488687201e-05,
+      "loss": 0.7943,
+      "step": 18025
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.2690427601337433,
+      "learning_rate": 6.18971606660061e-05,
+      "loss": 0.7951,
+      "step": 18030
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.3025277853012085,
+      "learning_rate": 6.181389790926224e-05,
+      "loss": 0.7487,
+      "step": 18035
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.2653854489326477,
+      "learning_rate": 6.173067665582659e-05,
+      "loss": 0.8685,
+      "step": 18040
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.2819320559501648,
+      "learning_rate": 6.164749694486579e-05,
+      "loss": 0.769,
+      "step": 18045
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.28492966294288635,
+      "learning_rate": 6.156435881552708e-05,
+      "loss": 0.821,
+      "step": 18050
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.24990375339984894,
+      "learning_rate": 6.148126230693796e-05,
+      "loss": 0.9141,
+      "step": 18055
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.2689098119735718,
+      "learning_rate": 6.139820745820643e-05,
+      "loss": 0.7734,
+      "step": 18060
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.2601291239261627,
+      "learning_rate": 6.131519430842083e-05,
+      "loss": 0.7885,
+      "step": 18065
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.26659342646598816,
+      "learning_rate": 6.123222289664993e-05,
+      "loss": 0.8574,
+      "step": 18070
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.260973185300827,
+      "learning_rate": 6.114929326194281e-05,
+      "loss": 0.7629,
+      "step": 18075
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.2958158850669861,
+      "learning_rate": 6.106640544332894e-05,
+      "loss": 0.8406,
+      "step": 18080
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.27698177099227905,
+      "learning_rate": 6.0983559479817986e-05,
+      "loss": 0.7944,
+      "step": 18085
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.26226866245269775,
+      "learning_rate": 6.090075541040015e-05,
+      "loss": 0.761,
+      "step": 18090
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.28929901123046875,
+      "learning_rate": 6.08179932740457e-05,
+      "loss": 0.7879,
+      "step": 18095
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.2864818871021271,
+      "learning_rate": 6.073527310970527e-05,
+      "loss": 0.7526,
+      "step": 18100
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.28640732169151306,
+      "learning_rate": 6.065259495630966e-05,
+      "loss": 0.7588,
+      "step": 18105
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.25879958271980286,
+      "learning_rate": 6.0569958852770026e-05,
+      "loss": 0.8613,
+      "step": 18110
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.26180025935173035,
+      "learning_rate": 6.048736483797765e-05,
+      "loss": 0.7988,
+      "step": 18115
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.28200724720954895,
+      "learning_rate": 6.040481295080402e-05,
+      "loss": 0.7577,
+      "step": 18120
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.27897992730140686,
+      "learning_rate": 6.0322303230100706e-05,
+      "loss": 0.7339,
+      "step": 18125
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.2738398313522339,
+      "learning_rate": 6.0239835714699656e-05,
+      "loss": 0.7929,
+      "step": 18130
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.2769078016281128,
+      "learning_rate": 6.015741044341282e-05,
+      "loss": 0.8469,
+      "step": 18135
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.2504939138889313,
+      "learning_rate": 6.0075027455032154e-05,
+      "loss": 0.806,
+      "step": 18140
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.27813395857810974,
+      "learning_rate": 5.999268678832982e-05,
+      "loss": 0.8322,
+      "step": 18145
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.25341561436653137,
+      "learning_rate": 5.9910388482058196e-05,
+      "loss": 0.8823,
+      "step": 18150
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.2485748827457428,
+      "learning_rate": 5.982813257494954e-05,
+      "loss": 0.7445,
+      "step": 18155
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.2577187120914459,
+      "learning_rate": 5.9745919105716195e-05,
+      "loss": 0.7756,
+      "step": 18160
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.28886187076568604,
+      "learning_rate": 5.966374811305051e-05,
+      "loss": 0.8127,
+      "step": 18165
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.27744728326797485,
+      "learning_rate": 5.9581619635625014e-05,
+      "loss": 0.815,
+      "step": 18170
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.2894884943962097,
+      "learning_rate": 5.9499533712092e-05,
+      "loss": 0.8005,
+      "step": 18175
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.3021494448184967,
+      "learning_rate": 5.941749038108385e-05,
+      "loss": 0.8278,
+      "step": 18180
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.2534330189228058,
+      "learning_rate": 5.9335489681212835e-05,
+      "loss": 0.8549,
+      "step": 18185
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.3106016516685486,
+      "learning_rate": 5.92535316510713e-05,
+      "loss": 0.6661,
+      "step": 18190
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.2976013422012329,
+      "learning_rate": 5.9171616329231364e-05,
+      "loss": 0.824,
+      "step": 18195
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.2501593232154846,
+      "learning_rate": 5.90897437542451e-05,
+      "loss": 0.787,
+      "step": 18200
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.28381872177124023,
+      "learning_rate": 5.900791396464445e-05,
+      "loss": 0.7354,
+      "step": 18205
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.259772390127182,
+      "learning_rate": 5.892612699894127e-05,
+      "loss": 0.789,
+      "step": 18210
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.24163563549518585,
+      "learning_rate": 5.884438289562717e-05,
+      "loss": 0.8713,
+      "step": 18215
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.2834208011627197,
+      "learning_rate": 5.8762681693173675e-05,
+      "loss": 0.841,
+      "step": 18220
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.2929078936576843,
+      "learning_rate": 5.868102343003201e-05,
+      "loss": 0.7945,
+      "step": 18225
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.30455702543258667,
+      "learning_rate": 5.8599408144633405e-05,
+      "loss": 0.7069,
+      "step": 18230
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.3452713191509247,
+      "learning_rate": 5.851783587538863e-05,
+      "loss": 0.8092,
+      "step": 18235
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.26134809851646423,
+      "learning_rate": 5.843630666068832e-05,
+      "loss": 0.8087,
+      "step": 18240
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.30902811884880066,
+      "learning_rate": 5.835482053890278e-05,
+      "loss": 0.8481,
+      "step": 18245
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.31992846727371216,
+      "learning_rate": 5.827337754838218e-05,
+      "loss": 0.7905,
+      "step": 18250
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.2618788182735443,
+      "learning_rate": 5.819197772745627e-05,
+      "loss": 0.7629,
+      "step": 18255
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.28527355194091797,
+      "learning_rate": 5.811062111443447e-05,
+      "loss": 0.826,
+      "step": 18260
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.31207770109176636,
+      "learning_rate": 5.8029307747605905e-05,
+      "loss": 0.7721,
+      "step": 18265
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.250207781791687,
+      "learning_rate": 5.794803766523939e-05,
+      "loss": 0.7803,
+      "step": 18270
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.29229018092155457,
+      "learning_rate": 5.786681090558332e-05,
+      "loss": 0.8023,
+      "step": 18275
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.273379385471344,
+      "learning_rate": 5.778562750686568e-05,
+      "loss": 0.7844,
+      "step": 18280
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.27281737327575684,
+      "learning_rate": 5.770448750729408e-05,
+      "loss": 0.8071,
+      "step": 18285
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.2896445095539093,
+      "learning_rate": 5.762339094505569e-05,
+      "loss": 0.7641,
+      "step": 18290
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.2542775571346283,
+      "learning_rate": 5.7542337858317257e-05,
+      "loss": 0.8674,
+      "step": 18295
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.2890925705432892,
+      "learning_rate": 5.746132828522506e-05,
+      "loss": 0.7799,
+      "step": 18300
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.303602397441864,
+      "learning_rate": 5.738036226390483e-05,
+      "loss": 0.7872,
+      "step": 18305
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.31776705384254456,
+      "learning_rate": 5.729943983246198e-05,
+      "loss": 0.8462,
+      "step": 18310
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.2585162818431854,
+      "learning_rate": 5.721856102898121e-05,
+      "loss": 0.8747,
+      "step": 18315
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.2587442696094513,
+      "learning_rate": 5.713772589152682e-05,
+      "loss": 0.8068,
+      "step": 18320
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.26109957695007324,
+      "learning_rate": 5.705693445814243e-05,
+      "loss": 0.7185,
+      "step": 18325
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.27592533826828003,
+      "learning_rate": 5.697618676685127e-05,
+      "loss": 0.8076,
+      "step": 18330
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.26613497734069824,
+      "learning_rate": 5.689548285565585e-05,
+      "loss": 0.803,
+      "step": 18335
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.2756407558917999,
+      "learning_rate": 5.681482276253811e-05,
+      "loss": 0.836,
+      "step": 18340
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.2865229845046997,
+      "learning_rate": 5.6734206525459355e-05,
+      "loss": 0.7962,
+      "step": 18345
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.2565017342567444,
+      "learning_rate": 5.6653634182360267e-05,
+      "loss": 0.8133,
+      "step": 18350
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.2665974497795105,
+      "learning_rate": 5.6573105771160875e-05,
+      "loss": 0.7833,
+      "step": 18355
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.3112146556377411,
+      "learning_rate": 5.6492621329760524e-05,
+      "loss": 0.8103,
+      "step": 18360
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.299140602350235,
+      "learning_rate": 5.641218089603779e-05,
+      "loss": 0.7744,
+      "step": 18365
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.25373366475105286,
+      "learning_rate": 5.6331784507850744e-05,
+      "loss": 0.8986,
+      "step": 18370
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.27519750595092773,
+      "learning_rate": 5.6251432203036544e-05,
+      "loss": 0.791,
+      "step": 18375
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.253791481256485,
+      "learning_rate": 5.617112401941163e-05,
+      "loss": 0.7265,
+      "step": 18380
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.26284411549568176,
+      "learning_rate": 5.609085999477166e-05,
+      "loss": 0.8349,
+      "step": 18385
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.26994234323501587,
+      "learning_rate": 5.601064016689165e-05,
+      "loss": 0.7947,
+      "step": 18390
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.3026762902736664,
+      "learning_rate": 5.593046457352568e-05,
+      "loss": 0.8011,
+      "step": 18395
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.2768292725086212,
+      "learning_rate": 5.585033325240704e-05,
+      "loss": 0.7135,
+      "step": 18400
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.28852131962776184,
+      "learning_rate": 5.5770246241248125e-05,
+      "loss": 0.8516,
+      "step": 18405
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.28244367241859436,
+      "learning_rate": 5.5690203577740654e-05,
+      "loss": 0.7431,
+      "step": 18410
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.2686423361301422,
+      "learning_rate": 5.561020529955531e-05,
+      "loss": 0.7382,
+      "step": 18415
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.2730972170829773,
+      "learning_rate": 5.5530251444341936e-05,
+      "loss": 0.8299,
+      "step": 18420
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.26289600133895874,
+      "learning_rate": 5.5450342049729475e-05,
+      "loss": 0.7141,
+      "step": 18425
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.2697264552116394,
+      "learning_rate": 5.5370477153325945e-05,
+      "loss": 0.7229,
+      "step": 18430
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.2662758231163025,
+      "learning_rate": 5.529065679271841e-05,
+      "loss": 0.8089,
+      "step": 18435
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.2873300313949585,
+      "learning_rate": 5.5210881005473e-05,
+      "loss": 0.8195,
+      "step": 18440
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.29123494029045105,
+      "learning_rate": 5.513114982913479e-05,
+      "loss": 0.8099,
+      "step": 18445
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.2564477324485779,
+      "learning_rate": 5.5051463301228e-05,
+      "loss": 0.675,
+      "step": 18450
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.29058778285980225,
+      "learning_rate": 5.497182145925575e-05,
+      "loss": 0.7853,
+      "step": 18455
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.32175374031066895,
+      "learning_rate": 5.489222434070009e-05,
+      "loss": 0.8176,
+      "step": 18460
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.291685551404953,
+      "learning_rate": 5.4812671983022046e-05,
+      "loss": 0.7884,
+      "step": 18465
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.25728997588157654,
+      "learning_rate": 5.473316442366167e-05,
+      "loss": 0.824,
+      "step": 18470
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.2623702883720398,
+      "learning_rate": 5.465370170003785e-05,
+      "loss": 0.8171,
+      "step": 18475
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.24209091067314148,
+      "learning_rate": 5.4574283849548354e-05,
+      "loss": 0.8101,
+      "step": 18480
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.2494385689496994,
+      "learning_rate": 5.449491090956982e-05,
+      "loss": 0.7616,
+      "step": 18485
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.307625949382782,
+      "learning_rate": 5.44155829174579e-05,
+      "loss": 0.8873,
+      "step": 18490
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.27069059014320374,
+      "learning_rate": 5.433629991054691e-05,
+      "loss": 0.8159,
+      "step": 18495
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.2697925567626953,
+      "learning_rate": 5.425706192615007e-05,
+      "loss": 0.8165,
+      "step": 18500
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.2592358887195587,
+      "learning_rate": 5.417786900155942e-05,
+      "loss": 0.7355,
+      "step": 18505
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.28814736008644104,
+      "learning_rate": 5.409872117404577e-05,
+      "loss": 0.8358,
+      "step": 18510
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.2974863350391388,
+      "learning_rate": 5.401961848085871e-05,
+      "loss": 0.7225,
+      "step": 18515
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.28759467601776123,
+      "learning_rate": 5.394056095922662e-05,
+      "loss": 0.74,
+      "step": 18520
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.3033956289291382,
+      "learning_rate": 5.3861548646356514e-05,
+      "loss": 0.7645,
+      "step": 18525
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.2899126410484314,
+      "learning_rate": 5.3782581579434325e-05,
+      "loss": 0.7309,
+      "step": 18530
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.2775956690311432,
+      "learning_rate": 5.370365979562453e-05,
+      "loss": 0.8238,
+      "step": 18535
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.2776409983634949,
+      "learning_rate": 5.362478333207034e-05,
+      "loss": 0.84,
+      "step": 18540
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.27701374888420105,
+      "learning_rate": 5.354595222589358e-05,
+      "loss": 0.7884,
+      "step": 18545
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.3034186065196991,
+      "learning_rate": 5.34671665141949e-05,
+      "loss": 0.7565,
+      "step": 18550
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.2622527480125427,
+      "learning_rate": 5.3388426234053414e-05,
+      "loss": 0.7612,
+      "step": 18555
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.28101763129234314,
+      "learning_rate": 5.330973142252691e-05,
+      "loss": 0.7192,
+      "step": 18560
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.2687520384788513,
+      "learning_rate": 5.323108211665178e-05,
+      "loss": 0.8082,
+      "step": 18565
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.31704583764076233,
+      "learning_rate": 5.3152478353443006e-05,
+      "loss": 0.6957,
+      "step": 18570
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.26655253767967224,
+      "learning_rate": 5.307392016989413e-05,
+      "loss": 0.8078,
+      "step": 18575
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.27396515011787415,
+      "learning_rate": 5.299540760297717e-05,
+      "loss": 0.8209,
+      "step": 18580
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.24590495228767395,
+      "learning_rate": 5.2916940689642887e-05,
+      "loss": 0.8348,
+      "step": 18585
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.23811030387878418,
+      "learning_rate": 5.283851946682033e-05,
+      "loss": 0.771,
+      "step": 18590
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.2999459505081177,
+      "learning_rate": 5.2760143971417155e-05,
+      "loss": 0.7364,
+      "step": 18595
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.257988840341568,
+      "learning_rate": 5.2681814240319415e-05,
+      "loss": 0.7833,
+      "step": 18600
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.27727702260017395,
+      "learning_rate": 5.260353031039177e-05,
+      "loss": 0.8248,
+      "step": 18605
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.24502165615558624,
+      "learning_rate": 5.252529221847719e-05,
+      "loss": 0.7083,
+      "step": 18610
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.2822337746620178,
+      "learning_rate": 5.244710000139714e-05,
+      "loss": 0.8521,
+      "step": 18615
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.2731589674949646,
+      "learning_rate": 5.2368953695951405e-05,
+      "loss": 0.8219,
+      "step": 18620
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.2679608464241028,
+      "learning_rate": 5.229085333891834e-05,
+      "loss": 0.73,
+      "step": 18625
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.2607372999191284,
+      "learning_rate": 5.221279896705452e-05,
+      "loss": 0.7845,
+      "step": 18630
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.2670726478099823,
+      "learning_rate": 5.213479061709492e-05,
+      "loss": 0.8545,
+      "step": 18635
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.29974955320358276,
+      "learning_rate": 5.2056828325752855e-05,
+      "loss": 0.8346,
+      "step": 18640
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.2804435193538666,
+      "learning_rate": 5.1978912129719956e-05,
+      "loss": 0.7781,
+      "step": 18645
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.27671873569488525,
+      "learning_rate": 5.1901042065666214e-05,
+      "loss": 0.8077,
+      "step": 18650
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.28457868099212646,
+      "learning_rate": 5.182321817023983e-05,
+      "loss": 0.8014,
+      "step": 18655
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.2779267728328705,
+      "learning_rate": 5.174544048006726e-05,
+      "loss": 0.8311,
+      "step": 18660
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.26867663860321045,
+      "learning_rate": 5.1667709031753406e-05,
+      "loss": 0.7116,
+      "step": 18665
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.27532848715782166,
+      "learning_rate": 5.159002386188118e-05,
+      "loss": 0.9608,
+      "step": 18670
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.2576589286327362,
+      "learning_rate": 5.151238500701184e-05,
+      "loss": 0.7806,
+      "step": 18675
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.29367679357528687,
+      "learning_rate": 5.1434792503684716e-05,
+      "loss": 0.7674,
+      "step": 18680
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.2842334806919098,
+      "learning_rate": 5.135724638841755e-05,
+      "loss": 0.7849,
+      "step": 18685
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.305401474237442,
+      "learning_rate": 5.1279746697706055e-05,
+      "loss": 0.7973,
+      "step": 18690
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.2906571328639984,
+      "learning_rate": 5.120229346802416e-05,
+      "loss": 0.808,
+      "step": 18695
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.2675495445728302,
+      "learning_rate": 5.112488673582389e-05,
+      "loss": 0.7582,
+      "step": 18700
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.2715294063091278,
+      "learning_rate": 5.1047526537535504e-05,
+      "loss": 0.779,
+      "step": 18705
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.29356223344802856,
+      "learning_rate": 5.0970212909567236e-05,
+      "loss": 0.8705,
+      "step": 18710
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.2562316656112671,
+      "learning_rate": 5.0892945888305446e-05,
+      "loss": 0.7458,
+      "step": 18715
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.31263595819473267,
+      "learning_rate": 5.081572551011458e-05,
+      "loss": 0.8115,
+      "step": 18720
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.31803205609321594,
+      "learning_rate": 5.07385518113371e-05,
+      "loss": 0.7623,
+      "step": 18725
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.25509113073349,
+      "learning_rate": 5.06614248282935e-05,
+      "loss": 0.8839,
+      "step": 18730
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.2617010474205017,
+      "learning_rate": 5.058434459728229e-05,
+      "loss": 0.7858,
+      "step": 18735
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.29557377099990845,
+      "learning_rate": 5.0507311154579976e-05,
+      "loss": 0.8331,
+      "step": 18740
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.28714311122894287,
+      "learning_rate": 5.043032453644113e-05,
+      "loss": 0.7831,
+      "step": 18745
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.27089929580688477,
+      "learning_rate": 5.035338477909817e-05,
+      "loss": 0.8846,
+      "step": 18750
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.2682959735393524,
+      "learning_rate": 5.027649191876147e-05,
+      "loss": 0.7482,
+      "step": 18755
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.26213210821151733,
+      "learning_rate": 5.019964599161935e-05,
+      "loss": 0.7949,
+      "step": 18760
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.30847010016441345,
+      "learning_rate": 5.0122847033838156e-05,
+      "loss": 0.7751,
+      "step": 18765
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.25752073526382446,
+      "learning_rate": 5.004609508156196e-05,
+      "loss": 0.8719,
+      "step": 18770
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.2857056260108948,
+      "learning_rate": 4.996939017091278e-05,
+      "loss": 0.7445,
+      "step": 18775
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.2874182164669037,
+      "learning_rate": 4.989273233799051e-05,
+      "loss": 0.7717,
+      "step": 18780
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.34927722811698914,
+      "learning_rate": 4.981612161887285e-05,
+      "loss": 0.7717,
+      "step": 18785
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.2875741422176361,
+      "learning_rate": 4.973955804961536e-05,
+      "loss": 0.8041,
+      "step": 18790
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.24138030409812927,
+      "learning_rate": 4.966304166625139e-05,
+      "loss": 0.8356,
+      "step": 18795
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.28111588954925537,
+      "learning_rate": 4.958657250479208e-05,
+      "loss": 0.7551,
+      "step": 18800
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.2540626525878906,
+      "learning_rate": 4.9510150601226394e-05,
+      "loss": 0.777,
+      "step": 18805
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.26399436593055725,
+      "learning_rate": 4.9433775991521006e-05,
+      "loss": 0.8436,
+      "step": 18810
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.27439209818840027,
+      "learning_rate": 4.9357448711620345e-05,
+      "loss": 0.7735,
+      "step": 18815
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.2720988988876343,
+      "learning_rate": 4.9281168797446504e-05,
+      "loss": 0.8395,
+      "step": 18820
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.30142033100128174,
+      "learning_rate": 4.920493628489946e-05,
+      "loss": 0.865,
+      "step": 18825
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.27852845191955566,
+      "learning_rate": 4.9128751209856684e-05,
+      "loss": 0.8549,
+      "step": 18830
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.2861536741256714,
+      "learning_rate": 4.9052613608173425e-05,
+      "loss": 0.77,
+      "step": 18835
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.26632633805274963,
+      "learning_rate": 4.8976523515682505e-05,
+      "loss": 0.8792,
+      "step": 18840
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.25601017475128174,
+      "learning_rate": 4.890048096819456e-05,
+      "loss": 0.6669,
+      "step": 18845
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.29297733306884766,
+      "learning_rate": 4.882448600149767e-05,
+      "loss": 0.8335,
+      "step": 18850
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.27074167132377625,
+      "learning_rate": 4.874853865135761e-05,
+      "loss": 0.7719,
+      "step": 18855
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.2736327648162842,
+      "learning_rate": 4.8672638953517724e-05,
+      "loss": 0.7569,
+      "step": 18860
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.2467767298221588,
+      "learning_rate": 4.859678694369892e-05,
+      "loss": 0.8277,
+      "step": 18865
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.26508867740631104,
+      "learning_rate": 4.852098265759969e-05,
+      "loss": 0.7552,
+      "step": 18870
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.295194149017334,
+      "learning_rate": 4.844522613089601e-05,
+      "loss": 0.7934,
+      "step": 18875
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.2907750606536865,
+      "learning_rate": 4.836951739924141e-05,
+      "loss": 0.8823,
+      "step": 18880
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.252069890499115,
+      "learning_rate": 4.829385649826702e-05,
+      "loss": 0.8101,
+      "step": 18885
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.263295978307724,
+      "learning_rate": 4.82182434635813e-05,
+      "loss": 0.8344,
+      "step": 18890
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.3141554892063141,
+      "learning_rate": 4.814267833077029e-05,
+      "loss": 0.8552,
+      "step": 18895
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.2977995276451111,
+      "learning_rate": 4.806716113539737e-05,
+      "loss": 0.7844,
+      "step": 18900
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.28713858127593994,
+      "learning_rate": 4.799169191300357e-05,
+      "loss": 0.7359,
+      "step": 18905
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.265524297952652,
+      "learning_rate": 4.791627069910713e-05,
+      "loss": 0.7948,
+      "step": 18910
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.26660722494125366,
+      "learning_rate": 4.784089752920381e-05,
+      "loss": 0.7537,
+      "step": 18915
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.3201920688152313,
+      "learning_rate": 4.776557243876665e-05,
+      "loss": 0.6509,
+      "step": 18920
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.25317713618278503,
+      "learning_rate": 4.7690295463246255e-05,
+      "loss": 0.8217,
+      "step": 18925
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.2925649881362915,
+      "learning_rate": 4.761506663807047e-05,
+      "loss": 0.8093,
+      "step": 18930
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.26825374364852905,
+      "learning_rate": 4.7539885998644365e-05,
+      "loss": 0.7968,
+      "step": 18935
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.2798418402671814,
+      "learning_rate": 4.746475358035046e-05,
+      "loss": 0.8392,
+      "step": 18940
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.25741979479789734,
+      "learning_rate": 4.738966941854866e-05,
+      "loss": 0.7922,
+      "step": 18945
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.27443379163742065,
+      "learning_rate": 4.731463354857602e-05,
+      "loss": 0.8369,
+      "step": 18950
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.2702893316745758,
+      "learning_rate": 4.7239646005746895e-05,
+      "loss": 0.7611,
+      "step": 18955
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.38152143359184265,
+      "learning_rate": 4.716470682535289e-05,
+      "loss": 0.8706,
+      "step": 18960
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.26766619086265564,
+      "learning_rate": 4.708981604266296e-05,
+      "loss": 0.8218,
+      "step": 18965
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.25141772627830505,
+      "learning_rate": 4.701497369292313e-05,
+      "loss": 0.8075,
+      "step": 18970
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.264624685049057,
+      "learning_rate": 4.694017981135671e-05,
+      "loss": 0.7692,
+      "step": 18975
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.2943498194217682,
+      "learning_rate": 4.6865434433164125e-05,
+      "loss": 0.7622,
+      "step": 18980
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.26615047454833984,
+      "learning_rate": 4.679073759352315e-05,
+      "loss": 0.7964,
+      "step": 18985
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.24245823919773102,
+      "learning_rate": 4.671608932758853e-05,
+      "loss": 0.7825,
+      "step": 18990
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.2933671772480011,
+      "learning_rate": 4.664148967049221e-05,
+      "loss": 0.8122,
+      "step": 18995
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.2749022841453552,
+      "learning_rate": 4.656693865734331e-05,
+      "loss": 0.7729,
+      "step": 19000
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.268289178609848,
+      "learning_rate": 4.649243632322796e-05,
+      "loss": 0.793,
+      "step": 19005
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.28177735209465027,
+      "learning_rate": 4.641798270320948e-05,
+      "loss": 0.8523,
+      "step": 19010
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.262329638004303,
+      "learning_rate": 4.6343577832328176e-05,
+      "loss": 0.8208,
+      "step": 19015
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.31374141573905945,
+      "learning_rate": 4.626922174560142e-05,
+      "loss": 0.8154,
+      "step": 19020
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.28644031286239624,
+      "learning_rate": 4.6194914478023754e-05,
+      "loss": 0.8359,
+      "step": 19025
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.2689576745033264,
+      "learning_rate": 4.6120656064566604e-05,
+      "loss": 0.9421,
+      "step": 19030
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.25812458992004395,
+      "learning_rate": 4.604644654017843e-05,
+      "loss": 0.8141,
+      "step": 19035
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.2505223751068115,
+      "learning_rate": 4.597228593978464e-05,
+      "loss": 0.8084,
+      "step": 19040
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.2971956431865692,
+      "learning_rate": 4.589817429828781e-05,
+      "loss": 0.7819,
+      "step": 19045
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.274215430021286,
+      "learning_rate": 4.5824111650567264e-05,
+      "loss": 0.7431,
+      "step": 19050
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.2821873724460602,
+      "learning_rate": 4.575009803147929e-05,
+      "loss": 0.8117,
+      "step": 19055
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.2926546037197113,
+      "learning_rate": 4.567613347585727e-05,
+      "loss": 0.8768,
+      "step": 19060
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.3123989701271057,
+      "learning_rate": 4.560221801851133e-05,
+      "loss": 0.7407,
+      "step": 19065
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.3110482394695282,
+      "learning_rate": 4.552835169422854e-05,
+      "loss": 0.7504,
+      "step": 19070
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.2745916545391083,
+      "learning_rate": 4.5454534537772825e-05,
+      "loss": 0.6989,
+      "step": 19075
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.2758617401123047,
+      "learning_rate": 4.5380766583885016e-05,
+      "loss": 0.7739,
+      "step": 19080
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.2954859733581543,
+      "learning_rate": 4.5307047867282734e-05,
+      "loss": 0.838,
+      "step": 19085
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.285125732421875,
+      "learning_rate": 4.523337842266047e-05,
+      "loss": 0.7301,
+      "step": 19090
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.2881282567977905,
+      "learning_rate": 4.515975828468949e-05,
+      "loss": 0.7589,
+      "step": 19095
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.30516520142555237,
+      "learning_rate": 4.508618748801793e-05,
+      "loss": 0.8142,
+      "step": 19100
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.315054714679718,
+      "learning_rate": 4.50126660672706e-05,
+      "loss": 0.7316,
+      "step": 19105
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.2825736999511719,
+      "learning_rate": 4.493919405704917e-05,
+      "loss": 0.777,
+      "step": 19110
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.2859044671058655,
+      "learning_rate": 4.486577149193191e-05,
+      "loss": 0.7306,
+      "step": 19115
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.29017677903175354,
+      "learning_rate": 4.479239840647405e-05,
+      "loss": 0.8264,
+      "step": 19120
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.28172457218170166,
+      "learning_rate": 4.471907483520732e-05,
+      "loss": 0.7622,
+      "step": 19125
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.28697964549064636,
+      "learning_rate": 4.464580081264026e-05,
+      "loss": 0.8655,
+      "step": 19130
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.290617972612381,
+      "learning_rate": 4.4572576373257986e-05,
+      "loss": 0.7366,
+      "step": 19135
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.2692755460739136,
+      "learning_rate": 4.4499401551522504e-05,
+      "loss": 0.7704,
+      "step": 19140
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.3180653750896454,
+      "learning_rate": 4.442627638187216e-05,
+      "loss": 0.7775,
+      "step": 19145
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.3238975703716278,
+      "learning_rate": 4.435320089872217e-05,
+      "loss": 0.7988,
+      "step": 19150
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.2899223864078522,
+      "learning_rate": 4.428017513646418e-05,
+      "loss": 0.7712,
+      "step": 19155
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.2541576027870178,
+      "learning_rate": 4.4207199129466685e-05,
+      "loss": 0.8302,
+      "step": 19160
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.3092675805091858,
+      "learning_rate": 4.4134272912074546e-05,
+      "loss": 0.7243,
+      "step": 19165
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.2940780818462372,
+      "learning_rate": 4.406139651860928e-05,
+      "loss": 0.7353,
+      "step": 19170
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.27896803617477417,
+      "learning_rate": 4.398856998336885e-05,
+      "loss": 0.7755,
+      "step": 19175
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.275955468416214,
+      "learning_rate": 4.391579334062798e-05,
+      "loss": 0.6845,
+      "step": 19180
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.27416709065437317,
+      "learning_rate": 4.3843066624637705e-05,
+      "loss": 0.8806,
+      "step": 19185
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.2871681749820709,
+      "learning_rate": 4.3770389869625654e-05,
+      "loss": 0.8377,
+      "step": 19190
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.2698437571525574,
+      "learning_rate": 4.3697763109795845e-05,
+      "loss": 0.7839,
+      "step": 19195
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.2793125510215759,
+      "learning_rate": 4.362518637932896e-05,
+      "loss": 0.7334,
+      "step": 19200
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.26299965381622314,
+      "learning_rate": 4.3552659712381965e-05,
+      "loss": 0.7914,
+      "step": 19205
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.29598289728164673,
+      "learning_rate": 4.34801831430883e-05,
+      "loss": 0.7296,
+      "step": 19210
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.30345529317855835,
+      "learning_rate": 4.340775670555787e-05,
+      "loss": 0.7574,
+      "step": 19215
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.27294403314590454,
+      "learning_rate": 4.333538043387695e-05,
+      "loss": 0.8409,
+      "step": 19220
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.2598150968551636,
+      "learning_rate": 4.326305436210823e-05,
+      "loss": 0.7326,
+      "step": 19225
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.24395301938056946,
+      "learning_rate": 4.3190778524290755e-05,
+      "loss": 0.8482,
+      "step": 19230
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.26434871554374695,
+      "learning_rate": 4.311855295443987e-05,
+      "loss": 0.7736,
+      "step": 19235
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.3252696394920349,
+      "learning_rate": 4.304637768654744e-05,
+      "loss": 0.7654,
+      "step": 19240
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.31384336948394775,
+      "learning_rate": 4.29742527545815e-05,
+      "loss": 0.7946,
+      "step": 19245
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.2744503617286682,
+      "learning_rate": 4.290217819248641e-05,
+      "loss": 0.8248,
+      "step": 19250
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.29785433411598206,
+      "learning_rate": 4.283015403418284e-05,
+      "loss": 0.777,
+      "step": 19255
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.26613983511924744,
+      "learning_rate": 4.275818031356783e-05,
+      "loss": 0.7121,
+      "step": 19260
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.2956700623035431,
+      "learning_rate": 4.268625706451454e-05,
+      "loss": 0.8807,
+      "step": 19265
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.2700934112071991,
+      "learning_rate": 4.261438432087247e-05,
+      "loss": 0.703,
+      "step": 19270
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.2920432984828949,
+      "learning_rate": 4.2542562116467245e-05,
+      "loss": 0.8218,
+      "step": 19275
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.2755813002586365,
+      "learning_rate": 4.247079048510089e-05,
+      "loss": 0.7852,
+      "step": 19280
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.3014310598373413,
+      "learning_rate": 4.239906946055148e-05,
+      "loss": 0.7117,
+      "step": 19285
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.32181209325790405,
+      "learning_rate": 4.2327399076573284e-05,
+      "loss": 0.8007,
+      "step": 19290
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.3137061297893524,
+      "learning_rate": 4.225577936689677e-05,
+      "loss": 0.7185,
+      "step": 19295
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.24845640361309052,
+      "learning_rate": 4.2184210365228575e-05,
+      "loss": 0.733,
+      "step": 19300
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.28396517038345337,
+      "learning_rate": 4.2112692105251446e-05,
+      "loss": 0.8673,
+      "step": 19305
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.2980553209781647,
+      "learning_rate": 4.204122462062422e-05,
+      "loss": 0.7282,
+      "step": 19310
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.31656941771507263,
+      "learning_rate": 4.196980794498185e-05,
+      "loss": 0.7511,
+      "step": 19315
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.26247191429138184,
+      "learning_rate": 4.189844211193548e-05,
+      "loss": 0.6768,
+      "step": 19320
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.26794061064720154,
+      "learning_rate": 4.182712715507217e-05,
+      "loss": 0.774,
+      "step": 19325
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.289065420627594,
+      "learning_rate": 4.175586310795515e-05,
+      "loss": 0.7939,
+      "step": 19330
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.27553194761276245,
+      "learning_rate": 4.168465000412359e-05,
+      "loss": 0.6774,
+      "step": 19335
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.29777318239212036,
+      "learning_rate": 4.161348787709282e-05,
+      "loss": 0.6694,
+      "step": 19340
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.289673388004303,
+      "learning_rate": 4.1542376760354045e-05,
+      "loss": 0.7335,
+      "step": 19345
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.27125170826911926,
+      "learning_rate": 4.1471316687374556e-05,
+      "loss": 0.8134,
+      "step": 19350
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.2760598063468933,
+      "learning_rate": 4.140030769159755e-05,
+      "loss": 0.8138,
+      "step": 19355
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.28096047043800354,
+      "learning_rate": 4.132934980644223e-05,
+      "loss": 0.7309,
+      "step": 19360
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.28885167837142944,
+      "learning_rate": 4.125844306530373e-05,
+      "loss": 0.752,
+      "step": 19365
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.2942561209201813,
+      "learning_rate": 4.118758750155311e-05,
+      "loss": 0.7993,
+      "step": 19370
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.24654746055603027,
+      "learning_rate": 4.111678314853732e-05,
+      "loss": 0.7266,
+      "step": 19375
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.2839394807815552,
+      "learning_rate": 4.104603003957934e-05,
+      "loss": 0.7632,
+      "step": 19380
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.27274009585380554,
+      "learning_rate": 4.097532820797786e-05,
+      "loss": 0.8847,
+      "step": 19385
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.2987731397151947,
+      "learning_rate": 4.0904677687007524e-05,
+      "loss": 0.7926,
+      "step": 19390
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.2836725413799286,
+      "learning_rate": 4.0834078509918756e-05,
+      "loss": 0.8595,
+      "step": 19395
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.3034559190273285,
+      "learning_rate": 4.076353070993799e-05,
+      "loss": 0.8338,
+      "step": 19400
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.3112902045249939,
+      "learning_rate": 4.0693034320267295e-05,
+      "loss": 0.7235,
+      "step": 19405
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.3092862367630005,
+      "learning_rate": 4.062258937408464e-05,
+      "loss": 0.7693,
+      "step": 19410
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.2575216591358185,
+      "learning_rate": 4.055219590454366e-05,
+      "loss": 0.769,
+      "step": 19415
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.28645190596580505,
+      "learning_rate": 4.0481853944774e-05,
+      "loss": 0.7985,
+      "step": 19420
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.29721811413764954,
+      "learning_rate": 4.0411563527880856e-05,
+      "loss": 0.878,
+      "step": 19425
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.27767476439476013,
+      "learning_rate": 4.0341324686945237e-05,
+      "loss": 0.8084,
+      "step": 19430
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.3142256736755371,
+      "learning_rate": 4.027113745502388e-05,
+      "loss": 0.7558,
+      "step": 19435
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.28294190764427185,
+      "learning_rate": 4.0201001865149206e-05,
+      "loss": 0.8359,
+      "step": 19440
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.31937623023986816,
+      "learning_rate": 4.013091795032937e-05,
+      "loss": 0.769,
+      "step": 19445
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.26449480652809143,
+      "learning_rate": 4.006088574354819e-05,
+      "loss": 0.7822,
+      "step": 19450
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.3096768856048584,
+      "learning_rate": 3.999090527776509e-05,
+      "loss": 0.7575,
+      "step": 19455
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.2941647171974182,
+      "learning_rate": 3.9920976585915296e-05,
+      "loss": 0.8384,
+      "step": 19460
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.3138192594051361,
+      "learning_rate": 3.985109970090956e-05,
+      "loss": 0.7938,
+      "step": 19465
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.24960915744304657,
+      "learning_rate": 3.9781274655634205e-05,
+      "loss": 0.7544,
+      "step": 19470
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.2531121075153351,
+      "learning_rate": 3.971150148295123e-05,
+      "loss": 0.9053,
+      "step": 19475
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.25435906648635864,
+      "learning_rate": 3.964178021569825e-05,
+      "loss": 0.7899,
+      "step": 19480
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.27135327458381653,
+      "learning_rate": 3.957211088668838e-05,
+      "loss": 0.7274,
+      "step": 19485
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.2866491675376892,
+      "learning_rate": 3.950249352871034e-05,
+      "loss": 0.7441,
+      "step": 19490
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.2988530397415161,
+      "learning_rate": 3.94329281745283e-05,
+      "loss": 0.7978,
+      "step": 19495
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.27263930439949036,
+      "learning_rate": 3.9363414856882126e-05,
+      "loss": 0.7804,
+      "step": 19500
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.2896611988544464,
+      "learning_rate": 3.929395360848704e-05,
+      "loss": 0.7589,
+      "step": 19505
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.28115859627723694,
+      "learning_rate": 3.9224544462033844e-05,
+      "loss": 0.7887,
+      "step": 19510
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.31314992904663086,
+      "learning_rate": 3.915518745018873e-05,
+      "loss": 0.7634,
+      "step": 19515
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.28734245896339417,
+      "learning_rate": 3.9085882605593485e-05,
+      "loss": 0.8887,
+      "step": 19520
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.26335301995277405,
+      "learning_rate": 3.901662996086519e-05,
+      "loss": 0.8303,
+      "step": 19525
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.2777862846851349,
+      "learning_rate": 3.8947429548596506e-05,
+      "loss": 0.7661,
+      "step": 19530
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.2924749553203583,
+      "learning_rate": 3.8878281401355366e-05,
+      "loss": 0.7203,
+      "step": 19535
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.265706866979599,
+      "learning_rate": 3.880918555168528e-05,
+      "loss": 0.8163,
+      "step": 19540
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.29131948947906494,
+      "learning_rate": 3.8740142032105e-05,
+      "loss": 0.8113,
+      "step": 19545
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.26033177971839905,
+      "learning_rate": 3.8671150875108715e-05,
+      "loss": 0.8122,
+      "step": 19550
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.2742202579975128,
+      "learning_rate": 3.8602212113165886e-05,
+      "loss": 0.8273,
+      "step": 19555
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.2769484221935272,
+      "learning_rate": 3.85333257787215e-05,
+      "loss": 0.7594,
+      "step": 19560
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.31669124960899353,
+      "learning_rate": 3.846449190419569e-05,
+      "loss": 0.8816,
+      "step": 19565
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.28259631991386414,
+      "learning_rate": 3.839571052198398e-05,
+      "loss": 0.7435,
+      "step": 19570
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.2604540288448334,
+      "learning_rate": 3.832698166445715e-05,
+      "loss": 0.7484,
+      "step": 19575
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.3139670491218567,
+      "learning_rate": 3.82583053639613e-05,
+      "loss": 0.7701,
+      "step": 19580
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.30153560638427734,
+      "learning_rate": 3.818968165281777e-05,
+      "loss": 0.8798,
+      "step": 19585
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.27177488803863525,
+      "learning_rate": 3.8121110563323115e-05,
+      "loss": 0.7226,
+      "step": 19590
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.27674952149391174,
+      "learning_rate": 3.8052592127749265e-05,
+      "loss": 0.8741,
+      "step": 19595
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.2683180570602417,
+      "learning_rate": 3.798412637834321e-05,
+      "loss": 0.767,
+      "step": 19600
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.25967836380004883,
+      "learning_rate": 3.7915713347327225e-05,
+      "loss": 0.7556,
+      "step": 19605
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.2514658570289612,
+      "learning_rate": 3.784735306689869e-05,
+      "loss": 0.7164,
+      "step": 19610
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.2914251983165741,
+      "learning_rate": 3.777904556923031e-05,
+      "loss": 0.9218,
+      "step": 19615
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.24623388051986694,
+      "learning_rate": 3.7710790886469837e-05,
+      "loss": 0.8447,
+      "step": 19620
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.295613169670105,
+      "learning_rate": 3.764258905074018e-05,
+      "loss": 0.735,
+      "step": 19625
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.3049282133579254,
+      "learning_rate": 3.7574440094139364e-05,
+      "loss": 0.7163,
+      "step": 19630
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.3349364101886749,
+      "learning_rate": 3.75063440487406e-05,
+      "loss": 0.8382,
+      "step": 19635
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.2821791172027588,
+      "learning_rate": 3.743830094659212e-05,
+      "loss": 0.7424,
+      "step": 19640
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.2976232171058655,
+      "learning_rate": 3.737031081971729e-05,
+      "loss": 0.7901,
+      "step": 19645
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.2489924281835556,
+      "learning_rate": 3.73023737001145e-05,
+      "loss": 0.7586,
+      "step": 19650
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.2726670503616333,
+      "learning_rate": 3.723448961975722e-05,
+      "loss": 0.8153,
+      "step": 19655
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.29014068841934204,
+      "learning_rate": 3.7166658610593955e-05,
+      "loss": 0.8317,
+      "step": 19660
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.28912895917892456,
+      "learning_rate": 3.709888070454822e-05,
+      "loss": 0.8345,
+      "step": 19665
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.2740095257759094,
+      "learning_rate": 3.70311559335185e-05,
+      "loss": 0.7405,
+      "step": 19670
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.3410658836364746,
+      "learning_rate": 3.696348432937845e-05,
+      "loss": 0.7129,
+      "step": 19675
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.2865118682384491,
+      "learning_rate": 3.689586592397647e-05,
+      "loss": 0.7907,
+      "step": 19680
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.26647239923477173,
+      "learning_rate": 3.6828300749136085e-05,
+      "loss": 0.8186,
+      "step": 19685
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.30425769090652466,
+      "learning_rate": 3.6760788836655624e-05,
+      "loss": 0.7776,
+      "step": 19690
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.28291577100753784,
+      "learning_rate": 3.669333021830854e-05,
+      "loss": 0.7774,
+      "step": 19695
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.294132262468338,
+      "learning_rate": 3.662592492584306e-05,
+      "loss": 0.7293,
+      "step": 19700
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.28489431738853455,
+      "learning_rate": 3.655857299098233e-05,
+      "loss": 0.7946,
+      "step": 19705
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.2683461308479309,
+      "learning_rate": 3.64912744454244e-05,
+      "loss": 0.7626,
+      "step": 19710
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.24661622941493988,
+      "learning_rate": 3.6424029320842265e-05,
+      "loss": 0.7492,
+      "step": 19715
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.266984760761261,
+      "learning_rate": 3.635683764888367e-05,
+      "loss": 0.7598,
+      "step": 19720
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.2817406952381134,
+      "learning_rate": 3.628969946117129e-05,
+      "loss": 0.852,
+      "step": 19725
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.2606806755065918,
+      "learning_rate": 3.6222614789302454e-05,
+      "loss": 0.863,
+      "step": 19730
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.2557438015937805,
+      "learning_rate": 3.615558366484958e-05,
+      "loss": 0.7349,
+      "step": 19735
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.27822768688201904,
+      "learning_rate": 3.608860611935969e-05,
+      "loss": 0.8482,
+      "step": 19740
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.2577543556690216,
+      "learning_rate": 3.602168218435464e-05,
+      "loss": 0.8088,
+      "step": 19745
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.3152078092098236,
+      "learning_rate": 3.595481189133101e-05,
+      "loss": 0.7895,
+      "step": 19750
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.2776154577732086,
+      "learning_rate": 3.588799527176026e-05,
+      "loss": 0.8115,
+      "step": 19755
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.2907949984073639,
+      "learning_rate": 3.5821232357088493e-05,
+      "loss": 0.8447,
+      "step": 19760
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.27817782759666443,
+      "learning_rate": 3.575452317873653e-05,
+      "loss": 0.7258,
+      "step": 19765
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.2644236087799072,
+      "learning_rate": 3.56878677680999e-05,
+      "loss": 0.7573,
+      "step": 19770
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.28008079528808594,
+      "learning_rate": 3.5621266156548925e-05,
+      "loss": 0.8521,
+      "step": 19775
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.2851678133010864,
+      "learning_rate": 3.55547183754285e-05,
+      "loss": 0.8946,
+      "step": 19780
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.29239290952682495,
+      "learning_rate": 3.5488224456058215e-05,
+      "loss": 0.7888,
+      "step": 19785
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.2597346305847168,
+      "learning_rate": 3.5421784429732316e-05,
+      "loss": 0.7592,
+      "step": 19790
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.2777891159057617,
+      "learning_rate": 3.535539832771971e-05,
+      "loss": 0.7478,
+      "step": 19795
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.2665190100669861,
+      "learning_rate": 3.528906618126385e-05,
+      "loss": 0.8196,
+      "step": 19800
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.2569584548473358,
+      "learning_rate": 3.52227880215829e-05,
+      "loss": 0.8045,
+      "step": 19805
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.2824585437774658,
+      "learning_rate": 3.5156563879869494e-05,
+      "loss": 0.7438,
+      "step": 19810
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.2764226496219635,
+      "learning_rate": 3.509039378729099e-05,
+      "loss": 0.7557,
+      "step": 19815
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.2752508521080017,
+      "learning_rate": 3.5024277774989203e-05,
+      "loss": 0.9046,
+      "step": 19820
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.28766295313835144,
+      "learning_rate": 3.4958215874080526e-05,
+      "loss": 0.7422,
+      "step": 19825
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.25225573778152466,
+      "learning_rate": 3.4892208115655837e-05,
+      "loss": 0.8565,
+      "step": 19830
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.3149223327636719,
+      "learning_rate": 3.482625453078065e-05,
+      "loss": 0.7516,
+      "step": 19835
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.30531787872314453,
+      "learning_rate": 3.4760355150494895e-05,
+      "loss": 0.7884,
+      "step": 19840
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.2835243046283722,
+      "learning_rate": 3.4694510005812996e-05,
+      "loss": 0.7924,
+      "step": 19845
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.2607024013996124,
+      "learning_rate": 3.462871912772382e-05,
+      "loss": 0.7911,
+      "step": 19850
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.29384374618530273,
+      "learning_rate": 3.4562982547190824e-05,
+      "loss": 0.8039,
+      "step": 19855
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.28320446610450745,
+      "learning_rate": 3.449730029515179e-05,
+      "loss": 0.7786,
+      "step": 19860
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.281463623046875,
+      "learning_rate": 3.4431672402518955e-05,
+      "loss": 0.7672,
+      "step": 19865
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.2654038369655609,
+      "learning_rate": 3.436609890017901e-05,
+      "loss": 0.8256,
+      "step": 19870
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.2744678556919098,
+      "learning_rate": 3.430057981899298e-05,
+      "loss": 0.7221,
+      "step": 19875
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.28920984268188477,
+      "learning_rate": 3.4235115189796375e-05,
+      "loss": 0.7532,
+      "step": 19880
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.3083685040473938,
+      "learning_rate": 3.416970504339897e-05,
+      "loss": 0.7365,
+      "step": 19885
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.25423839688301086,
+      "learning_rate": 3.410434941058495e-05,
+      "loss": 0.8097,
+      "step": 19890
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.25034499168395996,
+      "learning_rate": 3.4039048322112917e-05,
+      "loss": 0.7832,
+      "step": 19895
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.2709154188632965,
+      "learning_rate": 3.3973801808715676e-05,
+      "loss": 0.7133,
+      "step": 19900
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.2830277383327484,
+      "learning_rate": 3.390860990110042e-05,
+      "loss": 0.8309,
+      "step": 19905
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.26153650879859924,
+      "learning_rate": 3.384347262994858e-05,
+      "loss": 0.7983,
+      "step": 19910
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.23845182359218597,
+      "learning_rate": 3.377839002591599e-05,
+      "loss": 0.84,
+      "step": 19915
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.3157477080821991,
+      "learning_rate": 3.371336211963268e-05,
+      "loss": 0.7391,
+      "step": 19920
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.25639525055885315,
+      "learning_rate": 3.364838894170289e-05,
+      "loss": 0.7765,
+      "step": 19925
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.27470090985298157,
+      "learning_rate": 3.358347052270515e-05,
+      "loss": 0.7211,
+      "step": 19930
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.3165142238140106,
+      "learning_rate": 3.351860689319234e-05,
+      "loss": 0.7079,
+      "step": 19935
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.2669844329357147,
+      "learning_rate": 3.345379808369132e-05,
+      "loss": 0.7023,
+      "step": 19940
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.29056012630462646,
+      "learning_rate": 3.338904412470328e-05,
+      "loss": 0.8274,
+      "step": 19945
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.304496169090271,
+      "learning_rate": 3.332434504670358e-05,
+      "loss": 0.821,
+      "step": 19950
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.29382607340812683,
+      "learning_rate": 3.3259700880141815e-05,
+      "loss": 0.7403,
+      "step": 19955
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.27669087052345276,
+      "learning_rate": 3.319511165544165e-05,
+      "loss": 0.6459,
+      "step": 19960
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.2903481125831604,
+      "learning_rate": 3.313057740300089e-05,
+      "loss": 0.8414,
+      "step": 19965
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.2781134247779846,
+      "learning_rate": 3.30660981531915e-05,
+      "loss": 0.7587,
+      "step": 19970
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.24643659591674805,
+      "learning_rate": 3.3001673936359604e-05,
+      "loss": 0.7535,
+      "step": 19975
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.306083083152771,
+      "learning_rate": 3.293730478282534e-05,
+      "loss": 0.7106,
+      "step": 19980
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.26486045122146606,
+      "learning_rate": 3.2872990722882984e-05,
+      "loss": 0.8252,
+      "step": 19985
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.2603921890258789,
+      "learning_rate": 3.280873178680082e-05,
+      "loss": 0.7914,
+      "step": 19990
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.31112536787986755,
+      "learning_rate": 3.274452800482133e-05,
+      "loss": 0.7106,
+      "step": 19995
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.26250556111335754,
+      "learning_rate": 3.2680379407160886e-05,
+      "loss": 0.7859,
+      "step": 20000
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.28084659576416016,
+      "learning_rate": 3.261628602400995e-05,
+      "loss": 0.7564,
+      "step": 20005
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.2960032820701599,
+      "learning_rate": 3.2552247885533005e-05,
+      "loss": 0.8016,
+      "step": 20010
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.3024326264858246,
+      "learning_rate": 3.248826502186854e-05,
+      "loss": 0.8589,
+      "step": 20015
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.28151461482048035,
+      "learning_rate": 3.242433746312899e-05,
+      "loss": 0.8336,
+      "step": 20020
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.2718256711959839,
+      "learning_rate": 3.23604652394008e-05,
+      "loss": 0.6869,
+      "step": 20025
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.31024929881095886,
+      "learning_rate": 3.229664838074431e-05,
+      "loss": 0.9043,
+      "step": 20030
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.26419252157211304,
+      "learning_rate": 3.223288691719394e-05,
+      "loss": 0.7494,
+      "step": 20035
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.27617117762565613,
+      "learning_rate": 3.216918087875792e-05,
+      "loss": 0.7913,
+      "step": 20040
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.2638927102088928,
+      "learning_rate": 3.21055302954184e-05,
+      "loss": 0.7151,
+      "step": 20045
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.2976774573326111,
+      "learning_rate": 3.204193519713146e-05,
+      "loss": 0.7635,
+      "step": 20050
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.256143182516098,
+      "learning_rate": 3.197839561382711e-05,
+      "loss": 0.8376,
+      "step": 20055
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.2871249318122864,
+      "learning_rate": 3.1914911575409165e-05,
+      "loss": 0.707,
+      "step": 20060
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.269939661026001,
+      "learning_rate": 3.1851483111755276e-05,
+      "loss": 0.8975,
+      "step": 20065
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.2935774326324463,
+      "learning_rate": 3.178811025271705e-05,
+      "loss": 0.787,
+      "step": 20070
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.2772122621536255,
+      "learning_rate": 3.1724793028119846e-05,
+      "loss": 0.6818,
+      "step": 20075
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.3376406133174896,
+      "learning_rate": 3.1661531467762866e-05,
+      "loss": 0.6406,
+      "step": 20080
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.24912644922733307,
+      "learning_rate": 3.159832560141904e-05,
+      "loss": 0.8038,
+      "step": 20085
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.26904556155204773,
+      "learning_rate": 3.15351754588352e-05,
+      "loss": 0.8043,
+      "step": 20090
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.27094265818595886,
+      "learning_rate": 3.147208106973189e-05,
+      "loss": 0.8588,
+      "step": 20095
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.26303452253341675,
+      "learning_rate": 3.1409042463803406e-05,
+      "loss": 0.7485,
+      "step": 20100
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.29558247327804565,
+      "learning_rate": 3.1346059670717783e-05,
+      "loss": 0.8514,
+      "step": 20105
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.2617008090019226,
+      "learning_rate": 3.12831327201169e-05,
+      "loss": 0.7401,
+      "step": 20110
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.29769590497016907,
+      "learning_rate": 3.1220261641616226e-05,
+      "loss": 0.6958,
+      "step": 20115
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.2647980749607086,
+      "learning_rate": 3.1157446464804966e-05,
+      "loss": 0.824,
+      "step": 20120
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.2724788784980774,
+      "learning_rate": 3.1094687219246025e-05,
+      "loss": 0.8486,
+      "step": 20125
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.2676187753677368,
+      "learning_rate": 3.103198393447601e-05,
+      "loss": 0.8293,
+      "step": 20130
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.254800945520401,
+      "learning_rate": 3.096933664000518e-05,
+      "loss": 0.8403,
+      "step": 20135
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.2803385853767395,
+      "learning_rate": 3.090674536531741e-05,
+      "loss": 0.7535,
+      "step": 20140
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.3000488877296448,
+      "learning_rate": 3.084421013987023e-05,
+      "loss": 0.7952,
+      "step": 20145
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.24971972405910492,
+      "learning_rate": 3.07817309930948e-05,
+      "loss": 0.7537,
+      "step": 20150
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.29271334409713745,
+      "learning_rate": 3.0719307954395886e-05,
+      "loss": 0.8457,
+      "step": 20155
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.27046826481819153,
+      "learning_rate": 3.0656941053151846e-05,
+      "loss": 0.7313,
+      "step": 20160
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.3202589154243469,
+      "learning_rate": 3.059463031871456e-05,
+      "loss": 0.7813,
+      "step": 20165
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.2895718514919281,
+      "learning_rate": 3.05323757804096e-05,
+      "loss": 0.768,
+      "step": 20170
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.28640422224998474,
+      "learning_rate": 3.0470177467536006e-05,
+      "loss": 0.7901,
+      "step": 20175
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.2655580937862396,
+      "learning_rate": 3.040803540936637e-05,
+      "loss": 0.8789,
+      "step": 20180
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.26338696479797363,
+      "learning_rate": 3.034594963514674e-05,
+      "loss": 0.8176,
+      "step": 20185
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.27758362889289856,
+      "learning_rate": 3.028392017409685e-05,
+      "loss": 0.7612,
+      "step": 20190
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.31428468227386475,
+      "learning_rate": 3.0221947055409766e-05,
+      "loss": 0.7512,
+      "step": 20195
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.3056136965751648,
+      "learning_rate": 3.0160030308252113e-05,
+      "loss": 0.8479,
+      "step": 20200
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.2786805331707001,
+      "learning_rate": 3.0098169961763913e-05,
+      "loss": 0.8396,
+      "step": 20205
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.2735007405281067,
+      "learning_rate": 3.003636604505879e-05,
+      "loss": 0.7215,
+      "step": 20210
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.2933249771595001,
+      "learning_rate": 2.9974618587223665e-05,
+      "loss": 0.8251,
+      "step": 20215
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.24182629585266113,
+      "learning_rate": 2.991292761731897e-05,
+      "loss": 0.7686,
+      "step": 20220
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.31849405169487,
+      "learning_rate": 2.9851293164378488e-05,
+      "loss": 0.8349,
+      "step": 20225
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.2779251039028168,
+      "learning_rate": 2.9789715257409475e-05,
+      "loss": 0.7816,
+      "step": 20230
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.2377801239490509,
+      "learning_rate": 2.9728193925392502e-05,
+      "loss": 0.8013,
+      "step": 20235
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.29114991426467896,
+      "learning_rate": 2.966672919728159e-05,
+      "loss": 0.7782,
+      "step": 20240
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.2626573443412781,
+      "learning_rate": 2.960532110200402e-05,
+      "loss": 0.7895,
+      "step": 20245
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.26689302921295166,
+      "learning_rate": 2.9543969668460556e-05,
+      "loss": 0.7706,
+      "step": 20250
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.29479920864105225,
+      "learning_rate": 2.948267492552518e-05,
+      "loss": 0.8125,
+      "step": 20255
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.2591875195503235,
+      "learning_rate": 2.9421436902045255e-05,
+      "loss": 0.8254,
+      "step": 20260
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.23983192443847656,
+      "learning_rate": 2.9360255626841374e-05,
+      "loss": 0.731,
+      "step": 20265
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.27949729561805725,
+      "learning_rate": 2.9299131128707537e-05,
+      "loss": 0.7235,
+      "step": 20270
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.29695627093315125,
+      "learning_rate": 2.923806343641097e-05,
+      "loss": 0.8012,
+      "step": 20275
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.26276320219039917,
+      "learning_rate": 2.917705257869211e-05,
+      "loss": 0.7519,
+      "step": 20280
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.2703682482242584,
+      "learning_rate": 2.911609858426468e-05,
+      "loss": 0.8297,
+      "step": 20285
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.28486260771751404,
+      "learning_rate": 2.9055201481815727e-05,
+      "loss": 0.8341,
+      "step": 20290
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.266369104385376,
+      "learning_rate": 2.8994361300005407e-05,
+      "loss": 0.6976,
+      "step": 20295
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.2816215753555298,
+      "learning_rate": 2.8933578067467123e-05,
+      "loss": 0.846,
+      "step": 20300
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.273666650056839,
+      "learning_rate": 2.8872851812807502e-05,
+      "loss": 0.7309,
+      "step": 20305
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.30615535378456116,
+      "learning_rate": 2.8812182564606334e-05,
+      "loss": 0.7782,
+      "step": 20310
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.2994655966758728,
+      "learning_rate": 2.8751570351416564e-05,
+      "loss": 0.6678,
+      "step": 20315
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.2747027277946472,
+      "learning_rate": 2.8691015201764323e-05,
+      "loss": 0.7504,
+      "step": 20320
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.28770124912261963,
+      "learning_rate": 2.863051714414883e-05,
+      "loss": 0.7577,
+      "step": 20325
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.27923446893692017,
+      "learning_rate": 2.8570076207042574e-05,
+      "loss": 0.7865,
+      "step": 20330
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.26224157214164734,
+      "learning_rate": 2.8509692418891024e-05,
+      "loss": 0.7961,
+      "step": 20335
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.2644881308078766,
+      "learning_rate": 2.8449365808112773e-05,
+      "loss": 0.7063,
+      "step": 20340
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.2863686680793762,
+      "learning_rate": 2.8389096403099515e-05,
+      "loss": 0.8939,
+      "step": 20345
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.27196264266967773,
+      "learning_rate": 2.832888423221611e-05,
+      "loss": 0.803,
+      "step": 20350
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.2805080711841583,
+      "learning_rate": 2.826872932380034e-05,
+      "loss": 0.8347,
+      "step": 20355
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.27300190925598145,
+      "learning_rate": 2.820863170616312e-05,
+      "loss": 0.7662,
+      "step": 20360
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.2929331660270691,
+      "learning_rate": 2.81485914075884e-05,
+      "loss": 0.8332,
+      "step": 20365
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.3110920190811157,
+      "learning_rate": 2.808860845633312e-05,
+      "loss": 0.7865,
+      "step": 20370
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.2958769202232361,
+      "learning_rate": 2.8028682880627255e-05,
+      "loss": 0.7955,
+      "step": 20375
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.2582022547721863,
+      "learning_rate": 2.7968814708673758e-05,
+      "loss": 0.8322,
+      "step": 20380
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.27582964301109314,
+      "learning_rate": 2.7909003968648557e-05,
+      "loss": 0.8729,
+      "step": 20385
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.29320433735847473,
+      "learning_rate": 2.7849250688700637e-05,
+      "loss": 0.7828,
+      "step": 20390
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.2722160220146179,
+      "learning_rate": 2.7789554896951865e-05,
+      "loss": 0.7481,
+      "step": 20395
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.3223639130592346,
+      "learning_rate": 2.7729916621497022e-05,
+      "loss": 0.9109,
+      "step": 20400
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.2686358690261841,
+      "learning_rate": 2.7670335890403837e-05,
+      "loss": 0.7711,
+      "step": 20405
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.2783448398113251,
+      "learning_rate": 2.7610812731713084e-05,
+      "loss": 0.8384,
+      "step": 20410
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.3094173073768616,
+      "learning_rate": 2.7551347173438267e-05,
+      "loss": 0.7424,
+      "step": 20415
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.29645413160324097,
+      "learning_rate": 2.7491939243565862e-05,
+      "loss": 0.813,
+      "step": 20420
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.31017208099365234,
+      "learning_rate": 2.7432588970055182e-05,
+      "loss": 0.8065,
+      "step": 20425
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.29746440052986145,
+      "learning_rate": 2.7373296380838505e-05,
+      "loss": 0.7532,
+      "step": 20430
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.2688598930835724,
+      "learning_rate": 2.7314061503820865e-05,
+      "loss": 0.7811,
+      "step": 20435
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.30386409163475037,
+      "learning_rate": 2.7254884366880146e-05,
+      "loss": 0.7961,
+      "step": 20440
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.28410372138023376,
+      "learning_rate": 2.7195764997867085e-05,
+      "loss": 0.9115,
+      "step": 20445
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.2933676242828369,
+      "learning_rate": 2.7136703424605228e-05,
+      "loss": 0.8635,
+      "step": 20450
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.3018134832382202,
+      "learning_rate": 2.7077699674890903e-05,
+      "loss": 0.7459,
+      "step": 20455
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.2924741208553314,
+      "learning_rate": 2.7018753776493213e-05,
+      "loss": 0.8325,
+      "step": 20460
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.2926197946071625,
+      "learning_rate": 2.6959865757154054e-05,
+      "loss": 0.722,
+      "step": 20465
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.285312294960022,
+      "learning_rate": 2.6901035644588154e-05,
+      "loss": 0.8077,
+      "step": 20470
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.2824037969112396,
+      "learning_rate": 2.684226346648285e-05,
+      "loss": 0.6966,
+      "step": 20475
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.29435649514198303,
+      "learning_rate": 2.6783549250498315e-05,
+      "loss": 0.6975,
+      "step": 20480
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.26435035467147827,
+      "learning_rate": 2.6724893024267348e-05,
+      "loss": 0.8694,
+      "step": 20485
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.29340675473213196,
+      "learning_rate": 2.6666294815395612e-05,
+      "loss": 0.7117,
+      "step": 20490
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.30014726519584656,
+      "learning_rate": 2.6607754651461317e-05,
+      "loss": 0.7437,
+      "step": 20495
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.31490999460220337,
+      "learning_rate": 2.6549272560015406e-05,
+      "loss": 0.7764,
+      "step": 20500
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.26503077149391174,
+      "learning_rate": 2.6490848568581462e-05,
+      "loss": 0.7501,
+      "step": 20505
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.29323646426200867,
+      "learning_rate": 2.6432482704655845e-05,
+      "loss": 0.8158,
+      "step": 20510
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.2801787555217743,
+      "learning_rate": 2.637417499570741e-05,
+      "loss": 0.7489,
+      "step": 20515
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.2741035223007202,
+      "learning_rate": 2.631592546917774e-05,
+      "loss": 0.8378,
+      "step": 20520
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.300493448972702,
+      "learning_rate": 2.6257734152480893e-05,
+      "loss": 0.7904,
+      "step": 20525
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.2564448118209839,
+      "learning_rate": 2.619960107300374e-05,
+      "loss": 0.7423,
+      "step": 20530
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.26825183629989624,
+      "learning_rate": 2.6141526258105615e-05,
+      "loss": 0.7046,
+      "step": 20535
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.2867426872253418,
+      "learning_rate": 2.608350973511844e-05,
+      "loss": 0.8206,
+      "step": 20540
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.2587963938713074,
+      "learning_rate": 2.6025551531346688e-05,
+      "loss": 0.7434,
+      "step": 20545
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.3124673664569855,
+      "learning_rate": 2.5967651674067493e-05,
+      "loss": 0.7841,
+      "step": 20550
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.28737810254096985,
+      "learning_rate": 2.59098101905304e-05,
+      "loss": 0.7924,
+      "step": 20555
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.2854047417640686,
+      "learning_rate": 2.585202710795754e-05,
+      "loss": 0.7533,
+      "step": 20560
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.24524690210819244,
+      "learning_rate": 2.5794302453543524e-05,
+      "loss": 0.8201,
+      "step": 20565
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.30324557423591614,
+      "learning_rate": 2.573663625445554e-05,
+      "loss": 0.8067,
+      "step": 20570
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.29187995195388794,
+      "learning_rate": 2.567902853783319e-05,
+      "loss": 0.8313,
+      "step": 20575
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.26502034068107605,
+      "learning_rate": 2.562147933078859e-05,
+      "loss": 0.8097,
+      "step": 20580
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.32918286323547363,
+      "learning_rate": 2.556398866040631e-05,
+      "loss": 0.8703,
+      "step": 20585
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.3019461929798126,
+      "learning_rate": 2.5506556553743334e-05,
+      "loss": 0.8317,
+      "step": 20590
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.28017765283584595,
+      "learning_rate": 2.5449183037829163e-05,
+      "loss": 0.8315,
+      "step": 20595
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.31728842854499817,
+      "learning_rate": 2.5391868139665612e-05,
+      "loss": 0.8539,
+      "step": 20600
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.26865053176879883,
+      "learning_rate": 2.533461188622707e-05,
+      "loss": 0.7489,
+      "step": 20605
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.24447381496429443,
+      "learning_rate": 2.527741430446018e-05,
+      "loss": 0.7674,
+      "step": 20610
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.3278535306453705,
+      "learning_rate": 2.5220275421284015e-05,
+      "loss": 0.7342,
+      "step": 20615
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.30096837878227234,
+      "learning_rate": 2.5163195263590026e-05,
+      "loss": 0.7506,
+      "step": 20620
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.28471359610557556,
+      "learning_rate": 2.5106173858242073e-05,
+      "loss": 0.7999,
+      "step": 20625
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.2582080364227295,
+      "learning_rate": 2.5049211232076295e-05,
+      "loss": 0.8253,
+      "step": 20630
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.2530338168144226,
+      "learning_rate": 2.499230741190121e-05,
+      "loss": 0.8581,
+      "step": 20635
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.25447461009025574,
+      "learning_rate": 2.4935462424497576e-05,
+      "loss": 0.7793,
+      "step": 20640
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.24925333261489868,
+      "learning_rate": 2.487867629661865e-05,
+      "loss": 0.7587,
+      "step": 20645
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.2856590151786804,
+      "learning_rate": 2.4821949054989797e-05,
+      "loss": 0.7551,
+      "step": 20650
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.2937513589859009,
+      "learning_rate": 2.476528072630875e-05,
+      "loss": 0.7448,
+      "step": 20655
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.2708790600299835,
+      "learning_rate": 2.470867133724552e-05,
+      "loss": 0.7374,
+      "step": 20660
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.2769574820995331,
+      "learning_rate": 2.4652120914442352e-05,
+      "loss": 0.8323,
+      "step": 20665
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.2852955162525177,
+      "learning_rate": 2.459562948451375e-05,
+      "loss": 0.7135,
+      "step": 20670
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.2564711272716522,
+      "learning_rate": 2.4539197074046467e-05,
+      "loss": 0.7752,
+      "step": 20675
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.2757887840270996,
+      "learning_rate": 2.448282370959944e-05,
+      "loss": 0.7853,
+      "step": 20680
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.2913254201412201,
+      "learning_rate": 2.44265094177039e-05,
+      "loss": 0.8027,
+      "step": 20685
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.26618868112564087,
+      "learning_rate": 2.437025422486319e-05,
+      "loss": 0.7059,
+      "step": 20690
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.24177007377147675,
+      "learning_rate": 2.4314058157552885e-05,
+      "loss": 0.7975,
+      "step": 20695
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.26704952120780945,
+      "learning_rate": 2.4257921242220663e-05,
+      "loss": 0.751,
+      "step": 20700
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.28091442584991455,
+      "learning_rate": 2.4201843505286507e-05,
+      "loss": 0.8622,
+      "step": 20705
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.2535339891910553,
+      "learning_rate": 2.4145824973142435e-05,
+      "loss": 0.6663,
+      "step": 20710
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.26714247465133667,
+      "learning_rate": 2.4089865672152613e-05,
+      "loss": 0.759,
+      "step": 20715
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.2989547550678253,
+      "learning_rate": 2.4033965628653323e-05,
+      "loss": 0.8327,
+      "step": 20720
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.270155131816864,
+      "learning_rate": 2.3978124868953037e-05,
+      "loss": 0.7935,
+      "step": 20725
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.29163941740989685,
+      "learning_rate": 2.3922343419332247e-05,
+      "loss": 0.7799,
+      "step": 20730
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.27277231216430664,
+      "learning_rate": 2.386662130604359e-05,
+      "loss": 0.7378,
+      "step": 20735
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.308982253074646,
+      "learning_rate": 2.3810958555311647e-05,
+      "loss": 0.7067,
+      "step": 20740
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.24645307660102844,
+      "learning_rate": 2.375535519333324e-05,
+      "loss": 0.6404,
+      "step": 20745
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.30211979150772095,
+      "learning_rate": 2.3699811246277133e-05,
+      "loss": 0.7035,
+      "step": 20750
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.27554258704185486,
+      "learning_rate": 2.3644326740284147e-05,
+      "loss": 0.8621,
+      "step": 20755
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.24029062688350677,
+      "learning_rate": 2.358890170146711e-05,
+      "loss": 0.7073,
+      "step": 20760
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.2912255525588989,
+      "learning_rate": 2.353353615591096e-05,
+      "loss": 0.7371,
+      "step": 20765
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.29591822624206543,
+      "learning_rate": 2.3478230129672498e-05,
+      "loss": 0.6795,
+      "step": 20770
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.274454802274704,
+      "learning_rate": 2.3422983648780606e-05,
+      "loss": 0.8204,
+      "step": 20775
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.26619449257850647,
+      "learning_rate": 2.336779673923607e-05,
+      "loss": 0.8263,
+      "step": 20780
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.2655404806137085,
+      "learning_rate": 2.3312669427011748e-05,
+      "loss": 0.8028,
+      "step": 20785
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.2552018165588379,
+      "learning_rate": 2.3257601738052352e-05,
+      "loss": 0.8012,
+      "step": 20790
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.2989104390144348,
+      "learning_rate": 2.3202593698274546e-05,
+      "loss": 0.8466,
+      "step": 20795
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.25715914368629456,
+      "learning_rate": 2.3147645333566977e-05,
+      "loss": 0.7009,
+      "step": 20800
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.29836973547935486,
+      "learning_rate": 2.309275666979014e-05,
+      "loss": 0.7897,
+      "step": 20805
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.28271016478538513,
+      "learning_rate": 2.3037927732776472e-05,
+      "loss": 0.7738,
+      "step": 20810
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.31232964992523193,
+      "learning_rate": 2.29831585483303e-05,
+      "loss": 0.7341,
+      "step": 20815
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.29633277654647827,
+      "learning_rate": 2.292844914222777e-05,
+      "loss": 0.8116,
+      "step": 20820
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.2850092649459839,
+      "learning_rate": 2.287379954021703e-05,
+      "loss": 0.7608,
+      "step": 20825
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.2637297809123993,
+      "learning_rate": 2.2819209768017965e-05,
+      "loss": 0.7451,
+      "step": 20830
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.239261195063591,
+      "learning_rate": 2.276467985132232e-05,
+      "loss": 0.8302,
+      "step": 20835
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.3218871057033539,
+      "learning_rate": 2.271020981579365e-05,
+      "loss": 0.7533,
+      "step": 20840
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.26522281765937805,
+      "learning_rate": 2.265579968706745e-05,
+      "loss": 0.6176,
+      "step": 20845
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.3309444189071655,
+      "learning_rate": 2.2601449490750867e-05,
+      "loss": 0.68,
+      "step": 20850
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.295600026845932,
+      "learning_rate": 2.254715925242294e-05,
+      "loss": 0.7778,
+      "step": 20855
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.27975109219551086,
+      "learning_rate": 2.249292899763442e-05,
+      "loss": 0.8367,
+      "step": 20860
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.301899790763855,
+      "learning_rate": 2.2438758751907933e-05,
+      "loss": 0.7383,
+      "step": 20865
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.2855704724788666,
+      "learning_rate": 2.2384648540737736e-05,
+      "loss": 0.8029,
+      "step": 20870
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.31557753682136536,
+      "learning_rate": 2.2330598389589915e-05,
+      "loss": 0.7673,
+      "step": 20875
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.2771788239479065,
+      "learning_rate": 2.2276608323902266e-05,
+      "loss": 0.7175,
+      "step": 20880
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.305113285779953,
+      "learning_rate": 2.22226783690843e-05,
+      "loss": 0.7742,
+      "step": 20885
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.24871449172496796,
+      "learning_rate": 2.216880855051725e-05,
+      "loss": 0.678,
+      "step": 20890
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.2921499013900757,
+      "learning_rate": 2.2114998893554042e-05,
+      "loss": 0.7413,
+      "step": 20895
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.3036426603794098,
+      "learning_rate": 2.2061249423519244e-05,
+      "loss": 0.7398,
+      "step": 20900
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.27135762572288513,
+      "learning_rate": 2.200756016570922e-05,
+      "loss": 0.7812,
+      "step": 20905
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.2812870442867279,
+      "learning_rate": 2.1953931145391872e-05,
+      "loss": 0.7961,
+      "step": 20910
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.3006354570388794,
+      "learning_rate": 2.1900362387806812e-05,
+      "loss": 0.7816,
+      "step": 20915
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.27853453159332275,
+      "learning_rate": 2.1846853918165223e-05,
+      "loss": 0.7039,
+      "step": 20920
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.29205799102783203,
+      "learning_rate": 2.1793405761650062e-05,
+      "loss": 0.8397,
+      "step": 20925
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.27953484654426575,
+      "learning_rate": 2.1740017943415754e-05,
+      "loss": 0.8072,
+      "step": 20930
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.3323310315608978,
+      "learning_rate": 2.1686690488588387e-05,
+      "loss": 0.7661,
+      "step": 20935
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.33003172278404236,
+      "learning_rate": 2.16334234222656e-05,
+      "loss": 0.7778,
+      "step": 20940
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.24858631193637848,
+      "learning_rate": 2.1580216769516738e-05,
+      "loss": 0.914,
+      "step": 20945
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.2743765115737915,
+      "learning_rate": 2.152707055538251e-05,
+      "loss": 0.9042,
+      "step": 20950
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.283892959356308,
+      "learning_rate": 2.1473984804875332e-05,
+      "loss": 0.8793,
+      "step": 20955
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.28925827145576477,
+      "learning_rate": 2.1420959542979095e-05,
+      "loss": 0.7409,
+      "step": 20960
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.24598172307014465,
+      "learning_rate": 2.1367994794649312e-05,
+      "loss": 0.7843,
+      "step": 20965
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.2719762623310089,
+      "learning_rate": 2.1315090584812917e-05,
+      "loss": 0.9156,
+      "step": 20970
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.29581862688064575,
+      "learning_rate": 2.126224693836837e-05,
+      "loss": 0.7598,
+      "step": 20975
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.27417975664138794,
+      "learning_rate": 2.120946388018565e-05,
+      "loss": 0.8218,
+      "step": 20980
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.27194756269454956,
+      "learning_rate": 2.115674143510625e-05,
+      "loss": 0.8978,
+      "step": 20985
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.27011337876319885,
+      "learning_rate": 2.1104079627943087e-05,
+      "loss": 0.8051,
+      "step": 20990
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.26955413818359375,
+      "learning_rate": 2.1051478483480555e-05,
+      "loss": 0.7435,
+      "step": 20995
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.26142939925193787,
+      "learning_rate": 2.099893802647445e-05,
+      "loss": 0.7034,
+      "step": 21000
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.27008065581321716,
+      "learning_rate": 2.0946458281652155e-05,
+      "loss": 0.7948,
+      "step": 21005
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.2838139235973358,
+      "learning_rate": 2.089403927371232e-05,
+      "loss": 0.7694,
+      "step": 21010
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.2932431399822235,
+      "learning_rate": 2.0841681027325075e-05,
+      "loss": 0.8528,
+      "step": 21015
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.28293853998184204,
+      "learning_rate": 2.0789383567131955e-05,
+      "loss": 0.7757,
+      "step": 21020
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.27947354316711426,
+      "learning_rate": 2.0737146917745878e-05,
+      "loss": 0.7972,
+      "step": 21025
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.2797410190105438,
+      "learning_rate": 2.0684971103751158e-05,
+      "loss": 0.8079,
+      "step": 21030
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.2733669877052307,
+      "learning_rate": 2.0632856149703454e-05,
+      "loss": 0.7684,
+      "step": 21035
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.2936958074569702,
+      "learning_rate": 2.0580802080129744e-05,
+      "loss": 0.7623,
+      "step": 21040
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.2874435782432556,
+      "learning_rate": 2.05288089195285e-05,
+      "loss": 0.752,
+      "step": 21045
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.2876211404800415,
+      "learning_rate": 2.0476876692369398e-05,
+      "loss": 0.7191,
+      "step": 21050
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.3623277544975281,
+      "learning_rate": 2.0425005423093437e-05,
+      "loss": 0.7588,
+      "step": 21055
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.30501341819763184,
+      "learning_rate": 2.037319513611294e-05,
+      "loss": 0.723,
+      "step": 21060
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.29336845874786377,
+      "learning_rate": 2.0321445855811635e-05,
+      "loss": 0.8807,
+      "step": 21065
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.285491019487381,
+      "learning_rate": 2.026975760654439e-05,
+      "loss": 0.8104,
+      "step": 21070
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.27054548263549805,
+      "learning_rate": 2.0218130412637397e-05,
+      "loss": 0.7187,
+      "step": 21075
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.28753402829170227,
+      "learning_rate": 2.0166564298388182e-05,
+      "loss": 0.8717,
+      "step": 21080
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.25017327070236206,
+      "learning_rate": 2.0115059288065437e-05,
+      "loss": 0.7553,
+      "step": 21085
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.2623300850391388,
+      "learning_rate": 2.006361540590914e-05,
+      "loss": 0.8769,
+      "step": 21090
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.2736659348011017,
+      "learning_rate": 2.0012232676130503e-05,
+      "loss": 0.8399,
+      "step": 21095
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.3095528483390808,
+      "learning_rate": 1.9960911122911915e-05,
+      "loss": 0.6941,
+      "step": 21100
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.2801484763622284,
+      "learning_rate": 1.990965077040701e-05,
+      "loss": 0.8403,
+      "step": 21105
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.3449302017688751,
+      "learning_rate": 1.9858451642740637e-05,
+      "loss": 0.7571,
+      "step": 21110
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.27275073528289795,
+      "learning_rate": 1.9807313764008738e-05,
+      "loss": 0.8058,
+      "step": 21115
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.3031749129295349,
+      "learning_rate": 1.9756237158278593e-05,
+      "loss": 0.7587,
+      "step": 21120
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.27661368250846863,
+      "learning_rate": 1.9705221849588493e-05,
+      "loss": 0.8447,
+      "step": 21125
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.3269336223602295,
+      "learning_rate": 1.9654267861947946e-05,
+      "loss": 0.8247,
+      "step": 21130
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.2593838572502136,
+      "learning_rate": 1.9603375219337573e-05,
+      "loss": 0.8053,
+      "step": 21135
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.2844938039779663,
+      "learning_rate": 1.9552543945709183e-05,
+      "loss": 0.8455,
+      "step": 21140
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.28493261337280273,
+      "learning_rate": 1.950177406498563e-05,
+      "loss": 0.8471,
+      "step": 21145
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.2901061773300171,
+      "learning_rate": 1.9451065601060927e-05,
+      "loss": 0.6899,
+      "step": 21150
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.27742087841033936,
+      "learning_rate": 1.9400418577800148e-05,
+      "loss": 0.7318,
+      "step": 21155
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.2615971565246582,
+      "learning_rate": 1.9349833019039457e-05,
+      "loss": 0.8078,
+      "step": 21160
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.3021378219127655,
+      "learning_rate": 1.929930894858611e-05,
+      "loss": 0.812,
+      "step": 21165
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.29150012135505676,
+      "learning_rate": 1.924884639021841e-05,
+      "loss": 0.9014,
+      "step": 21170
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.2782303988933563,
+      "learning_rate": 1.9198445367685668e-05,
+      "loss": 0.7307,
+      "step": 21175
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.26657816767692566,
+      "learning_rate": 1.914810590470836e-05,
+      "loss": 0.8146,
+      "step": 21180
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.27062734961509705,
+      "learning_rate": 1.909782802497786e-05,
+      "loss": 0.8376,
+      "step": 21185
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.3097752630710602,
+      "learning_rate": 1.9047611752156628e-05,
+      "loss": 0.7168,
+      "step": 21190
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.2954844534397125,
+      "learning_rate": 1.8997457109878067e-05,
+      "loss": 0.6942,
+      "step": 21195
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.29877805709838867,
+      "learning_rate": 1.8947364121746677e-05,
+      "loss": 0.7444,
+      "step": 21200
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.303915798664093,
+      "learning_rate": 1.8897332811337868e-05,
+      "loss": 0.7517,
+      "step": 21205
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.2891033887863159,
+      "learning_rate": 1.8847363202198017e-05,
+      "loss": 0.7324,
+      "step": 21210
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.27175503969192505,
+      "learning_rate": 1.8797455317844452e-05,
+      "loss": 0.7523,
+      "step": 21215
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.32238414883613586,
+      "learning_rate": 1.8747609181765543e-05,
+      "loss": 0.7668,
+      "step": 21220
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.28903838992118835,
+      "learning_rate": 1.8697824817420514e-05,
+      "loss": 0.796,
+      "step": 21225
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.25564032793045044,
+      "learning_rate": 1.8648102248239537e-05,
+      "loss": 0.8034,
+      "step": 21230
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.30182021856307983,
+      "learning_rate": 1.859844149762371e-05,
+      "loss": 0.7384,
+      "step": 21235
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.33114859461784363,
+      "learning_rate": 1.8548842588945023e-05,
+      "loss": 0.8908,
+      "step": 21240
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.2683655917644501,
+      "learning_rate": 1.8499305545546367e-05,
+      "loss": 0.7694,
+      "step": 21245
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.30291497707366943,
+      "learning_rate": 1.8449830390741526e-05,
+      "loss": 0.8817,
+      "step": 21250
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.2633529007434845,
+      "learning_rate": 1.840041714781511e-05,
+      "loss": 0.737,
+      "step": 21255
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.27262598276138306,
+      "learning_rate": 1.8351065840022715e-05,
+      "loss": 0.7905,
+      "step": 21260
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.28199708461761475,
+      "learning_rate": 1.8301776490590664e-05,
+      "loss": 0.7542,
+      "step": 21265
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.3084707260131836,
+      "learning_rate": 1.8252549122716147e-05,
+      "loss": 0.6572,
+      "step": 21270
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.27886369824409485,
+      "learning_rate": 1.8203383759567186e-05,
+      "loss": 0.7748,
+      "step": 21275
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.2864547669887543,
+      "learning_rate": 1.8154280424282685e-05,
+      "loss": 0.8352,
+      "step": 21280
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.26778092980384827,
+      "learning_rate": 1.8105239139972277e-05,
+      "loss": 0.7049,
+      "step": 21285
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.28581252694129944,
+      "learning_rate": 1.8056259929716417e-05,
+      "loss": 0.7679,
+      "step": 21290
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.30422863364219666,
+      "learning_rate": 1.800734281656633e-05,
+      "loss": 0.7548,
+      "step": 21295
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.2808743715286255,
+      "learning_rate": 1.7958487823544082e-05,
+      "loss": 0.7986,
+      "step": 21300
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.3032390773296356,
+      "learning_rate": 1.790969497364242e-05,
+      "loss": 0.8185,
+      "step": 21305
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.2701072096824646,
+      "learning_rate": 1.7860964289824885e-05,
+      "loss": 0.7967,
+      "step": 21310
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.28802967071533203,
+      "learning_rate": 1.7812295795025773e-05,
+      "loss": 0.7764,
+      "step": 21315
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.2924157381057739,
+      "learning_rate": 1.776368951215007e-05,
+      "loss": 0.7171,
+      "step": 21320
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.2745770812034607,
+      "learning_rate": 1.7715145464073498e-05,
+      "loss": 0.806,
+      "step": 21325
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.2635575830936432,
+      "learning_rate": 1.7666663673642535e-05,
+      "loss": 0.7911,
+      "step": 21330
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.2602815330028534,
+      "learning_rate": 1.761824416367427e-05,
+      "loss": 0.8123,
+      "step": 21335
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.30612412095069885,
+      "learning_rate": 1.7569886956956585e-05,
+      "loss": 0.8126,
+      "step": 21340
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.2764699161052704,
+      "learning_rate": 1.752159207624797e-05,
+      "loss": 0.8279,
+      "step": 21345
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.2635042667388916,
+      "learning_rate": 1.7473359544277594e-05,
+      "loss": 0.6854,
+      "step": 21350
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.25210273265838623,
+      "learning_rate": 1.7425189383745265e-05,
+      "loss": 0.8533,
+      "step": 21355
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.26929739117622375,
+      "learning_rate": 1.7377081617321513e-05,
+      "loss": 0.7162,
+      "step": 21360
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.27640727162361145,
+      "learning_rate": 1.7329036267647417e-05,
+      "loss": 0.7429,
+      "step": 21365
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.2647760510444641,
+      "learning_rate": 1.7281053357334717e-05,
+      "loss": 0.8501,
+      "step": 21370
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.28919151425361633,
+      "learning_rate": 1.7233132908965784e-05,
+      "loss": 0.7655,
+      "step": 21375
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.295827180147171,
+      "learning_rate": 1.7185274945093547e-05,
+      "loss": 0.8074,
+      "step": 21380
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.2644157409667969,
+      "learning_rate": 1.713747948824155e-05,
+      "loss": 0.8458,
+      "step": 21385
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.2902376055717468,
+      "learning_rate": 1.7089746560903956e-05,
+      "loss": 0.8744,
+      "step": 21390
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.25895097851753235,
+      "learning_rate": 1.7042076185545388e-05,
+      "loss": 0.7331,
+      "step": 21395
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.2989761233329773,
+      "learning_rate": 1.6994468384601206e-05,
+      "loss": 0.7485,
+      "step": 21400
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.26505047082901,
+      "learning_rate": 1.6946923180477183e-05,
+      "loss": 0.7844,
+      "step": 21405
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.26280462741851807,
+      "learning_rate": 1.6899440595549674e-05,
+      "loss": 0.7854,
+      "step": 21410
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.2874717712402344,
+      "learning_rate": 1.6852020652165515e-05,
+      "loss": 0.7819,
+      "step": 21415
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.28164345026016235,
+      "learning_rate": 1.6804663372642175e-05,
+      "loss": 0.7315,
+      "step": 21420
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.2647458016872406,
+      "learning_rate": 1.6757368779267522e-05,
+      "loss": 0.8445,
+      "step": 21425
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.26932641863822937,
+      "learning_rate": 1.6710136894299987e-05,
+      "loss": 0.763,
+      "step": 21430
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.2817295491695404,
+      "learning_rate": 1.6662967739968402e-05,
+      "loss": 0.8031,
+      "step": 21435
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.303753137588501,
+      "learning_rate": 1.661586133847223e-05,
+      "loss": 0.6969,
+      "step": 21440
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.27712610363960266,
+      "learning_rate": 1.656881771198125e-05,
+      "loss": 0.8331,
+      "step": 21445
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.28993546962738037,
+      "learning_rate": 1.6521836882635775e-05,
+      "loss": 0.7937,
+      "step": 21450
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.2543039619922638,
+      "learning_rate": 1.6474918872546545e-05,
+      "loss": 0.7681,
+      "step": 21455
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.3048137426376343,
+      "learning_rate": 1.6428063703794714e-05,
+      "loss": 0.8169,
+      "step": 21460
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.30479833483695984,
+      "learning_rate": 1.6381271398431894e-05,
+      "loss": 0.7523,
+      "step": 21465
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.311480849981308,
+      "learning_rate": 1.6334541978480087e-05,
+      "loss": 0.7312,
+      "step": 21470
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.2890738844871521,
+      "learning_rate": 1.6287875465931698e-05,
+      "loss": 0.746,
+      "step": 21475
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.2721187472343445,
+      "learning_rate": 1.624127188274958e-05,
+      "loss": 0.827,
+      "step": 21480
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.2833477556705475,
+      "learning_rate": 1.61947312508669e-05,
+      "loss": 0.7932,
+      "step": 21485
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.28032922744750977,
+      "learning_rate": 1.614825359218724e-05,
+      "loss": 0.7861,
+      "step": 21490
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.2863568067550659,
+      "learning_rate": 1.6101838928584483e-05,
+      "loss": 0.6808,
+      "step": 21495
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.31328871846199036,
+      "learning_rate": 1.6055487281902973e-05,
+      "loss": 0.8056,
+      "step": 21500
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.28002050518989563,
+      "learning_rate": 1.6009198673957313e-05,
+      "loss": 0.84,
+      "step": 21505
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.28880271315574646,
+      "learning_rate": 1.5962973126532457e-05,
+      "loss": 0.8734,
+      "step": 21510
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.2959395945072174,
+      "learning_rate": 1.5916810661383638e-05,
+      "loss": 0.7527,
+      "step": 21515
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.28988808393478394,
+      "learning_rate": 1.5870711300236527e-05,
+      "loss": 0.6993,
+      "step": 21520
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.3125680983066559,
+      "learning_rate": 1.5824675064786968e-05,
+      "loss": 0.7884,
+      "step": 21525
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.31437787413597107,
+      "learning_rate": 1.577870197670118e-05,
+      "loss": 0.6855,
+      "step": 21530
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.2919071316719055,
+      "learning_rate": 1.573279205761554e-05,
+      "loss": 0.8805,
+      "step": 21535
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.288947194814682,
+      "learning_rate": 1.5686945329136865e-05,
+      "loss": 0.7189,
+      "step": 21540
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.28742465376853943,
+      "learning_rate": 1.5641161812842105e-05,
+      "loss": 0.8336,
+      "step": 21545
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.2684331238269806,
+      "learning_rate": 1.5595441530278517e-05,
+      "loss": 0.7574,
+      "step": 21550
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.29591086506843567,
+      "learning_rate": 1.5549784502963554e-05,
+      "loss": 0.7995,
+      "step": 21555
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.25457319617271423,
+      "learning_rate": 1.5504190752384987e-05,
+      "loss": 0.7078,
+      "step": 21560
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.28611642122268677,
+      "learning_rate": 1.5458660300000725e-05,
+      "loss": 0.7801,
+      "step": 21565
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.2762448489665985,
+      "learning_rate": 1.5413193167238908e-05,
+      "loss": 0.7721,
+      "step": 21570
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.29619264602661133,
+      "learning_rate": 1.5367789375497836e-05,
+      "loss": 0.8263,
+      "step": 21575
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.27710941433906555,
+      "learning_rate": 1.5322448946146114e-05,
+      "loss": 0.9136,
+      "step": 21580
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.2905004322528839,
+      "learning_rate": 1.5277171900522428e-05,
+      "loss": 0.7735,
+      "step": 21585
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.26848629117012024,
+      "learning_rate": 1.5231958259935646e-05,
+      "loss": 0.7816,
+      "step": 21590
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.28664660453796387,
+      "learning_rate": 1.5186808045664812e-05,
+      "loss": 0.7792,
+      "step": 21595
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.3380061984062195,
+      "learning_rate": 1.5141721278959112e-05,
+      "loss": 0.7895,
+      "step": 21600
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.26987069845199585,
+      "learning_rate": 1.5096697981037909e-05,
+      "loss": 0.7762,
+      "step": 21605
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.2945181131362915,
+      "learning_rate": 1.5051738173090582e-05,
+      "loss": 0.7818,
+      "step": 21610
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.30163273215293884,
+      "learning_rate": 1.500684187627681e-05,
+      "loss": 0.7005,
+      "step": 21615
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.28872594237327576,
+      "learning_rate": 1.496200911172622e-05,
+      "loss": 0.7485,
+      "step": 21620
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.2600472569465637,
+      "learning_rate": 1.491723990053862e-05,
+      "loss": 0.7653,
+      "step": 21625
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.2789260745048523,
+      "learning_rate": 1.487253426378387e-05,
+      "loss": 0.6609,
+      "step": 21630
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.23983633518218994,
+      "learning_rate": 1.482789222250194e-05,
+      "loss": 0.7225,
+      "step": 21635
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.27560335397720337,
+      "learning_rate": 1.4783313797702878e-05,
+      "loss": 0.8625,
+      "step": 21640
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.2681236267089844,
+      "learning_rate": 1.4738799010366747e-05,
+      "loss": 0.7163,
+      "step": 21645
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.2975192070007324,
+      "learning_rate": 1.4694347881443659e-05,
+      "loss": 0.7637,
+      "step": 21650
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.3074178397655487,
+      "learning_rate": 1.4649960431853842e-05,
+      "loss": 0.7499,
+      "step": 21655
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.28553348779678345,
+      "learning_rate": 1.4605636682487486e-05,
+      "loss": 0.8284,
+      "step": 21660
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.29183608293533325,
+      "learning_rate": 1.456137665420481e-05,
+      "loss": 0.7162,
+      "step": 21665
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.2632047235965729,
+      "learning_rate": 1.4517180367836062e-05,
+      "loss": 0.8519,
+      "step": 21670
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.28933554887771606,
+      "learning_rate": 1.4473047844181474e-05,
+      "loss": 0.7592,
+      "step": 21675
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.30451327562332153,
+      "learning_rate": 1.4428979104011295e-05,
+      "loss": 0.7692,
+      "step": 21680
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.27428263425827026,
+      "learning_rate": 1.4384974168065705e-05,
+      "loss": 0.7196,
+      "step": 21685
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.2668101191520691,
+      "learning_rate": 1.4341033057054885e-05,
+      "loss": 0.7442,
+      "step": 21690
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.27039745450019836,
+      "learning_rate": 1.4297155791659044e-05,
+      "loss": 0.7293,
+      "step": 21695
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.2836028039455414,
+      "learning_rate": 1.4253342392528227e-05,
+      "loss": 0.8122,
+      "step": 21700
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.26885733008384705,
+      "learning_rate": 1.4209592880282494e-05,
+      "loss": 0.6921,
+      "step": 21705
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.274857759475708,
+      "learning_rate": 1.4165907275511773e-05,
+      "loss": 0.7754,
+      "step": 21710
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.29404956102371216,
+      "learning_rate": 1.4122285598776035e-05,
+      "loss": 0.7997,
+      "step": 21715
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.29239895939826965,
+      "learning_rate": 1.4078727870605056e-05,
+      "loss": 0.6922,
+      "step": 21720
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.2807917892932892,
+      "learning_rate": 1.4035234111498539e-05,
+      "loss": 0.8521,
+      "step": 21725
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.26643988490104675,
+      "learning_rate": 1.3991804341926077e-05,
+      "loss": 0.8503,
+      "step": 21730
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.2644132375717163,
+      "learning_rate": 1.394843858232722e-05,
+      "loss": 0.7678,
+      "step": 21735
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.2644047141075134,
+      "learning_rate": 1.390513685311131e-05,
+      "loss": 0.8423,
+      "step": 21740
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.2851656675338745,
+      "learning_rate": 1.3861899174657542e-05,
+      "loss": 0.7707,
+      "step": 21745
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.2632659673690796,
+      "learning_rate": 1.381872556731501e-05,
+      "loss": 0.8024,
+      "step": 21750
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.2766389846801758,
+      "learning_rate": 1.3775616051402689e-05,
+      "loss": 0.7274,
+      "step": 21755
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.3009210228919983,
+      "learning_rate": 1.3732570647209334e-05,
+      "loss": 0.6989,
+      "step": 21760
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.2986798584461212,
+      "learning_rate": 1.3689589374993526e-05,
+      "loss": 0.7043,
+      "step": 21765
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.2332366406917572,
+      "learning_rate": 1.3646672254983649e-05,
+      "loss": 0.7085,
+      "step": 21770
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.2506479322910309,
+      "learning_rate": 1.3603819307378011e-05,
+      "loss": 0.8323,
+      "step": 21775
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.2667887806892395,
+      "learning_rate": 1.3561030552344566e-05,
+      "loss": 0.8099,
+      "step": 21780
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.26609328389167786,
+      "learning_rate": 1.3518306010021152e-05,
+      "loss": 0.805,
+      "step": 21785
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.2911878824234009,
+      "learning_rate": 1.3475645700515319e-05,
+      "loss": 0.8367,
+      "step": 21790
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.2831841707229614,
+      "learning_rate": 1.3433049643904476e-05,
+      "loss": 0.8377,
+      "step": 21795
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.2804235517978668,
+      "learning_rate": 1.3390517860235717e-05,
+      "loss": 0.8026,
+      "step": 21800
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.28793367743492126,
+      "learning_rate": 1.3348050369525931e-05,
+      "loss": 0.7157,
+      "step": 21805
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.28661012649536133,
+      "learning_rate": 1.3305647191761699e-05,
+      "loss": 0.729,
+      "step": 21810
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.2898370027542114,
+      "learning_rate": 1.3263308346899371e-05,
+      "loss": 0.7045,
+      "step": 21815
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.26269808411598206,
+      "learning_rate": 1.3221033854865027e-05,
+      "loss": 0.7042,
+      "step": 21820
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.29441672563552856,
+      "learning_rate": 1.3178823735554434e-05,
+      "loss": 0.7476,
+      "step": 21825
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.25631454586982727,
+      "learning_rate": 1.3136678008833069e-05,
+      "loss": 0.8577,
+      "step": 21830
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.32068613171577454,
+      "learning_rate": 1.309459669453613e-05,
+      "loss": 0.8105,
+      "step": 21835
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.23884128034114838,
+      "learning_rate": 1.3052579812468494e-05,
+      "loss": 0.7782,
+      "step": 21840
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.3259485363960266,
+      "learning_rate": 1.3010627382404675e-05,
+      "loss": 0.8539,
+      "step": 21845
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.30076032876968384,
+      "learning_rate": 1.2968739424088848e-05,
+      "loss": 0.818,
+      "step": 21850
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.26317545771598816,
+      "learning_rate": 1.2926915957234957e-05,
+      "loss": 0.766,
+      "step": 21855
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.23999077081680298,
+      "learning_rate": 1.2885157001526459e-05,
+      "loss": 0.7286,
+      "step": 21860
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.23871669173240662,
+      "learning_rate": 1.2843462576616531e-05,
+      "loss": 0.7398,
+      "step": 21865
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.25987398624420166,
+      "learning_rate": 1.2801832702127912e-05,
+      "loss": 0.7984,
+      "step": 21870
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.25374579429626465,
+      "learning_rate": 1.2760267397653063e-05,
+      "loss": 0.7778,
+      "step": 21875
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.3066677749156952,
+      "learning_rate": 1.2718766682753966e-05,
+      "loss": 0.8205,
+      "step": 21880
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.2812236547470093,
+      "learning_rate": 1.2677330576962235e-05,
+      "loss": 0.8151,
+      "step": 21885
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.2993893027305603,
+      "learning_rate": 1.263595909977907e-05,
+      "loss": 0.7551,
+      "step": 21890
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.2752331495285034,
+      "learning_rate": 1.2594652270675293e-05,
+      "loss": 0.8551,
+      "step": 21895
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.29263609647750854,
+      "learning_rate": 1.2553410109091221e-05,
+      "loss": 0.846,
+      "step": 21900
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.28177276253700256,
+      "learning_rate": 1.251223263443683e-05,
+      "loss": 0.7788,
+      "step": 21905
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.3226982355117798,
+      "learning_rate": 1.2479337242326714e-05,
+      "loss": 0.7536,
+      "step": 21910
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.29417189955711365,
+      "learning_rate": 1.2438276252961555e-05,
+      "loss": 0.7601,
+      "step": 21915
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.26856791973114014,
+      "learning_rate": 1.2397280004711845e-05,
+      "loss": 0.719,
+      "step": 21920
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.2941911518573761,
+      "learning_rate": 1.2356348516871839e-05,
+      "loss": 0.8129,
+      "step": 21925
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.3045237064361572,
+      "learning_rate": 1.2315481808705224e-05,
+      "loss": 0.8177,
+      "step": 21930
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.26458317041397095,
+      "learning_rate": 1.2274679899445234e-05,
+      "loss": 0.8406,
+      "step": 21935
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.3065710663795471,
+      "learning_rate": 1.2233942808294573e-05,
+      "loss": 0.7369,
+      "step": 21940
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.28398415446281433,
+      "learning_rate": 1.2193270554425521e-05,
+      "loss": 0.8903,
+      "step": 21945
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.29608169198036194,
+      "learning_rate": 1.215266315697977e-05,
+      "loss": 0.7931,
+      "step": 21950
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.2909848093986511,
+      "learning_rate": 1.2112120635068495e-05,
+      "loss": 0.8049,
+      "step": 21955
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.28937968611717224,
+      "learning_rate": 1.2071643007772353e-05,
+      "loss": 0.7373,
+      "step": 21960
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.2677673399448395,
+      "learning_rate": 1.2031230294141486e-05,
+      "loss": 0.7057,
+      "step": 21965
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.27190762758255005,
+      "learning_rate": 1.199088251319545e-05,
+      "loss": 0.7522,
+      "step": 21970
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.2613365650177002,
+      "learning_rate": 1.1950599683923234e-05,
+      "loss": 0.8177,
+      "step": 21975
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.27314963936805725,
+      "learning_rate": 1.1910381825283294e-05,
+      "loss": 0.7326,
+      "step": 21980
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.23859882354736328,
+      "learning_rate": 1.1870228956203487e-05,
+      "loss": 0.7208,
+      "step": 21985
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.2868862450122833,
+      "learning_rate": 1.183014109558108e-05,
+      "loss": 0.7957,
+      "step": 21990
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.2559562027454376,
+      "learning_rate": 1.1790118262282715e-05,
+      "loss": 0.7123,
+      "step": 21995
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.30129274725914,
+      "learning_rate": 1.1750160475144543e-05,
+      "loss": 0.7947,
+      "step": 22000
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.26090654730796814,
+      "learning_rate": 1.171026775297197e-05,
+      "loss": 0.6892,
+      "step": 22005
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.28930899500846863,
+      "learning_rate": 1.1670440114539858e-05,
+      "loss": 0.6845,
+      "step": 22010
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.3468032777309418,
+      "learning_rate": 1.1630677578592401e-05,
+      "loss": 0.7648,
+      "step": 22015
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.2959340810775757,
+      "learning_rate": 1.1590980163843194e-05,
+      "loss": 0.8748,
+      "step": 22020
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.28836220502853394,
+      "learning_rate": 1.1551347888975126e-05,
+      "loss": 0.7935,
+      "step": 22025
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.29400843381881714,
+      "learning_rate": 1.1511780772640494e-05,
+      "loss": 0.8172,
+      "step": 22030
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.28731659054756165,
+      "learning_rate": 1.1472278833460886e-05,
+      "loss": 0.7621,
+      "step": 22035
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.26251596212387085,
+      "learning_rate": 1.1432842090027227e-05,
+      "loss": 0.7288,
+      "step": 22040
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.2831023037433624,
+      "learning_rate": 1.1393470560899742e-05,
+      "loss": 0.799,
+      "step": 22045
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.28475290536880493,
+      "learning_rate": 1.135416426460799e-05,
+      "loss": 0.8506,
+      "step": 22050
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.31778624653816223,
+      "learning_rate": 1.1314923219650807e-05,
+      "loss": 0.7712,
+      "step": 22055
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.28575217723846436,
+      "learning_rate": 1.1275747444496353e-05,
+      "loss": 0.783,
+      "step": 22060
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.3009990155696869,
+      "learning_rate": 1.1236636957582062e-05,
+      "loss": 0.7318,
+      "step": 22065
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.27844470739364624,
+      "learning_rate": 1.1197591777314597e-05,
+      "loss": 0.8126,
+      "step": 22070
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.2680925130844116,
+      "learning_rate": 1.1158611922069904e-05,
+      "loss": 0.7173,
+      "step": 22075
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.2805362343788147,
+      "learning_rate": 1.1119697410193246e-05,
+      "loss": 0.8169,
+      "step": 22080
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.2496698796749115,
+      "learning_rate": 1.1080848259999054e-05,
+      "loss": 0.7369,
+      "step": 22085
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.2835550606250763,
+      "learning_rate": 1.1042064489771035e-05,
+      "loss": 0.7064,
+      "step": 22090
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.29828065633773804,
+      "learning_rate": 1.100334611776209e-05,
+      "loss": 0.7352,
+      "step": 22095
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.28538069128990173,
+      "learning_rate": 1.0964693162194427e-05,
+      "loss": 0.6496,
+      "step": 22100
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.2641838490962982,
+      "learning_rate": 1.0926105641259392e-05,
+      "loss": 0.7463,
+      "step": 22105
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.2766837775707245,
+      "learning_rate": 1.0887583573117526e-05,
+      "loss": 0.806,
+      "step": 22110
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.28771060705184937,
+      "learning_rate": 1.0849126975898626e-05,
+      "loss": 0.9518,
+      "step": 22115
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.2707420587539673,
+      "learning_rate": 1.0810735867701614e-05,
+      "loss": 0.7166,
+      "step": 22120
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.2899439036846161,
+      "learning_rate": 1.077241026659464e-05,
+      "loss": 0.7041,
+      "step": 22125
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.25429069995880127,
+      "learning_rate": 1.0734150190615005e-05,
+      "loss": 0.7543,
+      "step": 22130
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.29701635241508484,
+      "learning_rate": 1.069595565776914e-05,
+      "loss": 0.7908,
+      "step": 22135
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.27902600169181824,
+      "learning_rate": 1.06578266860327e-05,
+      "loss": 0.8859,
+      "step": 22140
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.29597392678260803,
+      "learning_rate": 1.0619763293350447e-05,
+      "loss": 0.865,
+      "step": 22145
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.24774394929409027,
+      "learning_rate": 1.0581765497636253e-05,
+      "loss": 0.767,
+      "step": 22150
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.2774536907672882,
+      "learning_rate": 1.0543833316773127e-05,
+      "loss": 0.8292,
+      "step": 22155
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.2544390857219696,
+      "learning_rate": 1.0505966768613273e-05,
+      "loss": 0.8584,
+      "step": 22160
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.2902020215988159,
+      "learning_rate": 1.0468165870977901e-05,
+      "loss": 0.6859,
+      "step": 22165
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.292767196893692,
+      "learning_rate": 1.0430430641657383e-05,
+      "loss": 0.8134,
+      "step": 22170
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.2644735872745514,
+      "learning_rate": 1.0392761098411146e-05,
+      "loss": 0.6928,
+      "step": 22175
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.26982855796813965,
+      "learning_rate": 1.0355157258967772e-05,
+      "loss": 0.7652,
+      "step": 22180
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.2815554141998291,
+      "learning_rate": 1.0317619141024858e-05,
+      "loss": 0.8082,
+      "step": 22185
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.25484582781791687,
+      "learning_rate": 1.028014676224907e-05,
+      "loss": 0.7836,
+      "step": 22190
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.26840999722480774,
+      "learning_rate": 1.0242740140276185e-05,
+      "loss": 0.7332,
+      "step": 22195
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.27440059185028076,
+      "learning_rate": 1.0205399292710969e-05,
+      "loss": 0.7333,
+      "step": 22200
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.28358691930770874,
+      "learning_rate": 1.0168124237127301e-05,
+      "loss": 0.7612,
+      "step": 22205
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.2975500226020813,
+      "learning_rate": 1.0130914991068028e-05,
+      "loss": 0.7863,
+      "step": 22210
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.3028092682361603,
+      "learning_rate": 1.0093771572045045e-05,
+      "loss": 0.6993,
+      "step": 22215
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.28230762481689453,
+      "learning_rate": 1.0056693997539317e-05,
+      "loss": 0.7698,
+      "step": 22220
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.2743557393550873,
+      "learning_rate": 1.0019682285000785e-05,
+      "loss": 0.7347,
+      "step": 22225
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.25398826599121094,
+      "learning_rate": 9.982736451848367e-06,
+      "loss": 0.7027,
+      "step": 22230
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.2597667872905731,
+      "learning_rate": 9.94585651546997e-06,
+      "loss": 0.7767,
+      "step": 22235
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.2996768355369568,
+      "learning_rate": 9.90904249322259e-06,
+      "loss": 0.8047,
+      "step": 22240
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.3190852105617523,
+      "learning_rate": 9.872294402432074e-06,
+      "loss": 0.7218,
+      "step": 22245
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.27263718843460083,
+      "learning_rate": 9.835612260393316e-06,
+      "loss": 0.8415,
+      "step": 22250
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.29362040758132935,
+      "learning_rate": 9.798996084370143e-06,
+      "loss": 0.8255,
+      "step": 22255
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.30126357078552246,
+      "learning_rate": 9.76244589159535e-06,
+      "loss": 0.726,
+      "step": 22260
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.28834068775177,
+      "learning_rate": 9.725961699270662e-06,
+      "loss": 0.8232,
+      "step": 22265
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.2588998079299927,
+      "learning_rate": 9.689543524566746e-06,
+      "loss": 0.8129,
+      "step": 22270
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.24092987179756165,
+      "learning_rate": 9.653191384623204e-06,
+      "loss": 0.7258,
+      "step": 22275
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.2920580506324768,
+      "learning_rate": 9.616905296548588e-06,
+      "loss": 0.863,
+      "step": 22280
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.2646152079105377,
+      "learning_rate": 9.58068527742032e-06,
+      "loss": 0.7227,
+      "step": 22285
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.26156240701675415,
+      "learning_rate": 9.544531344284745e-06,
+      "loss": 0.7387,
+      "step": 22290
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.259880930185318,
+      "learning_rate": 9.50844351415707e-06,
+      "loss": 0.8297,
+      "step": 22295
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.2808079421520233,
+      "learning_rate": 9.47242180402148e-06,
+      "loss": 0.8567,
+      "step": 22300
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.2962208688259125,
+      "learning_rate": 9.436466230830958e-06,
+      "loss": 0.6969,
+      "step": 22305
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.30381715297698975,
+      "learning_rate": 9.40057681150742e-06,
+      "loss": 0.6078,
+      "step": 22310
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.29196277260780334,
+      "learning_rate": 9.364753562941556e-06,
+      "loss": 0.7343,
+      "step": 22315
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.2667737603187561,
+      "learning_rate": 9.328996501993047e-06,
+      "loss": 0.8018,
+      "step": 22320
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.27238935232162476,
+      "learning_rate": 9.29330564549033e-06,
+      "loss": 0.7812,
+      "step": 22325
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.28509876132011414,
+      "learning_rate": 9.257681010230683e-06,
+      "loss": 0.8695,
+      "step": 22330
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.2799687683582306,
+      "learning_rate": 9.222122612980281e-06,
+      "loss": 0.8006,
+      "step": 22335
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.2903033494949341,
+      "learning_rate": 9.186630470474054e-06,
+      "loss": 0.8644,
+      "step": 22340
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.2779753506183624,
+      "learning_rate": 9.15120459941579e-06,
+      "loss": 0.7324,
+      "step": 22345
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.29276126623153687,
+      "learning_rate": 9.1158450164781e-06,
+      "loss": 0.7755,
+      "step": 22350
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.2905157804489136,
+      "learning_rate": 9.080551738302328e-06,
+      "loss": 0.7065,
+      "step": 22355
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.2716805636882782,
+      "learning_rate": 9.04532478149873e-06,
+      "loss": 0.8302,
+      "step": 22360
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.2544827461242676,
+      "learning_rate": 9.010164162646249e-06,
+      "loss": 0.8856,
+      "step": 22365
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.30418241024017334,
+      "learning_rate": 8.975069898292647e-06,
+      "loss": 0.8106,
+      "step": 22370
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.2704138457775116,
+      "learning_rate": 8.940042004954412e-06,
+      "loss": 0.8389,
+      "step": 22375
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.28807222843170166,
+      "learning_rate": 8.90508049911689e-06,
+      "loss": 0.7696,
+      "step": 22380
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.28206029534339905,
+      "learning_rate": 8.870185397234086e-06,
+      "loss": 0.7857,
+      "step": 22385
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.26948267221450806,
+      "learning_rate": 8.83535671572883e-06,
+      "loss": 0.8041,
+      "step": 22390
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.2844690680503845,
+      "learning_rate": 8.800594470992611e-06,
+      "loss": 0.8143,
+      "step": 22395
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.28157171607017517,
+      "learning_rate": 8.765898679385742e-06,
+      "loss": 0.7778,
+      "step": 22400
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.28640446066856384,
+      "learning_rate": 8.731269357237192e-06,
+      "loss": 0.8495,
+      "step": 22405
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.28528541326522827,
+      "learning_rate": 8.696706520844693e-06,
+      "loss": 0.8586,
+      "step": 22410
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.2665455937385559,
+      "learning_rate": 8.6622101864746e-06,
+      "loss": 0.759,
+      "step": 22415
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.2542749047279358,
+      "learning_rate": 8.627780370362108e-06,
+      "loss": 0.7185,
+      "step": 22420
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.2605261504650116,
+      "learning_rate": 8.593417088710992e-06,
+      "loss": 0.8301,
+      "step": 22425
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.27719512581825256,
+      "learning_rate": 8.55912035769376e-06,
+      "loss": 0.8109,
+      "step": 22430
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.27308791875839233,
+      "learning_rate": 8.524890193451573e-06,
+      "loss": 0.722,
+      "step": 22435
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.26642659306526184,
+      "learning_rate": 8.490726612094323e-06,
+      "loss": 0.7022,
+      "step": 22440
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.2815400958061218,
+      "learning_rate": 8.456629629700518e-06,
+      "loss": 0.887,
+      "step": 22445
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.27426496148109436,
+      "learning_rate": 8.422599262317303e-06,
+      "loss": 0.7021,
+      "step": 22450
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.27988290786743164,
+      "learning_rate": 8.388635525960503e-06,
+      "loss": 0.7661,
+      "step": 22455
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.31152406334877014,
+      "learning_rate": 8.35473843661461e-06,
+      "loss": 0.7652,
+      "step": 22460
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.2869008481502533,
+      "learning_rate": 8.320908010232702e-06,
+      "loss": 0.7633,
+      "step": 22465
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.28940802812576294,
+      "learning_rate": 8.287144262736506e-06,
+      "loss": 0.749,
+      "step": 22470
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.27460524439811707,
+      "learning_rate": 8.253447210016363e-06,
+      "loss": 0.7667,
+      "step": 22475
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.2848700284957886,
+      "learning_rate": 8.219816867931218e-06,
+      "loss": 0.7629,
+      "step": 22480
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.2719096541404724,
+      "learning_rate": 8.18625325230861e-06,
+      "loss": 0.7007,
+      "step": 22485
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.2773694396018982,
+      "learning_rate": 8.152756378944708e-06,
+      "loss": 0.7708,
+      "step": 22490
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.2669026851654053,
+      "learning_rate": 8.119326263604281e-06,
+      "loss": 0.8003,
+      "step": 22495
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.27180686593055725,
+      "learning_rate": 8.085962922020611e-06,
+      "loss": 0.8165,
+      "step": 22500
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.2707700729370117,
+      "learning_rate": 8.052666369895622e-06,
+      "loss": 0.7817,
+      "step": 22505
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.2820728123188019,
+      "learning_rate": 8.019436622899727e-06,
+      "loss": 0.7616,
+      "step": 22510
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.27186763286590576,
+      "learning_rate": 7.986273696672019e-06,
+      "loss": 0.7928,
+      "step": 22515
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.29205748438835144,
+      "learning_rate": 7.953177606820044e-06,
+      "loss": 0.8722,
+      "step": 22520
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.2753778398036957,
+      "learning_rate": 7.92014836891991e-06,
+      "loss": 0.6978,
+      "step": 22525
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.2816515564918518,
+      "learning_rate": 7.887185998516266e-06,
+      "loss": 0.7766,
+      "step": 22530
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.2690742313861847,
+      "learning_rate": 7.85429051112232e-06,
+      "loss": 0.7389,
+      "step": 22535
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.26442471146583557,
+      "learning_rate": 7.821461922219769e-06,
+      "loss": 0.8625,
+      "step": 22540
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.2712923288345337,
+      "learning_rate": 7.788700247258855e-06,
+      "loss": 0.734,
+      "step": 22545
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.2940865159034729,
+      "learning_rate": 7.756005501658297e-06,
+      "loss": 0.7485,
+      "step": 22550
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.25925499200820923,
+      "learning_rate": 7.723377700805316e-06,
+      "loss": 0.7913,
+      "step": 22555
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.23880963027477264,
+      "learning_rate": 7.690816860055648e-06,
+      "loss": 0.7763,
+      "step": 22560
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.2328793704509735,
+      "learning_rate": 7.658322994733517e-06,
+      "loss": 0.7291,
+      "step": 22565
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.23459170758724213,
+      "learning_rate": 7.625896120131575e-06,
+      "loss": 0.8022,
+      "step": 22570
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.26686781644821167,
+      "learning_rate": 7.593536251511045e-06,
+      "loss": 0.751,
+      "step": 22575
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.2697281837463379,
+      "learning_rate": 7.5612434041015305e-06,
+      "loss": 0.737,
+      "step": 22580
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.30781927704811096,
+      "learning_rate": 7.529017593101105e-06,
+      "loss": 0.8416,
+      "step": 22585
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.2810058891773224,
+      "learning_rate": 7.496858833676306e-06,
+      "loss": 0.7729,
+      "step": 22590
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.24234548211097717,
+      "learning_rate": 7.464767140962124e-06,
+      "loss": 0.7394,
+      "step": 22595
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.263785719871521,
+      "learning_rate": 7.432742530061997e-06,
+      "loss": 0.8007,
+      "step": 22600
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.27168524265289307,
+      "learning_rate": 7.400785016047733e-06,
+      "loss": 0.8045,
+      "step": 22605
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.2800976634025574,
+      "learning_rate": 7.3688946139596025e-06,
+      "loss": 0.8114,
+      "step": 22610
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.29521191120147705,
+      "learning_rate": 7.3370713388063165e-06,
+      "loss": 0.7672,
+      "step": 22615
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.27929726243019104,
+      "learning_rate": 7.30531520556496e-06,
+      "loss": 0.7258,
+      "step": 22620
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.25764504075050354,
+      "learning_rate": 7.273626229180995e-06,
+      "loss": 0.8177,
+      "step": 22625
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.26366397738456726,
+      "learning_rate": 7.242004424568309e-06,
+      "loss": 0.7124,
+      "step": 22630
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.30640193819999695,
+      "learning_rate": 7.210449806609197e-06,
+      "loss": 0.8062,
+      "step": 22635
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.2728465795516968,
+      "learning_rate": 7.178962390154314e-06,
+      "loss": 0.8824,
+      "step": 22640
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.3235665559768677,
+      "learning_rate": 7.1475421900226705e-06,
+      "loss": 0.6664,
+      "step": 22645
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.3118913471698761,
+      "learning_rate": 7.116189221001622e-06,
+      "loss": 0.7975,
+      "step": 22650
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.2510865032672882,
+      "learning_rate": 7.084903497846983e-06,
+      "loss": 0.7969,
+      "step": 22655
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.28625842928886414,
+      "learning_rate": 7.053685035282808e-06,
+      "loss": 0.7862,
+      "step": 22660
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.30692657828330994,
+      "learning_rate": 7.02253384800156e-06,
+      "loss": 0.6927,
+      "step": 22665
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.30167925357818604,
+      "learning_rate": 6.9914499506640135e-06,
+      "loss": 0.7441,
+      "step": 22670
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.30318981409072876,
+      "learning_rate": 6.960433357899281e-06,
+      "loss": 0.8208,
+      "step": 22675
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.268376886844635,
+      "learning_rate": 6.929484084304837e-06,
+      "loss": 0.732,
+      "step": 22680
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.26787877082824707,
+      "learning_rate": 6.89860214444638e-06,
+      "loss": 0.7454,
+      "step": 22685
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.29671844840049744,
+      "learning_rate": 6.8677875528580176e-06,
+      "loss": 0.767,
+      "step": 22690
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.27219846844673157,
+      "learning_rate": 6.8370403240421146e-06,
+      "loss": 0.7894,
+      "step": 22695
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.3147854804992676,
+      "learning_rate": 6.806360472469313e-06,
+      "loss": 0.8088,
+      "step": 22700
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.27097561955451965,
+      "learning_rate": 6.775748012578597e-06,
+      "loss": 0.756,
+      "step": 22705
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.25546738505363464,
+      "learning_rate": 6.745202958777174e-06,
+      "loss": 0.8471,
+      "step": 22710
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.27763062715530396,
+      "learning_rate": 6.714725325440595e-06,
+      "loss": 0.7465,
+      "step": 22715
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.265898197889328,
+      "learning_rate": 6.684315126912654e-06,
+      "loss": 0.7873,
+      "step": 22720
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.292896032333374,
+      "learning_rate": 6.653972377505368e-06,
+      "loss": 0.7363,
+      "step": 22725
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.2958504557609558,
+      "learning_rate": 6.623697091499031e-06,
+      "loss": 0.7821,
+      "step": 22730
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.28079283237457275,
+      "learning_rate": 6.5934892831422616e-06,
+      "loss": 0.7395,
+      "step": 22735
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.26290813088417053,
+      "learning_rate": 6.563348966651805e-06,
+      "loss": 0.7934,
+      "step": 22740
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.3071194291114807,
+      "learning_rate": 6.533276156212697e-06,
+      "loss": 0.8421,
+      "step": 22745
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.2782292366027832,
+      "learning_rate": 6.503270865978216e-06,
+      "loss": 0.749,
+      "step": 22750
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.2784491181373596,
+      "learning_rate": 6.47333311006985e-06,
+      "loss": 0.7861,
+      "step": 22755
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.2715974450111389,
+      "learning_rate": 6.443462902577296e-06,
+      "loss": 0.8529,
+      "step": 22760
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.2865915894508362,
+      "learning_rate": 6.413660257558473e-06,
+      "loss": 0.8501,
+      "step": 22765
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.30358344316482544,
+      "learning_rate": 6.383925189039479e-06,
+      "loss": 0.7397,
+      "step": 22770
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.2950074076652527,
+      "learning_rate": 6.3542577110146375e-06,
+      "loss": 0.781,
+      "step": 22775
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.2862589955329895,
+      "learning_rate": 6.324657837446445e-06,
+      "loss": 0.7687,
+      "step": 22780
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.2698042392730713,
+      "learning_rate": 6.295125582265575e-06,
+      "loss": 0.8329,
+      "step": 22785
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.2783750295639038,
+      "learning_rate": 6.265660959370894e-06,
+      "loss": 0.7668,
+      "step": 22790
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.2662694752216339,
+      "learning_rate": 6.236263982629441e-06,
+      "loss": 0.7561,
+      "step": 22795
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.35889652371406555,
+      "learning_rate": 6.206934665876417e-06,
+      "loss": 0.8031,
+      "step": 22800
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.2936451733112335,
+      "learning_rate": 6.177673022915147e-06,
+      "loss": 0.7609,
+      "step": 22805
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.299941748380661,
+      "learning_rate": 6.1484790675171315e-06,
+      "loss": 0.6878,
+      "step": 22810
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.30326518416404724,
+      "learning_rate": 6.119352813422046e-06,
+      "loss": 0.8334,
+      "step": 22815
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.2659267783164978,
+      "learning_rate": 6.09029427433766e-06,
+      "loss": 0.7561,
+      "step": 22820
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.2844538986682892,
+      "learning_rate": 6.061303463939882e-06,
+      "loss": 0.8201,
+      "step": 22825
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.274046391248703,
+      "learning_rate": 6.032380395872732e-06,
+      "loss": 0.7778,
+      "step": 22830
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.25698041915893555,
+      "learning_rate": 6.003525083748406e-06,
+      "loss": 0.7376,
+      "step": 22835
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.2826564908027649,
+      "learning_rate": 5.97473754114714e-06,
+      "loss": 0.8031,
+      "step": 22840
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.2923043370246887,
+      "learning_rate": 5.946017781617329e-06,
+      "loss": 0.7715,
+      "step": 22845
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.2728569507598877,
+      "learning_rate": 5.917365818675396e-06,
+      "loss": 0.8772,
+      "step": 22850
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.2695605754852295,
+      "learning_rate": 5.888781665805986e-06,
+      "loss": 0.9072,
+      "step": 22855
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.25165680050849915,
+      "learning_rate": 5.860265336461689e-06,
+      "loss": 0.6898,
+      "step": 22860
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.2680535614490509,
+      "learning_rate": 5.8318168440632695e-06,
+      "loss": 0.8437,
+      "step": 22865
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.28424060344696045,
+      "learning_rate": 5.8034362019995e-06,
+      "loss": 0.7131,
+      "step": 22870
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.25301888585090637,
+      "learning_rate": 5.775123423627298e-06,
+      "loss": 0.7976,
+      "step": 22875
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.27347350120544434,
+      "learning_rate": 5.746878522271553e-06,
+      "loss": 0.711,
+      "step": 22880
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.26733747124671936,
+      "learning_rate": 5.718701511225299e-06,
+      "loss": 0.6827,
+      "step": 22885
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.2814013361930847,
+      "learning_rate": 5.690592403749511e-06,
+      "loss": 0.7927,
+      "step": 22890
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.264848530292511,
+      "learning_rate": 5.662551213073324e-06,
+      "loss": 0.9032,
+      "step": 22895
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.28370270133018494,
+      "learning_rate": 5.634577952393848e-06,
+      "loss": 0.7843,
+      "step": 22900
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.2731534242630005,
+      "learning_rate": 5.606672634876203e-06,
+      "loss": 0.8433,
+      "step": 22905
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.24351301789283752,
+      "learning_rate": 5.578835273653581e-06,
+      "loss": 0.7116,
+      "step": 22910
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.29447489976882935,
+      "learning_rate": 5.551065881827138e-06,
+      "loss": 0.7504,
+      "step": 22915
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.2870446443557739,
+      "learning_rate": 5.523364472466118e-06,
+      "loss": 0.7319,
+      "step": 22920
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.2608989179134369,
+      "learning_rate": 5.495731058607677e-06,
+      "loss": 0.8248,
+      "step": 22925
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.2600563168525696,
+      "learning_rate": 5.468165653257028e-06,
+      "loss": 0.8477,
+      "step": 22930
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.2777158319950104,
+      "learning_rate": 5.440668269387394e-06,
+      "loss": 0.7868,
+      "step": 22935
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.2984352707862854,
+      "learning_rate": 5.4132389199399384e-06,
+      "loss": 0.7796,
+      "step": 22940
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.3199765682220459,
+      "learning_rate": 5.385877617823819e-06,
+      "loss": 0.8571,
+      "step": 22945
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.2830100953578949,
+      "learning_rate": 5.35858437591617e-06,
+      "loss": 0.775,
+      "step": 22950
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.23226149380207062,
+      "learning_rate": 5.33135920706213e-06,
+      "loss": 0.706,
+      "step": 22955
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.28286078572273254,
+      "learning_rate": 5.304202124074736e-06,
+      "loss": 0.7191,
+      "step": 22960
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.3073217570781708,
+      "learning_rate": 5.277113139735012e-06,
+      "loss": 0.7106,
+      "step": 22965
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.2731364667415619,
+      "learning_rate": 5.250092266791944e-06,
+      "loss": 0.7427,
+      "step": 22970
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.2613232135772705,
+      "learning_rate": 5.22313951796246e-06,
+      "loss": 0.7188,
+      "step": 22975
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.25765460729599,
+      "learning_rate": 5.196254905931413e-06,
+      "loss": 0.9263,
+      "step": 22980
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.29852619767189026,
+      "learning_rate": 5.169438443351581e-06,
+      "loss": 0.6889,
+      "step": 22985
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.2636108100414276,
+      "learning_rate": 5.142690142843703e-06,
+      "loss": 0.6934,
+      "step": 22990
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.2629612684249878,
+      "learning_rate": 5.116010016996392e-06,
+      "loss": 0.8041,
+      "step": 22995
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.2794385552406311,
+      "learning_rate": 5.089398078366219e-06,
+      "loss": 0.8116,
+      "step": 23000
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.2664497494697571,
+      "learning_rate": 5.062854339477634e-06,
+      "loss": 0.8767,
+      "step": 23005
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.30518075823783875,
+      "learning_rate": 5.036378812823028e-06,
+      "loss": 0.7819,
+      "step": 23010
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.2723018229007721,
+      "learning_rate": 5.0099715108626485e-06,
+      "loss": 0.82,
+      "step": 23015
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.27372339367866516,
+      "learning_rate": 4.983632446024638e-06,
+      "loss": 0.8301,
+      "step": 23020
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.29022863507270813,
+      "learning_rate": 4.957361630705031e-06,
+      "loss": 0.7167,
+      "step": 23025
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.2727348804473877,
+      "learning_rate": 4.931159077267771e-06,
+      "loss": 0.8002,
+      "step": 23030
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.2919345200061798,
+      "learning_rate": 4.905024798044627e-06,
+      "loss": 0.7175,
+      "step": 23035
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.25965866446495056,
+      "learning_rate": 4.878958805335276e-06,
+      "loss": 0.7755,
+      "step": 23040
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.4287550151348114,
+      "learning_rate": 4.852961111407239e-06,
+      "loss": 0.787,
+      "step": 23045
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.30621278285980225,
+      "learning_rate": 4.827031728495878e-06,
+      "loss": 0.8296,
+      "step": 23050
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.29937830567359924,
+      "learning_rate": 4.801170668804433e-06,
+      "loss": 0.7147,
+      "step": 23055
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.2686162292957306,
+      "learning_rate": 4.775377944503983e-06,
+      "loss": 0.8084,
+      "step": 23060
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.2906171381473541,
+      "learning_rate": 4.749653567733402e-06,
+      "loss": 0.7937,
+      "step": 23065
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.3064448833465576,
+      "learning_rate": 4.72399755059949e-06,
+      "loss": 0.758,
+      "step": 23070
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.3103048801422119,
+      "learning_rate": 4.698409905176803e-06,
+      "loss": 0.7798,
+      "step": 23075
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.3002449870109558,
+      "learning_rate": 4.672890643507727e-06,
+      "loss": 0.8077,
+      "step": 23080
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.26185083389282227,
+      "learning_rate": 4.647439777602469e-06,
+      "loss": 0.6943,
+      "step": 23085
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.2598731517791748,
+      "learning_rate": 4.6220573194390655e-06,
+      "loss": 0.7912,
+      "step": 23090
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.283272385597229,
+      "learning_rate": 4.596743280963344e-06,
+      "loss": 0.7841,
+      "step": 23095
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.29078394174575806,
+      "learning_rate": 4.571497674088925e-06,
+      "loss": 0.8184,
+      "step": 23100
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.26592543721199036,
+      "learning_rate": 4.546320510697221e-06,
+      "loss": 0.7237,
+      "step": 23105
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.2742617130279541,
+      "learning_rate": 4.52121180263747e-06,
+      "loss": 0.7937,
+      "step": 23110
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.2708317041397095,
+      "learning_rate": 4.496171561726636e-06,
+      "loss": 0.8146,
+      "step": 23115
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.2601563036441803,
+      "learning_rate": 4.471199799749508e-06,
+      "loss": 0.719,
+      "step": 23120
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.2519415318965912,
+      "learning_rate": 4.446296528458604e-06,
+      "loss": 0.8397,
+      "step": 23125
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.2795381247997284,
+      "learning_rate": 4.421461759574247e-06,
+      "loss": 0.708,
+      "step": 23130
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.2789604067802429,
+      "learning_rate": 4.396695504784503e-06,
+      "loss": 0.7077,
+      "step": 23135
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.30929040908813477,
+      "learning_rate": 4.371997775745184e-06,
+      "loss": 0.6999,
+      "step": 23140
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.2771584689617157,
+      "learning_rate": 4.347368584079858e-06,
+      "loss": 0.8967,
+      "step": 23145
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.290021687746048,
+      "learning_rate": 4.322807941379869e-06,
+      "loss": 0.7981,
+      "step": 23150
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.2715446352958679,
+      "learning_rate": 4.298315859204254e-06,
+      "loss": 0.7353,
+      "step": 23155
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.27552729845046997,
+      "learning_rate": 4.273892349079794e-06,
+      "loss": 0.8716,
+      "step": 23160
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.2995644509792328,
+      "learning_rate": 4.249537422500992e-06,
+      "loss": 0.7564,
+      "step": 23165
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.3228721022605896,
+      "learning_rate": 4.225251090930132e-06,
+      "loss": 0.7422,
+      "step": 23170
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.2784271538257599,
+      "learning_rate": 4.201033365797119e-06,
+      "loss": 0.7439,
+      "step": 23175
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.2869255840778351,
+      "learning_rate": 4.176884258499652e-06,
+      "loss": 0.8764,
+      "step": 23180
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.2846572697162628,
+      "learning_rate": 4.152803780403058e-06,
+      "loss": 0.8168,
+      "step": 23185
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.2854968011379242,
+      "learning_rate": 4.12879194284047e-06,
+      "loss": 0.84,
+      "step": 23190
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.2596077620983124,
+      "learning_rate": 4.104848757112616e-06,
+      "loss": 0.7741,
+      "step": 23195
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.2705884575843811,
+      "learning_rate": 4.080974234487966e-06,
+      "loss": 0.6468,
+      "step": 23200
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.2820831835269928,
+      "learning_rate": 4.057168386202681e-06,
+      "loss": 0.7343,
+      "step": 23205
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.2777920663356781,
+      "learning_rate": 4.033431223460548e-06,
+      "loss": 0.825,
+      "step": 23210
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.25113433599472046,
+      "learning_rate": 4.0097627574330825e-06,
+      "loss": 0.8486,
+      "step": 23215
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.26045823097229004,
+      "learning_rate": 3.9861629992594405e-06,
+      "loss": 0.8892,
+      "step": 23220
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.2702331840991974,
+      "learning_rate": 3.962631960046453e-06,
+      "loss": 0.8179,
+      "step": 23225
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.2872564494609833,
+      "learning_rate": 3.939169650868645e-06,
+      "loss": 0.7651,
+      "step": 23230
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.28898364305496216,
+      "learning_rate": 3.915776082768118e-06,
+      "loss": 0.7968,
+      "step": 23235
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.2918056845664978,
+      "learning_rate": 3.8924512667546645e-06,
+      "loss": 0.7513,
+      "step": 23240
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.28438860177993774,
+      "learning_rate": 3.86919521380572e-06,
+      "loss": 0.8384,
+      "step": 23245
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.29474344849586487,
+      "learning_rate": 3.8460079348663795e-06,
+      "loss": 0.8156,
+      "step": 23250
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.28143933415412903,
+      "learning_rate": 3.82288944084933e-06,
+      "loss": 0.7344,
+      "step": 23255
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.28403791785240173,
+      "learning_rate": 3.7998397426349024e-06,
+      "loss": 0.7604,
+      "step": 23260
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.29978764057159424,
+      "learning_rate": 3.7768588510710353e-06,
+      "loss": 0.7076,
+      "step": 23265
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.2770501375198364,
+      "learning_rate": 3.753946776973327e-06,
+      "loss": 0.8365,
+      "step": 23270
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.27618712186813354,
+      "learning_rate": 3.7311035311249348e-06,
+      "loss": 0.7706,
+      "step": 23275
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.2640831768512726,
+      "learning_rate": 3.7083291242766764e-06,
+      "loss": 0.7948,
+      "step": 23280
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.2883791923522949,
+      "learning_rate": 3.685623567146895e-06,
+      "loss": 0.8043,
+      "step": 23285
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.280841201543808,
+      "learning_rate": 3.6629868704216415e-06,
+      "loss": 0.9078,
+      "step": 23290
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.25826311111450195,
+      "learning_rate": 3.640419044754461e-06,
+      "loss": 0.8755,
+      "step": 23295
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.2830309271812439,
+      "learning_rate": 3.6179201007665413e-06,
+      "loss": 0.8145,
+      "step": 23300
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.290726900100708,
+      "learning_rate": 3.5954900490465956e-06,
+      "loss": 0.8466,
+      "step": 23305
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.28319695591926575,
+      "learning_rate": 3.5731289001509954e-06,
+      "loss": 0.7302,
+      "step": 23310
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.27678796648979187,
+      "learning_rate": 3.5508366646036236e-06,
+      "loss": 0.8587,
+      "step": 23315
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.26364395022392273,
+      "learning_rate": 3.528613352895937e-06,
+      "loss": 0.858,
+      "step": 23320
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.2955693006515503,
+      "learning_rate": 3.506458975486953e-06,
+      "loss": 0.6728,
+      "step": 23325
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.272214412689209,
+      "learning_rate": 3.484373542803298e-06,
+      "loss": 0.794,
+      "step": 23330
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.2752835154533386,
+      "learning_rate": 3.4623570652390743e-06,
+      "loss": 0.6453,
+      "step": 23335
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.28152868151664734,
+      "learning_rate": 3.440409553155993e-06,
+      "loss": 0.7392,
+      "step": 23340
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.25148865580558777,
+      "learning_rate": 3.4185310168832748e-06,
+      "loss": 0.777,
+      "step": 23345
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.2435861974954605,
+      "learning_rate": 3.3967214667176666e-06,
+      "loss": 0.7149,
+      "step": 23350
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.33531370759010315,
+      "learning_rate": 3.3749809129234752e-06,
+      "loss": 0.7284,
+      "step": 23355
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.30802983045578003,
+      "learning_rate": 3.353309365732548e-06,
+      "loss": 0.714,
+      "step": 23360
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.2919432818889618,
+      "learning_rate": 3.3317068353441945e-06,
+      "loss": 0.8152,
+      "step": 23365
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.2593708634376526,
+      "learning_rate": 3.310173331925331e-06,
+      "loss": 0.6821,
+      "step": 23370
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.27010101079940796,
+      "learning_rate": 3.288708865610318e-06,
+      "loss": 0.753,
+      "step": 23375
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.28530481457710266,
+      "learning_rate": 3.267313446501041e-06,
+      "loss": 0.7806,
+      "step": 23380
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.27056190371513367,
+      "learning_rate": 3.245987084666879e-06,
+      "loss": 0.6671,
+      "step": 23385
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.33442938327789307,
+      "learning_rate": 3.2247297901447534e-06,
+      "loss": 0.7138,
+      "step": 23390
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.2769376039505005,
+      "learning_rate": 3.2035415729390613e-06,
+      "loss": 0.838,
+      "step": 23395
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.29188603162765503,
+      "learning_rate": 3.1824224430216425e-06,
+      "loss": 0.7758,
+      "step": 23400
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.27181747555732727,
+      "learning_rate": 3.161372410331897e-06,
+      "loss": 0.8393,
+      "step": 23405
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.2749415338039398,
+      "learning_rate": 3.1403914847766497e-06,
+      "loss": 0.8015,
+      "step": 23410
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.28852447867393494,
+      "learning_rate": 3.1194796762302353e-06,
+      "loss": 0.8061,
+      "step": 23415
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.2778397798538208,
+      "learning_rate": 3.0986369945344312e-06,
+      "loss": 0.829,
+      "step": 23420
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.3023031949996948,
+      "learning_rate": 3.0778634494984912e-06,
+      "loss": 0.7389,
+      "step": 23425
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.3010731339454651,
+      "learning_rate": 3.0571590508991607e-06,
+      "loss": 0.8619,
+      "step": 23430
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.28016921877861023,
+      "learning_rate": 3.0365238084805955e-06,
+      "loss": 0.7831,
+      "step": 23435
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.2581021189689636,
+      "learning_rate": 3.015957731954427e-06,
+      "loss": 0.7122,
+      "step": 23440
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.2601596713066101,
+      "learning_rate": 2.9954608309997296e-06,
+      "loss": 0.784,
+      "step": 23445
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.2719314992427826,
+      "learning_rate": 2.9750331152630535e-06,
+      "loss": 0.7701,
+      "step": 23450
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.25035297870635986,
+      "learning_rate": 2.9546745943583418e-06,
+      "loss": 0.785,
+      "step": 23455
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.2835271656513214,
+      "learning_rate": 2.93438527786698e-06,
+      "loss": 0.7819,
+      "step": 23460
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.2592954635620117,
+      "learning_rate": 2.914165175337796e-06,
+      "loss": 0.7281,
+      "step": 23465
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.29438310861587524,
+      "learning_rate": 2.8940142962870784e-06,
+      "loss": 0.7564,
+      "step": 23470
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.2590184807777405,
+      "learning_rate": 2.873932650198457e-06,
+      "loss": 0.862,
+      "step": 23475
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.2566207945346832,
+      "learning_rate": 2.853920246523023e-06,
+      "loss": 0.7873,
+      "step": 23480
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.2697911560535431,
+      "learning_rate": 2.8339770946793073e-06,
+      "loss": 0.8254,
+      "step": 23485
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.2645513117313385,
+      "learning_rate": 2.814103204053203e-06,
+      "loss": 0.7198,
+      "step": 23490
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.294327974319458,
+      "learning_rate": 2.794298583998028e-06,
+      "loss": 0.8079,
+      "step": 23495
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.26449063420295715,
+      "learning_rate": 2.77456324383446e-06,
+      "loss": 0.6838,
+      "step": 23500
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.27684205770492554,
+      "learning_rate": 2.7548971928506693e-06,
+      "loss": 0.7777,
+      "step": 23505
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.29637137055397034,
+      "learning_rate": 2.7353004403021017e-06,
+      "loss": 0.8073,
+      "step": 23510
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.297203004360199,
+      "learning_rate": 2.7157729954116634e-06,
+      "loss": 0.671,
+      "step": 23515
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.27864888310432434,
+      "learning_rate": 2.6963148673696034e-06,
+      "loss": 0.7304,
+      "step": 23520
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.30311015248298645,
+      "learning_rate": 2.676926065333562e-06,
+      "loss": 0.7151,
+      "step": 23525
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.2594775855541229,
+      "learning_rate": 2.657606598428591e-06,
+      "loss": 0.8637,
+      "step": 23530
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.2853897213935852,
+      "learning_rate": 2.6383564757470168e-06,
+      "loss": 0.7382,
+      "step": 23535
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.27156689763069153,
+      "learning_rate": 2.6191757063486252e-06,
+      "loss": 0.7312,
+      "step": 23540
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.26867052912712097,
+      "learning_rate": 2.6000642992605127e-06,
+      "loss": 0.7705,
+      "step": 23545
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.27880826592445374,
+      "learning_rate": 2.581022263477134e-06,
+      "loss": 0.703,
+      "step": 23550
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.26525866985321045,
+      "learning_rate": 2.5620496079603205e-06,
+      "loss": 0.7129,
+      "step": 23555
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.36034145951271057,
+      "learning_rate": 2.5431463416392296e-06,
+      "loss": 0.7513,
+      "step": 23560
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.28856807947158813,
+      "learning_rate": 2.5243124734103616e-06,
+      "loss": 0.7044,
+      "step": 23565
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.27370092272758484,
+      "learning_rate": 2.5055480121375426e-06,
+      "loss": 0.7817,
+      "step": 23570
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.28846797347068787,
+      "learning_rate": 2.486852966651992e-06,
+      "loss": 0.7372,
+      "step": 23575
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.26241225004196167,
+      "learning_rate": 2.468227345752155e-06,
+      "loss": 0.7602,
+      "step": 23580
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.30139365792274475,
+      "learning_rate": 2.4496711582039365e-06,
+      "loss": 0.7023,
+      "step": 23585
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.29070574045181274,
+      "learning_rate": 2.4311844127404668e-06,
+      "loss": 0.7737,
+      "step": 23590
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.3289770781993866,
+      "learning_rate": 2.41276711806222e-06,
+      "loss": 0.8992,
+      "step": 23595
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.2710113525390625,
+      "learning_rate": 2.394419282836979e-06,
+      "loss": 0.7763,
+      "step": 23600
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.31186643242836,
+      "learning_rate": 2.3761409156998532e-06,
+      "loss": 0.7533,
+      "step": 23605
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.2570800185203552,
+      "learning_rate": 2.357932025253262e-06,
+      "loss": 0.7039,
+      "step": 23610
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.2939087450504303,
+      "learning_rate": 2.3397926200668994e-06,
+      "loss": 0.8794,
+      "step": 23615
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.26302385330200195,
+      "learning_rate": 2.3217227086777533e-06,
+      "loss": 0.6978,
+      "step": 23620
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.27785399556159973,
+      "learning_rate": 2.3037222995901716e-06,
+      "loss": 0.7772,
+      "step": 23625
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.2980126738548279,
+      "learning_rate": 2.2857914012757107e-06,
+      "loss": 0.7843,
+      "step": 23630
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.30063748359680176,
+      "learning_rate": 2.267930022173253e-06,
+      "loss": 0.7381,
+      "step": 23635
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.2854771316051483,
+      "learning_rate": 2.250138170688942e-06,
+      "loss": 0.7182,
+      "step": 23640
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.2724713683128357,
+      "learning_rate": 2.2324158551962457e-06,
+      "loss": 0.8062,
+      "step": 23645
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.2653485834598541,
+      "learning_rate": 2.214763084035842e-06,
+      "loss": 0.6749,
+      "step": 23650
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.27311810851097107,
+      "learning_rate": 2.197179865515736e-06,
+      "loss": 0.6745,
+      "step": 23655
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.3039811849594116,
+      "learning_rate": 2.179666207911157e-06,
+      "loss": 0.7771,
+      "step": 23660
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.30827796459198,
+      "learning_rate": 2.1622221194646294e-06,
+      "loss": 0.7983,
+      "step": 23665
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.27435705065727234,
+      "learning_rate": 2.144847608385919e-06,
+      "loss": 0.9009,
+      "step": 23670
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.2926265299320221,
+      "learning_rate": 2.1275426828520347e-06,
+      "loss": 0.7032,
+      "step": 23675
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.30964577198028564,
+      "learning_rate": 2.110307351007262e-06,
+      "loss": 0.7401,
+      "step": 23680
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.2811949849128723,
+      "learning_rate": 2.0931416209631126e-06,
+      "loss": 0.6413,
+      "step": 23685
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.2701903283596039,
+      "learning_rate": 2.07604550079834e-06,
+      "loss": 0.8361,
+      "step": 23690
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.26354488730430603,
+      "learning_rate": 2.0590189985589755e-06,
+      "loss": 0.8271,
+      "step": 23695
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.2724680006504059,
+      "learning_rate": 2.0420621222582255e-06,
+      "loss": 0.7632,
+      "step": 23700
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.28979775309562683,
+      "learning_rate": 2.02517487987659e-06,
+      "loss": 0.7434,
+      "step": 23705
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.2960839867591858,
+      "learning_rate": 2.0083572793617274e-06,
+      "loss": 0.717,
+      "step": 23710
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.25892457365989685,
+      "learning_rate": 1.9916093286285904e-06,
+      "loss": 0.8393,
+      "step": 23715
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.2716827690601349,
+      "learning_rate": 1.9749310355592907e-06,
+      "loss": 0.8344,
+      "step": 23720
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.25879597663879395,
+      "learning_rate": 1.958322408003232e-06,
+      "loss": 0.7689,
+      "step": 23725
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.2955610752105713,
+      "learning_rate": 1.9417834537769463e-06,
+      "loss": 0.7773,
+      "step": 23730
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.2859145402908325,
+      "learning_rate": 1.925314180664239e-06,
+      "loss": 0.7632,
+      "step": 23735
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.28816118836402893,
+      "learning_rate": 1.9089145964160614e-06,
+      "loss": 0.8108,
+      "step": 23740
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.2830142080783844,
+      "learning_rate": 1.892584708750655e-06,
+      "loss": 0.7739,
+      "step": 23745
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.28230804204940796,
+      "learning_rate": 1.876324525353373e-06,
+      "loss": 0.8458,
+      "step": 23750
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.28245803713798523,
+      "learning_rate": 1.8601340538767938e-06,
+      "loss": 0.7592,
+      "step": 23755
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.29182082414627075,
+      "learning_rate": 1.8440133019407056e-06,
+      "loss": 0.701,
+      "step": 23760
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.2928227186203003,
+      "learning_rate": 1.8279622771320723e-06,
+      "loss": 0.7389,
+      "step": 23765
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.2605980932712555,
+      "learning_rate": 1.8119809870050351e-06,
+      "loss": 0.8737,
+      "step": 23770
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.27855974435806274,
+      "learning_rate": 1.7960694390809105e-06,
+      "loss": 0.8722,
+      "step": 23775
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.2521946430206299,
+      "learning_rate": 1.7802276408482086e-06,
+      "loss": 0.8727,
+      "step": 23780
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.2708735167980194,
+      "learning_rate": 1.7644555997626153e-06,
+      "loss": 0.844,
+      "step": 23785
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.26479655504226685,
+      "learning_rate": 1.7487533232469597e-06,
+      "loss": 0.801,
+      "step": 23790
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.32824069261550903,
+      "learning_rate": 1.7331208186912472e-06,
+      "loss": 0.7282,
+      "step": 23795
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.2727472186088562,
+      "learning_rate": 1.7175580934526756e-06,
+      "loss": 0.7332,
+      "step": 23800
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.2697487771511078,
+      "learning_rate": 1.7020651548555863e-06,
+      "loss": 0.8685,
+      "step": 23805
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.3158542513847351,
+      "learning_rate": 1.6866420101914468e-06,
+      "loss": 0.7077,
+      "step": 23810
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.2666032612323761,
+      "learning_rate": 1.6712886667189173e-06,
+      "loss": 0.784,
+      "step": 23815
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.28167471289634705,
+      "learning_rate": 1.6560051316637678e-06,
+      "loss": 0.6687,
+      "step": 23820
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.2855972945690155,
+      "learning_rate": 1.640791412218978e-06,
+      "loss": 0.8986,
+      "step": 23825
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.28779542446136475,
+      "learning_rate": 1.6256475155446037e-06,
+      "loss": 0.7325,
+      "step": 23830
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.27318522334098816,
+      "learning_rate": 1.61057344876786e-06,
+      "loss": 0.783,
+      "step": 23835
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.29986461997032166,
+      "learning_rate": 1.595569218983106e-06,
+      "loss": 0.8758,
+      "step": 23840
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.2557889223098755,
+      "learning_rate": 1.5806348332518593e-06,
+      "loss": 0.78,
+      "step": 23845
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.26995155215263367,
+      "learning_rate": 1.5657702986026976e-06,
+      "loss": 0.8988,
+      "step": 23850
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.2958948612213135,
+      "learning_rate": 1.550975622031375e-06,
+      "loss": 0.7705,
+      "step": 23855
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.27343350648880005,
+      "learning_rate": 1.5362508105007721e-06,
+      "loss": 0.8335,
+      "step": 23860
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.34859418869018555,
+      "learning_rate": 1.5215958709408783e-06,
+      "loss": 0.8144,
+      "step": 23865
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.26644057035446167,
+      "learning_rate": 1.50701081024876e-06,
+      "loss": 0.7872,
+      "step": 23870
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.26625731587409973,
+      "learning_rate": 1.4924956352886763e-06,
+      "loss": 0.769,
+      "step": 23875
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.28103673458099365,
+      "learning_rate": 1.478050352891913e-06,
+      "loss": 0.8232,
+      "step": 23880
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.3763957619667053,
+      "learning_rate": 1.463674969856915e-06,
+      "loss": 0.7883,
+      "step": 23885
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.28736910223960876,
+      "learning_rate": 1.4493694929492206e-06,
+      "loss": 0.722,
+      "step": 23890
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.2688366770744324,
+      "learning_rate": 1.4351339289014608e-06,
+      "loss": 0.9024,
+      "step": 23895
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.27338284254074097,
+      "learning_rate": 1.420968284413343e-06,
+      "loss": 0.8657,
+      "step": 23900
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.2942239046096802,
+      "learning_rate": 1.4068725661517343e-06,
+      "loss": 0.7344,
+      "step": 23905
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.26615390181541443,
+      "learning_rate": 1.392846780750495e-06,
+      "loss": 0.7641,
+      "step": 23910
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.27061372995376587,
+      "learning_rate": 1.378890934810678e-06,
+      "loss": 0.7982,
+      "step": 23915
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.2560684084892273,
+      "learning_rate": 1.3650050349003294e-06,
+      "loss": 0.7434,
+      "step": 23920
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.2993808388710022,
+      "learning_rate": 1.3511890875546217e-06,
+      "loss": 0.7741,
+      "step": 23925
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.2594783902168274,
+      "learning_rate": 1.3374430992758033e-06,
+      "loss": 0.827,
+      "step": 23930
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.31000006198883057,
+      "learning_rate": 1.3237670765331998e-06,
+      "loss": 0.7506,
+      "step": 23935
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.3279038071632385,
+      "learning_rate": 1.310161025763179e-06,
+      "loss": 0.7283,
+      "step": 23940
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.2829456925392151,
+      "learning_rate": 1.2966249533692352e-06,
+      "loss": 0.7266,
+      "step": 23945
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.2958797812461853,
+      "learning_rate": 1.2831588657218728e-06,
+      "loss": 0.6592,
+      "step": 23950
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.2799309194087982,
+      "learning_rate": 1.2697627691586887e-06,
+      "loss": 0.7966,
+      "step": 23955
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.3030032217502594,
+      "learning_rate": 1.2564366699842899e-06,
+      "loss": 0.8914,
+      "step": 23960
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.32187244296073914,
+      "learning_rate": 1.2431805744704426e-06,
+      "loss": 0.8033,
+      "step": 23965
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.2817818820476532,
+      "learning_rate": 1.2299944888558732e-06,
+      "loss": 0.781,
+      "step": 23970
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.27116790413856506,
+      "learning_rate": 1.2194960316631197e-06,
+      "loss": 0.7533,
+      "step": 23975
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.3104703426361084,
+      "learning_rate": 1.2064359794837553e-06,
+      "loss": 0.8843,
+      "step": 23980
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.27435311675071716,
+      "learning_rate": 1.1934459544969023e-06,
+      "loss": 0.8259,
+      "step": 23985
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.29726967215538025,
+      "learning_rate": 1.1805259628160867e-06,
+      "loss": 0.7628,
+      "step": 23990
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.2689824104309082,
+      "learning_rate": 1.167676010521912e-06,
+      "loss": 0.8803,
+      "step": 23995
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.27188640832901,
+      "learning_rate": 1.154896103661973e-06,
+      "loss": 0.8005,
+      "step": 24000
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.2570241689682007,
+      "learning_rate": 1.1421862482509093e-06,
+      "loss": 0.7015,
+      "step": 24005
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.27499955892562866,
+      "learning_rate": 1.1295464502704187e-06,
+      "loss": 0.8028,
+      "step": 24010
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.30428576469421387,
+      "learning_rate": 1.11697671566921e-06,
+      "loss": 0.8024,
+      "step": 24015
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.3098774254322052,
+      "learning_rate": 1.1044770503630008e-06,
+      "loss": 0.7861,
+      "step": 24020
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.2888442575931549,
+      "learning_rate": 1.092047460234552e-06,
+      "loss": 0.7146,
+      "step": 24025
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.2686418294906616,
+      "learning_rate": 1.0796879511336676e-06,
+      "loss": 0.6743,
+      "step": 24030
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.26199212670326233,
+      "learning_rate": 1.0673985288771114e-06,
+      "loss": 0.7769,
+      "step": 24035
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.27240923047065735,
+      "learning_rate": 1.0551791992487068e-06,
+      "loss": 0.8166,
+      "step": 24040
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.2600533366203308,
+      "learning_rate": 1.0430299679992704e-06,
+      "loss": 0.7604,
+      "step": 24045
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.2539452612400055,
+      "learning_rate": 1.0309508408466617e-06,
+      "loss": 0.8341,
+      "step": 24050
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.23374611139297485,
+      "learning_rate": 1.018941823475683e-06,
+      "loss": 0.8019,
+      "step": 24055
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.2858569025993347,
+      "learning_rate": 1.0070029215381803e-06,
+      "loss": 0.8227,
+      "step": 24060
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.3056763708591461,
+      "learning_rate": 9.951341406530088e-07,
+      "loss": 0.7465,
+      "step": 24065
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.2973862290382385,
+      "learning_rate": 9.833354864060171e-07,
+      "loss": 0.6763,
+      "step": 24070
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.30157533288002014,
+      "learning_rate": 9.716069643500467e-07,
+      "loss": 0.8366,
+      "step": 24075
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.2873646318912506,
+      "learning_rate": 9.59948580004899e-07,
+      "loss": 0.6718,
+      "step": 24080
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.2701966464519501,
+      "learning_rate": 9.483603388574345e-07,
+      "loss": 0.8791,
+      "step": 24085
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.283039927482605,
+      "learning_rate": 9.368422463614411e-07,
+      "loss": 0.7677,
+      "step": 24090
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.3029240071773529,
+      "learning_rate": 9.253943079377157e-07,
+      "loss": 0.8087,
+      "step": 24095
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.2594771683216095,
+      "learning_rate": 9.14016528974032e-07,
+      "loss": 0.7987,
+      "step": 24100
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.2657313048839569,
+      "learning_rate": 9.027089148251731e-07,
+      "loss": 0.7931,
+      "step": 24105
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.26724809408187866,
+      "learning_rate": 8.914714708128657e-07,
+      "loss": 0.7604,
+      "step": 24110
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.2993227541446686,
+      "learning_rate": 8.803042022258289e-07,
+      "loss": 0.769,
+      "step": 24115
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.27348792552948,
+      "learning_rate": 8.692071143197588e-07,
+      "loss": 0.7721,
+      "step": 24120
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.2768844664096832,
+      "learning_rate": 8.581802123172776e-07,
+      "loss": 0.7656,
+      "step": 24125
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.2716030478477478,
+      "learning_rate": 8.472235014080508e-07,
+      "loss": 0.8586,
+      "step": 24130
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.2838304340839386,
+      "learning_rate": 8.363369867486369e-07,
+      "loss": 0.6991,
+      "step": 24135
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.2540549039840698,
+      "learning_rate": 8.255206734626207e-07,
+      "loss": 0.8098,
+      "step": 24140
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.30254021286964417,
+      "learning_rate": 8.147745666405137e-07,
+      "loss": 0.8074,
+      "step": 24145
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.33351510763168335,
+      "learning_rate": 8.040986713397867e-07,
+      "loss": 0.756,
+      "step": 24150
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.28228092193603516,
+      "learning_rate": 7.934929925848543e-07,
+      "loss": 0.724,
+      "step": 24155
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.290088951587677,
+      "learning_rate": 7.829575353671235e-07,
+      "loss": 0.7698,
+      "step": 24160
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.29970574378967285,
+      "learning_rate": 7.724923046449117e-07,
+      "loss": 0.7149,
+      "step": 24165
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.2955015301704407,
+      "learning_rate": 7.62097305343512e-07,
+      "loss": 0.8158,
+      "step": 24170
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.31175607442855835,
+      "learning_rate": 7.517725423551613e-07,
+      "loss": 0.6608,
+      "step": 24175
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.3113124966621399,
+      "learning_rate": 7.415180205390392e-07,
+      "loss": 0.7003,
+      "step": 24180
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.25491559505462646,
+      "learning_rate": 7.313337447212519e-07,
+      "loss": 0.777,
+      "step": 24185
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.3029264509677887,
+      "learning_rate": 7.212197196948655e-07,
+      "loss": 0.7743,
+      "step": 24190
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.25185316801071167,
+      "learning_rate": 7.111759502198721e-07,
+      "loss": 0.7702,
+      "step": 24195
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.29424992203712463,
+      "learning_rate": 7.012024410232076e-07,
+      "loss": 0.7636,
+      "step": 24200
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.25971463322639465,
+      "learning_rate": 6.912991967987169e-07,
+      "loss": 0.792,
+      "step": 24205
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.2720402479171753,
+      "learning_rate": 6.814662222072387e-07,
+      "loss": 0.7822,
+      "step": 24210
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.2848743498325348,
+      "learning_rate": 6.717035218764543e-07,
+      "loss": 0.7481,
+      "step": 24215
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.31447309255599976,
+      "learning_rate": 6.620111004010387e-07,
+      "loss": 0.7175,
+      "step": 24220
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.2868044376373291,
+      "learning_rate": 6.52388962342576e-07,
+      "loss": 0.7447,
+      "step": 24225
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.281565397977829,
+      "learning_rate": 6.428371122295273e-07,
+      "loss": 0.739,
+      "step": 24230
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.27926504611968994,
+      "learning_rate": 6.333555545573299e-07,
+      "loss": 0.7863,
+      "step": 24235
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.2796558439731598,
+      "learning_rate": 6.23944293788331e-07,
+      "loss": 0.7584,
+      "step": 24240
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.27483847737312317,
+      "learning_rate": 6.146033343517709e-07,
+      "loss": 0.777,
+      "step": 24245
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.27021974325180054,
+      "learning_rate": 6.053326806438163e-07,
+      "loss": 0.8146,
+      "step": 24250
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.28549179434776306,
+      "learning_rate": 5.961323370275439e-07,
+      "loss": 0.8211,
+      "step": 24255
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.2960338592529297,
+      "learning_rate": 5.870023078329567e-07,
+      "loss": 0.73,
+      "step": 24260
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.2839416563510895,
+      "learning_rate": 5.779425973569174e-07,
+      "loss": 0.791,
+      "step": 24265
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.28548988699913025,
+      "learning_rate": 5.689532098632487e-07,
+      "loss": 0.7879,
+      "step": 24270
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.289668470621109,
+      "learning_rate": 5.6003414958265e-07,
+      "loss": 0.7007,
+      "step": 24275
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.2637322247028351,
+      "learning_rate": 5.511854207127298e-07,
+      "loss": 0.8707,
+      "step": 24280
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.27071821689605713,
+      "learning_rate": 5.424070274179904e-07,
+      "loss": 0.7963,
+      "step": 24285
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.2783946692943573,
+      "learning_rate": 5.336989738298436e-07,
+      "loss": 0.7763,
+      "step": 24290
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.24435043334960938,
+      "learning_rate": 5.250612640465779e-07,
+      "loss": 0.7297,
+      "step": 24295
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.26858675479888916,
+      "learning_rate": 5.164939021334081e-07,
+      "loss": 0.7109,
+      "step": 24300
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.28042492270469666,
+      "learning_rate": 5.079968921223754e-07,
+      "loss": 0.7229,
+      "step": 24305
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.26023703813552856,
+      "learning_rate": 4.995702380124977e-07,
+      "loss": 0.695,
+      "step": 24310
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.27599677443504333,
+      "learning_rate": 4.912139437696195e-07,
+      "loss": 0.7659,
+      "step": 24315
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.2630544602870941,
+      "learning_rate": 4.829280133264779e-07,
+      "loss": 0.7636,
+      "step": 24320
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.30303630232810974,
+      "learning_rate": 4.747124505827371e-07,
+      "loss": 0.6761,
+      "step": 24325
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.26069948077201843,
+      "learning_rate": 4.665672594048875e-07,
+      "loss": 0.8025,
+      "step": 24330
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.2887049615383148,
+      "learning_rate": 4.5849244362634597e-07,
+      "loss": 0.8476,
+      "step": 24335
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.2869950532913208,
+      "learning_rate": 4.504880070473726e-07,
+      "loss": 0.8227,
+      "step": 24340
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.2907213866710663,
+      "learning_rate": 4.4255395343513744e-07,
+      "loss": 0.8843,
+      "step": 24345
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.27622485160827637,
+      "learning_rate": 4.3469028652365346e-07,
+      "loss": 0.7809,
+      "step": 24350
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.3061695098876953,
+      "learning_rate": 4.268970100138269e-07,
+      "loss": 0.6801,
+      "step": 24355
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.26443079113960266,
+      "learning_rate": 4.1917412757345724e-07,
+      "loss": 0.7058,
+      "step": 24360
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.30546635389328003,
+      "learning_rate": 4.1152164283715373e-07,
+      "loss": 0.8242,
+      "step": 24365
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.265575647354126,
+      "learning_rate": 4.039395594064521e-07,
+      "loss": 0.7344,
+      "step": 24370
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.29234451055526733,
+      "learning_rate": 3.9642788084971454e-07,
+      "loss": 0.7067,
+      "step": 24375
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.25120893120765686,
+      "learning_rate": 3.889866107021966e-07,
+      "loss": 0.7673,
+      "step": 24380
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.32437393069267273,
+      "learning_rate": 3.816157524660135e-07,
+      "loss": 0.6869,
+      "step": 24385
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.2962343096733093,
+      "learning_rate": 3.7431530961014034e-07,
+      "loss": 0.7815,
+      "step": 24390
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.2750551700592041,
+      "learning_rate": 3.6708528557037873e-07,
+      "loss": 0.8175,
+      "step": 24395
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.2620263993740082,
+      "learning_rate": 3.5992568374945685e-07,
+      "loss": 0.7162,
+      "step": 24400
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.26457253098487854,
+      "learning_rate": 3.528365075168793e-07,
+      "loss": 0.6064,
+      "step": 24405
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.29997846484184265,
+      "learning_rate": 3.4581776020907725e-07,
+      "loss": 0.8372,
+      "step": 24410
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.2950671315193176,
+      "learning_rate": 3.388694451293084e-07,
+      "loss": 0.7008,
+      "step": 24415
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.28241240978240967,
+      "learning_rate": 3.3199156554765684e-07,
+      "loss": 0.7895,
+      "step": 24420
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.2898365557193756,
+      "learning_rate": 3.251841247011167e-07,
+      "loss": 0.6871,
+      "step": 24425
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.2402508705854416,
+      "learning_rate": 3.1844712579345843e-07,
+      "loss": 0.8315,
+      "step": 24430
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.2538807988166809,
+      "learning_rate": 3.1178057199536256e-07,
+      "loss": 0.7951,
+      "step": 24435
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.24774649739265442,
+      "learning_rate": 3.051844664443026e-07,
+      "loss": 0.7521,
+      "step": 24440
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.29270678758621216,
+      "learning_rate": 2.9865881224466206e-07,
+      "loss": 0.744,
+      "step": 24445
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.2553872764110565,
+      "learning_rate": 2.9220361246761746e-07,
+      "loss": 0.8013,
+      "step": 24450
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.24886707961559296,
+      "learning_rate": 2.858188701511721e-07,
+      "loss": 0.7174,
+      "step": 24455
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.23313623666763306,
+      "learning_rate": 2.7950458830023893e-07,
+      "loss": 0.7931,
+      "step": 24460
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.25668811798095703,
+      "learning_rate": 2.7326076988650747e-07,
+      "loss": 0.7746,
+      "step": 24465
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.2648046016693115,
+      "learning_rate": 2.6708741784852716e-07,
+      "loss": 0.8661,
+      "step": 24470
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.2649308741092682,
+      "learning_rate": 2.6098453509167394e-07,
+      "loss": 0.8434,
+      "step": 24475
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.29732733964920044,
+      "learning_rate": 2.549521244881836e-07,
+      "loss": 0.8796,
+      "step": 24480
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.3012838065624237,
+      "learning_rate": 2.489901888771184e-07,
+      "loss": 0.7311,
+      "step": 24485
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.3150104880332947,
+      "learning_rate": 2.4309873106433395e-07,
+      "loss": 0.75,
+      "step": 24490
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.260047048330307,
+      "learning_rate": 2.3727775382256232e-07,
+      "loss": 0.7938,
+      "step": 24495
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.26424214243888855,
+      "learning_rate": 2.3152725989132869e-07,
+      "loss": 0.8173,
+      "step": 24500
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.29962679743766785,
+      "learning_rate": 2.2584725197705156e-07,
+      "loss": 0.7686,
+      "step": 24505
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.2528585195541382,
+      "learning_rate": 2.202377327528926e-07,
+      "loss": 0.7052,
+      "step": 24510
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.31817159056663513,
+      "learning_rate": 2.1469870485888995e-07,
+      "loss": 0.8064,
+      "step": 24515
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.26599112153053284,
+      "learning_rate": 2.092301709018751e-07,
+      "loss": 0.8348,
+      "step": 24520
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.26814529299736023,
+      "learning_rate": 2.0383213345555595e-07,
+      "loss": 0.8049,
+      "step": 24525
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.2638113796710968,
+      "learning_rate": 1.9850459506041693e-07,
+      "loss": 0.8862,
+      "step": 24530
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.2369757890701294,
+      "learning_rate": 1.9324755822375248e-07,
+      "loss": 0.8741,
+      "step": 24535
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.28597578406333923,
+      "learning_rate": 1.8806102541973346e-07,
+      "loss": 0.6238,
+      "step": 24540
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.30425626039505005,
+      "learning_rate": 1.829449990892906e-07,
+      "loss": 0.788,
+      "step": 24545
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.29849773645401,
+      "learning_rate": 1.7789948164019795e-07,
+      "loss": 0.769,
+      "step": 24550
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.27328890562057495,
+      "learning_rate": 1.7292447544703937e-07,
+      "loss": 0.7794,
+      "step": 24555
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.26762855052948,
+      "learning_rate": 1.6801998285124186e-07,
+      "loss": 0.7316,
+      "step": 24560
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.2702507972717285,
+      "learning_rate": 1.6318600616099241e-07,
+      "loss": 0.8332,
+      "step": 24565
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.2567251920700073,
+      "learning_rate": 1.5842254765135453e-07,
+      "loss": 0.7298,
+      "step": 24570
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.2939104437828064,
+      "learning_rate": 1.5372960956413493e-07,
+      "loss": 0.688,
+      "step": 24575
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.28854793310165405,
+      "learning_rate": 1.491071941080002e-07,
+      "loss": 0.7518,
+      "step": 24580
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.2733674645423889,
+      "learning_rate": 1.445553034584268e-07,
+      "loss": 0.7832,
+      "step": 24585
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.27166128158569336,
+      "learning_rate": 1.4007393975766777e-07,
+      "loss": 0.7416,
+      "step": 24590
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.260300874710083,
+      "learning_rate": 1.3566310511480272e-07,
+      "loss": 0.7861,
+      "step": 24595
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.22208818793296814,
+      "learning_rate": 1.3132280160572105e-07,
+      "loss": 0.7546,
+      "step": 24600
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.2763347029685974,
+      "learning_rate": 1.270530312731055e-07,
+      "loss": 0.7988,
+      "step": 24605
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.2827538847923279,
+      "learning_rate": 1.2285379612648195e-07,
+      "loss": 0.672,
+      "step": 24610
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.27377375960350037,
+      "learning_rate": 1.187250981421195e-07,
+      "loss": 0.7166,
+      "step": 24615
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.26959091424942017,
+      "learning_rate": 1.1466693926311388e-07,
+      "loss": 0.8364,
+      "step": 24620
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.29500630497932434,
+      "learning_rate": 1.1067932139940394e-07,
+      "loss": 0.7982,
+      "step": 24625
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.24503514170646667,
+      "learning_rate": 1.0676224642767184e-07,
+      "loss": 0.884,
+      "step": 24630
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.30973246693611145,
+      "learning_rate": 1.0291571619142624e-07,
+      "loss": 0.7196,
+      "step": 24635
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.30464237928390503,
+      "learning_rate": 9.913973250096907e-08,
+      "loss": 0.7846,
+      "step": 24640
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.2849758565425873,
+      "learning_rate": 9.543429713339546e-08,
+      "loss": 0.8509,
+      "step": 24645
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.2636762261390686,
+      "learning_rate": 9.179941183264372e-08,
+      "loss": 0.7391,
+      "step": 24650
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.29409074783325195,
+      "learning_rate": 8.823507830936216e-08,
+      "loss": 0.8165,
+      "step": 24655
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.2603693902492523,
+      "learning_rate": 8.474129824109222e-08,
+      "loss": 0.8494,
+      "step": 24660
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.27702707052230835,
+      "learning_rate": 8.131807327208529e-08,
+      "loss": 0.7467,
+      "step": 24665
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.2720186710357666,
+      "learning_rate": 7.796540501343595e-08,
+      "loss": 0.8084,
+      "step": 24670
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.27940112352371216,
+      "learning_rate": 7.468329504303206e-08,
+      "loss": 0.8123,
+      "step": 24675
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.2809736132621765,
+      "learning_rate": 7.147174490552132e-08,
+      "loss": 0.7652,
+      "step": 24680
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.29643842577934265,
+      "learning_rate": 6.833075611239469e-08,
+      "loss": 0.8162,
+      "step": 24685
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.329057514667511,
+      "learning_rate": 6.526033014188637e-08,
+      "loss": 0.883,
+      "step": 24690
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.2862626612186432,
+      "learning_rate": 6.226046843904042e-08,
+      "loss": 0.7266,
+      "step": 24695
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.28131476044654846,
+      "learning_rate": 5.933117241571084e-08,
+      "loss": 0.8039,
+      "step": 24700
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.2740655839443207,
+      "learning_rate": 5.647244345049484e-08,
+      "loss": 0.7991,
+      "step": 24705
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.2963801324367523,
+      "learning_rate": 5.368428288881621e-08,
+      "loss": 0.7506,
+      "step": 24710
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.2843016982078552,
+      "learning_rate": 5.096669204287529e-08,
+      "loss": 0.7785,
+      "step": 24715
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.2702232897281647,
+      "learning_rate": 4.831967219168231e-08,
+      "loss": 0.7964,
+      "step": 24720
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.3016199767589569,
+      "learning_rate": 4.574322458097413e-08,
+      "loss": 0.7077,
+      "step": 24725
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.31466734409332275,
+      "learning_rate": 4.3237350423330766e-08,
+      "loss": 0.775,
+      "step": 24730
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.256837397813797,
+      "learning_rate": 4.080205089810884e-08,
+      "loss": 0.7682,
+      "step": 24735
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.28616052865982056,
+      "learning_rate": 3.843732715142489e-08,
+      "loss": 0.7311,
+      "step": 24740
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.24874311685562134,
+      "learning_rate": 3.614318029620533e-08,
+      "loss": 0.7582,
+      "step": 24745
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.27316519618034363,
+      "learning_rate": 3.3919611412153156e-08,
+      "loss": 0.8188,
+      "step": 24750
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.2963423728942871,
+      "learning_rate": 3.1766621545764594e-08,
+      "loss": 0.8122,
+      "step": 24755
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.27651023864746094,
+      "learning_rate": 2.9684211710279128e-08,
+      "loss": 0.774,
+      "step": 24760
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.28626778721809387,
+      "learning_rate": 2.7672382885762788e-08,
+      "loss": 0.8288,
+      "step": 24765
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.2841629683971405,
+      "learning_rate": 2.5731136019058186e-08,
+      "loss": 0.7956,
+      "step": 24770
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.27702802419662476,
+      "learning_rate": 2.3860472023767842e-08,
+      "loss": 0.7156,
+      "step": 24775
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.3124939799308777,
+      "learning_rate": 2.206039178028751e-08,
+      "loss": 0.7434,
+      "step": 24780
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.2639085650444031,
+      "learning_rate": 2.0330896135806184e-08,
+      "loss": 0.7925,
+      "step": 24785
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.28985705971717834,
+      "learning_rate": 1.8671985904289422e-08,
+      "loss": 0.798,
+      "step": 24790
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.29289746284484863,
+      "learning_rate": 1.7083661866446053e-08,
+      "loss": 0.8071,
+      "step": 24795
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.25807228684425354,
+      "learning_rate": 1.5565924769811446e-08,
+      "loss": 0.8328,
+      "step": 24800
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.2961677610874176,
+      "learning_rate": 1.4118775328680886e-08,
+      "loss": 0.7459,
+      "step": 24805
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.28014975786209106,
+      "learning_rate": 1.2742214224126246e-08,
+      "loss": 0.8192,
+      "step": 24810
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.26734206080436707,
+      "learning_rate": 1.143624210402927e-08,
+      "loss": 0.8212,
+      "step": 24815
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.2634017765522003,
+      "learning_rate": 1.0200859582981669e-08,
+      "loss": 0.7237,
+      "step": 24820
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.27349719405174255,
+      "learning_rate": 9.036067242418343e-09,
+      "loss": 0.8578,
+      "step": 24825
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.2613513171672821,
+      "learning_rate": 7.941865630534116e-09,
+      "loss": 0.7642,
+      "step": 24830
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.2503609359264374,
+      "learning_rate": 6.918255262300387e-09,
+      "loss": 0.7935,
+      "step": 24835
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.27115345001220703,
+      "learning_rate": 5.9652366194318235e-09,
+      "loss": 0.7417,
+      "step": 24840
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.29795727133750916,
+      "learning_rate": 5.082810150486283e-09,
+      "loss": 0.7581,
+      "step": 24845
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.2834166884422302,
+      "learning_rate": 4.270976270731585e-09,
+      "loss": 0.685,
+      "step": 24850
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.2606295347213745,
+      "learning_rate": 3.5297353622787406e-09,
+      "loss": 0.821,
+      "step": 24855
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.27347010374069214,
+      "learning_rate": 2.8590877739487205e-09,
+      "loss": 0.7438,
+      "step": 24860
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.2770770788192749,
+      "learning_rate": 2.2590338213890334e-09,
+      "loss": 0.8011,
+      "step": 24865
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.27099478244781494,
+      "learning_rate": 1.729573786990457e-09,
+      "loss": 0.8102,
+      "step": 24870
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.31159690022468567,
+      "learning_rate": 1.2707079199536507e-09,
+      "loss": 0.8513,
+      "step": 24875
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.2697480618953705,
+      "learning_rate": 8.824364362225445e-10,
+      "loss": 0.6736,
+      "step": 24880
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.25522977113723755,
+      "learning_rate": 5.647595185176435e-10,
+      "loss": 0.7454,
+      "step": 24885
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.26916685700416565,
+      "learning_rate": 3.1767731638598916e-10,
+      "loss": 0.8128,
+      "step": 24890
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.27489572763442993,
+      "learning_rate": 1.4118994608458556e-10,
+      "loss": 0.8325,
+      "step": 24895
+    },
+    {
+      "epoch": 3.0,
+      "step": 24897,
+      "total_flos": 3.2538134341012685e+19,
+      "train_loss": 0.8796718681596458,
+      "train_runtime": 1082714.0985,
+      "train_samples_per_second": 0.368,
+      "train_steps_per_second": 0.023
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 24897,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "total_flos": 3.2538134341012685e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}